The 34th Conference on Priorities in Perinatal Care in South Africa was held under the auspices of the Priorities in Perinatal Care Association and sponsored by

Abbvie (Pty) Ltd.

i

Editor’s Note:

The articles included in these Proceedings were, mostly, received electronically and have been included as submitted by the presenter/author. Faxed articles have been retyped.Abstracts were included where articles were not submitted.Articles have not been included for presentations which were withdrawn and not presented at Priorities.In some cases, hyperlinks have been provided to .pdf files as this is how the articles have been submitted for the Proceedings. Click on hyperlink (different colour) in the index to take you to the article as the article is not in the Proceedings.Late submissions received after the Proceedings had been compiled and passwords allocated are included at the end of the Proceedings.

ii

INDEX

QUALITY OF CARE AUDIT AND IMPROVING PERINATAL MORTALITY. DOES IT WORK? Emma Allanson

WHAT IS THE IMPACT OF MULTI-PROFESSIONAL EMERGENCY OBSTETRIC AND NEONATAL CARE TRAINING? Anne-Marie Bergh

ESSENTIAL STEPS IN MANAGING OBSTETRIC EMERGENCIES (ESMOE): IS ON-SITE SATURATION TRAINING EQUIVALENT TO OFF-SITE SATURATION TRAINING? JK Marcus

IS SATURATION TRAINING ESSENTIAL? PRELIMINARY ANALYSIS OF THE ESMOE-EOST SCALE-UP PROGRAMME (abstract). RC Pattinson

CAUSES OF PERINATAL MORTALITY AND ASSOCIATED MATERNAL COMPLICATIONS IN A SOUTH AFRICAN PROVINCE   : CHALLENGES IN PREDICTING POOR OUTCOMES. Emma Allanson

STILLBIRTH TRENDS IN 2013 (abstract). KA Kgomo

SAVING BABIES 2012-2013: NINTH PERINATAL CARE SURVEY OF SOUTH AFRICA (abstract). RC Pattinson (www.ppip.co.za)

PRELIMINARY ADVOCACY FINDINGS ON THE CHOICE OF TERMINATION OF PREGNANCY ACT, COMMON LAW MURDER, CONCEALMENT OF BIRTH OF NEWLY BORN CHILD AND CASE LAW (S V MSHUMPA) (abstract). Luke Lamprecht

BARRIERS TO EARLY ANTENATAL CARE IN SOUTH AFRICA. JD Makin

ELECTRONIC VERSION OF THE NATIONAL BIRTH REGISTER (abstract). CM Bezuidenhout

THE USE OF PHOTOGRAPHS TO EXPLORE THE QUALITY OF CARE FOR NEWBORN INFANTS IN NEONATAL NURSERIES IN KWAZULU-NATAL (abstract). L Haskins

A PROTOCOL FOR THE MEDICAL INVESTIGATION AND MANAGEMENT OF SUSPECTED INFLICTED INFANT HEAD INJURIES (IIHI) (abstract). Lorna Jacklin

HYPERNATRAEMIA IN VERY LOW BIRTH WEIGHT INFANTS ADMITTED AT CHRIS HANI BARAGWANATH ACADEMIC HOSPITAL: INCIDENCE, FLUID MANAGEMENT AND OUTCOME. Kim Barnard

PATHOGENS ISOLATED FROM BLOOD STREAM OF NEONATES DIAGNOSED WITH HEALTHCARE ASSOCIATED INFECTIONS: ANTIMICROBIAL SUSCEPTIBILITY AND CASE FATALITY RATES. Sithembiso Velaphi

RANDOMISED TRIAL TO EVALUATE THE EFFECTS OF BIFIDOBACTERIUM LACTIS IN THE PREVENTION OF NECROTIZING ENTEROCOLITIS IN PRETERM INFANTS. Peter Cooper

MATERNAL HUMAN IMMUNODEFICIENCY VIRUS STATUS AND MORTALITY OF VERY LOW BIRTH WEIGHT INFANTS. Mayowa M. Tiam

iii

THE CUMULATIVE INCIDENCE OF HIV INFECTION IN HIV-EXPOSED INFANTS WITH A BIRTH WEIGHT OF ≤1500G RECEIVING BREAST MILK AND DAILY NEVIRAPINE. Melantha Coetzee

POSTNATAL MOTHER-TO-CHILD TRANSMISSION OF HIV AND HIV-FREE SURVIVAL IN AN HIV-EXPOSED NATIONAL COHORT, SOUTH AFRICA, DECEMBER 2012-SEPTEMBER 2014 (abstract). Jackson DJ

FEAR OF KNOWING THE CHILD IS HIV POSITIVE AND STIGMA FROM DISCLOSURE REDUCE ACCESS TO EARLY INFANT DIAGNOSIS IN THE RURAL COMMUNITIES OF OR TAMBO DISTRICT, SOUTH AFRICA: A QUALITATIVE EXPLORATION OF MATERNAL PERSPECTIVE (abstract). VINCENT OLADELE ADENIYI

ANAEMIA IN PREGNANCY (abstract). K Tunkyi

THE IMPACT OF HIV INFECTION ON OBSTETRIC HAEMORRHAGE AND BLOOD TRANSFUSION IN SOUTH AFRICA. S Fawcus

SYPHILIS IN HIV-INFECTED MOTHERS AND INFANTS: RESULTS FROM THE NICHD/ HPTN 040 STUDY. GB Theron

DEVELOPMENT OF A LOW COST VITAL SIGNS DEVICE TO DETECT PRE-ECLAMPSIA AND SHOCK (abstract). Hannah L Nathan

THE EFFECT OF CALCIUM SUPPLEMENTATION ON BLOOD PRESSURE IN NON-PREGNANT WOMEN WITH PREVIOUS PRE-ECLAMPSIA: AN EXPLORATORY, RANDOMIZED PLACEBO CONTROLLED STUDY (WHO STUDY A65750). M Singata

EFFECT OF DIFFERENT POSITIONS ON MATERNAL AND FETAL HEART RATES AT TERM (abstract). HJ Odendaal

A BASELINE EVALUATION OF CARE PROVIDED AT NEONATAL NURSERIES IN KWAZULU-NATAL (abstract). C. Horwood

KWAZULU NATAL INITIATIVE FOR NEWBORN CARE: A QUALITATIVE IMPROVEMENT PROGRAMME FOR NEONATAL CARE IN DISTRICT HOSPITALS. Dolly Nyasulu

THE IMPACT OF AN EDUCATIONAL COUNSELLING ON ACCEPTABILITY OF DONATED BREAST MILK AT A SEMI-RURAL REGIONAL HOSPITAL, NORTHERN KWAZULU-NATAL. N. Kapongo

LATE NEONATAL DEATHS AFTER A NORMAL HOSPITAL STAY AND DISCHARGE. FS BONDI

CHANGING PATTERN OF NEONATAL DEATHS IN MADADENI DISTRICT HOSPITAL- A REVIEW OF 1071 DEATHS. FS BONDI

EVERY NEWBORN: GLOBAL RESULTS & SOUTH AFRICA’S CONTRIBUTION TO ENDING PREVENTABLE DEATHS. M Kinney

iv

A COMPARISON OF BOOKED AND UNBOOKED PREGNANT WOMEN PRESENTING IN LABOUR AT THREE HEALTH FACILITIES IN REGION F, INNER CITY OF JOHANNESBURG. Siphamandla GumedePRACTICE OF OPERATIVE VAGINAL DELIVERY: A QUESTIONNAIRE BASED STUDY (abstract). JM Devjee

AN AUDIT OF VACUUM DELIVERIES AT ZITHULELE HOSPITAL. CB Gaunt

SEVERE MATERNAL OUTCOME FROM CAESAREAN SECTION RELATED HAEMORRHAGE IN SOUTHERN GAUTENG, SOUTH AFRICA. Maswime TS

EVALUATION OF HEALTHCARE SYSTEMS IN THE DECISION TO INCISION INTERVAL FOR CAESAREAN SECTIONS AT CHBAH (abstract). Williams M

NEONATAL DEATHS IN CHILDREN’S WARDS: THREE TRIENNIA OF CHILD PIP DATA 2005-2013 (abstract). Dr CR Stephen

SAVING MOTHERS 2011-2013: SIXTH REPORT ON CONFIDENTIAL ENQUIRIES INTO MATERNAL DEATHS. OVERVIEW (abstract). RC Pattinson for NCCEMD

DEATHS DUE TO ECTOPIC PREGNANCY AND MISCARRIAGE IN SOUTH AFRICA: FINDINGS AND RECOMMENDATIONS FROM SAVING MOTHERS 2011-13. N.Moran

AUDIT OF MATERNAL DEATH AT NATALSPRUIT HOSPITAL (abstract). Dr B Uzabakiriho

REDUCTION IN INDIRECT OBSTETRIC DEATHS WITH INCREASED HIV TREATMENT IN SOUTH AFRICA: A 5-YEAR AUDIT (abstract). Dr Vivian Black

MATERNAL NEAR MISS AND MATERNAL DEATH IN THE PRETORIA ACADEMIC COMPLEX – A POPULATION-BASED STUDY. L Langa-Mlambo

RETHINKING IUCD USE AS A CONTRACEPTIVE OPTION IN KWAZULU-NATAL, SOUTH AFRICA (abstract). Pinky Phungula

THE PAEDIATRIC ART DATA MANAGEMENT TOOL (PADMT): USING POINT-OF-CARE TECHNOLOGY IN THE FORM OF AN ELECTRONIC MEDICAL RECORD TO ENHANCE CLINICAL QUALITY SHORTFALLS IN RESOURCE-POOR PAEDIATRIC ARV SETTINGS (abstract). Kim Harper

SEX DIFFERENCES IN EEG ACTIVITY BETWEEN HEALTHY NEWBORNS DURING THE FIRST DAYS OF LIFE (abstract). WP Fifer

EXPERIENCE WITH INDUCED HYPOTHERMIA AT THE CHRIS HANI BARAGWANATH ACADEMIC HOSPITAL: PRELIMINARY RESULTS. Firdose Nakwa

THE LIFELONG EFFECTS OF EARLY NEONATAL AND CHILDHOOD ADVERSITY AND TOXIC STRESS. - MATERNAL INFANT SEPARATION AND THE BUSY NICU ENVIRONMENT (abstract). Vanessa Booysen

A DESCRIPTIVE STUDY OF THE PROFILE OF SUSPECTED INFLICTED CHILD DEATHS IN JOHANNESBURG OVER A FOUR YEAR PERIOD (abstract). Luke Lamprecht

v

LONG TERM DEVELOPMENTAL OUTCOMES OF CHILD VICTIMS OF INFLICTED INFANT HEAD INJURIES OVER A 10 YEAR PERIOD IN A JOHANNESBURG CHILD ABUSE CLINIC. Lorna Jacklin

vi

WHAT IS THE IMPACT OF MULTI-PROFESSIONAL EMERGENCY OBSTETRIC AND NEONATAL CARE TRAINING?

Anne-Marie Bergh, Shisana Baloyi, Bob PattinsonMRC Research Unit for Maternal and Infant Health Care Strategies, University of Pretoria

BackgroundIn many low- and middle-income countries there has been insufficient progress with the Millennium Development Goals 4 and 5 aimed at reducing maternal and under-five (including neonatal) mortality. In high-income countries like the UK and the US there has been concern about substandard obstetric care and a high incidence of medical errors.

In recent years there has been a training shift from individual technical perfection to better team coordination for patient safety. There has also been a proliferation of studies in emergency obstetric and neonatal care (EmONC) training and there was a need to map the EmONC training landscape. The research question was whether studies have shown a change in provider behaviour and/or patient outcomes (morbidity and/or mortality) after EmONC training.

MethodA review was done of peer-reviewed articles in English. Other important inclusion criteria were that the training had to have been multi-professional and should have covered at least three emergency types. The search included a variety of databases and yielded 4,235 hits. Eventually 72 papers were identified for detailed analysis and ultimately 35 papers representing 23 studies or trials were included in the review.

A modified version of Kirkpatrick’s model1 for the evaluation of training programmes was used for categorising the effect of EmONC training programmes included in the review. This is depicted in Table 1.

1

Table 1. Adapted Kirkpatrick model used for evaluating the studies

Level

Kirkpatrick Adapted Level

Description

1 Reaction Satisfaction 1 Same

Indi

vidu

al

2 Learning Effectiveness

2a Attitudes & perceptions

2b Knowledge2c Skills

3 Behaviour Efficiency 3a Perceptions of change3b Skill retention3c Change in practice

Team4 Outcome Impact 4a Organisation

4b Processes4c Patient outcomes

ResultsOf the 23 studies, 13 were from high-income countries (HICs) and 10 from low- and middle-income countries (LMICs). The HICs include the US (n=6), the UK (n=3), Europe (n=3) and Australia (n=1). For the studies from LMICs, 7 were from sub-Saharan Africa, 1 from Latin America and 2 from Asia. The majority of studies (n=16) were before-after observational studies. There were 2 quasi-experimental studies and 5 randomised controlled trials (RCTs). The range of hospitals included in the studies ranged from tertiary teaching hospitals to community and field hospitals. Only 2 RCTs included more than 20 hospitals.

In the studies included in the review there were three implicit types of research questions posed around the effect of training:

1. Does EmONC skill training work (regardless of training method)? (n = 16)2. Which skill training methods work better? (n = 2) (Skill simulation and

‘fire drills’/ scenarios plus repeats/refreshers, with the focus individual obstetric skills)

3. How can EmONC skill training be complemented to improve processes and reduce errors? (n = 5) (Teamwork and communication training, with a focus that goes beyond individual skills)

In HIC studies the focus of training was on reducing errors to improve patient safety, on reducing morbidity and on minimising litigation. The LMIC studies

2

focused on improved capacity for safe clinical skills in order to reduce maternal and neonatal mortality and morbidity.

The demonstrated impact of EmONC training at level 4c (patient outcomes) can be divided into three categories:

1. Adverse Outcomes Index (weighted maternal & neonatal outcomes) (n=4)2. Significant infant outcomes:

Reduction in low Apgar scores (n=4)Reduction in hypoxic-ischaemic encephalopathy (n=1)Reduction in cord pH <7 (n=1)Increase in cord lactates (n=1)Reduction in birth trauma (n=3)Reduction in neonatal mortality (in certain types of hospitals) (n=1) No effect on stillbirths (n=2)

3. Significant maternal outcomes:Reduction in postpartum haemorrhage (n=2)Reduction in maternal mortality (in certain types of hospitals) (n=1)

ConclusionKirkpatrick’s model of evaluating training programmes was useful for demonstrating the impact of EmoNC training. Our review has found that EmONC skill training does make a difference right down to patient-outcome level. There is a preference for continuous onsite training in local ‘in-house’ environments, using realistic, low-tech equipment for simulations. Therefore, all facilities providing maternity services should regularly conduct ‘fire drills’ and regular EmONC refresher training. However, scaling up a training package and maintaining high quality are problematic in all countries. In order for regular onsite training and drills to be sustainable, there is a need to look at quality assurance and monitoring and evaluation mechanisms beyond the short term that should include all levels of the health system and the individual institutions.Our review, using the Kirkpatrick model, focused on the question “Does training work?” There is a need to consider other models that will enable us to answer questions like “How does the training work?” and “Why does the training work?” This is necessary for understanding the underlying mechanisms that hinder or enable achievement of higher-level programme outcomes.

3

Reference1 Kirkpatrick D. Evaluating training programs: the four levels. 2nd ed. San Francisco, CA: Berrett-Kochler Publishers; 1998.

4

ESSENTIAL STEPS IN MANAGING OBSTETRIC EMERGENCIES (ESMOE): IS ON-SITE SATURATION TRAINING EQUIVALENT TO OFF-SITE SATURATION TRAINING?

JK Marcus, JD Makin, A-M Bergh, CM Bezuidenhout, RC PattinsonMRC Maternal and Infant Health Care Strategies Unit Department of Obstetrics and Gynaecology, University of Pretoria

Introduction Emergency Obstetric Care training has shown to improve the knowledge and skills of maternity care providers. The Essential Steps in Managing Obstetric Emergencies (ESMOE) programme has demonstrated this improvement in the South African context (Frank et al, 2009). The scale up of this training is being done using either an on-site or an off-site saturation training approach. The on-site training entails the training of a doctor-midwife dyad as facilitator-mentors from each facility who then train other providers in their facility using Emergency Obstetric Simulation Training (EOST). Off-site training entails all maternity care providers at a facility being trained at venues away from their facilities over either 2 or 3 days. It was previously shown that off-site training is 5 times more costly than onsite training (Bezuidenhout et al, 2014). On-going engagement with emergency drills is needed to sustain the effect of the initial training and to maintain the skills and team work within the facility (Siassakos et al, 2009).

AimThe aim of the study was to compare the ability of health care professionals to manage an obstetric emergency who have had the either the on-site or off-site saturation training in their district.

MethodAssessment of the ability of health care professionals to manage an obstetric emergency was assessed and an unannounced EOST drill was conducted by a national ESMOE trainer (JKM), assisted by either a member of the district MCWH team or a DCST member. The facilities were scored using the standard EOST exercise scoring instrument which takes into account the technical aspects of executing the clinical skill, communication and team work. Enablers and barriers to performing the drills were also investigated and these were correlated to the scores at facility level and district level.

5

6

Results 30 district hospitals were visited from 8 districts throughout South Africa. Fifteen of the facilities had previously received off-site saturation training and the other 15 received on-site saturation training. Results show that in hospitals where the on-site saturation training had taken place scored higher on average for the unannounced drills (45% vs 32%) than the off-site saturation training hospitals. The Student’s T-test showed a P-value of 0.023, which is a significant difference in the performance of the two groups, as shown in Figure 1. The range of score for the on-site trained districts was 24-61% (n=15) and for the off-site trained districts the range was 10-72% (n=15).Figure 1

Off-site On-site0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

32%

45%

Mean scores

There was no statistically significant difference in the mean scores of the districts who had the two different types of training as demonstrated in Figure 2.

7

Figure 2

1 2 30.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

100.00

0 0 0 0 0

28

43

5359 61

Off-site training district average %On-site training district average %

P=0.29

When comparing the differences in the clinical aspects of the drills there was no statistically significant difference amongst the districts which had either of the training interventions (p=0.07). The same holds true for how the drill was executed taking into account communication, teamwork, documentation and sequence of the drill (p=0.17). These finding are tabulated in Table 1.Table 1

Clinical Teamwork and communication

Off-site 33% 35%

On-site 44% 44 %

P=0.07 P=0.17

DiscussionIt was evident that facilities in both groups where there were frequent, documented drills scored higher than facilities where drills were not frequently performed or not done at all. Districts where there was good support from local DCST members and/or MCWH co-ordinators performed better in the drills. The presence of dedicated, enthusiastic clinicians at facility level appears to positively influence the technical correctness of knowledge and skills as well as the team work aspects of the drills in addition to its frequency. This was more so the case in the facilities where a doctor-midwife dyad attended the on-site

8

training who would take the ultimate responsibility for the drills to be done. In facilities where off-site training was received, there was a lesser likelihood of an identified person or people who would assume this responsibility. The existence of in-house will and leadership to appear to positively influence situating the EOST drills on the institutional agenda. The performance of EOST drills is a component of the key performance indicators of hospital managers/CEOs. It is important to not “set up” the drill in order to maintain high fidelity of the exercise. Staff need to retrieve the required care equipment and supplies as the “emergency” unfolds to create awareness of time sensitive actions such as starting intravenous infusions and administering oxygen and drugs. These are opportunities to test the system of responding to an emergency and whether the appropriate supplies, drugs and equipment are available.

Conclusions The data suggests on-site saturation training is at least equivalent to, if not better than, off-site saturation training with respect to health care professionals managing an obstetric emergency. In on-site saturation training there is less service disruption as personnel are not required to be away from work stations en masse for prolonged periods of time. Adequate on-going support from DCST members or MCWH coordinators as well as institutional managers is essential if either approach is to be successful in bringing about the change in the behaviours of maternity care providers with a view to reducing maternal and perinatal morbidity and mortality.

ReferencesBezuidenhout CM, Makin JD, Pattinson RC. Scaling up of ESMOE in South Africa: What is it going to cost and what type of training are we going to use? Proceedings of the 33rd Conference on Priorities in Perinatal Care in South Africa, 11th-14th March 2014, Cape Town, South Africa.Crofts JF, Ellis D, Draycott TJ, Winter C, Hunt LP, Akande VA. Change in knowledge of midwives and obstetricians following obstetric emergency training: a randomised controlled trial of local hospital, simulation centre and teamwork. BJOG 2007; 114: 1534-1541.

9

Draycott T, Sibanda T, Owen L, Akande V, Winter C, Reading S, Whitelaw A. Does training in obstetric emergencies improve neonatal outcome? BJOG 2006; 113:177–182Frank K, Lombaard H, Pattinson RC. Does completion of the Essential Steps in Managing Obstetrics Emergencies (ESMOE) training package result in improved knowledge and skills in managing obstetric emergencies? SAJOG. 2009; 15(3):94-99.Osman H, Campbell OMR, Nassar AH. Using emergency obstetric drills in maternity units as a performance improvement tool. Birth 36:1 March 2009: 43-50 Siassakos D, Crofts JF, Winter C, Weiner CP, Draycott TJ. The active components of effective training in obstetric emergencies. BJOG 2009; 116: 1028-1032.

10

IS SATURATION TRAINING ESSENTIAL? PRELIMINARY ANALYSIS OF THE ESMOE-EOST SCALE-UP PROGRAMME

RC Pattinson # , JD Makin#, C Bezuidenhout#, J Marcus#, J Lambert*, Y Pillay$, N van den Broek*, J Moodley@ #MRC Maternal and Infant Health Care Strategies unit, University of Pretoria, $National Department of Health, * Liverpool School of Tropical Medicine, ESMOE Board@

Objective: To assess whether training eighty percent plus of health care professionals in maternity unit in emergency obstetric and neonatal care reduces maternal and perinatal mortality

Method: 12 “most in need districts” were selected for off-site saturation training of health care professionals in ESMOE-EOST using a stepped wedge design. A preliminary analysis was conducted of the first 6 districts to have completed their saturation training.

Results: In every district the target of training 80% or more health care professionals involved in the care of pregnant women and their babies was achieved. Most delegates were very or extremely happy with the ESMOE course and felt the course had been beneficial to them. The knowledge as measured by a multiple choice questionnaire significantly increased for all professional levels, (doctors, 49-67%; advanced midwives (33-48%) and professional nurses with midwifery 27-42%). The skills were also significantly improved (doctors 44-78%; nurses 40-66%).The number of births in the pre-saturation and post-saturation training group at the time of analysis was 29564 and 22444 births respectively. The stillbirth rates before and after were 27.5/1000 births and 25.5/1000 births respectively (NS). The neonatal death rates before and after 25.5/1000 live births and 16.5/1000 live births respectively (p<0.001). There were 93 maternal deaths before saturation training and 49 after saturation training. The institutional Maternal Mortality Ratio’s before and after were 323.5/100000 live births and 224.0/100000 live births, (p<0.001).

Conclusion: Preliminary information suggests that saturation training using the ESMOE-EOST programme significantly improves knowledge and skills of the health professionals and is associated with a reduction in neonatal and maternal mortality.

11

STILLBIRTH TRENDS IN 2013

KA Kgomo, Y Adam

Introduction:There are several causes of stillbirths and this contributes to the Perinatal Mortality. Understanding the trends over time may help in determining the causes of SB and the future directions for research. We postulate that there may be a seasonal difference to the SBR over the epidemiological weeks and that this may mirror the seasonal variations in viral infections. The Objective of this study was to evaluate the trends of Stillbirths over the “epidemiological weeks” in 2013.

Methods:This is a cross sectional study using routinely collected data in the Department of Obstetrics & Gynaecology, CHBAH.

Results:There were 22603 deliveries in 2013. The LBW rate (>500g) was 19.19, the LBW rate (>1000g) was 17.05. The SBR (>500g) was 26.06, the SBR (>1000g) was 14.47. THE PMR (>500g) was 38.80 and the PMR (>1000g) was 24.78. The peak in LBW rates were highest for weeks 22-24. The peak in SBR was in week 10. The SBR (>500g) was 42.89 at week 10. The SBR (>1000g) was 36.12 at week 10.

Conclusion:The SBR is at its highest at week 10. It is noteworthy that this coincides with the peak incidence of Respiratory syncytial virus infection. Similarly, the LBW rates peak at weeks 22-23 and this mirrors the Influenza season. It remains to be seen whether these concurrences are of epidemiological significance, or merely coincidental.

12

SAVING BABIES 2012-2013: NINTH PERINATAL CARE SURVEY OF SOUTH AFRICA

RC Pattinson#, N Rhoda* for the PPIP users# MRC Maternal and Infant Health Care Strategies Unit, University of Pretoria, *Futures Group

Full reports available on PPIP website: www.ppip.co.za

Objective: To assess the quality of perinatal care in South Africa

Method: Data entered into PPIP from 1/1/2012 to 31/12/2013 by PPIP users was analysed.

Results: PPIP gave good quality of care assessments in Free State, Mpumalanga, North West, and Western Cape; moderate in Limpopo and Northern Cape and poor in Eastern Cape, Gauteng and KwaZulu-Natal.Perinatal care index highest in district hospitals and the mortality rates per birth weight category are higher than expected for district hospitals in the 1000g-1999g categories, especially in early neonatal deaths. Unexplained stillbirth remains the largest category of perinatal deaths. Intrapartum birth asphyxia is the most common category in fresh stillbirths in CHCs and district hospitals. Almost half of the deaths due to intrapartum asphyxia were thought to be probably preventable; the common problems being with fetal monitoring, use of the partogram and the second stage of labour. Inadequate facilities were the most common avoidable factor in spontaneous preterm labour. In almost a third of babies dying due to complications of hypertension, hypertension was detected but NOT acted upon. In almost one in five deaths which were unexplained stillbirths the patient was reported as not responding to poor fetal movements. Most perinatal deaths occur in district hospitals, as well as most births; mortality rates were the highest in district hospitals. BUT for the average district hospital delivering 2000 babies per year a live born baby between 1000g and 1999g will be delivered once every 5 days, and 12% (one in eight will be an early neonatal death). If the hospital delivers 500 live babies per year, then they will deliver a live born between 1000g and 1999g about once every 20 days and two will die per year. Most caesarean sections are performed in district and regional hospitals with the regional hospitals averaging 4 per day and the district hospitals averaging 1 per day. If the hospital does less than 500 deliveries per year, it will likely do less than 1 caesarean section per week.

Discussion: The issue of what resources and what level of resources should be available for what work load for safe and cost-effective maternity care will be discussed.

13

PRELIMINARY ADVOCACY FINDINGS ON THE CHOICE OF TERMINATION OF PREGNANCY ACT, COMMON LAW MURDER, CONCEALMENT OF BIRTH OF NEWLY BORN CHILD AND CASE LAW (S V MSHUMPA)

Luke Lamprecht*, Sheri Errington, Lorna Jacklin

In work with female reproductive rights and health there is a complex child rights issue that emerges that appears to have fallen outside of the main discourse on child safety. The Choice of Termination of Pregnancy Act provides conditions whereby TOP’s can be legally performed and also the penalty when they are not. A challenge to the penalty clauses was brought about by the findings in S v Mshumpa where the court found that until a foetus takes a breath it is not a legal person and therefore the common law of murder cannot apply to cases of TOP. As a result we are left with a rather obscure law that belies the true extent and severity of cases of possible infanticide. In South Africa, section 113 of the General Law Amendment Act, 1935, amended by the Judicial Matters Amendment Act 66, 2008, creates the offence of concealing the birth of a child: 113. (1) Any person who, without a lawful burial order, disposes of the body of any newly born child with intent to conceal the fact of its birth, whether the child died before, during or after birth, shall be guilty of an offence and liable on conviction to a fine or to imprisonment for a period not exceeding three years. (2) A person may be convicted under subsection (1) although it has not been proved that the child in question died before its body was disposed of.

In reviewing these cases from the Johannesburg Mortuary it became clear that some of the cases recorded under this section never enter the child death statistics and many of these dead and disposed of babies are of a gestational age and weight and may have been murdered.

This paper will present the preliminary findings with a full overview of all cases under section 113 for one year and extract those where the infants were viable and question the ability of the law to deal with these cases in any constructive way. From this an advocacy agenda will be developed to make visible these thrown away babies.

14

BARRIERS TO EARLY ANTENATAL CARE IN SOUTH AFRICA

Diane N. Haddad*a, b, Jennifer D. Makin b , Robert C. Pattinsonb, and Brian W. Forsyth b, c

a. Doris Duke International Clinical Research Fellowshipb. University of Pretoria MRC Maternal and Infant Healthcare Strategies, Pretoria, South Africa c. Yale University Department of Pediatrics, New Haven, Connecticut, United States

Introduction: Antenatal care provides an important opportunity to address major causes of maternal and infant mortality in sub-Saharan Africa [1]. Antenatal care (ANC) allows for necessary HIV screening and early ANC allows for initiation of antiretroviral (ARV) therapy. ANC also provides an opportunity to screen for conditions such as preeclampsia syphilis and, Rh disease and to provide appropriate guidance regarding pregnancy-warning signs

However, the majority of women (56%) in South Africa do not present for care before twenty weeks gestational age2. A number of factors have been identified as potential barriers to care including transportation3, household commitments4, under-resourced clinics with increased waiting times4, and lack of perceived benefit5 in addition to delayed booking at clinic presentation6. Additionally in sub-Saharan Africa, cultural superstitions about jealousy and bewitching are reported to delay presentation.4

The aim of this study was to further explore the barriers encountered by women when seeking ANC in an environment of changing healthcare policy.

Materials and Methods: In order to further understand the complex factors influencing a woman’s decision

to seek care, individual qualitative interviews were conducted with pregnant women presenting for antenatal care at Phomolong clinic near Kalafong Hospital in Pretoria in November-December 2013.The results were presented at Priorities 2014. From the qualitative results a quantitative questionnaire was developed and administered in the post-natal wards at Kalafong Hospital.

Results Surveys were conducted in the post-natal wards with 204 women aged 18-42

(mean 28.5) Seventy seven percent of women had been pregnant before.

Table 1: Patient Survey Demographics (n=204)Age, mean (SD) 28.5 (6.4)Gravidity, median [IQR] 2 [1,3]Parity, median [IQR] 1 [0,2]Level of Schooling, Grade, mean (SD) 11 (1.5)

15

Socioeconomic Score [1-5], mean (SD) 4 (1.5)

The average presentation at first ANC was 19.1 weeks although the average first knowledge of pregnancy was 9.3 weeks (Table 2). The majority of women were aware that they should present before 12 weeks. Eighteen percent reported coming to the clinic to book and subsequently being turned away to come back later. The average delay in booking in these cases was 2.5 weeks. Nineteen percent described traveling to the clinic as difficult One hundred percent were tested for HIV during pregnancy or hospital delivery. The majority had previously been tested for HIV. Twenty six percent interviewed were HIV positive. The majority reported knowing they would be tested for HIV at the clinic but only 45.8% reported knowledge of opt-out practices.

The mean Patient-Provider Relationship Scale7 was 44.3.The majority of pregnancies were unplanned and 20% of women had contemplated termination during the pregnancy.

Table 2: Antenatal Care Statistics from Survey Data (n=203)Attended at least 1 clinic visit, n (%) 200 (98.5)GA at first ANC clinic visit, mean (SD) 19.1 (7.7)GA at first pregnancy knowledge, mean (SD) 9.3 (6.4)Knowledge of 12 weeks recommendation, n (%) 123(65)Patients described transport as ‘difficult’, n (%) 39 (19.1)Patients who experienced booking delay, n (%) 36 (17.6)Average booking delay (weeks), mean (SD) 2.5(2.2)Pt provider relationship scale, mean (SD) 44.3 (9.43)Patients tested for HIV, n (%) 203 (100)HIV Positive, n (%) 53(26.1)Patients Previously Tested for HIV, n (%) 175 (86.2)Knowledge of HIV testing guidelines, n (%) 181(88.7)Knowledge of Provider Initiated Testing, n (%) 91 (46)Unplanned pregnancy, n (%) 125 (61.9)Contemplated Termination, n (%) 40 (19.8)

Bivariate analysis exploring factors associated with earlier gestational age at first presentation is shown in Table 3. Women were more likely to present for care early if the pregnancy was planned (p=0.013) and less likely if they had at any point contemplated termination (p=0.021). Earlier presentation was also significantly associated with age less than 21 years and higher levels of education. HIV status had no effect on first presentation to antenatal care. However, if women knew they would be tested for HIV at antenatal care, they were significantly less likely to present before 14 weeks gestational age (p=0.013)

16

Table 4: Factors Influencing Gestational Age at First GA at 1st ANC Value P-value Age>=21, mean (SD) 18.5 (7.7) 0.021<21, mean (SD) 22.5(8.4) HIV Status HIV positive, mean (SD) 18.6 (7.3) 0.673HIV negative, mean (SD) 19.2 (8.0)Knowledge of HIV Testing at ANC Yes, mean (SD) 19.2 (7.7) 0.280No, mean (SD) 17.2 (8.5) Women presenting after 14 weeks Yes, frequency (%) 131(93.6) 0.013Planned vs. Unplanned Planned, mean (SD) 17.3 (7.2) 0.013Unplanned, mean (SD) 20.4 (8.1) Women presenting after 20 weeks Unplanned, frequency (%) 67 (69.1) 0.012Contemplated TerminationYes, mean (SD) 21.8(7.7) 0.021No, mean (SD ) 15.5 (7.8)

The prevalence of previously identified community held beliefs also is shown in Table 4. Most women endorsed community perceptions that knowledge of HIV positive status could cause psychological stress harmful to the baby. Forty nine percent of women confirmed the presence of a community belief that it is better not to know their HIV status if they are feeling healthy.

Table 4: Community Perceptions It is better not to know HIV status if feeling healthy 49%Knowing HIV status can cause unnecessary stress which can harm the baby 56%Only reason to go to clinic to get antenatal card to deliver baby at hospital 45%It is better not to go to clinic early so other people do not see the pregnancy 41%If others see me at clinic early, they may become jealous and harm my baby 49%

Conclusions1. The majority of women despite being aware that they should attend ANC early do not

do so.2. Women present earlier to care if the pregnancy is planned and wanted.3. Younger women and those with less schooling tend to present late.4. Negative perceptions regarding the benefit to be derived from presentation persist in

a substantial proportion of women.5. Fears regarding HIV –related stigma and psychological stress still persist and hinder

access to care.6. It appears that clinic-related factors (good patient provider relationship scores and

reduced booking delays) do not appear to play such a large role anymore.7. Cultural concerns about harm and jealousy still prevail in the community.

References:

17

1. Trends in Maternal Mortality: 1990 to 2010 (WHO, UNICEF, UNFPA and The World Bank). South Africa: Country Profile Available from:<http://www.who.int/gho/countries/zaf/country_profiles/en/.>

2. District Health Information System (DHIS) Database Extracted January 2014 ed. South Africa 2012/2013. <http://hisp.org>

3. Gabrysch, S. and O.M. Campbell, Still too far to walk: literature review of the determinants of delivery service use. BMC Pregnancy Childbirth, 2009. 9: p. 34.

4 Brighton, A., et al., Perceptions of prenatal and obstetric care in Sub-Saharan Africa. International Journal of Gynecology & Obstetrics, 2013. 120(3): p. 224-227.

5 Finlayson, K. and S. Downe, Why do women not use antenatal services in low- and middle-income countries? A meta-synthesis of qualitative studies. PLoS Med, 2013. 10(1): p. e1001373.

6 Solarin, I. and V. Black, "They told me to come back": women's antenatal care booking experience in inner-city Johannesburg. Matern Child Health J, 2013. 17(2): p. 359-67.

7. Barry, O.M., et al., Development of a measure of the patient-provider relationship in antenatal care and its importance in PMTCT. AIDS Care, 2012. 24(6): p. 680-6.

18

ELECTRONIC VERSION OF THE NATIONAL BIRTH REGISTER

CM Bezuidenhout, JD Makin, RC Pattinson, T Liabsuetrakul* & T Prappre*MRC Maternal and Infant Health Care Strategies Research Unit and Obstetrics and Gynaecology Department, University of Pretoria*Epidemiology Unit, Faculty of Medicine,Prince of Songkla University, THAILAND

BACKGROUND AND AIM: The National Birth Register has been introduced to all health facilities in South Africa. One of the aims of the birth register is to establish a one-stop site for all data which include: DHIS, HAST and PPIP.An electronic version of the national birth register has been developed (which is a mirror of the book version). Required monthly data that needs to be submitted to DHIS and entered into PPIP can be printed as well as individual patient records. This presentation is to demonstrate the computerised birth register.

METHOD:The register was adapted from a WHO version which was developed in conjunction with the Prince of Songkla University, Thailand. The system is a web-based tool that works through the internet network. It can be operated with a notebook, netbook, desktop computer or mobile phone. The system has been developed to support mobile devices using Android. It also works with Windows or Mac computers/notebooks.All possible measures were put in place to ensure security of maternal records.

RESULTS:The program is running at Kalafong and Mamelodi hospitals and Stanza Bopape MOU.

The future:Plans to link up with DHIS so that accurate data will be available 24/7.

If you would like to be part of the testing phase, please e-mail cathy.bezuidenhout@up.ac.za

19

THE USE OF PHOTOGRAPHS TO EXPLORE THE QUALITY OF CARE FOR NEWBORN INFANTS IN NEONATAL NURSERIES IN KWAZULU-NATAL.

L. Haskins, D. Nyasulu, Z. Mbatha and C. HorwoodUniversity of KwaZulu-Natal, Centre for Rural Health (CRH)

BackgroundThe KwaZulu-Natal Initiative for Newborn Care (KINC) is a programme implemented in all 39 district hospitals in KZN. This approach includes training health workers in the KINC guidelines, providing ongoing mentoring and support, peer-to-peer learning, and accreditation of hospitals. An evaluation at baseline, mid-point and post intervention is underway. The evaluation consists of a facility review, review of quality of care of neonates in the nursery and a record review. In addition, photographs of all the neonatal nurseries are taken. Information captured by photographs adds to the detail about quality of care in the neonatal nursery that could not be captured through qualitative or quantitative methods.

MethodologyAt baseline evaluation, photographs were taken in all neonatal nurseries in all district hospitals. The photographs were then printed for analysis. An analysis framework was developed by three neonatologists using a subset of photographs. The framework looked at aspects of staffing including circulation space, administration space and infection prevention; patient aspects including naming of babies, neurodevelopmental cares, clinical records, individual infection prevention, and a space to accommodate the mother at the bedside; aspects of how the beds and equipment is set up and other aspects e.g. cleanliness, resuscitation, etc. This framework was then applied to analyse the photographs from all district hospitals in KZN.

ResultsA total of 211 photographs were taken and analysed. Both positive and negative aspects of care were observed. Some nurseries were well laid out with space and equipment available to render good care. Attention to neurodevelopment of infants were observed. Conversely, incubators were incorrectly aligned in 24 of the 39 hospital nurseries and there was poor spacing between nursery beds in 26 hospitals (see photographs). Nursery equipment was not ready for use in many hospitals and in some nurseries large equipment was stored in the nursery contributing to the overcrowding (see photographs).

ConclusionsUse of photograpahs revealed deficiencies in provision of care to newborn that were not identified using other data collection methods. It is possible through photographs to identify shortcoming in care rendered in neonatal nurseries.

20

A PROTOCOL FOR THE MEDICAL INVESTIGATION AND MANAGEMENT OF SUSPECTED INFLICTED INFANT HEAD INJURIES (IIHI)

Lorna Jacklin*, Luke Lamprecht, Sheri Errington

Inflicted Infant Head Injuries (IIHI)) have been recognised in literature since Caffey in 1946i and results from the violent shaking of infants by caregivers. The resultant brain injuries may cause haemorrhages, oedema, increased intracranial pressure, anoxia, diffuse axonal injury and atrophy. In addition to the intracranial injuries diagnosis may include retinal bleeding and a range of fractures in the infant.ii Despite a vast amount of literature on IIHI, major documented controversies exist. These range from disputing the existence of shaken baby syndromeiii and the mechanism of the injuryiv to the severity and number of symptoms that need to be present for a definitive diagnosis. While there are various differential diagnoses for the injuries in isolation, IIHI is seen as a syndrome resulting in a wide spectrum of symptoms. A review of the literature shows that IIHI exists but that its true prevalence is unknown. It is often misdiagnosed or over/underdiagnosedv with various acute and chronic symptoms, injuries and prognosesvi. The result is perilous as IIHI may cause acute traumaviii, varying degrees of disabilityviii or deathix. The aim of this paper is to describe the mechanism of, and injuries to, the developing brain in IIHI. Brain injury in childhood results in a variety of potential consequences for the child’s subsequent development. These cases need to be investigated from a medico-legal perspective and this paper will present a protocol for the medical investigation that can be applied to both the medical management and the forensic investigation process.

i. Monteleone, J.A. Child Maltreatment: A Clinical Reference Guide (2nd Edition). St Louis. G.W. Medical Publishing, Inc. 1998 Chpt 6

ii. Peinkofer, J.R. Silenced Angels. Westport: Auburn House: 2002 pg 67-84iii. Glueck and Cihak, New Scientific Evidence Refutes Existence of Shaken

Baby Syndrome. Thursday, Oct. 28, 2004. Electronic citation @ http://newsmax.com/archives/articles/2004/10/28/100609.shtml

iv. Letter submitted to the American Journal of Forensic Medicine and Pathology (1998) by John Plunkett, M.D.: Shaken Baby Syndrome and Other Mysteries. Electronic version cited from http://www.potia.org/chapter08/mystery.html.

v. Jayakumar, I et al Case Reports: Shaken Baby Syndrome (2004). Indian Pediatrics, 41, 2004 pp 280-282

vi. (Author not cited) Wrong Diagnosis: Prognosis of Shaken Baby Syndrome. Electronic version @ http://www.wrongdiagnosis.com/s/shaken_baby_syndrome/prognosis.htm

vii. King. W.J. et al. Research:Shaken Baby Syndrome in Canada: Clinical characteristics and outcomes of hospital cases. Canadian Medical Journal. 168(2) January 21, 2003. Electronic version @ http://www.cmaj.ca/cgi/content/full/168/2/155.

viii. Diamond and Jaudes ***Med Studentsix. Ashraf, A.F et al. Case Reports:Fatal Physical Child Abuse in two children of

a family (1999). Annals of Saudi Medicine, Vol. 19, No 2, 1999. Pp. 120-123

21

HYPERNATRAEMIA IN VERY LOW BIRTH WEIGHT INFANTS ADMITTED AT CHRIS HANI BARAGWANATH ACADEMIC HOSPITAL: INCIDENCE, FLUID MANAGEMENT AND OUTCOME

Kim Barnard, Sithembiso VelaphiDepartment of Paediatrics, Chris Hani Baragwanath Academic HospitalFaculty of Health Sciences, University of the Witwatersrand

BACKGROUNDOngoing advances in neonatal medicine have significantly improved the survival rate of very low birth weight infants (VLBWI),1 which are defined as those neonates with a birth weight less than 1500g. VLBWI being preterm are at risk of having fluid and electrolyte disturbances, including hypernatraemia, due to the prematurity of multiple systems that allow for increased insensible water loss (IWL).2–4 Hypernatraemia is an electrolyte abnormality with a reported incidence of 30-40% in VLBWI, usually occurring within the first 72 hours of life, and is associated with high morbidity and mortality in these infants.2 It is most commonly caused by excessive fluid loss or decreased fluid intake, rather than increased sodium intake, highlighting the importance of fluid management in the first few days of life.2,3,5,6 A number of environmental factors namely incubators, radiant warmers and phototherapy may worsen this water loss during the first 72 hours of life.1,7,8 Hypernatraemia alone or the management thereof is associated with certain morbidities and mortality, therefore its management is critical in survival of preterm infants with or without morbidities. The morbidity and mortality associated with hypernatraemia is mostly due to dehydration and the subsequent fluid and electrolyte shifts between compartments. 9 The goal of management of hypernatraemia in preterm infants is to reduce sodium levels by correcting fluid deficit and avoidance of giving extra sodium.6 Correcting the free water deficit is done by calculating the amount of free water needed to reduce the sodium level by 0.5mmol/hour.10 The aims of the study were therefore to determine the incidence, fluid management and mortality rate in infants with hypernatraemia and factors associated with its severity.

STUDY DESIGN AND METHODS This was a retrospective descriptive study based at the Neonatal Division of Paediatrics at Chris Hani Baragwanath Academic Hospital (CHBAH). All VLBWI born at <37 weeks gestation, who were admitted at CHBAH between 1 June and 31 December 2012, and were diagnosed with hypernatraemia within 72 hours of life were eligible for the study. Hypernatraemia was defined as serum sodium

22

greater than 150 mmol/L. Medical records of VLBWI with a diagnosis of hypernatraemia were reviewed. Data on maternal demographics, infant characteristics, serum sodium levels, type and amount of fluids infants received at diagnosis and in response to hypernatraemia, as well as the number of patients who died within the first 28 days of life was collected and recorded in an appropriately designed data collection sheet

RESULTSFour hundred and forty three VLBWI were delivered at CHBAH, between 1 June 2012 and 31 December 2012. Files for 370 VLBWI, born during this period, were retrieved, reflecting an 83.5% recovery rate of files. Patients were excluded for various reasons leaving 125 patients to be included in calculation of incidence but only 120 for further analysis. The findings from this study are that the incidence of hypernatraemia in VLBWI admitted at CHBAH is very high at 33.8%. The majority of VLBWI who developed hypernatraemia were of gestational age less than 30 weeks. Only 30% of infants received antenatal steroids. All infants were started on a sodium containing fluid (Potassium-free Neonatalyte containing 33 mmol/L of sodium). The volume of fluids infants were started on at birth in ml/kg/day was almost similar for all weight categories and gestational ages at 80 mls/kg/day (Table 1).

Table 2: Total amount of fluids the infants were receiving at time of diagnosis of hypernatraemia

Volume (Median) ml/kg/day

Volume (Ranges) ml/kg/day

Birth Weight< 1000g (n=29)1000g - 1499g

(n=91)

8080

75 - 10060 - 110

Gestational Age<26 wks (n=4)

26 - 28 wks (n=48)29 - 30 wks (n=35)31 -34 wks (n=31)

> 34 wks (n=2)

8780808080

80 - 10070 -10070 - 10060 - 110

80

At time of diagnosis the severity of hypernatraemia was not related to gestational age, birth weight or to the amount of fluid or sodium that the infant was receiving (Figure 1a and 1b).

23

24

Figure 3: Serum sodium across the different weight categories (A) gestational ages (B) at diagnosis

A.

B.

In the majority of patients (75%) the amount of fluid used to correct hypernatraemia was calculated based on a formula for free water-deficit. The method used to determine amount of fluids did not affect rate of correction of hypernatraemia. The rate of correction of hypernatraemia was appropriate in the

25

majority of cases, as only 8.4% had their sodium levels reducing by >1 mmol/hour (Figure 2).

Figure 2: Rate of correction of hypernatraemia(mmol/hr) according to method used to calculate fluid volume increases

The environment under which infants were nursed did not affect the severity of hypernatraemia. Among the infants who had cranial ultrasounds done and results recorded, 15% had grade III IVH. The all-cause overall mortality rate in infants with hypernatraemia during the neonatal period was 22%, and the proportion of infants who died was greater in infants with smaller birth weight and younger gestational age.

DISCUSSIONThe high incidence of hypernatraemia in this cohort could be related to the use of sodium containing fluids on admission. There is a need for a prospective randomised control trial with infants being allocated to either a sodium free solution (5 or 10% dextrose) or Potassium-free Neonatalyte. Once hypernatraemia was diagnosed, the majority of patients were managed with an increase in fluid volume with either a sodium-containing or sodium-free solution. The fact that the hypernatraemia resolved by adding free water alone, without restricting sodium intake support findings from previous studies that hypernatraemia in preterm or VLBWI is almost always due to increased IWL.2,3,7 A formula that calculates water deficit to correct hypernatraemia has been

26

suggested, and this formula was used in this study and appeared to be safe. 11 The mortality rate in this cohort of babies was not much different from the overall mortality rate of 75% in VLBW infants in this neonatal unit. In conclusion, hypernatraemia is a common problem among VLBWI admitted in the neonatal division at CHBAH. Factors that might be contributing to high incidence of hypernatraemia could be related to use of sodium containing fluid from birth. The rate of correction of hypernatraemia was appropriate for most patients. Using a formula to correct free water deficit appears to be safe but a prospective study to confirm this is recommended as numbers in this study were small. Though correction of hypernatraemia appears to be appropriate in most patients, it is not clear as to what is the best fluid to use to correct hypernatraemia and how to calculate the volume. Therefore it is recommend that prospective studies are conducted to assess the effect of using a sodium-free solution at birth on incidence of hypernatraemia, use of free-water deficit formula to correct hypernatraemia and lastly to accurately look at morbidities associated with hypernatraemia.

REFERENCES

1. Wada M, Kusuda S, Takahashi N, Nishida H. Fluid and electrolyte balance in extremely preterm infants <24 weeks of gestation in the first week of life. Pediatr Int Off J Jpn Pediatr Soc. 2008 Jun;50(3):331–6. 2. Gawlowski Z, Aladangady N, Coen PG. Hypernatraemia in preterm infants born at less than 27 weeks gestation. J Paediatr Child Health. 2006 Dec;42(12):771–4. 3. Omar SA, DeCristofaro JD, Agarwal BI, La Gamma EF. Effects of prenatal steroids on water and sodium homeostasis in extremely low birth weight neonates. Pediatrics. 1999 Sep;104(3 Pt 1):482–8. 4. Lim W-H, Lien R, Chiang M-C, Fu R-H, Lin J-J, Chu S-M, et al. Hypernatremia and grade III/IV intraventricular hemorrhage among extremely low birth weight infants. J Perinatol Off J Calif Perinat Assoc. 2011 Mar;31(3):193–8. 5. Oddie SJ, Craven V, Deakin K, Westman J, Scally A. Severe neonatal hypernatraemia: a population based study. Arch Dis Child Fetal Neonatal Ed. 2013 Mar 19; 6. Moritz ML, Ayus JC. Disorders of water metabolism in children: hyponatremia and hypernatremia. Pediatr Rev Am Acad Pediatr. 2002 Nov;23(11):371–80. 7. Bhatia J. Fluid and electrolyte management in the very low birth weight neonate. J Perinatol Off J Calif Perinat Assoc. 2006 May;26 Suppl 1:S19–21. 8. Meyers R. Pediatric Fluid and Electrolyte Therapy. J. 2009;14(4):204–11. 9. Barnette AR, Myers BJ, Berg CS, Inder TE. Sodium intake and intraventricular hemorrhage in the preterm infant. Ann Neurol. 2010 Jun;67(6):817–23. 10. Van Amerongen RH, Moretta AC, Gaeta TJ. Severe hypernatremic dehydration and death in a breast-fed infant. Pediatr Emerg Care. 2001 Jun;17(3):175–80.

27

11. Shann F. Drug Doses Intensive Care Unit Royal Children’s Hospital Australia. fourteenth. 2008.

28

PATHOGENS ISOLATED FROM BLOOD STREAM OF NEONATES DIAGNOSED WITH HEALTHCARE ASSOCIATED INFECTIONS: ANTIMICROBIAL SUSCEPTIBILITY AND CASE FATALITY RATES

Sithembiso Velaphi 1 , Jeannette Wadula2, Shabir Madhi31Department of Paediatrics, Chris Hani Baragwanath Hospital and the University of the Witwatersrand2Department of Microbiology, National Health Laboratory Services, Chris Hani Baragwanath Hospital and the University of the Witwatersrand3National Institute of Communicable Diseases, National Health Laboratory Services, Johannesburg

INTRODUCTIONHealthcare associated infections (HAI) are the major cause of mortality during the neonatal period. They are defined as infection occurring after a patient has been in a healthcare facility for more than 48-72 hours. Neonates are at risk of HAI because of immature immune function, especially preterm infants or very low birth weight infants. Pathogens causing HAI vary among healthcare facilities or countries, for example, in developed countries pathogens reported to cause HAIs are mainly Gram positive bacteria (70%), namely coagulase negative staphylococcus (CONS), staphylococcus aureus, enterococcus species, then gram negatives (18-20%) and fungi (Candida) (12%), while in developing countries the common pathogens are Gram negative bacteria (58%)1,2. The susceptibilities of the pathogens causing HAIs also vary depending on types of pathogens and healthcare facilities. Mortality associated with HAIs vary from country to country or facility to facility also depending on how early the diagnosis is made, use of appropriate antibiotics, type of organisms and their susceptibility and availability of resources. In this study we sought to determine pathogens causing neonatal sepsis at Chris Hani Baragwanath Academic Hospital, their susceptibility and case fatality rates.

METHODSThis was a retrospective descriptive study. Data on positive blood cultures in the neonatal unit from 2008 to 2011 was retrieved from a computerized microbiology database from the National Health Laboratory Services. The

29

following organisms were considered to be contaminants, Coagulase Negative Staphylococcus, Micrococcus, Corynebacteria, and Bacillus and therefore were not included in the analysis. Hospital files or bedletters of infants who were admitted at the neonatal unit at Chris Hani Baragwanath Academic Hospital from 2008 to 2011 with positive blood culture were retrieved and reviewed for the following; demographic characteristics, anthropometry, clinical signs, diagnosis, laboratory findings, and whether infant survived or not. Organisms were grouped on whether they were multi-drug resistant or not based on antimicrobial susceptibility results from the microbiology laboratory. Case fatality rate for bacteria was defined as percentage of patients who died within 7 days of positive cultures among those who had a positive blood culture due to bacteria, and case fatality rate for Candida was percentage of patients who died within 14 days among those who had a positive blood culture due to Candida.

RESULTSThere were 2747 positive cultures retrieved from the computerized database, including 1277 that were considered to be contaminants and excluded. Clinical records were available for 974 (66%) of the remaining 1470 positive cultures, of which 701 were associated with HAIs. The HAI pathogens included 46.1% Gram negative bacteria (GNB), 30.2% Gram positive bacteria (GPB) and 25.4% Candida species. The common GNB were Acinetobacter sp. (51.7%), Klebsiella sp. (25.4%) and Enterobacter sp. (11.5%) (Table 1).

Table 1: Organisms isolated from blood stream of infants with hospital acquired infectionGroup of organisms Number (%)Gram Negative Bacilli

- Acinetobacter sp.- Klebsiella sp.- Enterobacter sp.- Escherichia coli- Pseudomonas sp.- Others

323 (46.1)167 (51.7) 82 (25.4) 37 (11.5) 15 (4.6) 8 (2.5) 14 (4.3)

30

Gram Positive Cocci- Staphylococcus Aureus [MRSA, MSSA]- Enterococcus sp.- Viridans Streptococcus- Streptococci Agalactiae- Others

212 (30.2)105 (59.4) [94 (44.3), 11 (5.1)] 83 (39.1) 13 (6.1) 7 (3.3) 4 (1.9)

Candida- Candida Parapsillosis- Candida Albicans- Others

166 (25.4) 77 (46.4) 74 (44.6) 15 (9.0)

Overall, 40% of GNB were MDR, including 14% of Acinetobacter sp. that were pan-resistant (only sensitive to Colistin), whilst 75.6% and 86.5% of Klebsiella sp. and Enterobacter sp. respectively were extended spectrum beta lactamases (ESBLs) producing isolates (Table 2). Ninety percent (94/105) of Staphylococcus aureus isolates were Methicillin-resistant (MRSA), whilst 33.7% of Candida parapsillosis and 1.4% of Candida albicans isolates were resistant to Fluconazole.

Table 2. Proportion of organisms that are multi-drug resistantOrganisms Number (%)Acinetobacter sp.

- Sensitive- Pan Resistant Acinetobacter sp.

167143 (85.6) 24 (14.4)

Klebsiella sp- Sensitive- ESBL

8220 (24.4)62 (75.6)

Enterobacter sp.- Sensitive- ESBL

37 5 (13.5)32 (86.5)

Escherichia coli- Sensitive- ESBL

1510 (66.7) 5 (33.3)

Pseudomonas sp.- Sensitive- Pan Resistant Pseudomonas sp.

8 7 (87.5) 1 (13.5)

Overall the case fatality rate was 21.4%; being highest with Acinetobacter sp (31.7%), Candida albicans (29.7% and Candida parapsillosis (27.3%) (Tables 3 & 4).

Table 3. Case fatality rates of Gram negative bacilliOrganisms Total Number Number Died (CFR, %)Acinetobacter sp.

- Sensitive- Pan Resistant Acinetobacter sp.

167143 24

53 (31.7)47 (32.9) 6 (25.0)

Klebsiella sp- Sensitive- ESBL

8220 62

13 (15.9) 4 (20.0) 9 (14.5)

Enterobacter sp.- Sensitive- ESBL

375 32

9 (24.3) 0 – 5 (28.1)

31

Escherichia coli- Sensitive- ESBL

1510 5

2 (13.3) 1 (10.0) 1 (20.0)

Pseudomonas sp.- Sensitive- Pan Resistant Pseudomonas sp.

87 1

1 (12.5) 1 (14.3) 0

32

Table 4. Case fatality rates of gram positive and candida speciesOrganisms Total Number Number Died (CFR, %)Staphylococcus Aureus

- Methicillin Sensitive- Methicillin Resistant

1051194

18 (17.1) 1 (9.1)17 (18.1)

Enterococcus sp. 83 6 (7.2)Streptococcus Agalactiae 7 0Viridans Streptococcus 13 2 (15)CandidaCandida AlbicansCandida ParapsillosisOthers

166747715

45 (27.1)22 (29.7)21 (27.3) 2 (13.3)

CONCLUSIONThe common pathogens isolated from sterile sites of neonates with a diagnosis of HAI at CHBAH are Gram negative bacilli. A significant number of these are multi-drug resistant. HAIs are associated with high mortality rates especially for Acinetobacter sp. and Candida infections. Reducing infections in the neonatal unit should be one of priorities as we continue with efforts to improve outcomes of neonates in our hospitals.

REFERENCES1. Srivastava S, Shetty N. Healthcare associated infections in neonatal

units: lessons from contrasting worlds. Journal of Hospital Infection 2007;65:292-306

2. Zaidi A, Huskins C, Thaver D, et al. Hospital acquired neonatal infections in developing countries. Lancet 2005;365:1175-88

33

RANDOMIZED TRIAL TO EVALUATE THE EFFECTS OF BIFIDOBACTERIUM LACTIS IN THE PREVENTION OF NECROTIZING ENTEROCOLITIS IN PRETERM INFANTS

Peter Cooper, Keith Bolton, Sithembiso VelaphiDept of Paediatrics, University of the WitwatersrandPhilippe Steenhout Nestlé Clinical Development Unit, Vevey Switzerland

BackgroundNecrotizing enterocolitis (NEC) is the most serious gastrointestinal complication in premature infants and is a leading cause of morbidity and mortality. Complications include bowel perforation, sepsis and short bowel syndrome. While it is a disease largely of very low birth weight infants (<1500g at birth) and those <32 weeks gestation, it may occur in more mature and larger newborn infants. Survivors of NEC have also been shown to be at higher risk for long term neurodevelopmental deficits.

PreventionFeeding premature infants with mothers own milk or banked human milk has been shown to reduce the incidence of NEC by two-thirds and is currently the most effective preventive strategy. Studies investigating the timing of beginning enteral feeds and/or the rate of increase of enteral feeds have been unable to show any difference in the incidence of NEC. Recent studies have suggested that probiotics may reduce the risk of NEC in premature infants. A meta-analysis of randomized controlled trials in 2010 which included nine studies reported that four of these studies showed a significant reduction in NEC. However the meta-analysis showed a significant reduction with a relative risk of 0.35 favoring probiotics (95% CI 0.23-0.55).

ObjectiveThe primary object of this randomized, double-blind, multicenter study was to evaluate the effect of Bifidobacterium lactis (B.lactis) on the

34

incidence of NEC in preterm infants. Secondary objectives included the incidence of nosocomial sepsis and mortality

MethodsPreterm infants with birth weight between 800-1500 g who were tolerating enteral feeding at 48 hours postnatal age were randomly assigned to receive a daily capsule containing either B. lactis (1E9

cfu/capsule) or maltodextrin (placebo). The contents of the capsules were added to one feed per day of either preterm formula or breast milk. Infants received 1 capsule/day for a maximum of 6 weeks beginning ≤60 hours after birth until either discharge or first occurrence of NEC (stage II or higher, modified Bell’s criteria). The primary outcome was incidence of stage II or III NEC. Based on a power of 80% to detect a two-thirds reduction in NEC in the B. lactis group using a baseline incidence of 6%, it was calculated that 425 infants would be needed in each group. Allowing for drop outs, a total of 500 infants in each group were to be enrolled. An interim analysis for efficacy was planned after 300 infants had been enrolled.

ResultsAfter 300 infants were randomized, the Independent Data Monitoring Committee (IDMC) performed an interim analysis. By the time the analysis was complete a total of 318 infants had been randomized. Of these 309 could be evaluated 158 in the B. lactis group and 151 in the placebo group. The flow of enrollments of these infants can be seen in the figure below.

35

Baseline characteristics were similar in both groups. Mean gestational age was 30.9±2.3 weeks in the B. lactis group and 30.8±2.3 weeks in the placebo group and mean birth weights were 1.26kg±0.18 vs 1.24kg±0.18 respectively. There were no differences in the male to female ratio, exposure to antenatal steroids or 5 minute Apgar scores between the two groups.As regards the primary outcome there was no difference in the number of infants who developed stage II or III NEC according to the modified Bell’s criteria as can be seen in the table. There were also no significant differences found between B. Lactis and placebo for incidence of the individual NEC stages. There were no differences in the number of infants who developed nosocomial sepsis or in the number of infants who died. As regards NEC related deaths (3 in B. lactis, 4 in placebo; OR 0.61, 95% CI 0.11-3.49), there was also no significant difference. On the basis of the interim analysis of these results the IDMC recommended that the study be terminated due to futility and this was accordingly done.

36

37

B. lactis (n=158) Placebo (n=151) P ValueNEC (No of infants)

14 (8.9%) 13 (8.6%) 1.0

Nosocomial Sepsis (No of infants )

46 (29.1) 47 (31.1%) 0.79

Deaths (No of infants)

13 (8.2%) 11 (7.3%) 0.92

DiscussionThis study showed no beneficial effects with regard to incidence of NEC, nosocomial sepsis or mortality with the administration of a daily dose of B. lactis to premature infants. This is in contrast to the findings of the meta-analysis done in 2010 and a more recent Cochrane review published in 2014. However, although both of these meta-analyses showed benefit, a number of individual studies included in these did not show benefit. There has also been no uniformity in the type of probiotic used or the doses used.

ConclusionsAlthough meta-analyses have shown benefit from the administration of probiotics to premature for the prevention of NEC, the choice of probiotic, the dose and the most appropriate target group of premature infants remain to be defined.

38

MATERNAL HUMAN IMMUNODEFICIENCY VIRUS STATUS AND MORTALITY OF VERY LOW BIRTH WEIGHT INFANTS

Mayowa M. Tiam 1 , SithembisoVelaphi21Baylor College of Medicine Children’s Foundation, Lesotho2Department of Paediatrics, Chris Hani Baragwanath Academic Hospital

BACKGROUNDHuman immunodeficiency virus (HIV) remains a common cause of infection in pregnant mothers with transmission risk to the baby. Various strategies on prevention of mother to child transmission have led to a reduction in transmission rates to less than 5%. Studies have reported that infants born to HIV-infected women have an increased risk for low birth weight, preterm birth and high mortality rate1,2. The high mortality rate could possibly be related to infections due to impaired immune function. A number of immune abnormalities have been reported in HIV exposed uninfected infants, therefore putting them at risk of developing nosocomial infections3. These include the lower specific antibody response lower percentage of naïve T-cells quantitative and qualitative changes in neonatal dendritic cells increased percentage of activated CD8+ with a decreased CD4+/CD8+ ratio immature T-cell function, in HIV exposed uninfected infants compared to HIV unexposed infants. It is not very clear if these impaired immune functions translate to clinical infections or poor outcomes4,5. Very low birth weight infants are often admitted in hospitals for long periods and therefore at risk of developing nosocomial infection5,6. Considering the possible impaired immunity in HIV-exposed infants, it is possible that HIV-exposed VLBWI might have an increased incidence of nosocomial sepsis and associated high mortality.

OBJECTIVETo determine sepsis rates and mortality rates in VLBWI according to maternal HIV status

METHODSDesign: Retrospective record review

39

Study Setting: Chris Hani Baragwanath Academic Hospital, a public sector tertiary hospitalStudy population: VLBWI admitted at CHBAH, January to June 2011 Outcomes: Sepsis episodes, NEC, IVH and mortality at hospital discharge Analysis: Descriptive- demographics, clinical and laboratory findings, and mortality rates, comparative- above outcomes between HIV exposed and unexposed

RESULTSAmong the total of 302 patients reviewed, 103 (34.1%) were born to HIV-positive mothers (HIV-exposed). Overall, almost all the infants were born before 36 weeks gestational age. There was no significant difference between HIV exposed and HIV unexposed infants in terms of gestational age (p-0.224). The median birth weight was 1220 grams. There were more babies with birth weight less than 1000 grams in the HIV exposed infants compared to HIV-unexposed (41.1 vs 23.1%) (p-0.001). There were more infants with Apgar score <7 at 1 minute (p-0.015) and at 5 minutes (p-0.030) in the HIV exposed infants. Sixty three percent of VLBW infants were exclusively breast fed (Table 1).

Table 1: Characteristics of very low birth weight infants admitted at CHBAH over the 6 month period

Total

(N=302)

HIV-Positive

(N=103)

HIV-Negative

(N=199)

p-value

Mode of delivery***Vaginal Abdominal

N = 276134 (48.6%)142 (51.4%)

N = 9141 (45.1%)50 (54.9%)

N = 18593 (50.3%)92 (49.7%)

0.4151

Mean Gestational age (weeks)*<2828-32>32

63 (21)196 (65.3) 41 (13.7)

29.4 ±2.9 28 (27.7) 60 (59.4) 13 (12.9)

29.82 ± 2.7 35 (17.6) 136 (68.3) 28 (14.1)

0.224

Female 164 (54.5) 59 (57.8) 105 ( 52.8) 0.400Birth Weight** Number weighing <1000 gramsNumber weighing 1000-1499 grams

88 (29.2)213 (70.8)

42 (41.2) 60 (58.8)

46 (23.1) 153 (76.9)

0.001

Body length at birth (cm) 37.5 ± 3.9 36.6± 4.0 38.0 ± 3.9 0.006Head Circumference at birth (cm)

27.5 ± 2.3 26.9± 2.4 27.7 ±2.2 0.003

Apgar score at 1 minute0-6

N = 267122 (45.7)

N = 9150(55.0)

N = 176 72 (40.9) 0.015

40

7-10 145 (54.3) 41(45.0) 104 (59.1)Apgar score at 5 minutes0-67-10

N = 256 38 (14.9)218 (85.1)

N = 8718 (20.7) 69 (79.3)

N = 169 20 (13.4) 149 (86.6)

0.030

There were no differences in incidence of NEC or severe NEC between HIV-exposed and unexposed VLBW infants (17.2 vs 21.2%, p-0.465). Of the 63 infants who had records of cranial ultrasound having been done to exclude IVH, about two thirds (70%) had abnormal findings (Grade I-IV IVH) with HIV exposed infants having significantly more severe grades of IVH compared to HIV unexposed infants (p <0.001) (Table 2).

Table 2: Morbidity and mortality outcomes of very low birth weight infantsVariables Total

N=302 (%)

HIV exposed n=103 (%)

HIV unexposed n=199 (%)

p-value

Congenital abnormalities*NoYes

286 (96.6) 11 (3.4)

98 (97.0) 3 (3.0)

188 (95.9) 8 (4.1)

0.631

Admission diagnosisRespiratory Distress SyndromeCongenital PneumoniaBoth of the above

220 (72.8) 4 (1.3) 78 (25.9)

79 (35.9) 0 (0.0)24 (30.8)

141 (70.9) 4 (2.0) 54 (27.1)

0.250

White blood cell count (x 109/L)No. with count <5 x 109/ LNo. with count 5-25 x 109 /LNo. with count >25 x 109 /L

n = 28834 (11.8)

245 (85.1)9 (3.1)

n = 9411 (11.7) 81 (86.2) 2 (2.1)

n = 19423 (11.9)

164 (84.5)7 (3.6)

0.734

Platelets count (x 109/L)No. platelets <100 x 109 /LNo. platelets 100-150 x 109/LNo. platelets ≥150 x 109 /L

n = 26618 (6.7)

62 (23.3)206 (77.4)

n = 936 (6.4)

22 (23.6) 66 (71.0)

n = 17312 (6.9)40 (23.1)

140 (80.9)

0.350

CRP No. with CRP <10 mg/dL No. with CRP ≥10 mg/dL

241(94.5) 14 (5.5)

80 (94.1) 5 (5.9)

161(94.7) 9 (5.3)

0.934

Blood culture at birth +NegativePositive Contaminants

255 (90.7) 16 (5.7) 10 (3.6)

84 (85.8) 7 (7.1) 7 (7.1)

171 (93.5) 9 (4.9) 3 (1.6)

0.270

Nosocomial sepsis No Yes

101 (37.1)171 (62.9)

32 (36.8)55 (63.2)

69 (37.3)116 (62.7)

0.144

Necrotizing enterocolitis- No- Stage 1 - Stage 2- Stage 3

240 (80.3) 49 (16.1) 7 (2.3) 4 (1.3 )

84 (82.4) 13 (12.3) 2 (2.0) 3 (2.9)

156 (78.8) 36 (18.2 5 (2.5) 1 (0.5)

0.464

Intraventricular haemorrhage 17 (27.0) 4 (14.3) 13 (37.1)

<0.001

41

NormalGrade 1-2Grade 3Grade 4

26 (41.3) 11 (17.5) 9 (14.3)

8 (28.6) 9 (32.1) 7 (25.0)

18 (51.4) 2 (5.7) 2 (5.7)

Alive at hospital dischargeYesNo

216 (73.0) 80 (27.0)

62 (62.4)39 (38.6)

154 (79.0) 41(21.0)

0.012

Among the 296 with recorded outcome, 216 (73.0%) infants were discharged alive, while 80/296 (27.0%) infants died before discharge. There were more deaths in the HIV exposed infants than those who were not and this difference was statistically significant (38.6 vs 21%) (p<0.001) (Table 2). In the univariate analysis the factors associated with mortality were maternal HIV status (p-0.001), infant birth weight (p<0.001), gestational age (p<0.001), and Apgar score at 1 (p<0.001) and 5 minutes (p<0.001), and presence of congenital abnormalities (p-0.006) but in a multivariate analysis, the only predictor of mortality was infant birth weight (p<0.001). Among the HIV exposed infants the univariate analysis revealed that factors that were associated with increased risk for mortality were birth weight (p<0.001), gestational age (p<0.001), Apgar score at 1 minute (p<0.001), and 5 minutes (p<0.001), and diagnosis at birth (p<0.010). This was the same for the HIV unexposed (data not shown). Although in univariate, the predictors of mortality were birth weight, gestational age, APGAR at 1&5 minutes and congenital abnormalities; in multivariate the predictors were birth weight and gestational age.

CONCLUSIONMaternal HIV status is not associated with increased risk for nosocomial sepsis, NEC and mortality. Birth weight remains a good predictor of outcome irrespective of HIV exposure.Maternal HIV status should not be used in predicting outcomes of VLBWI. Role of HIV exposure on IVH needs to be studied further

References:1. Reitter A, Stücker AU, Buxmann H, Herrmann E, Haberl AE,

Schlößer R, et al. Prenatal Ultrasound Screening for Fetal Anomalies and Outcomes in High-Risk Pregnancies due to Maternal HIV Infection: A Retrospective Study. Infectious Diseases in Obstetrics

42

and Gynecology 2013, 208482;10. Available: http://dx.doi.org/10.1155/2013/208482. [Accessed 21 February 2014]

2. Health Systems Trust. Health Indicators: recently updated indicators. Durban: HST, 2011. Available: http://indicators.hst.org.za/healthstats/314/data [Accessed 25 October 2011]

3. De Maria A, Cirillo C, Moretta L. Occurrence of human immunodeficiency virus type 1 (HIV-1)-specific cytolytic T cell activity in apparently uninfected children born to HIV-1-infected mothers. J Infect Dis 1994; 170:1296–1299.

4. Brocklehurst P, French R. The association between maternal HIV infection and perinatal outcome: a systematic review of the literature and meta-analysis. Br J. Obstet. Gynaecol 1998; 105:836-848.

5. Shah PS, Kaufman DA. Antistaphylococcal immunoglobulins to prevent staphylococcal infection in very low birth weight infants. Cochrane Database of Systematic Reviews 2009;2:006449. Available:http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD006449.pub2/pdf [Accessed 23 December 2013]

6. Legrand FA, Nixon DF, Loo CP, Ono E, Chapman JM, Miyamoto M, et al. Strong HIV-1-Specific T Cell Responses in HIV-1-Exposed Uninfected Infants and Neonates Revealed after Regulatory T Cell Removal. PLoS ONE 2006; 1(1): e102. Available: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1762312/pdf/pone.0000102.pdf Accessed 27 February 2014.

43

THE CUMULATIVE INCIDENCE OF HIV INFECTION IN HIV-EXPOSED INFANTS WITH A BIRTH WEIGHT OF ≤1500G RECEIVING BREAST MILK AND DAILY NEVIRAPINE.

Melantha Coetzee, Steve Biko Academic Hospital, University of Pretoria.Suzanne Delport, Kalafong Hospital, University of Pretoria.Piet Becker, SA Medical Research Council / University of Pretoria.

BackgroundVery low birth weight (VLBW) and premature infants are at an increased risk of mother-to-child transmission (MTCT) of HIV due to immature gastrointestinal barrier function. Pasteurisation of expressed breast milk (EBM) has been used to decrease MTCT of HIV in these infants. After the introduction of the 2010 prevention of mother-to-child transmission (PMTCT) guideline recommending daily infant nevirapine (NVP) for the duration of exposure to HIV-infected milk, fewer institutions are practising pasteurisation, although the incidence of MTCT of HIV under these circumstances remains unknown.

IntroductionHIV transmission involves a complex interaction between maternal transmittability (antiretroviral use, HIV viral load, CD4 count and HIV-related co-morbidities) and infant susceptibility (prematurity, birth weight and gastrointestinal immaturity).1,2,3,4 MTCT of HIV in developed countries has been reduced to <2% with the introduction of antiretroviral (ARV) therapy and avoidance of breastfeeding.1 Factors contributing to MTCT of HIV in resource-poor settings include poor accessibility to ARVs for all pregnant women and avoidance of breastfeeding is not always acceptable, feasible, affordable, sustainable and safe (AFASS).

A study performed at Kalafong Hospital, South Africa, in 2006 and 2007 showed that the risk of becoming HIV infected during the peripartum period was 10% by 6 weeks of age in infants with a birth weight of ≤1500g. Only 50% of the mothers received any type of PMTCT. Infants received single dose NVP within 72 hours of birth as standard of care, and exclusive feeding with Pretoria Pasteurised expressed breast milk was the norm.5

A study performed at Groote Schuur Hospital, South Africa, in 2010/2011 reported a HIV transmission rate of 2.7% by 6 weeks in extremely low birth

44

weight (ELBW) infants (95% CI: 0.7 – 14.1%). All infants received pasteurised milk or donor expressed breast milk (DEBM) and all infants received appropriate PMTCT. Infant mortality of 27% before 6 weeks of age may have influenced the transmission rate of HIV in this cohort of ELBW infants.6

In 2011 the MTCT rate in a cohort of predominantly term infants receiving nevirapine, was 2.7% (95% CI: 2.1% – 3.2%) by 4-8 weeks of age. HIV-exposure prevalence was 32.2% and 93.9% of mother/infant pairs received appropriate PMTCT. Nearly 36% of infants were exclusively breastfed, 47% were exclusively formula fed and 17% were mixed feeding. These results cannot however be extrapolated to the infant weighing ≤1500g at birth, as only 12% of these infants were LBW.7

The primary objective of this study was to assess the cumulative incidence and MTCT rate of peripartum HIV infection by 4 weeks of age in infants weighing ≤1500g at birth that received raw breast milk and daily NVP prophylaxis. The secondary objectives were to investigate the impact of certain maternal and infant factors on MTCT of HIV.

MethodEthical consent was obtained to undertake a retrospective, descriptive study on HIV-exposed infants weighing ≤1500g at birth. Data was collected on infants admitted from 1 March 2010 to 28 February 2013. Infants were identified using the ward statistics sheet that document both birth weight and HIV-exposure. Hospital files were retrieved to collect the appropriate data. Only infants admitted to Kalafong Hospital within 72 hours of birth were included. Infants who demised within the first 4 weeks of life were excluded as well as infants that received zidovudine prophylaxis, those that received exclusive DEBM or formula feeds and those without a HIV DNA-PCR test at ≥4 weeks. HIV DNA-PCR test results were acquired from the National Health Laboratory Service (NHLS) database.

Analysis of DataA total of 3790 infants were admitted to the neonatal unit at Kalafong Hospital from 01 March 2010 to 28 February 2013. Of the 3790 infants admitted, 690

45

infants weighed ≤1500g at birth and of these 219 were identified as being HIV-exposed (31.7%).

Seventy-two of 219 HIV-exposed infants weighing ≤1500g were excluded for the following reasons: 39 infants (17.8%) demised prior to 4 weeks of age, 23 infants received AZT, 7 infants received exclusive DEBM or formula milk, 2 infants were referred >72 hours after birth and one infant was confirmed to have intrauterine HIV infection. Of the 39 infants that demised, twenty-nine infants (74%) weighed <1000g at birth and twenty-four infants (62%) demised before 72 hours of life. These deaths were likely related to complications of prematurity. Eight of 219 infants (3.7%) demised after 7 days of life, which may be a confounder in the results. Of the 147/219 remaining infants, a further 13 were excluded as their files could not be traced or were incomplete, and an additional 54 were excluded, as no HIV DNA-PCR test ≥4 weeks of age were available. HIV DNA-PCR tests were available for the remaining 80 infants.

Data was analysed for 72 mothers (corrected for 8 twin pregnancies). Eleven mothers (15.3%) did not attend antenatal care clinic. Twenty-six mothers had intact membranes up to delivery, 7 had rupture of membranes (ROM) <24 hours, 11 had ROM ≥24 hours and the state of the membranes was not documented in 28 mothers. Fourteen mothers (20%) did not receive any antiretrovirals during pregnancy, twenty-three mothers (32%) received dual therapy and 34 mothers (48%) received lifelong HAART. ARV use was not recorded for one mother. Total maternal ARV uptake during pregnancy was 80% with an average duration of antiretroviral use prior to delivery of 12 weeks. The median maternal CD4 count was 272 cells/mm3 (range: 8-1097) with 42 mothers (59%) having a CD4 count <350 cells/mm3. Seven mothers (9.7%) were treated for tuberculosis during pregnancy.

The median birth weight of the infants was 1130g (510-1500g). Single dose NVP was documented in 74 infants (92.5%). There were no discordant HIV results for the 8 multiple pregnancies. Of the 80 infants included in the study, 2 HIV DNA-PCR tests were positive at 9 days and 20 days, respectively.

Infant 1 Infant 21st positive HIV DNA-PCR

Day 9 of life Day 20 of life

46

Maternal age 18 years Not recordedAntenatal Care No NoMaternal CD4 count 1097 cells/mm3 124 cells/mm3

Maternal PMTCT* None NoneMaternal Tuberculosis No NoMethod of delivery NVD Not recordedRupture of membranes Not documented Not documentedBirth weight 1120g 1400gInfant sdNVP Yes Not documented

* Using Fishers exact test it was determined that the absence of maternal PMTCT (2/15, 13.3%) versus administration of maternal PMTCT (0/64, 0%) is a risk factor for peripartum MTCT of HIV in VLBW infants (p=0.034).

Since an early HIV DNA-PCR test (≤72 hours) was not available for either of these infants, perinatal transmission could not be distinguished from intrauterine transmission. The peripartum HIV transmission rate was calculated, firstly, assuming that both of these infants were HIV DNA-PCR positive ≤72 hours of life and, secondly, assuming that both were HIV DNA-PCR negative ≤72 hours of life. Therefore, the peripartum MTCT rate of HIV is between 0 – 2.5% (95% CI: 0% - 6%) at 4 weeks of age. The cumulative incidence of peripartum HIV by 4 weeks of age is 0 - 25/1000 HIV-exposed infants.

ConclusionBy comparing the HIV transmission rate of this study (0 - 2.5%) to the MTCT rate of 10% at Kalafong Hospital in 2006/2007, 2.7% in ELBW infants at Groote Schuur Hospital in 2010/2011, and the National MTCT rate in term infants of 2.7% in 2011, raw breast milk seems a safe feeding option in vulnerable infants ≤1500g at birth in the presence of daily NVP prophylaxis. Direct contact with the mother’s breast will facilitate mother-infant bonding and optimal lactation in a group of infants whom previously received pasteurised EBM to prevent MTCT of HIV through breastfeeding. Absence of maternal PMTCT was a significant factor for peripartum MTCT of HIV in this study (p=0.03).

Small patient numbers, missing data from patient files, untraceable files and unavailability of HIV DNA-PCR tests by 4 weeks of age are the limitations of this retrospective study. Additionally, viral suppression caused by NVP prophylaxis used for PMTCT may be a confounder in the definitive diagnosis of HIV by 4-6 weeks of age.

47

References1. Newell ML. Current issues in the prevention of mother-to-child

transmission of HIV-1 infection. Trans R Soc Trop Med Hyg 2006;100:1-5.2. Kourtis AP, Bulterys M. Mother-to-child transmission of HIV: pathogenesis,

mechanisms and pathways. Clin Perinat 2010;37:721-37.3. Prestes-Carneiro LE, Spir PRN, Ribeiro AA, Goncalves VLMA. HIV-1 mother-

to-child transmission and associated characteristics in a public maternity unit in Presidente Prudente, Brazil. Rev Inst Med Trop Sao Paulo 2012;(54)1:25-9.

4. National Department of Health, South Africa. Guidelines for the Management of HIV in Children, 2nd ed. Department of Health Republic of South Africa; 2010.

5. Delport SD. HIV and the VLBW infant. Unpublished data.6. Tooke L, Horn AR, Harrison MC. HIV transmission to extremely low birth

weight infants. Pediatr Infect Dis J 2013;(32)1:36-38.7. Goga AE, Dinh TH, Jackson DJ for the SAPMTCTE study group. Early (4-8

weeks post-delivery) Population-level Effectiveness of WHO PMTCT Option A, South Africa, 2011. South African Medical Research Council, National Department of Health of South Africa and PEPFAR/US Centers for Disease Control and Prevention. 2013.

48

POSTNATAL MOTHER-TO-CHILD TRANSMISSION OF HIV AND HIV-FREE SURVIVAL IN AN HIV-EXPOSED NATIONAL COHORT, SOUTH AFRICA, DECEMBER 2012-SEPTEMBER 2014

Goga AE (1, 2), Jackson DJ (3, 4), Dinh TH (5), Ramokolo V (1), Ngandu N (1), Noveve N (1), Singh Y (1), Magasana V (1), Ramraj T (1), Sherman G (6,7), Puren A (6, 8), Doherty T (1, 3, 9), Nsibande D (1), Bhardwaj S (10), Shaffer N (11), Chopra M (4), Cheyip M (12), Mogoshoa M (12), Pillay Y (13), Lombard C (14, 15) for the SAPMTCTE study group.

(1) Health Systems Research Unit, South African Medical Research Council, South Africa (SA)(2) Department of Paediatrics and Child Health, University of Pretoria, SA(3) School of Public Health, University of the Western Cape, SA(4) UNICEF, New York(5) Centers for Disease Control and Prevention, Center for Global Health, Division of Global HIV/AIDS,

United, States of America (USA)(6) Centre for HIV and STI, National Institute of Communicable Diseases, SA(7) Department of Paediatrics and Child Health, Faculty of Health Sciences, University of Witwatersrand,(8) Division of Virology and Communicable Diseases, School of Pathology, University of the Witwatersrand

Medical School(9) School of Public Health, University of Witwatersrand, SA(10) UNICEF, SA(11) World Health Organisation, Geneva(12) Centers for Disease Control and Prevention, South Africa(13) National Department of Health, SA(14) Biostatistics Unit, South African Medical Research Council, SA(15) School of Public Health and Family Medicine, University of Cape Town, Cape Town 7935, SA

Background: Infant HIV-free survival is the gold standard for measuring the effectiveness of programmes to prevent mother-to-child transmission of HIV (PMTCT). Global targets for eliminating mother-to-child transmission of HIV (MTCT) are <2% by six weeks and <5% by 18 months postpartum. In the context of monitoring progress, we measured population-level MTCT and infant HIV-free survival (HFS) at 4-8 weeks (when Option A was policy) and 3, 6, 9, 12, 15 and 18 months (during the policy transition to Option B) in South Africa, a high HIV prevalence, middle income country.Method: A nationally representative cross-sectional survey was conducted in 580 public health facilities randomly selected after multistage probability proportional to size methodology. We interviewed caregivers of systematically or consecutively sampled, consented infants aged 4-8 weeks receiving their six week immunisation and tested infant dried blood spots for HIV exposure (antibody) and infection. A closed cohort of infants (antibody, or maternal self-reported, positive) was invited for follow-up at 3, 6, 9, 12, 15 and 18 months. At each visit we interviewed primary caregivers and tested infants for HIV infection. Results are weighted for sample ascertainment and population live births.Results: At 4-8 weeks, data from 9120 infants were analysed, showing 33.1% (95% Confidence Interval – CI - 31.8-34.3%) infant HIV exposure and 2.6% (CI 2.0-3.2%) early MTCT. Of 2787 infants in the cohort, 1069 (38.3%) were lost to follow-up (LTFU) by 18 months. Cumulative (from birth) preliminary unweighted MTCT by 3, 6, 9, 12, 15, and 18 months was 2.65%, 3.41%, 3.73%, 3.93%, 4.08% and 4.19%, respectively; corresponding preliminary HFS was 97.3%, 95.6%, 94.4%, 93.7% 93.2% and 92.9%. By 18 months 71 of 1000 HIV-exposed infants were dead or HIV infected. The first three months posed the largest risk of HIV/death. Conclusions: Although preliminary unweighted 18-month MTCT was <5%, there is potential that global targets may not have been achieved as our data excludes early infections/deaths prior to 6 weeks postpartum and outcomes amongst LTFU infants. Our findings highlight the critical need for continued improvements in

49

antenatal and perinatal PMTCT programmes and intense postnatal follow-up to remove bottlenecks and reach global targets.

50

FEAR OF KNOWING THE CHILD IS HIV POSITIVE AND STIGMA FROM DISCLOSURE REDUCE ACCESS TO EARLY INFANT DIAGNOSIS IN THE RURAL COMMUNITIES OF OR TAMBO DISTRICT, SOUTH AFRICA: A QUALITATIVE EXPLORATION OF MATERNAL PERSPECTIVE

VINCENT OLADELE ADENIYI, OV (AAHIVS), PhD Candidate, University of Fort Hare, East London, South AfricaELZA THOMSON, University of Stellenbosch, South AfricaDANIEL TER GOON, University of Fort Hare, South AfricaAJAYI IDOWU ANTHONY, University of Fort Hare, East London, South Africa

Background: Despite the overwhelming evidence confirming the morbidity and mortality benefits of early initiation of highly active anti-retroviral therapy (HAART) in HIV-infected infants, some children are still disadvantaged from gaining access to care. The understanding of the maternal perspective on early infant HIV diagnosis and prompt initiation of HAART has not been adequately explored, especially in the rural communities of South Africa. This study examines the perspectives of mothers of HIV-exposed infants with regard to early infant diagnosis (EID) through a lens of social and structural barriers to accessing primary healthcare in OR Tambo district, Eastern Cape Province, South Africa.

MethodsIn this qualitative study, we conducted semi-structured interviews at two primary healthcare centres in the King Sabata Dalindyebo Municipality of the OR Tambo district, South Africa. Twenty-four purposefully selected mothers of HIV-exposed infants took part in the study. Interviews were tape-recorded, transcribed and field notes were obtained. The findings were triangulated with two focus group discussions in order to enrich and validate the qualitative data. Thematic content analysis was employed to analyse the data.

ResultsThe participants have fairly good knowledge of mother-to-child transmission of HIV and the risks during pregnancy, delivery and breastfeeding. The majority of participants were confident of the protection offered by anti-retroviral drugs provided during pregnancy. Knowledge of optimal time for early infant diagnosis of HIV was lacking among the study participants. Reasons for not accessing EID included fear of finding out that their child is HIV positive, feelings of guilt and/or shame and embarrassment with respect to raising an HIV infected infant. Personal experiences of HIV diagnosis and HAART were associated with participants’ attitudes and beliefs toward care-seeking behaviours. Stigma resulting from their own disclosure to others reduced their likelihood of recommending EID to other members of their communities.

ConclusionDespite the good knowledge of mothers about infant HIV infection and the availability of treatment, the knowledge of the optimal time for early infant diagnosis is lacking. Fear of infant HIV diagnosis and stigma are challenges for universal coverage of early infant diagnosis in these rural communities. Hence, community education and intensive counselling of pregnant women about early infant diagnosis are urgently needed.

51

ANAEMIA IN PREGNANCY

K Tunkyi, *J Moodley Department of O & G, Addington Hospital *Women's Health and HIV Research Group, Department of O & G, University of KwaZulu-Natal, Durban.

IntroductionAnaemia is a common condition in poor countries; it is of particular importance during pregnancy because it is associated with both maternal and perinatal morbidity and mortality. There is little data on anaemia in pregnancy in South Africa, this despite the fact that it is a country with antenatal HIV rates of approximately 30%. HIV and the antiviral agents used to treat it are associated with anaemia. We therefore decided to establish the incidence of anaemia in 2000 women attending an antenatal clinic of a regional/district hospital.

MethodsFollowing institutional regulatory permissions (ethics and hospital) informed consent was obtained and venous blood samples taken for complete blood counts. All women except for those who were HIV infected were not on any medication at the time of entry to the study.

ResultsOf the 2000 women included in the study, 996 (49.8%) had haemoglobin (Hb) levels of ≥ 11 g/dl, therefore 1004 (50.2%) were anaemic (Hb levels < 11 g/dl) according to the WHO classification. The main types of anaemia in 1004 women were hypochromic microcytic (n=11; 1.1%), normochromic normocytic (751; 74.8%), hypochromic normocytic (n=168; 16.7%) and normochromic microcytic (n=74; 7.4%). Six hundred and nine (60.6%) were HIV infected and 395 (39.4%) were uninfected. In HIV infected women (n=609), the main types of anaemia were hypochromic microcytic (n=7; 1.1%), normochromic normocytic (n=424; 69.6%), hypochromic normocytic (n=108; 17.7%) and normochromic microcytic (n=70; 11.5%).The number of women with Hb levels of <10g/dl was 25.3% (n=506).The overall age of the 2000 women was 27.6 ± 7.6; the mean age of the women with Hb levels > 11 was 28.4 ± 5.6 years and those with Hb levels <11/g/dl was 28.2 ± 3.7 years.The overall mean parity was 3 while the mean parity in the HIV infected was 2.The mean gestational age at first visit was 24 ± 2.2 weeks (HIV infected 27± 3.1). Four hundred and twenty nine (70.4%) of the 609 infected women were on HAART.DiscussionThere is a paucity of data on anaemia in pregnancy in South Africa. Our study confirms the clinical impression that anaemia is a common problem in antenatal attendees in a South African urban population; 50.2 % had Hb levels of < 11g/dl at their first antenatal visit thus verifying the Saving Mother’s Reports which have consistently indicated that anaemia is present in 40% of all maternal deaths. The incidence of anaemia is also higher in HIV infected women irrespective of treatment.Given the high incidence of maternal deaths associated with obstetric haemorrhage, more attention must be paid to providing information on healthy diets and appropriate nutritional supplements.

52

THE IMPACT OF HIV INFECTION ON OBSTETRIC HAEMORRHAGE AND BLOOD TRANSFUSION IN SOUTH AFRICA.

EM Bloch1,2, MBChB, R Crookes, MBChB3 , J Hull4,5 , MBChB, S Fawcus, MBChB 6,7, R Gangaram8,9 MBChB, , J Anthony10,7, MBChB, C Ingram3, MBChB, S Ngcobo3, MBChB, J Croxford10, MPH, D Creel10, EL Murphy 2,1, MD, MPH for the International Component of the NHLBI Recipient Epidemiology and Donor Evaluation Study-III (REDS-III)1Blood Systems Research Institute2University of California, San Francisco3South African National Blood Service4Chris-Hani Baragwanath Hospital5 University of Witwatersrand6Mowbray Maternity Hospital7University of Cape Town8King Edward VIII Hospital9University of Kwazulu-Natal10Groote Schuur Hospital11Research Triangle International (RTI)

BackgroundBoth globally and within South Africa, obstetric hemorrhage (OH) is a leading cause of obstetric mortality. The South Africa CEMD report for 2011-2013¹ show OH to be the second most common cause of maternal death and it has increased in the last decade. Bleeding associated with caesarean section (CS) accounts for a third of OH deaths, and OH MMR is higher in HIV positive women compared to HIV negative². While data exists on OH incidence and maternal mortality, there are little data available on blood transfusion practices and associated outcomes. Problems with availability of blood for transfusion is cited as contributing to 26% of maternal deaths in developing countries³ and 13.6 % in South Africa¹.There is also little data on the impact of HIV infection on OH and on blood transfusion in the very high prevalence setting of South Africa. Aims To determine the incidence of OH and peripartum blood transfusion in four South African hospitals and to compare these rates in HIV positive and HIV negative women.

53

Methods We conducted a cross-sectional review of the medical records on all peripartum women who delivered at four major urban referral hospitals in South Africa: Chris Hani Baragwanath Hospital (Soweto), King Edward VIII Hospital (Durban), Mowbray Maternity and Groote Schuur Hospital (both Cape Town) over a four-month study period (April to August 2012). We collected limited demographic and clinical data on all peripartum (delivery period plus 48 hours before or 48 hours after) obstetric patients admitted during this period. For the subset of patients who sustained OH (WHO definition of >= 500mls after vaginal delivery and >=1000mls after CS); and/or received a blood transfusion, we collected more detailed information on antecedent risk factors for OH, obstetric management, HIV disease, blood transfusion, and maternal and infant outcomes.

Results: We surveyed 15,725 women over the study period of whom 25.2% were HIV positive and 95.7% had received at least some antenatal care; 58.0% had normal vaginal deliveries, 40.4% had Caesarean sections and 189 (1.2%) delivered before arrival at hospital. A total of 387 (2.5%) sustained OH and 447 (2.8%) were transfused; of these, 213 (1.4%) women had both OH and transfusion.

54

Figure 1. Relation between Obstetric haemorrhage and blood transfusion

Transfused Not Transfused Total

Missing 9 46 55

Hemorrhage 213 174 387 (2.5%)

No Hemorrhage 225 15,058 15,283

Total 447 (2.8%) 15,278 15,725

There was no significant difference in OH incidence between HIV positive (2.8%) and HIV negative women (2.3%) . In contrast, the incidence of blood transfusion was significantly higher among HIV positive (3.8%) compared to HIV negative women (2.5%) .

Figure 2a Relation between HIV and OH Hemorrhage YES NO

HIV POSITIVE 112 (2.8%) 3,841

HIV NEGATIVE 264 (2.3%) 11,074

OR = 1.13, 95% CI 0.89-1.43 (p=0.3355

55

Figure 2b Relation between HIV and Blood transfusion Transfusion YES NO

HIV POSITIVE 150 (3.8%) 3,819

HIV NEGATIVE 280 (2.5%) 11,091

OR = 1.54, 95% CI 1.24-1.92 (p<0.0001)Controlled for age, parity, mode of delivery, and hospital

Although CS was significantly more common in HIV positive women (44%) compared to HIV negative (41%) (p= 0.0008); logistic regression found it not to be a significant determinant of peripartum blood transfusion

Mean hemoglobin on hospital admission was 11.4 g/dL in HIV negative and 11.0 g/dL in HIV positive patients (p< 0.0001). 46% of HIV positive women were anaemic (HB<11gms/dl)37% of HIV negative women were anaemic p< 0.0001

Figure 3. Anaemia, blood transfusion and HIV status

56

ConclusionsThe incidence of OH in South Africa of 2.5% is similar to that in the United States (2.3-2.9%)⁴ and is not associated with HIV status. In contrast, the incidence of blood transfusion (2.8%) is tenfold higher than in the United States (0.24-0.46%)⁵ and is significantly associated with HIV infection. Based upon the observed difference in hemoglobin values between HIV positive and negative patients, we hypothesize that a higher prevalence of antenatal anemia in HIV infected patients accounts in part for higher blood utilization in South Africa. HIV related coagulopathy and institutional or physician specific variability in transfusion practice also remain plausible explanations. Further analysis of clinical parameters is in progress.Research is in progress (the TIPS study) on antepartum and peripartum transfusion practices as well as antenatal anaemia. In addition the study is being extended to rural settings.

References1. Saving Mothers 2011-2013. The sixth report of the National Committee for

Confidential Enquiry into Maternal Deaths in South Africa. Short report. Dept Health, Pretoria 2015.

2. Saving Mothers 2005-2007. The fourth report of the National Committee for Confidential Enquiry into Maternal Deaths in South Africa. Comprehensive report. Dept Health, Pretoria 2008.

3. Bates I, Chapotera GK, McKew S, van den Broek N. Maternal mortality in sub-Saharan Africa: the contribution of ineffective blood transfusion services. BJOG 2008;115: 1331-9.

4. Callaghan WM, Kuklina EV, Berg CJ. Trends in postpartum hemorrhage: United States, 1994-2006. Am J Obstet Gynecol 2010;202: 353 e1-6.

5. Kuklina EV, Meikle SF, Jamieson DJ, et al. Severe obstetric morbidity in the United States: 1998-2005. Obstet Gynecol 2009;113: 293-9.

57

SYPHILIS IN HIV-INFECTED MOTHERS AND INFANTS: RESULTS FROM THE NICHD/ HPTN 040 STUDY

The NICHD HPTN 040 Study Team, presenting author Gerhard Theron, Department of Obstetrics and Gynaecology, University of Stellenbosch and Tygerberg Hospital.

Background: Untreated syphilis during pregnancy is associated with spontaneous abortion, stillbirth, prematurity and infant mortality. Syphilis may facilitate HIV transmission, which is especially concerning in low and middle income countries where both diseases are common.

Methods: We performed an analysis of data available from NICHD/HPTN 040 (P1043), a study focused on the prevention of intrapartum HIV transmission to 1684 infants born to 1664 untreated HIV-infected women. The present analysis evaluates risk factors and outcomes associated with a syphilis diagnosis in this cohort of HIV-infected women and their infants.

Results: Approximately 10% (n=171) of women enrolled had serological evidence of syphilis without adequate treatment documented and 1.4% (n=24) infants were dually HIV and syphilis infected. Factors associated with confirmed or presumed syphilis in HIV infected mothers and HIV transmission to infant are summarised in Table 1. Risk factors associated with infant syphilis and HIV co-infection are summarised in Table 2. Multivariate logistic analysis showed that compared to HIV-infected women, co-infected women were significantly more likely to self-identified ethnicity (AOR 2.5, 95% CI 1.5-4.2), to consume alcohol during pregnancy (AOR 1.5, 95% CI 1.1-2.1) and to transmit HIV to their infants (AOR 2.1, 95% CI 1.3-3.4), with 88% of HIV infections being acquired in-utero. As compared to HIV infected or HIV exposed infants, co-infected infants were significantly more likely to be born to mothers with VDRL titers >1:16 (AOR 3, 95% CI 1.1-8.2) and higher viral loads (AOR 1.5 95% CI 1.1-1.9). Of 6 newborns with symptomatic syphilis, 2 expired shortly after birth, and 2 were HIV-infected. Description of infants symptomatic of syphilis infection shortly after birth is provided in Table 3.

Conclusion: Syphilis continues to be a common co-infection in HIV-infected women and can facilitate in utero transmission of HIV to infants. Most infants are asymptomatic at birth, but those with symptoms have high mortality rates.

58

59

Table 1: Factors associated with confirmed or presumed syphilis in HIV infected mothers and HIV transmission to infant

SyphilisN=171

No syphilisN=1493

Odds ratio (95% CI)

Adjusted Odds ratio (95% CI)a

Age (mean, SD) 28 (+6.3) 27 (+6.3) 1.03 (1-1.1)* 1.03 (1-1.06)*Ethnicity (non-white) 153 (89%) 1166

(78%)2.4 (1.4-3.9)* 2.5 (1.5-4.2)*

Illegal substance use (yes) 24 (14%) 123 (8%) 1.8 (1.1-2.9)* 1.5 (0.88-2.7)Alcohol use (yes) 79 (47%) 511 (35%) 1.6 (1.2-2.3)* 1.5 (1.1-2.1)*Tobacco use (yes) 61 (36%) 472 (32%) 1.2 (0.87-1.7) 1 (0.67-1.5)CD4 count (cells/mm3, SD) 506 (+300) 516

(+310)1 1

Log10 viral load (copies/ml, SD)

9.61 (+2) 9.5 (+1.9) 1.04 (0.96-1.1)

1.01 (0.93-1.1)

Gestational age at delivery (mean, SD)

38.5 (+1.7) 38.6 (+1.7)

0.96 (0.87-1) 1

No prenatal Care 76 (45%) 554 (37%) 1.4 (1-2)* 1.5 (0.9-2.12)Region- (South Africa) 46 (27%) 420 (28%) 0.9 (0.63-1.3) 0.68 (0.44-1)HIV transmitted to infant 24 (14%) 116 (8%) 2 (1.2-3.2)* 2.1 (1.3-3.4)*

HIV transmitted in uteron (% of total HIV cases)

21 (88%) 72 (62%) 4.3 (1.2-15.2)*

Table 2: Risk factors associated with infant syphilis and HIV co-infection Co-infected infants with HIV & syphilis (n=24)

Infants not co-infected (n=1660)

Odds ratio (95% CI)

Adjusted odds ratio (95% CI)a

Birth weight (mean, SD) 2.735 kg (+430g)

2.991 kg (+522g)

0.99 (0.99-.1)* 1 (0.998-1)

Gestational wk (mean, SD)

38.3 (+2) 38.6 (+1.7) 0.88 (0.71-1.1)

Maternal VDRL Titers# 3.7 (1.5-9.1)* 3 (1.1-8.2)*1:1-1:8 12 (52%) 117 (80%)>1:16 11 (48%) 29 (20%)

Maternal log10 viral load (copies/ml, SD)

10.8 (+1.6) 9.5 (+2) 1.5 (1.2-1.9)* 1.5 (1.1-1.9)*

HIV Treatment arm 1 (0.44-2.4)Zidovudine only 8 (33%) 557 (34 %)

Two and three drug group

16 (67%) 1103 (66%)

Delivery type-vaginal 11 (46%) 1073 (65%) 0.45 (0.2-1)No use of zidovudine during delivery

12 (50%) 980 (59%) 0.69 (0.3-1.5)

Region- Born in South Africa

3 (13%) 472 (28%) 0.36 (0.11-1.2)

0.3 (0.07-1.1)

Mothers not referred for syphilis treatment during L&D admission

2 (8%) 25 (17%) .44 (0.1-2)

#3 unknown titer levels *P value <0.05a Adjusted for birthweight, maternal VDRL titers, maternal log10 viral load, regionSD: Standard Deviation, HIV Human Immunodeficiency Virus, VDRL: Venereal Disease Research Laboratory, L&D: Labor and Delivery

60

Table 3: Description of infants symptomatic of syphilis infection shortly after birthCase Mother’s

VDRLMother’s Confirmatory testing

Country

Infant’s titer

Symptoms Death

HIV Status

1 1:256 Reactive Brazil 1:256 Premature, low birth weight, pneumonia cardiorespiratory arrest

Yes Unknown

2 1:256 Reactive South Africa

No titer Hepatomegaly, sudden death

Yes Unknown

3 1:64 Reactive Brazil 1:16 Thrombocytopenia, hepatomegaly

No Infected in utero

4 Reactive, no titer performed

Not performed

Brazil No titer performed

Elevated transaminases, thrombocytopenia, jaundice, sepsis

No Infected in utero

5 1:1024 Reactive Brazil 1:64 Fetal distress, neonatal jaundice, anemia

No Uninfected

6 Nonreactive

Reactive Brazil 1:4 Rash on extremities No Uninfected

61

DEVELOPMENT OF A LOW COST VITAL SIGNS DEVICE TO DETECT PRE-ECLAMPSIA AND SHOCK

Hannah L Nathan, Natasha L Hezelgrave, Lucy C Chappell, Andrew H ShennanWomen’s Health Academic Centre, King’s College London, London, UK

Introduction Obstetric haemorrhage, pre-eclampsia and sepsis contribute to more than half of all maternal deaths worldwide; all are associated with abnormal vital signs. Following a successful pilot study, we developed, validated and evaluated a vital sign-monitoring device suitable for use in low- and middle-income countries (LMICs).

Methods1. Device development; incorporating ‘traffic-light’ early warning systems

(EWS), suitable for minimally-trained healthcare providers.2. Prospective pregnancy hypertension (including pre-eclampsia) and

hypotension validations in South Africa (British Hypertension Society protocol).

3. Retrospective analysis of two large datasets of women with PPH (UK, Nigeria, Zimbabwe, Zambia and Egypt) evaluating vital sign prediction of adverse maternal outcomes, determining optimal predictor thresholds for the EWS shock triggers.

4. Mixed-methods, multi-centre prospective evaluation of vital sign-monitoring device and predictive value of vital sign thresholds of traffic-light EWS within the device at three South African institutions.

ResultsThe CRADLE device achieved a B/A grade for pregnancy, including pre-eclampsia (n=45) and an A/A grade for pregnant women with low BP (n=30). Shock index (SI), ratio of HR to systolic BP, was the most consistent predictor of adverse outcomes, compared to conventional vital signs. SI had the highest AUC value for predicting ICU admission (0.75, 0.63-0.76), significantly higher than SBP (p=0.023). Using optimal sensitivity and specificity SI values thresholds of ≥0.9 and ≥1.7 were selected for amber and red lights respectively. Established hypertension triggers were used for the hypertension thresholds.Twelve-month prospective evaluation of vital sign-monitoring device and predictive value of vital sign thresholds of traffic-light EWS, according to adverse maternal outcome, has commenced at three South African institutions.

Conclusion The CRADLE device is an accurate, low-cost and user-friendly device incorporating a traffic light EWS, alerting clinicians to the need for urgent referral or intervention, potentially reducing maternal mortality and morbidity. Prospective facility-based evaluation is underway at three sites in South Africa. Funded by Bill and Melinda Gates Foundation

62

THE EFFECT OF CALCIUM SUPPLEMENTATION ON BLOOD PRESSURE IN NON-PREGNANT WOMEN WITH PREVIOUS PRE-ECLAMPSIA: AN EXPLORATORY, RANDOMIZED PLACEBO CONTROLLED STUDY (WHO STUDY A65750)

On behalf of the calcium and Pre-eclampsia Study Group*-Hofmeyr GJ, Seuc A, etrán AP, Purnat T, Ciganda A, Manyame S, Munjanja SP, Singata M, Fawcus S, Frank K, Hall DR, Cormick G, Roberts J, Belizan J, Bergel E, Drebit S, Von Dadelszen P

Background: Epidemiological findings suggest that the link between poverty and pre-eclampsia might be dietary calcium deficiency. Calcium supplementation in the second half of pregnancy is associated with a modest reduction in pre-eclampsia. In the general (non-pregnant) population, low dietary calcium intake is associated with hypertension with some evidence that calcium supplementation may reduce blood pressure. If calcium deficiency is a factor in the genesis of pre-eclampsia, it is feasible that women predisposed to pre-eclampsia may be more susceptible to the effects of calcium on blood pressure, particularly those with a history of severe pre-eclampsia.Methods: The WHO long-term calcium supplementation in women at high risk of pre-eclampsia (CAP) Study is a multi-country randomized, double-blind placebo-controlled clinical trial to test the hypothesis that calcium deficiency in early pregnancy may play a role in the genesis of pre-eclampsia. Non-pregnant women who had pre-eclampsia or eclampsia in their most recent pregnancy are randomized to receive either 500 mg/day elemental calcium or placebo. The trial started in 2011 and is on-going.This is an exploratory sub-study to determine the effect of calcium supplementation on the blood pressure of non-pregnant women with previous pre-eclampsia, and the relative effect on women with previous severe versus non-severe pre-eclampsia. Women recruited in the trial at least 6 weeks after the preceding pregnancy with blood pressure data available for baseline and one or two subsequent follow-up trial visits while still not pregnant were included. Follow-up visits are scheduled at approximately 12 and 24 weeks after randomization.

63

Results: Of 836 women randomized in the study database dated 9 September 2014, relevant data were available at the 1st visit (12 week follow-up visit) in 367 women of whom 217 had previously had severe pre-eclampsia, and at the 2nd visit (24 week follow-up visit) in 201 women. There was an overall trend to reduced blood pressure at the follow-up visits with consistently greater reduction in the calcium than in the placebo groups, however the differences were small (1 to 2.5 mmHg) and not statistically significant. In the subgroup with previous severe pre-eclampsia, the mean reduction in diastolic blood pressure in the calcium group (-2.6 mmHg) was statistically larger than the corresponding change in the placebo group (+0.8 mmHg), (mean difference -3.4, 95% confidence interval -0.4 to -6.4; p=0.025). The effect of calcium on diastolic pressure change at 12 weeks was greater in the group with previous severe pre-eclampsia than that with previous non-severe pre-eclampsia (p=0.020, ANOVA analysis).Conclusions: Due to the exploratory nature of this sub-study and the multiple comparisons performed, the results need to be interpreted with caution. The findings are consistent with our hypothesis that women with previous severe forms of pre-eclampsia are more sensitive to the effects of calcium supplementation on diastolic blood pressure than women with previous non-severe pre-eclampsia. The possibility that women with a history of severe pre-eclampsia are uniquely sensitive to calcium deficiency warrants further research.

BackgroundCalcium intake and pre-eclampsiaHypertension has been estimated to complicate 5% of all pregnancies and 11% of first pregnancies, half of these being associated with pre-eclampsia, and accounting for 14% of 343 000 maternal deaths annually (1). Pre-eclampsia, defined as high blood pressure and proteinuria occurring after the 20th week of pregnancy, is considerably more prevalent in poor than in wealthy communities. Two striking exceptions have been identified. More than 50 years ago, a low prevalence of pre-

64

eclampsia was reported from Ethiopia where the diet, among other features, contained high levels of calcium (2). The observation in 1980 that Mayan Indians in Guatemala, who traditionally soaked their corn in lime before cooking, had a low incidence of pre-eclampsia and eclampsia (3), stimulated interest in the concept that the link between poverty and pre-eclampsia might be dietary calcium deficiency.The hypothesis that an increase in calcium intake during pregnancy might reduce the incidence of pre-eclampsia was tested in several randomized trials commencing in the late 1980s. Our systematic review showed that calcium supplementation of at least 1g daily, commencing around mid-pregnancy, was associated with a modest reduction in pre-eclampsia, and notably a reduction in its severe manifestations, particularly among women at increased risk, or with low dietary calcium intake (4). We recently reviewed evidence from randomized trials of lower doses of calcium (<1000 mg, mainly 500 mg daily) (5). The limited evidence suggested a similar reduction in pre-eclampsia to that found with larger doses.A trial nested within the large WHO trial of calcium supplementation (1.5 g daily from approximately 20 weeks’ gestation) in pregnant women with low dietary calcium intake, failed to demonstrate an effect of calcium supplementation on biochemical measures commonly elevated in pre-eclampsia: serum urate, platelet count, and urine protein/creatinine ratio (6). To reconcile the considerable evidence for reduced pre-eclampsia with calcium supplementation, with the absence of evidence of an effect on proteinuria and other markers for pre-eclampsia, we proposed the hypothesis that calcium supplementation in the second half of pregnancy reduces blood pressure and thus the diagnosis and severe manifestations of pre-eclampsia, without a significant effect on the underlying pathology, and the epidemiological association of low dietary calcium with pre-eclampsia might be due to an effect in the first half of pregnancy. To test this hypothesis we are conducting a placebo-controlled randomized trial of calcium supplementation commencing before pregnancy in women with previous pre-eclampsia (7).

65

Calcium intake and hypertensionLow dietary calcium intake is associated with hypertension in the general population (8). In animal studies, a high calcium diet reduces hypertension associated with oral contraceptive treatment, by improving diuresis and vasorelaxant responses (9). A high dairy diet has been found to reduce systolic and diastolic blood pressure by about 2 mmHg, an effect correlated with reduction in intracellular calcium (10). An alternative mechanism whereby calcium supplementation might lower blood pressure is by mitigating the hypertensive effect of sodium chloride (common salt) (11). Another possible mechanism is via changes in vitamin D and parathyroid hormone levels (12, 13). Calcium supplementation has also been shown to improve insulin sensitivity in women with Type 2 diabetes and hypertension (14). In the latter study, blood pressure was not significantly reduced (15).A meta-analysis published in 1999 of the effect of dietary or non-dietary calcium supplementation on blood pressure found an overall reduction in systolic and diastolic blood pressure of 1,44 and 0.84 mmHg, respectively (4560 participants) (16). A subsequent randomized placebo controlled trial in 1471 non-hypertensive women with average age 74 years found that 1g calcium daily was associated with a small reduction in systolic blood pressure at 6 months only. For a subgroup of women with low dietary calcium, the reduction was greater and persisted (17).Whether calcium supplementation reduces blood pressure in hypertensive patients is not clear. A Cochrane review in 2006 including 13 small trials in men and women (total participants 485, largest trial 90, calcium dosage 400 mg to 2 g, median follow up 8 weeks) concluded that: “Due to poor quality of included trials and heterogeneity between trials, the evidence in favour of causal association between calcium supplementation and blood pressure reduction is weak and is probably due to bias” (18). We are not aware of any previous randomized trials of calcium supplementation in non-pregnant women with previous pre-eclampsia. If calcium deficiency is a factor in the genesis of pre-eclampsia, it is feasible that women predisposed to pre-eclampsia may be more susceptible to the

66

effects of calcium on blood pressure. This manuscript presents the results of a sub-analysis of the above-mentioned ongoing WHO trial: long term calcium supplementation in women at high risk of pre-eclampsia (7) with the objective to assess the effect of calcium supplementation on systolic and diastolic blood pressure in non-pregnant women with previous pre-eclampsia. Our hypotheses are that (1) calcium supplementation in non-pregnant women with previous pre-eclampsia is associated with reduced systolic and diastolic blood pressure; and (2) the effect of calcium supplementation on blood pressure is greater in women with previous severe pre-eclampsia/eclampsia.

MethodsThe WHO long term calcium supplementation in women at high risk of pre-eclampsia trial is a multicenter randomized, double-blind placebo-controlled clinical trial. Non-pregnant women who had pre-eclampsia or eclampsia in their most recent pregnancy are randomized to receive either 500 mg/day elemental calcium or placebo. The trial started in 2011 and is on-going. The methodology of the trial has been described in the published protocol (7). A brief description is presented below.SettingsThis is a multi-centre, multi-country trial in hospitals in South Africa, Zimbabwe and Argentina. The sites in Africa are government secondary or tertiary urban referral hospitals with large obstetric units serving urban and rural population. The South African hospitals are located in Cape Town, East London, Johannesburg and Stellenbosch. In Zimbabwe the two maternity units included are in Harare. In Argentina, the site comprises one maternity-hospital in Tucuman, two in Buenos Aires and one in the province of Buenos Aires. Dietary surveys in the East London, Mdantsane and Western Johannesburg populations prior to participation in the 2006 WHO trial of calcium supplementation during pregnancy to prevent pre-eclampsia, found that the median daily dietary calcium of primiparous women was about 600 mg (19). In the current on-going trial, a nutritional interview is conducted

67

among all women enrolled in the trial who reach 20 weeks gestation to confirm low calcium intake in the study population. ParticipantsWomen are eligible if they had pre-eclampsia or eclampsia in their most recent pregnancy, if they are in a sexual relationship, not pregnant, not using contraception and if they give informed consent. Exclusion criteria are: less than18 years of age; chronic hypertension with persistent proteinuria; calcium supplement intake; and history or symptoms of urolithiasis, renal disease or parathyroid disease.The interventionWomen in the intervention group take one chewable tablet containing elemental calcium 500 mg daily from enrolment (before pregnancy) until 20 weeks’ gestation. The women are asked to chew the tablet during the day, not close in time to taking food or iron supplements. Women in the control group take placebo tablets identical in shape, colour and taste to the intervention tablet. The women are encouraged not to take any additional calcium supplements. All women take unblinded calcium supplementation (1.5 g) from 20 weeks’ gestation until delivery, according to WHO recommendations (20).

68

Data acquisition for blood pressure measurementBlood pressure measurement was standardized at the beginning of the trial according to recommendations by the British Hypertension Society (http://www.bhsoc.org/latest-guidelines/how-to-measure-blood-pressure/). Blood pressure is measured and recorded for all women at recruitment (baseline) and in all subsequent trial visits. The visits are scheduled every 12 weeks until the woman becomes pregnant, then at 8, 20, and 32 weeks of pregnancy.Blood pressure is measured with a standard mercury sphyngomanometer at the sites in Africa and with an automated sphyngomanometer at the sites in Argentina.Sub-study population and methods For the current sub-study, routine data collected during the trial visits were analysed. Women recruited in the trial with blood pressure data available for baseline and at least one subsequent follow-up trial visit while still not pregnant up to September 2014 were included. To eliminate the effect of recent pre-eclampsia, women with their previous delivery at less than six weeks prior to baseline (randomization) were excluded. Baseline data were compared between the calcium and placebo groups for the study population to ensure no selective loss to follow-up. Due to the high number of comparisons we used 0.025 as indicating statistical significance. We used an intention-to-treat (ITT) approach for this analysis.Two comparisons were computed: between baseline and 1st trial visit (approximately 12 weeks after baseline); and between baseline and 2nd

trial visit (approximately 24 weeks after baseline). For each woman the changes in systolic and diastolic blood pressures were calculated by subtracting the measurement at baseline from the measurement at the first and second follow-up visit. The mean change in systolic and in diastolic blood pressures were compared between the calcium and placebo groups for the first follow-up visit minus baseline and for the second follow-up visit minus baseline. The mean changes were expressed as mean differences with 95% confidence intervals. For each woman, the 1st visit (or 12 week visit) was considered as the visit occurring between 6

69

and 18 weeks, and the 2nd visit (or 24 week visit) was considered as the visit occurring between 18 and 30 weeks.Sub-group analyses were performed to compare the effect of calcium supplementation on blood pressure at the first visit after admission between women with a history of severe pre-eclampsia and women without such a history. A history of severe pre-eclampsia in the previous pregnancy was defined as: i) eclampsia, or ii) HELLP syndrome i, or iii) systolic blood pressure higher than 160 or iv) diastolic blood pressure higher than 110 or v) onset of pre-eclampsia earlier than 28 weeks, or vi) ICU admission. An ANOVA analysis on both systolic and diastolic blood pressures was conducted controlling for treatment received (placebo vs. calcium) and history of severe pre-eclampsia (yes/no). Using preliminary information, we had decided that for this secondary analysis we would need no less than 180 subjects per arm (360 patients in total) to be able to detect a systolic blood pressure difference of 5 mmHg between the placebo and Calcium groups (130 and 125 mmHg respectively), and with a common SD of 16 mmHg. With an alpha value of 0.05, this sample size would allow us to detect this difference with 90% power. SPSS version 20 was used for all analyses.

ResultsA total of 836 women were randomized in the CAP study database dated 9 September 2014 (figure 1). Blood pressure at admission was compared with blood pressure in the 1st follow-up visit (approximately 12 weeks after admission) in 367 women, and with blood pressure in the 2nd follow-up visit (approximately 24 weeks after admission) in 201 women. In the 12 week sub-group with severe pre-eclampsia, 217/367 women versus 150/376 women without severe preeclampsia were analyzed.Figure 1 shows the trial profile of the sub-study. From 836 women randomised, 469 were excluded from the 1st visit vs. admission comparison for the following reasons: 168 had a previous pregnancy less than six weeks before enrolment and 388 had blood pressure missing at admission and/or in the 1st trial visit (e.g. not yet due for 12 week visit, or

70

missed visit). None were pregnant at the 1st visit. For the comparison between 2nd visit vs. admission, an additional 166 women were excluded from the analysis because blood pressure was missing at this visit (e.g. not yet due for visit or missed visit). Baseline data comparing the calcium and placebo groups for women included or excluded from the comparisons at 1st and 2nd visit are shown in Table 1. Results are presented as means or percentages, with p values There were not statistically significant differences at the p <0.025 level between the included calcium versus placebo groups. We did not quantified statistically the differences at baseline between women included in the sub-study vs. women excluded but Table 1 shows the characteristic of both groups.Table 2 shows the changes in systolic and diastolic blood pressure in the 1st and 2nd visit stratified by calcium and placebo groups. There was an overall trend to reduced blood pressure at the follow-up visits, except for the placebo group diastolic pressure in the 1st visit. The reductions in blood pressures were consistently greater in the calcium than in the placebo groups, however the differences were small (1 to 2.5 mmHg) and none were statistically significant.Table 3 shows the results for the subgroup with history of severe pre-eclampsia. The mean reduction in diastolic blood pressure in the calcium group (-2.6 mmHg) is statistically larger than the corresponding change in the placebo group (+0.8 mmHg), (mean difference -3.4, 95% confidence interval -0.4 to -6.4; p=0.025) but the other differences are not statistically significant. In the ANOVA analysis controlling for treatment received (placebo vs. calcium) and history of severe pre-eclampsia (yes/no), the only statistically significant result was the interaction between the two predictors for DBP (p=0.020).We used an intention-to-treat (ITT) approach for this analysis. Nevertheless, compliance was similar and above 80% in both groups. At 1st visit, compliance was 81.1% and 82.4% in the calcium and placebo group, respectively. At 2nd visit, compliance was 83.2% and 82% in the calcium and placebo group, respectively.

71

DiscussionDue to the exploratory nature of this sub-study and the multiple comparisons performed, the results need to be interpreted with caution. This sub-study did not confirm an overall effect of calcium supplementation on blood pressure. We found a consistent trend to greater blood pressure reductions in the calcium than the placebo group in both trial visits (at 12 weeks and 24 weeks after randomization) which is consistent with such an effect, but the differences were not statistically significant except for diastolic pressure in the sub-group of women with previous severe pre-eclampsia. These women with previous severe forms of pre-eclampsia (including eclampsia and HELLP syndrome) may be more sensitive to the effects of calcium supplementation on diastolic blood pressure than women with previous non-severe pre-eclampsia. The later finding is consistent with our hypothesis that women who are susceptible to pre-eclampsia may be uniquely sensitive to calcium deficiency, either because of greater dietary deficiency or because of an inherent sensitivity to calcium deficiency. The CAP study once completed should provide sufficient data to test this hypothesis in an independent dataset.While emphasizing the need for caution in interpreting the results of exploratory research with several outcomes measured, these findings are consistent with the possibility of a role for calcium deficiency in the aetiology of pre-eclampsia and justify further research in this field.

AcknowledgementThe calcium and pre-eclampsia (CAP) study is part of the PRE-EMPT (Pre-eclampsia-Eclampsia, Monitoring, Prevention and Treatment) study and is supported by the University of British Columbia, a grantee of the Bill & Melinda Gates Foundation; the World Health Organisation; and Argentina Fund for Horizontal Cooperation of the Argentinean Ministry of Foreign Affairs.

References1.Say L, Chou D, Gemmill A, Tuncalp O, Moller AB, Daniels J, et al. Global causes of maternal death: a WHO systematic analysis. The Lancet Global health. 2014;2(6):e323-33.2.Hamlin RH. The prevention of eclampsia and pre-eclampsia. Lancet. 1952;1(6698):64-8.

72

3.Belizan JM, Villar J. The relationship between calcium intake and edema-, proteinuria-, and hypertension-getosis: an hypothesis. AmJClinNutr. 1980;33(10):2202-10.4.Hofmeyr GJ, Atallah AN, Duley L. Calcium supplementation during pregnancy for preventing hypertensive disorders and related problems. Cochrane Database Syst Rev. 2006;3:CD001059.5.Hofmeyr GJ, Belizan JM, von Dadelszen P, Calcium, Pre-eclampsia Study G. Low-dose calcium supplementation for preventing pre-eclampsia: a systematic review and commentary. BJOG. 2014;121(8):951-7.6.Hofmeyr GJ, Mlokoti Z, Nikodem VC, Mangesi L, Ferreira S, Singata M, et al. Calcium supplementation during pregnancy for preventing hypertensive disorders is not associated with changes in platelet count, urate, and urinary protein: a randomized control trial. Hypertens Pregnancy. 2008;27(3):299-304.7.Protocol 11PRT/4028: Long term calcium supplementation in women at high risk of pre-eclampsia: a randomised, placebo-controlled trial (PACTR201105000267371). Lancet.8.Centeno V, de Barboza GD, Marchionatti A, Rodriguez V, de Talamoni NT. Molecular mechanisms triggered by low-calcium diets. NutrResRev. 2009;22(2):163-74.9.Olatunji LA, Soladoye AO. High-calcium diet reduces blood pressure, blood volume and preserves vasorelaxation in oral contraceptive-treated female rats. VasculPharmacol. 2010;52(1-2):95-100.10.Hilpert KF, West SG, Bagshaw DM, Fishell V, Barnhart L, Lefevre M, et al. Effects of dairy products on intracellular calcium and blood pressure in adults with essential hypertension. Journal of the American College of Nutrition. 2009;28(2):142-9.11.Kotchen TA, McCarron DA. Dietary electrolytes and blood pressure: a statement for healthcare professionals from the American Heart Association Nutrition Committee. Circulation. 1998;98(6):613-7.12.Belizan JM, Villar J, Repke J. The relationship between calcium intake and pregnancy-induced hypertension: up-to-date evidence. Am J Obstet Gynecol. 1988;158(4):898-902.13.Webb RC. Smooth muscle contraction and relaxation. Adv Physiol Educ. 2003;27:201-6.14.Pikilidou MI, Lasaridis AN, Sarafidis PA, Koliakos GG, Tziolas IM, Kazakos KA, et al. Insulin sensitivity increase after calcium supplementation and change in intraplatelet calcium and sodium-hydrogen exchange in hypertensive women with Type 2 diabetes. Diabetic medicine : a journal of the British Diabetic Association. 2009;26(3):211-9.15.Pikilidou MI, Befani CD, Sarafidis PA, Nilsson PM, Koliakos GG, Tziolas IM, et al. Oral calcium supplementation ambulatory blood pressure and relation to changes in intracellular ions and sodium-hydrogen exchange. American journal of hypertension. 2009;22(12):1263-9.16.Griffith LE, Guyatt GH, Cook RJ, Bucher HC, Cook DJ. The influence of dietary and nondietary calcium supplementation on blood pressure: an updated metaanalysis of randomized controlled trials. Am J Hypertens. 1999;12(1 Pt 1):84-92.17.Reid IR, Horne A, Mason B, Ames R, Bava U, Gamble GD. Effects of calcium supplementation on body weight and blood pressure in normal older women: a randomized controlled trial. The Journal of clinical endocrinology and metabolism. 2005;90(7):3824-9.18.Dickinson HO, Nicolson DJ, Cook JV, Campbell F, Beyer FR, Ford GA, et al. Calcium supplementation for the management of primary hypertension in adults. Cochrane Database Syst Rev. 2006(2):CD004639.

73

19.Villar J, Abdel-Aleem H, Merialdi M, Mathai M, Ali MM, Zavaleta N, et al. World Health Organization randomized trial of calcium supplementation among low calcium intake pregnant women. American journal of obstetrics and gynecology. 2006;194(3):639-49.20.WHO. Guideline: Calcium Supplementation in pregnant women. Geneva: World Health Organization, 2013.

74

EFFECT OF DIFFERENT POSITIONS ON MATERNAL AND FETAL HEART RATES AT TERM

HJ ODENDAAL 1 , S IBRAHIM2, E JAREFORS3, DG NEL4, CA GROENEWALD5,L VOLLMER6

1, 5, 6 Department of Obstetrics and Gynaecology, Stellenbosch University2, 3 Medical student, University of Linkōping, Sweden4 Department of Statistics and Actuarial Science, Stellenbosch University

IntroductionA key assessment in the Safe Passage Study (SPS) (www.safepassagestudy.org) is the transabdominal recording of maternal and fetal ECG data by the Monica AN24TM device near term. Raw data on uterine activity (electrohisterogram) are collected simultaneously. Extracting the maternal heart rate (MHR) from the raw data demonstrated regular decelerations associated with uterine activity. As MHR decelerations during periods of uterine activity could be indicative of aortocaval compression, this prospective study was done to assess different maternal positions at term on maternal blood pressure (BP), MHR and fetal heart rate (FHR).

ObjectiveTo determine the effect of maternal position on periodic changes of the maternal heart rate.

MethodsThe study was done on the day before elective caesarean section. Four different maternal positions were studied, lateral (right or left), supine, lateral (the opposite from the first) and supine again, each for 15 minutes at a time. Raw electrical data were obtained from 5 ECG electrodes on the maternal abdominal and blood pressure every 5 minutes from the left arm and leg. The raw data were downloaded on a dedicated laptop and analyzed with the DK1.9 programme supplied by Monica Health CareTM. The Dawes and Redman criteria were used to analyze the maternal and FHR patterns in detail. STATISTICA version 12 was used for the analyses. 

ResultsHundred-and-fifty seven women were approached for the study, 122 consented and 119 reached completion. No decline in blood pressure was noted when women were changed to the supine position. No parameter of the FHR changed significantly during position changes. Large decelerations of the MHR, exceeding 20 bpm for at least 4 occasions were observed in 13 (10.9%) of participants. No significant differences were seen when these 13 participants were compared to the remaining 106.

ConclusionsSupine position for short periods of time has no influence on the MHR, FHR and BP in pregnant women at term.

75

A BASELINE EVALUATION OF CARE PROVIDED AT NEONATAL NURSERIES IN KWAZULU-NATAL

C. Horwood, Z. Mbatha, S. Phakathi, R. Brown and L. Haskins.

BackgroundMore than 23000 newborn babies die every year in South Africa; as a result, improving outcomes for these infants is a priority. The KINC intervention aims to improve neonatal care using a multi-pronged approach of training, mentoring, and peer-to-peer learning. We conducted a quantitative baseline evaluation of care provided in neonatal nurseries (NNN) in KZN.

MethodologyAn observational cross-sectional study was undertaken in 39 district hospitals in KZN.Structured data collection tools were used to collect observational data about staffing, infrastructure, equipment, consumables, record keeping and resuscitation equipment. Policies and administration were reviewed, and knowledge, skill and care practices were assessed. Observational data and questionnaires were entered and analysed using Access database. The record review was manually scored out of 10.ResultsThere was a dedicated doctor in 12 (31%) of NNN but only 5 (12.8%) doctors were trained to use the KINC guidelines. In 28 (71.7%) NNN there was a KINC trained professional nurse and in 3 (7.6%) there was a KINC trained enrolled nurse. Some district hospitals lacked the infrastructure to support the allocated number of high and intermediate care beds (Table 1). Thirty three (84.6%) NNN lacked the number of wall oxygen points and 26 (66.6%) did not have functioning portable oxygen available. Only 14 (35.8%) NNN had a CPCP machine and 15 (38.4%) NNN had piped medical air or a compressor (but 2 of these did not have a CPAP machine). Of the 13 NNN with a functioning CPAP machine, 5 had not used the machine in the last 3 months.Many NNN lack essential equipment to function effectively (Table 2). Essential resuscitation equipment is not available in some NNN, labour wards, post-natal wards and operating theatres (Table 3).The record review showed the full complement of all 5 records were unavailable in 7 hospitals. The scores for the hospitals ranged from 1.4 to 9.2. Conclusion At baseline we found a number of district hospitals lacked basic infrastructure and essential equipment for the NNN to function effectively. Moreover, resuscitation equipment was deficient in many district hospitals. In addition, staff lacked the training to use the KINC/SAINC guidelines. It is essential to have all the necessary infrastructure and equipment to resuscitate and manage neonates should they need it as well as staff trained to do so.

76

Table 1: Infrastructure to support neonatal nurseries in district hospitals in KwaZulu-Natal Beds/Infrastructure KZN Norm Complian

tToo Many

Non-Complian

tMinimum HC beds set up and ready

2 8 15 16

Wall oxygen points 2 per HC bed

1 per IC bed

2 4 33

Has portable oxygen functioning

1 per nursery

13 NA 26

Suction points 2 per HC bed

1 per IC bed

2 7 30

Portable suction 1 per nursery

31 NA 7*

Electrical points 12 per HC bed

1 3 34*

Medical air or compressor 1 per nursery

15 24

CPAP machine 1 per nursery

14 25

Wall thermometer 1 per nursery

16 22

*missing data

Table 2: Equipment levels in neonatal nurseries in KwaZulu-Natal district hospitals Equipment KZN Norm Compliant Too Many Non-

CompliantMultifunction monitors 1 per HC

bed7 10 21*

BP cuffs (both term and preterm)

12 26

Oxygen blenders 1 per HC bed

1 38

Head boxes 0,5 per O2 point

7 32

Glucometers 1 per cubicle

33 6

Stethoscopes 1 per HC 4 35

77

bedInfusion pumps 2 per HC

bed4 4 30*

Syringe drivers 4 per HC bed

0 38*

*missing data

78

Table 3: Essential neonatal resuscitation equipment

  In the neonatal nursery

In the labour ward

In the operating theatre

In the post natal ward

 

Com

plia

nt

Part

ially

co

mpl

iant

Non

-co

mpl

iant

Com

plia

nt

Part

ially

co

mpl

iant

Non

-co

mpl

iant

Com

plia

nt

Part

ially

co

mpl

iant

Non

-co

mpl

iant

Com

plia

nt

Part

ially

co

mpl

iant

Non

-co

mpl

iant

Resuscitaire working with both heat and light

      15   23* 21   16* 1   36*

Neonatal ambubag and mask functioning

32   6* 32   6* 31   6* 3   34*

All mask sizes present (0, 00, 000) 0   39 4 7 27* 2 7 28* 2   35*

Laryngoscope set up and functioning

22 9 8 28   10* 27   9* 5   32*

Blades for laryngoscope 22 9 8 13 4 20* 14 6 17* 5   32*Spare batteries or second laryngoscope

17   22 16   21* 16   21* 3   34*

ET tubes (sizes 2.5, 3.0 and 3.5) 19 4 16 12 5 21* 14 4 9* 2 1 34*

Adrenaline 34   5 33   5* 28   4* 4   33**Missing data

79

KWAZULU NATAL INITIATIVE FOR NEWBORN CARE: A QUALITATIVE IMPROVEMENT PROGRAMME FOR NEONATAL CARE IN DISTRICT HOSPITALS.

Nyasulu, Dolly, Horwood, Christiane, Haskins, Lyn 1 Centre for Rural Health, School of Nursing and Public Health, University of KwaZulu Natal, Durban, South Africa

80

1.1 IntroductionMillenium Development Goals (MDGs) 3 and 4 commit nations to reduce maternal mortality by three-quarters, and mortality among children under 5 years by two-thirds, between 1990 and 2015 1. Despite the introduction of many health policies and programmes in South Africa since 1994, and substantial investment in public health infrastructure, little progress has been made towards achieving the improvements in child and maternal mortality required by the MDGs.South Africa has been highlighted as one of only twelve countries with an increasing level of child mortality since the setting of the MDGs in 1990 2. Neonatal death rates have risen in resource-limited settings, and there has been no measurable change in mortality in the first week after birth 3. The South African under-five mortality rate is estimated at 68 per 1,000 live births, or over 50,000 children per year 4, and 30% of these under-five deaths are neonatal, occurring in the first 28 days of life 5. The following factors: infections, complications of preterm birth and birth asphyxia has been identified to be the cause of approximately 90% of global neonatal deaths 6. The University of KwaZulu Natal (UKZN) Centre for Rural Health (CRH) recognises the need and the urgency for creating effective interventions that can be successfully implemented within the existing constraints in the health system. CRH is a health care organisation which is committed to work with the Department of Heath (DoH) to improve the quality of newborn care and thus reduce mortality and morbidity. CRH is commited to the improvement of neonatal care in KZN as demonstrated in their recent research studied 8, 9. KwaZulu Natal Initiative for Newborn Care (KINC) one of the flagship programmes of the CRH is central to improving the quality of newborn care particiualary in district hospitals.

1.2 Brief Background of the KINC Programme The KINC programme sterms from the Limpopo Iniative for Newborn Care (LINC) which was implemented in Limpopo for over 10 years. Centre for

81

Rural health was involved in the LINC programme and experiences and lessons learnt from LINK were used to develop a programme called the Zululand Iniative for Newborn Care (ZINC). CRH successfully implemented ZINC in one district in KZN before rolling out the programme to all KZN districts. Currently the programme is recognised by National Deprtment of Health and has now been adopted as a National Programme called South African Initiative for Newborn Care (SA _INC).

1.3 Aims of the KINC programmeThe aim of the KINC programme is building capacity of health care providers using evidence based new born care practices to improve the quality of neonatal care in district hospitals in KZN. Furthermore CRHs aims to promote an enabling environment within which optimal newborn care can be provided at all times.

1.4 Objectives of KINCIn order to fulfil these aims the following objectives have been outlined:

To provide KINC training to health workers in the maternity units in district hospitals in KZN, in partnership with district clinical specialist teams and regional hospital paediatric consultants.

To provide site based mentoring and support of health workers in district hospitals to ensure implementation of KINC guidelines.

To undertake an accreditation process for all district hospitals in KZN

To undertake an impact evaluation of KINC implementation in KZN

1.5 KINC Training ApproachKINC Training is district based and is implemented in all 39 KZN district hosptials. Training targets the doctors, professional nurses and enrolled nurses working in the neonatal nurseries of the district hospitals. Doctors and professional nurses are trained together in order to build strong working relationships between the two, and to minimise conflict that can

82

arise in the work place from misunderstanding of the roles expected from each of the team members.Enrolled nurses have been found to be a more stable workforce than doctors and professional nurses hence they have been trained in the KINC approach to ensure a stable capacity in nurseries. In addition to training doctors, professional nurses and enrolled nurses, KINC trains a group of master trainers to enable them to rollout the KINC programme in the districts and to ensure sustainability of the programme beyond CRH participation. These master trainers consist of DCSTs Paediatricians and some Obstetricians. District Clinical Team Members are also trained in KINC with a special focus on KINC facilitation and supervisory skills in order for the DCST to be able to support the work in the neonatal nursery.KINC guidelines are introduced to Nursery Unit Managers and they are oriented on the KINC standards of practice. This is done in order to facilitate institutionalisation and sustainability of the program.KINC training is mostly held in a regional hospital to enable exposure of participants who are based at district hospitals where some facilities and some conditions are usually not available. During the training regional hospital consultants assist with the training and clinical demonstrations of selected procedures essential for care of sick neonates. The use of regional hospitals also promotes good working relationships between the district hospitals and their referral hospitals. Referral patterns are discussed during the training in order to provide clarity related to regional hospitals not accepting babies referred to them from the district hospitals.

1.6 KINC training guidelines and methodologyThe KINC training guidelines are evidence based and are derived from recommendations listed in the South African Saving Babies Reports. The colour coding model similar to IMCI has been used in the design of the learning materials.The KINC training program follows the principles of adult learning and competency based training methods. Participants are encouraged to use

83

their experiences to contribute meangfully to group discussions and demonstration of skills. A pre- and post-course assessment of knowledge are conducted as a form of assessing participants’ baseline knowledge and knowledge gained during training.The training package consists of Newborn Care Charttbooks; Trainee Manuals; Exercise Manuals and Facilitators Manuals. Each participant is provided with a chart books, a trainee manual and an exercise manual. The KINC facilitator’s manual is utilised to organise training and to facilitate training and learning.

1.7 KINC ActivitiesKINC training activities are evidence based and include the following:• Resuscitation for birth Asphyxia including Helping Babaies Breath

( HBB)

• Emergency care for life threatening conditions of the newborn

• Immediate care and subsequent care for life threatening conditions of the newborn

• Urgent specialised and non specialised care including referral to next level of care

• Monitoring in the immediate newborn period• Routine care, thermal protection, eye care and cord care • Early initiation of breastfeeding without pre-lacteal feeds and

exclusive breastfeeding on demand• Counselling for care and actions at home, advice for early detection

of danger signs and appropriate care-seeking

1.8 Achievements The programme is currently in its third and the last year which will expire at the end of March 2016. Since the start of the programme Todate four hundred and fifty six (456) health care professionals have been trained. All 39 KwaZuluNatal district hospitals were visited by KINC

84

team for supervision. Each hospital received at least four visits. This translated to a total of 135 menetoring visits understaken by the KINC team. The table below shows the number of trained health care works from the start of the programme up to 06 March 2015.

Table 1: No of categories of KINC trained personnel Doctors

Professional Nurses

Enrolled Nurses

Unit managers

KINC Master Trainers

HBB Master Trainers

Total Trained

59 103 141 53 36 64 456

Training of operational managers on Action Learning has been set up in all 11. A baseline evaluation was conducted before the start of the programme. The midterm evaluation has already been undertaken and this will be followed by the end programme evaluation later in the year.

1.9 Strengths of the KINC programme

Project communicationStrong partnerships have been established with the Department of Health during this first year of the KINC project. A KINC task team was set up to guide the project. Meetings are heald every three months. Dr McKerrow, Head of Paediatrics from KZN DoH, chairs the meetings and representatives include DCSTs in all districts, Departments of Paediatrics at regional and tertiary hospitals, and KZN DoHand staff from the MCWH directorate. Five meetings of the KINC task team were held during the reporting period. The KINC team at CRH has established strong working relationshsips with the District Clinical Specialist Teams, particularly in terms of the mentoring visits which are always conducted together with a representative of district management.Support of the KZN The KZN Department of Health

85

KZN DOH has provided strong leadership and support for the KINC project. Dr McKerrow has been available at all times to support KINC activities, and to work with the CRH team to ensure that activities are aligned with DoH priorities. He is always readily available to provide the logistical support required to ensure that planned activities are carried out as planned. Support from Porvincial DOH province has strengthened the buy-in from DoH management at district level which is crucial to ensure that activities involving DoH staff can be successfully arranged with appropriate participants being released to participate in training. 1.11 ChallengesAs with all projects there are challenges experienced during implementation of the KINC programme. However, as mentioned above, the strong partnership and support of DOH makes it much easier to find solutions to these challenges. The most common challenges are:• High staff turn over

The rapid staff turnover makes it difficult to standardise practice when the experienced staff leave. • Funding of accommodation for residential KINC training

Although funding for accommodation is available, it is not always possible to get suitable accommodation for participants at a reasonable cost. This has resulted to burning of available funds at a faster rate than planned. • Availability of master trainers during training sessions

Master Trainers are a valuable assert to their ditricts and as such they are involved in many projects making it difficult for them to be always available when required for training.• Managing and maintenance of Equipment. Requests for

repairs and purchace of essential equipment.

Policies related to purchase of essential eaquipment, servicing equipment and repairing old equipment are cumbersome and often have a very long turna around time. Many a times works have to do without essential equipment and this impacts on the quality of care provided.

86

• Structural constraints.

Some Units are too small to accommodate equipment required. This makes it difficult for facilities to comply with DOH Equipment norms. It has also been noted that facilities are always reluctant to embark on major changes especially if there are plans for renovations and or building new facilities.

87

1.12 Conclusion:In conclusion the KINC programme has highlighted newborn care in district hospitals. All districts have well established KINC teams with each district hospital having KINC trained personnel in nursery. The strong teamwork at district and province level has strengthened the programmeFinally a robust end programme evaluation will determine the impact of KINC on quality of care and neonatal care provided in district hospitals. It is hoped DOH members trained on KINC will ensure sustainability of KINC within their districts.

References[1] Assembly, U. G. (2000) United Nations Millennium Declaration: Resolution, UN.[2] Tomlinson, M., Doherty, T., Jackson, D., Lawn, J. E., Ijumba, P., Colvin, M., Nkonki, L.,

Daviaud, E., Goga, A., Sanders, D., Lombard, C., Persson, L. A., Ndaba, T., Snetro, G., and Chopra, M. (2011) An effectiveness study of an integrated, community-based package for maternal, newborn, child and HIV care in South Africa: study protocol for a randomized controlled trial, Trials 12, 236.

[3] Lawn, J. E., Kerber, K., Enweronu-Laryea, C., and Massee Bateman, O. (2009) Newborn survival in low resource settings--are we delivering?, BJOG 116 Suppl 1, 49-59.

[4] UN Inter-agency Group for Child Mortality Estimation, U., WHO, The World Bank, United Nations DESA/Population Division. (2010) Levels & Trends in Child Mortality, Report 2010.

[5] Chopra, M., Daviaud, E., Pattinson, R., Fonn, S., and Lawn, J. E. (2009) Saving the lives of South Africa's mothers, babies, and children: can the health system deliver?, Lancet 374, 835-846.

[6] Black, R. E., Cousens, S., Johnson, H. L., Lawn, J. E., Rudan, I., Bassani, D. G., Jha, P., Campbell, H., Walker, C. F., Cibulskis, R., Eisele, T., Liu, L., and Mathers, C. (2010) Global, regional, and national causes of child mortality in 2008: a systematic analysis, Lancet 375, 1969-1987.

[7] Pattinson, R. (2003) Why babies die-a perinatal care survey of South Africa, 2000-2002: original article, South African Medical Journal 93, p. 445-450.

[8] Horwood, C., Butler, L. M., Haskins, L., Phakathi, S., and Rollins, N. (2013) HIV-infected adolescent mothers and their infants: low coverage of HIV services and high risk of HIV transmission in KwaZulu-Natal, South Africa, PloS one 8, e74568.

[9] Horwood, C., Vermaak, K., Butler, L., Haskins, L., Phakathi, S., and Rollins, N. (2012) Elimination of paediatric HIV in KwaZulu-Natal, South Africa: large-scale assessment of interventions for the prevention of mother-to-child transmission, Bulletin of the World Health Organization 90, 168-175.

88

THE IMPACT OF AN EDUCATIONAL COUNSELLING ON ACCEPTABILITY OF DONATED BREAST MILK AT A SEMI-RURAL REGIONAL HOSPITAL, NORTHERN KWAZULU-NATAL.

N. Kapongo*, T. Kalala* Z.* Duze, M. Moelo(*): Paediatric Department, Neonatal Unit, Lower Umfolozi, Regional War Memorial Hospital (LURWMH).1. INTRODUCTION

1.1 Background and Literature review

The importance of breast milk for infant’s growth, development and overall health has been well established (1). Breast milk is the ideal food for newborns and infants. It gives infants all nutrients they need for healthy development. It is safe and contains antibodies and other immunological factors that help to protect infants from common childhood illnesses- such as diarrhoea and pneumonia, the two primary causes of child mortality worldwide (2). Breastfeeding also benefits mothers. The practice when done exclusively is associated with a natural method of birth control (98% of protection in first 6 months after birth). It reduces risks of breast and ovarian cancer later in life, helps women return to their pre-pregnancy weight faster, and lowers rate of obesity(1). The World Health Organization (WHO) endorsement of human milk banking has remained consistently positive and in 1992 the WHO recommends that for infants who cannot receive breast milk from their mothers, the next preferred option is donated breast milk (donor milk) (5).

There are multiples obstacles to the acceptability of donor milk: lack of awareness/ familiarity with the processes around the donor milk, psychological concerns, and lack of trust in health system, lack of government and health worker endorsement and commitment. These obstacles can be addressed through education (13). In Nigeria, Ighogboja et al. (14) studied mothers’ attitudes towards donated breast milk, human banking, and breastfeeding in the event of HIV positivity. About 71 per cent would not accept donated breast milk for their baby while the rest would consent only if the donors were close family relatives, owing to fear

89

of diseases (28%), fear of transfer of genetic traits (22%) and religious and cultural taboos (14%). However 60 per cent were willing to donate breast milk. Only 38 per cent would accept milk from a breast milk bank. None would breastfeed in case of HIV positivity.

A lack of familiarity with the processing of donor breast milk - including screening, pasteurization, and cold storage – is the largest impediment to greater acceptability. Also ensuring the effectiveness of screening methods and pasteurization procedures is critical to achieving more widespread acceptability. Education campaigns to increase acceptability should begin with antenatal care and should be targeted at increase the Knowledge of the dangers associated with formula feeding and of the lifesaving properties of breast milk. To ensure wide acceptability of human milk banking in South Africa, the practice needs to be supported by the country’s national and local departments of health. With the support of the Department, health professionals will be in the best position to strengthen the public’s confidence in the safety of breast milk collection and pasteurization (15).

LUDWMH is a regional hospital situated in Northern Kwazulu Natal province with catchment population of 2.3 million. It is an accredited Baby friendly hospital initiative (BFHI) since 2006. In 2011, the hospital put in place a process to set up a breast milk bank. Two phases were targeted , the first being the provision of safe breast milk to HIV exposed infants using the flash heating method and the second, the establishment of breast milk bank. The level of acceptability of donated breast milk at LUDWMH is unknown.

The present study was undertaken with the aim to determine the level of donated breast milk acceptability among antenatal clinic attendants and post natal mothers. Secondly, we want to develop an instrument to measure acquisition of knowledge on donated breast milk. Thirdly we sought to assess the short term impact of such acquired knowledge on donated breast milk acceptability.

90

2. METHODS

This was a prospective, descriptive, short term-interventional, hospital based study on donated human breast milk acceptability. The targeted study populations were: All pregnant mothers attending antenatal clinic at LUDWMH, l mothers in the post natal wards and those in KMC unit or Nursery lodger’s mothers.

2.3 Settings Empangeni Neonatal unit is part of a child and Maternal Hospital, Lower Umfolozi District War Memorial Hospital (LUDWMH) in North-East of Kwazulu-Natal, South Africa. This is the only maternity and neonatal regional referral centre for an area with a population estimated at 2, 3 Millions. According to census 2001, poverty rate and unemployment rates stand at 63.5% and 53.7% respectively. The proportion of households with access to safe water (32%) and sanitation (24%) are far below the national figures (79% and 62%, respectively). Fifty thousand (50 000) live births occur in the entire area annually .The neonatal service was introduced at the hospital in 1998 with a 15 unit beds without intensive care facilities. Between 1999 and 2008 the unit was expanded to 92 bed neonatal units: 16 NICU beds, 40 high care, 16 special care beds and a Kangaroo mother care (KMC) unit (20 beds). In 2012, there were 9975 live births at LUDWMH. The early neonatal mortality rate (ENMR) and the neonatal mortality rate (NMR) were 16/1000 and 25/1000 live births, respectively. Very low birth weights (<1500 g) infants represented 67, 3% of the neonatal deaths. Extreme prematurity, hyaline membrane disease and infections are the main causes of death in this birth-weight category. Up take of breast milk in the unit is estimated to be 85% according to administrative records. It is believed in the unit that increase this up take to 100% will have significant impact on mortality and will change the microbial profile of Nosocomial infection which is a significant contributor to morbidity and mortality.

2.4 Study Participants

91

At each location site within the hospital, potential participants were identified through active screening.

o Inclusion criteriao Any pregnant woman attending ante-natal clinic or any mother

in post natal setting at LUDWMH and willing and able to give written informed consent to participate to the study.

o Exclusion criteriao Any potential participant unwilling and unable to give written

informed consent.

2.5 Data Collection

2.5.1 Pre- intervention data

Study participant were interviewed using a structured pre-tested questionnaire. Socio-demographic data was obtained in addition to the participant’s willingness to be recipient or donors of breast milk. Reasons were solicited for the negative responses. Participants were specifically questioned in relation with transmissible diseases such as HIV. Their attitudes towards breast milk bank and offering stored breast milk to their babies were ascertained. Pre-intervention test to assess participant knowledge on donated milk processing and the benefits of breast milk was conducted.

2.5.2 The intervention

This was an educational counseling with the aim to increase the familiarity of participants with the processing of donor breast milk – including donor screening, pasteurization, and cold storage. The educational counseling addressed the dangers associated with formula feeding, the lifesaving properties of breast milk and the issues around disease transmission. The intervention was conducted by a pre-trained nurse, member of the hospital breastfeeding committee very familiar with breast milk

92

counseling. She was blind to the participant pre-intervention responses with regards to attitudes towards donated breast milk.

Selection of relevant information for educational intervention

To enhance participant understanding of issues around processing of donor breast milk, it was first necessary to develop and select appropriate educational interventions. The neonatal Unit team developed , by consensus, a list of minimal information detailing specific issues relevant to concerns about donor milk as reported by PATH with regard to increasing acceptability of donated breast milk . Five categories have been identified:

Modalities of providing breast milk to babies Benefits of breast milk Negative health consequences of formula Safety of donated breast milk

Transparency of the process of donated breast milk

Development and testing of the measure of educational material acceptability (EMA)

We incorporated 5 domains into the EMA: Clarity, accuracy, content, reading level and technical quality. The response in this multi-item instrument used a five-point scale ranging from strongly disagree (1) to strongly agree (5).

Development of the knowledge acquisition questionnaire (KAQ)

The instrument to measure knowledge was multidimensional with each dimension representing a specific aspect of the educational process related to each of the minimal criteria used in selecting the education (see annex 1). Thus the questionnaire

93

incorporates core areas of knowledge that provided participants with the minimum information required. A pilot study was conducted to assess these 2 instruments.

3. Statistical 1. Patient rating of booklets

AcceptabilityEach of the 10 items of the EMA were assessed using a repeated measure analysis of variance across the two booklets analysed using the Epi- info program 2002.2. Pre-post-tests of knowledge

Reliability and responsivenessThe internal consistency of the instrument was measured using Cronbach’s alpha. For pre and post-test KAQ questionnaires, the object of measurement was the participant mother. Reliability was calculated pooling all mothers who received pre and post-test questionnaires. Responsiveness refers to the ability of the instrument to detect changes in participant status, even if those changes are small in score from post-test to pre-test divided by the standard deviation of change.

4. RESULTS1. The measure of educational material acceptability (EMA)

Rating of the two booklets was performed on 20 participants. For the 10-items EMA instrument, the overall composite EMA mean scores from thembalethu EMA and Empangeni EMA were similar , 42.35 ( std 5.74) vs 42.45 (std 5.99) , respectively. For item 2 which relates to the size print of the booklet being large enough for easy reading, the Empangeni EMA scored better than the thembalethu one (mean score 4.4 vs 3.3; 95% CI of difference -1.3724<-1.1<-0.8276; p=0.0000). There was a difference in mean score (4.45 vs 3.95) in favour of the thembalethu booklet with regard to item 7 (the booklet has motivated me to make changes in how I view and understand donated breast milk). But this difference was not statistically significant (p=0.9999).

94

2. Participant characteristicsTable 2 presents the demographics and general characteristics of 100 participants who consent to a pre and post educational counselling questionnaire to assess their views and knowledge on breastfeeding and donated breast milk. The mean age was 25.7 and the majority (75%) was in the age group 20 to 35 years. So were those in the parity group 0-1 (51%). Of the hundred participant 81 per cent completed at least grade 8. But only 20% of them completed grade 12 at the time of the interview. One third (30%) of participant had HIV positive status.

3. The knowledge acquisition questionnaireThe responsiveness of the KAQ instrument developed in our study was 0.84/ 2.1 =0.4. The pre-and post –mean scores are shown in table 3. The questions that the participants consistently scored less post educational counselling, were related to the safety of donated breast milk (pooled mean score of 1.97, STD =0.94; median =2; out of a perfect score of 3.) and the transparency of breast milk bank processes( pooled score of 1.19 ; STD 0.58; median =1 ; out of a perfect score of 2). This information is summarized in table 4.

4. Measuring acceptability of donated breast milkThe pre-educational counselling (pre-KAQ) acceptability to donate and to receive donated breast milk was 57% and 74%, respectively. The post-educational counselling (post KAQ) acceptability rates were 80% to donate breast milk and 84% to receive. A responsiveness of 0.40 to KAQ yielded significant increase in donated breast milk acceptability. The composite acceptability mean score (combined score of 2 acceptability rates and the positive attitude towards breast feeding for a mother with HIV positive status) showed significant increased rate post intervention (mean score difference =1.84; 95% CI: 1.13<1.84<2.54).

5. DISCUSSIONAscertaining the acceptance and usefulness of education materials requires the ability to measure (1) the patient’s perception (acceptability) of the material ; and (2) measure knowledge gained (acquisition) as a

95

result of the educational intervention. At the bottom line, ascertain how the knowledge acquisition has influenced or is associated with positive behaviour change / attitude, significant enough to yield positive impact on health wellbeing. We have developed instruments that help both health care practitioners and lodger mothers (1) educational interventions one can get the most information on donated breast milk, (2) provide a measure of how well intervention contributes to patient knowledge about donated breast milk and finally showed how the knowledge acquisition changed the acceptability of donated breast milk. Measuring acceptability of booklet (EMA)The internal consistency of the educational material (EMA), using Cronbach’s alpha was 0.75 (a value > 0.70 indicate good internal consistency). Only one of ten items questionnaire gave superior acceptability to Empangeni EMA if compared to Thembalethu EMA. This is item number 2 which is related to the print size being large enough for easy reading.Testing knowledge gained and assessing acceptability on donated breast milk.Our primary finding was that the KAQ instrument was poorly responsive (0.40). The ability of an instrument to detect change over time is necessary for the evaluation of educational programs. Participants scored poorly with regard to knowledge related to safety of donated breast milk and the transparency of breast milk bank. However, despite this low responsiveness, significant improvement in acceptability of donated breast milk was achieved as short-term impact of short educational counselling on donated breast milk among young mothers with mid high school level education and from a poor socio-economic background in majority. One should be realistic how much 30 minutes educational counselling can achieve in term of knowledge acquisition in such socio-economic environment. The 80 to 84 % acceptability of donated breast milk is high enough to sustain breast milk donation to our milk bank from the pool of lodger mothers of the neonatal Unit. This is also an acceptable

96

baseline level to sustain and improve breast milk up take in the Unit. As part of quality assurance project, allocating more time to safety and transparency aspect of the EMA during educational counselling has a potential to improve the short- term knowledge acquisition score and may be escalate the acceptability rates. One could argue that the possible limitation of the KAQ is the possible ceiling effect in patient who has been previously educated. We used the framework of the health education model, positing the following causal steps: mothers will hear, understand and learn information about donated breast milk and the importance of breast milk which will subsequently lead to improved breastfeeding rate and up take of donated breast milk. We are aware that acceptability is a complex decision process where knowledge acquisition is only one determinant among multiple others, with some responding to rules beyond the rationale boundaries, and among those, personal preferences, religious beliefs, and cultural practices are commonly reported.Ultimately, a number of complex interactions between patients and health care providers or information they receive from various sources will influence their decision making process and behaviour. This is the basis of the Health education Model which constitutes the first step in dealing with issues related to non-compliant and acceptability behaviour. Ensuring optimal knowledge acquisition requires that educational materials be understandable and appealing to mothers, and we, clinicians need to measure the extent to which mothers have integrated their educational messages. Our two instruments may help providers utilize information that patients understand and evaluate the impact of such knowledge on acceptability rates. The strength of our work lies in (1) the development of an instrument to measure patient acceptability of education, (2) an instrument of educational counselling on donated breast milk, (3) the measure of knowledge acquisition among a sample of mid high school mothers of semi-rural, poor socio-economic background, (4) and assess

97

change in donated breast milk following short term small knowledge acquisition on donated breast milk. Measuring acceptability of donated breast milkWe have established for the first time in the hospital, what is our baseline acceptability of donated breast milk among lodger mothers, post Natal ward and pregnant women attending antenatal clinic. The pre- and post-intervention acceptability to donate and receive breast milk of 57% and 74%, respectively may be already a result of our routine breast feeding education campaign in the Unit. The increase to 80% and 84% level post intervention, respectively, is remarkable despite a low level responsiveness (0.40) of acquisition knowledge observed in the study. This observation of little increase in knowledge acquisition coupled to significant positive attitude change toward donated breast milk is encouraging. It may be possible with little routine educational effort and perseverance to maintain this positive attitude. Here again it is possible that a cycle of action research aimed at maximizing knowledge acquisition especially in the safety and transparency component of donated breast will go a long way to push the acceptability rates around the ceiling levels achievable trough the health educational model. Even the more psychological concerns would also likely be reduced over time as these educational efforts progress and donated breast milk practices cross the critical boundaries to become part of institutional memories to be passing on from one generation of health care workers to the next. Mothers are more likely to accept to receive than to donate breast milk in this observational study. Part of this fact may be explain by the fact that mothers with HIV positive status are more likely to say to be unfavourable to donate breast milk and this opinion did not change post intervention. But they are willing to receive breast milk as long as the educational interventions address the concern of safety of the milk.Assessing the benefit of the study on the Unit

The EMA we developed for the purpose of the study has been changed many time through multiple cycles of research actions during the course of the project, from the initial planning , the pilot

98

study, the incorporation of maternal feedback and lessons learn from poor performance on knowledge acquisition with respect to safety and transparency components. The current format and current content culminated into an official pamphlet we use to communicate with mothers in the lodger mothers and KMC facilities in the Neonatal Unit at LUDWMH. This pamphlet has become part of the hand out resource package on admission and discharge.

The KAQ document has become a model of the audit tool to assess the neonatal Unit campaign on breast feeding in the hospital

The acceptability assessment tool will become our model to monitor acceptability of donated breast milk as we establish our breast milk bank.

Going forward our understanding of mothers characteristics and the determinants of acceptability of donated breast milk among the peculiar young and low level educated mother’s population will be of great help to sustain the breast milk bank

The baseline acceptability rates are of great help to assess the impact of future quality improvement projects on donated breast milk.

The accumulated knowledge, practical lessons learnt, basic skills in planning, leadership, communication, information search and implementation expertise have enabled me to participate actively in breast milk forum at hospital level and our input in formulating the neonatal Unit’s milk bank operational protocol was highly appreciated in the Unit.

6. CONCLUSIONShort educational counselling on donated breast milk can influence donated breast milk acceptability. Donated breast milk programs need instruments that help evaluate specific aspects of program. The instruments we constructed to evaluate mother’s satisfaction with education and knowledge about breastfeeding and donated breast milk and the impact on acceptability rates can be used in units with donated

99

breast milk program to sustain and improve acceptability of donated breast milk. Further, the KAQ’s ability to measure knowledge gained, offer a prospect to further research to explore the relationship between knowledge change and acceptability rate change and hopefully optimal compliance and clinical outcomes.

7. References

1. WHO Collaborative study Team on the role of breastfeeding on the prevention of infant mortality: Effect of breastfeeding on infant and child mortality due to infectious diseases in less developed countries: a pooled analysis. Lancet 2000, 355: 451-455.

2. Quigley MA, Kelly YJ, Sacker A., Breastfeeding and hospitalization for diarrheal and respiratory infection in the United Kingdom Millennium Cohort Study. Pediatrics 2007;119: e837-42

3. WHO: Infant and young child feeding: Model Chapter for textbooks for Medical Students and Allied Health Professionals. Geneva, Switzerland: World Health Organization; 2003

4. Irene Coutsoudis, Alissa Petrites, Anna Coutsoudis. Acceptability of donated breast milk in a resource limited South African setting. International breastfeeding Journal 2011, 6:3.

5. Ighogboja IS et al. Mothers’ attitudes towards donated breast milk in Jos, Nigeria www.ncbi.nlm.gov/pubmed/7619300.

6. PATH. Increasing acceptability. Http//www. Hmbasa.org.za

100

8. Results tables

Table 2 Participant demographics and general characteristicsAge (years)

1. Mean age (range) : 25.7 ( 16-44)2. Age groups : n = 100

< 20 yrs: 15 (15%) 20-35 yrs: 75 (75%) > 35 yrs: 10 (10%)

Parity1. Mean (range) : 1.6 (0-6)2. Parity group: n= 100

0-1: 51 (51%) 2-4: 46 (46%) > 4: 3 (3%)

Level of education ( n=100) No schooling: 1 (1%) Grade 1-7: 9 (9%) Grade 8-12: 81 (81%) Higher education: 9 (9%)

HIV status n=100 Negative: 30 (30%) Positive: 69(69%) Unknown: 1 (1%)

101

Table 3 Pooled composite mean KAQ scores and responsiveness of the KAQPooled KAQ scores N perfect score Mean SD Pre-test 100 15 10.98 2.68Post-test 100 15 11.82 2.14Difference - - 0.84KAQ responsiveness(Post -test Mean – pre-test Mean) / SD : 11.82-10.98/ 2.1 = 0.40

Table 4 pooled mean KAQ scores per individual question itemsQuestion items Perfect score Mean STD Median

1. Why Breast milk is good for babies 3 2.93 0.29 32. What are the modalities of B/F 4 3.1 0.90 33. What are the side effects of formula 3 2.5 0.67 34. How safe is donated breast milk 3 1.97 0.94 25. Transparency of breast milk bank 2 1.19 0.58 1

Table 5. acceptability of donated breast milk1. Acceptability to donate breast milk

N Yes No Pre- educational counselling 100 57 (57%) 43 (43%)Post- educational counselling 100 80 (80%) 20 (20%) OR=3.02. ( 1.54< OR< 5.96) p< 0.00008

2. Acceptability to receive donated breast milk

N Yes NoPre-educational counselling 100 74 (74%) 26 (26%)Post-educational counselling 100 84 (84%) 16 (16%)

OR= 1.84. ( 0.87< OR< 3.93 ) ; p=0.053. Composite acceptability score ( to donate, to receive, acceptability of a HIV positive

mother to breast) N Perfect score Mean score 95% CI STD

Pre-educational counselling 100 15 10.98 9.77-10.78 2.6Post-educational counselling 100 15 12.12 11.63-12. 61 2.5

Mean score difference= 1.84 ( 95% CI : 1.13< 1.84 < 2.54 ) Responsiveness = 12.12-10.98/2.5 = 0.45

102

LATE NEONATAL DEATHS AFTER A NORMAL HOSPITAL STAY AND DISCHARGE

FS BONDI

INTRODUCTIONAbout 40% of under 5 mortality (U5M) in South Africa (SA) is made up of neonates(1) and thus if SA is to accelerate reduction in U5M it must concentrate on the neonatal period ( 0 to 28 days).However, accurate data on neonatal mortality rate (NMR) is a challenge as there are variations in levels depending on the data source. The perinatal problem identification programme (PPIP) continue to provide large amount of data on NMR at all levels of care in public facilities (2). However, a notable shortcoming of PIPP data is that it concentrates on early neonatal deaths (ENND). These are deaths that occur in labour ward, operating theatre (at the time of caesarean section) and in the nursery. Thus, PIPP tend to omit late neonatal deaths (LNND) or deaths in neonates aged 8 to 28 days.There is thus a need for comprehensive information on neonatal deaths. This is because the 2006 Child PIP data reveals that 5% of the paediatric deaths occur in infants aged 8 to 28 days (3). Since 2002, Madadeni Hospital has provided PIPP data and, the hospital has been part of CHIP since 2007. We found out that when both data were compared, LNND deaths that were documented in CHIP were not reflected in PIPP, as these deaths occurred in the paediatric wards and not the Maternity Wing of Madadeni Hospital. This thus underestimates the neonatal deaths reported from Madadeni to the national PPIP data base.

2. OBJECTIVESTo investigate the causes of neonatal admissions and deaths in the general paediatrics wards of Madadeni Hospital from 2007 to 2012 with the aim of providing the corrected neonatal deaths (NND) figures for PPIP.STUDY SETTINGMadadeni hospital and its nine allied obstetric clinics, Amajuba District, KwaZulunatal.

103

STUDY PERIODJanuary 2007 to December, 2012.DESIGNThis was a retrospective study of neonatal deaths in the general paediatric wards of Madadeni Hospital.The study population consisted of 476 infants aged 28 days or below who were born to parents who were residents of Amajuba district. After a normal admission and discharge (no medical condition at discharge) from the nursery at Madadeni Hospital they were subsequently admitted in the general paediatric wards at Madadeni with various illnesses.Patients were identified from the admission and death registers of general wards 9 and 10 and their case notes were retrieved.The following descriptive information was extracted: birth weight, present weight, gender, age at readmission, social circumstances, main reason for readmission and outcome (survived or died).

3. RESULTSThe profile of 476 neonatal readmissions at Madadeni from 2007 to 2012 is shown below (Table).Almost all (96%) the 476 late neonatal admissions were normal deliveries and low birth weight was present in 179 (62%). Age at readmission revealed that 130 (27.3%) babies were aged below 7 days and the remaining 346 were older.When the socioeconomic context was examined, 223 (47%) came from the surrounding farms, household income was less than R 2000 in 287 (60.3%) and the number of dependents on this income were at least three in 176 (37%) of the 476 infants.The commonest diagnosis was jaundice in 156 (33%). Jaundice was moderate (>300mmol/l) in 81 and severe (>400) in 14. The other two commonest diagnoses were infection and pneumonia.The outcome showed that 54 of 476 (11.3%) re-admissions demised.

DISCUSSION

104

When the early neonatal (ENN) admissions were compared with the late neonatal admissions (LNN) three striking findings were noted:1. The commonest diagnoses amongst the ENN were prematurity, birth asphyxia and infection(4). The corresponding diagnoses in this series of the LNN were jaundice, infection and pneumonia.2. The prevalence of LBW was 38% (179/476) for LNN cases as against 75.6% (811/1071) for ENN cases.3. The overall IHMR for ENN admissions is much lower than that LNN admissions, 6.1% as against 11.3%. Furthermore when the late deaths are added to early deaths the NMR increases. This is an important reason to count every newborn.

105

Number 476 neonates

Age at presentation

<2500g 179(38%)

<7 days 130(27%)

>2500g 297 7-14 days 134Mode of del 15-21

days58

caesarean 20(4%) 21-28 days

54(20%)

NVD 454 HIV statusAssisted 2 exposed 94(19.8%)GENDER Unknown/-

ve/ no results

282

MALES 242(51%)

Time of admission

FEMALES 234 08:00-16:00

223(47%)

16:00-08:00

253

106

Socioeconomic circumstances

ResidenceFarms 223(47%

)Monthly family income

Township 168 <R500 60R500-2000

227

Dependents R2000+ 871 102 No

regular income

102(21.4%)

2 98 Diagnoses

3 53 Jaundice 156(32.8%)

4 35 Sepsis 1225+ 88 Pneumoni

a65

GE 25OTHER 8

Peak TSB LEVELS

Out of 156 jaundiced

Duration of admission

<200 13 1 to 2 days 136 cases200-299 48 3 to 4 days 99 cases300-399 81 5days+ 141(29.6%)

cases400+ 14

OUTCOME Total live births from 2007 to2012

33,602

54 died NND reported by ppip from 2007-2012

443 (NMR 13/100

422 survived When 54 NND in G9&G10 are added

497 NND or NMR 15/1000 LIVE BIRTHS

107

CHANGING PATTERN OF NEONATAL DEATHS IN MADADENI DISTRICT HOSPITAL- A REVIEW OF 1071 DEATHS

FS BONDI

INTRODUCTIONNeonatal (0 to 28 days) mortality is declining globally. Since 1990, neonatal mortality rate (NMR) fell from 33 to 20 per 1000 live births (1).However, in 2013, 2.8 million babies still died in the neonatal period and the proportion of under-five mortality (U5M) deaths during the neonatal period has increased in every region of the world. In South Africa (SA), infants below one month of age account for 40% U5M (2)In order to reduce neonatal deaths (NND), SA has to focus on a better understanding of the causes and circumstances of deaths in the rural and semirural areas, where a majority NND occur in district hospitals(3).It is with this in mind that we undertook this investigation of NND in a large district hospital in Amajuba District, KwaZulu Natal(KZN), South Africa.SETTINGThe study site is Madadeni Hospital and its nine obstetrics clinics, which serve a population of 500,000 in Amajuba District.DESIGNThis was a retrospective, descriptive study of all neonates 500g or more that were delivered in Madadeni Hospital and its annexed clinics between 2003 and 2012. Information pertaining to the patients was extracted from the case notes and transferred to

1. a specially designed form, which was completed soon after the infant demised. Additional details were obtained from the admission and death registers of the participating hospital and clinics .Finally, each death was discussed during our monthly mortality meeting and the allocation of cause of death was based on the outcome of the discussion. The meeetings were supervised by a senior medical officer or consultant paediatrician.

OBJECTIVES

108

To document causes and trend of NND in Madadeni Hospital and its allied clinics.DEFINITIONSSimple and practical definitions were used for the following causes of death:DISSIMINATED INTRAVASCULAR COAGULOPATHY- bleeding from multiple orifices that cannot be reversed by transfusion with FDP.MASSIVE PULMONARY HAEMORRHAGE- sudden deterioration of pulmonary state associated with fresh blood in ET tube or other sites.INRAVENTRICULAR HAEMORRHAGE-sudden deterioration in clinical state, pallor and uniformly blood stained CSF.NECROTISING ENTEROCOLITIS- feeding intolerance, abdominal distension and abnormal gas pattern on abdominal X-ray.SEPSIS- history of chorioamniotis, abnormal WBC with or without a positive blood culture. Meningitis- acutely ill infant with CSF WBC>20 with or without a positive culture.Hypothermia- admission temperature below 35 CPERINATAL ASPHXIA- peripartum complications, depressed baby and Apgar 4 or less at 5 minutes.RESPIRATORY DISTRESS- pulmonary or non-pulmonary cause leading to recession and grunt or sats <88.RESPIRATORY DITRESS SYNDROME- early onset respiratory failure with reticulogranular pattern on CXR.MECONIUM ASPIRATION SYNDROME- particulate meconium in liquor, resp. distress and coarse granular pattern on CXR.PNEUMONIA- respiratory distress, chorioamniotis, unstable tempt, abnormal WBC plus streaks or mottled densities on CXR. Acquired pneumonia in infant older than 7 days.IMMATURITY- infants below 1500g in whom other causes of death have been excluded.COT DEATH-unexplained sudden death, a diagnosis of exclusion.

109

DYSMORPHISM-multiple or complex congenital abnormalities usually of cardiac or neural origin.

110

RESULTSThere were 60,983 live births of which 17,607 (28.9%) were admitted. There were 1071 deaths amongst the 17,607 (6.1%) admissions. The weight categories of the 1071 NND is depicted in table 1 whilst the yearly trend in mortality from 2003 to 2012 is shown in table 2. Prematurity,birth asphyxia and sepsis/meningitis were the commonest causes of death (Table 3).4. The major findings were (a)The NMR has been on the decline since 2007 but stalled in 2010; (b) 51% of the infants who died weighed below 1500g but almost 25% weighed above 2500g; (c)the leading cause of death in this series is prematurity(316 cases) which was further sub-classified into respiratory distress syndrome (108 cases), immaturity (68cases), hypothermia (64cases), necrotising enterocolitis (54cases) and intraventricular haemorrhage (24 cases).It is of interest that dysmorphism, cot death and pulmonary haemorrhage/ disseminated intravascular coagulopathy occurred in 31, 19, 32 cases respectively.

DISCUSSIONThe two important findings in this series are in accordance with national report as reflected in the “saving babies 2012-2013” report. However, the levels and trends and well the proportionate representative of the causes vary and PIPP does not go into the details of each facility. In this study, 21% of the deaths were in infants less than 1000g and a further 31% weighed between 1000 and 1499g.These are the groups of newborns that then to require intensive care. Also, not infrequently, these babies are too ill or unstable to undertake long journey to PMB or Durban for tertiary care. There is thus a need for decentralisation of NICU services in KZN. Madadeni and Newcastle have benefitted from such decentralisation since 2012 .

REFERENCES1. Committing to child survival: a promise renewed. Progress Report 2014.UNICEF.2. Velapi S. Reducing neonatal deaths in South Africa. Are we there yet, and what can be done? S Afri JSH 2012;3:67-67.

111

3. Pattinson R C, Rhoda N. Saving babies 2012-2013. Ninth report on perinatal care in South Africa. Tshepesa Press, Pretoria, 2014.

112

TABLE 1weight 2003-

20052006-2008

2009-2011

2012 Total Per cent

500-999 72 79 60 12 223 20.81000-1499

160 116 43 8 327 30.5

1500-1999

89 57 27 4 177 16.5

2000-2499

37 29 13 5 84 7.8

2500+ 116 93 36 15 260 24.4

Total 474 374 179 44 1,071 100.0

113

TABLE 2.Year All live

birthsSCBU admissions

NND NMR ENMR IHMR

2003 6616 2977 146 22 17 4.92004 6870 2658 185 27 13 6.92005 7357 2846 143 22 19 5.02006 6538 2924 162 25 22 5.52007 6713 2908 115 17 17 3.92008 6593 790 97 15 14 12.32009 6007 544 89 15 12 16.32010 5064 590 46 9 7 7.82011 4978 707 46 9 7 6.52012 4203 663 44 10 6 6.6TOTAL/average 60,939 17,607 1,071 17.1 13.4 7.5

114

TABLE 3.DIAGNOSIS 2003-

20052006-2008

2009-2011

2012 TOTAL PER CENT

PREMATURITY 147 168 57 16 388 36.2

BIRTH ASPHYXIA 104 107 26 7 244 22.8

SEPSIS/MENINGITIS 59 63 20 8 155 14.5

RESP. DISTRESS 82 80 30 8 195 18.2

DYSMORPHISM 13 10 5 3 31 2.9

UNEXPLAINED/SIDS 13 3 7 0 19 1.8

OTHER 11 8 14 6 39 3.6

TOTAL 429 439 159 44 1,071 100.00

115

A COMPARISON OF BOOKED AND UNBOOKED PREGNANT WOMEN PRESENTING IN LABOUR AT THREE HEALTH FACILITIES IN REGION F, INNER CITY OF JOHANNESBURG

Siphamandla Gumede, Vivian Black, Nicolette NaidooInstitution: Wits Reproductive Health and HIV Institute

BackgroundThe global health care strategy is driven by the Millennium Development Goals (MGDs) of which the child and maternal health (MDG 4 and 5) and HIV prevention related goals (MDG 6) emphasize the need to reduce child and maternal mortality and reduction of deaths from HIV/AIDS by 2015.

As part of child and maternal health strengthening, the World Health Organization (WHO) has encouraged development of policies and guidelines that emphasize the importance of antenatal care and ensuring integration of antenatal, PMTCT and maternal services within antenatal care (1).

Although health care services have been oriented towards the delivery of integrated patient care, booking and antenatal attendance remains poor. In Low and Middle Income Countries (LMIC) unbooked women account for 17% to 29% of births (2). To further confirm challenges associated with antenatal booking, the WHO (3) and Demographic and Health Survey (4) reported that out of all pregnant women attending antenatal care during their pregnancy, globally only 53% of women are able to attend four antenatal visits and markedly fewer (36%) in LMIC.

According to a study conducted by Joshi, et al. (5), completing all recommended antenatal visits reduces chances of suffering from pregnancy induced hypertension and pre-term labour. Ideally, a minimum of four antenatal visits are important in ensuring that the pregnancy is well monitored throughout and that all diagnostic assessments are completed under supervision of skilled health care workers.

There is limited information on booking status and outcomes among highly mobile communities or adolescents in communities with a high HIV prevalence. Therefore, this study aimed to determine the prevalence of women who did not attend antenatal care clinic during their current pregnancy (unbooked women).

116

This study also aimed to compare birth outcomes for the booked group of women to the unbooked group. The study also allowed comparison of birth outcomes for booked and unbooked women separated into categories including adolescents and adults. For this study, adolescents referred to women between 10-19 years old.

MethodsStudy design We conducted a cross sectional study using records of births recorded between January 2008 and December 2009. Primary data was collected through review of labour ward registers and information was collected as part of a labour ward birth audit that was conducted by Wits RHI at Charlotte Maxeke Johannesburg Academic Hospital, Hillbrow Community Health Centre and South Rand Hospital. Not all registers were reviewed as some of the birth registers were missing. A sample of registers from CMJAH for 2008-2009 was collected. Because of sampling, CMJAH provided fewer data entries compared to HCHC and SRH.

Study settingThe primary data was collected in the labour wards of Charlotte Maxeke Johannesburg Academic Hospital (CMJAH), South Rand Hospital (SRH) and Hillbrow Community Health Centre (HCHC). These three facilities provide services to the population of the inner city of Johannesburg (region F) and outlying areas.

Inclusion criteriaAll births recorded at CMJAH, SRH and HCHC between January 2008 and December 2009 was included. This included all deliveries at the facility and BBAs that were subsequently managed at the facility and recorded in the birth register.

Sample sizeA total of 12,455 records over a two year period i.e. between January 2008-December 2009 from the three maternal and obstetric units were included in the sample for analysis. Of 12,455 records collected; 7,543 patients were from HCHC, 4,113 records were from SRH and 799 records were from CMJAH.

117

Data collectionUsing a structured data collection tool, labour ward birth registers were reviewed and data was entered into Microsoft Excel worksheet.Data AnalysisFor analysis, data was coded and analysed using STATA version 12 (STATA Corporation, College Station, Texas, USA). Both descriptive and inferential statistics were performed. Descriptive and summary statistics are presented with the 95% confidence interval (CI). Descriptive analysis included determining the frequencies and means with 95% confidence interval and or standard deviation of unbooked women accessing all three facilities in region F. Chi squared test or Fischer Exact test (for observations below 5) were used for analysing categorical variables. Inferential statistics included logistic regression to determine factors associated with booking status, delivery method and infant birth outcome. A p-value indicating significance of ≤0.05 was also used. Lastly, logistic regression analyses were conducted for all factors associated with birth outcomes. Ethical ConsiderationsEthics clearance for the study was obtained from the Human Research Ethics Committee of the University of the Witwatersrand (M121148). Patients’ names or any information linking to patient identities were not collected.

FundingPrimary analysis was part of operational research for health systems strengthening grant within Wits RHI funded by President's Emergency Plan for AIDS Relief (PEPFAR).

Results

Prevalence of booking statusOverall, the majority, 82.0%, (95% CI 79.5%-84.2%) of women had been booked and seen for antenatal care during their pregnancy whilst 1,817 (14.6%) were unbooked. Booking prevalence varied by facility; CMJAH had the highest booking rate (92.1%) whilst HCHC had the lowest booking prevalence at 77.0%. HCHC had the highest proportion of women who were unbooked (n=1728, 22.9%, p<0.001).

118

119

Table 1: Factors associated with booking status

Booking statusVariables n (%, CI) Booked,

(n=10,210, 84.9, 84.2-85.5)

Unbooked,(n=1,817, 15.1, 14.5-

15.8)

p value,N (%)

Perinatal mortality n (%, CI)n=12,027p<0.001

Alive 10,010 (98.0, 97.8-98.3) 1,786 (98.3, 97.6-98.8)Dead 37 (0.4, 0.3-0.5) 20 (1.1, 0.7-1.7)Unknown 163 (1.6, 1.4-1.9) 11 (0.6, 0.3-1.1)Duration of pregnancy, n (%, CI)

n=12,027p=0.070Term 9,978 (97.7, 97.4-98.8) 1,771 (97.5, 96.6-98.1)

Pre-term 211 (2.1, 1.8-2.4) 46 (2.5, 1.9-3.4)Post-term 21 (0.2, 0.1-0.3) 0HIV status, n (%, CI)

n=12,027p<0.001

Negative 6,241 (61.1, 60.2-62.1) 63 (3.5, 2.7-4.4)Positive 2,800 (27.4, 26.6-28.3) 101 (5.6, 4.6-6.7)Unknown 1,169 (11.4, 10.8-12.1) 1,653 (91.0, 89.6-92.3)

Age category, n (%, CI)

6,518 (78.6, 77.8-79.6) 1,767 (21.3, 20.4-22.2)n=8,285p=0.027Adolescent (10-19

years)680 (10.4, 9.7-11.2) 217 (12.3, 10.8-13.9)

Adult (>19) 5,838 (89.6, 88.8-90.3) 1,550 (87.7, 86.1-89.2)

Booking status and perinatal mortality? Our study showed that birth outcome was associated with booking status (p<0.001). From the total 10,210 women booked; 10,010 (98.0%) infants were recorded as alive, 37 (0.4%) were recorded as dead and 163 (1.6%) had unknown viability or birth outcome. Out of the 1,817 total unbooked women, 1,786 (98.3%) were recorded as alive, 20 (1.1%) were recorded as dead and 11 (0.6%) had unknown viability or birth outcome.

Booking status and duration of pregnancyOur study showed no association between duration of pregnancy and booking status (p=0.070).

Of the 10,210 total women booked, 9,978 (97.7%) had term deliveries (n= 9,978, 97.7%), 211 (2.1%) were pre-term deliveries and 21 (0.2%) recorded as post-term. Out of the total 1,817 unbooked women; 1,771 (97.5%) were term deliveries, 46 (2.5%) were pre-term deliveries.

Booking status and maternal HIV status

120

Our study showed a significant association between HIV status and booking status (p<0.001).

From the total 10,210 women booked; 6,241 (61.1%) were HIV negative, 2,800 (27.4%) were HIV positive which was in line with existing routine DHIS data within antenatal care in the public health setting. However, a high proportion of women with unknown HIV status within the category of booked women was also observed (n=1,169, 11.4%).

From the total of unbooked women (1,817); the majority had an unknown HIV status (n=1,653, 91.0%) in comparison to 63 (3.5%) with a HIV negative status and 101 (5.6%) who were recorded as HIV positive.

Booking status and age (adults vs. adolescents)Our study showed a significant association between booking status and age category (p=0.027).

Of 6,518 (78.7%) booked women with recorded age; 680 (10.4%) were adolescents and 5,838 (89.6%) were adults. Out of all 1,767 (21.3%) unbooked women with recorded age; 217 (12.3%) were adolescents while 1,550 (87.7%) were adults. Data showed that there were more adolescents who were recorded as unbooked as compared to those recorded as booked (12.3% unbooked adolescents versus 10.4% booked adolescents, p=0.027).

121

Table 2: Factors associated with pregnancy outcome (viability and duration of pregnancy)

Pregnancy outcomeViability Duration of pregnancy at delivery

Variables Alive Dead TOTAL p-valuePre-term delivery Term delivery TOTAL p-value

Duration of pregnancy, n (%, CI)

12,218 (99.5, 99.4-99.6)

59 (0.5, 0.3-0.6) 12,277

n=12,277

p<0.001

- --

-

Term 11,988 (98.1, 97.9-98.4)

39 (66.1, 52.6-77.9)

12,027 (98.0, 97.3-98.2 ) - - -

Pre-term 212 (1.7, 1.5-2.0) 20 (33.9, 22.1-47.4)

232 (1.9, 1.7-2.1 ) - - -

Post-term 18 (0.1, 0.08-0.2) 0 18 (0.1, 0.08-0.2) - - -        Age category, n (%, CI)

8,077 (99.5, 99.3-99.6)

43 (0.5, 0.3-0.7) 8,120 n=8,12

0 p=0.86

7

216 (2.6, 2.3-3.0)

8,062 (97.4, 97.0-97.7) 8,278 n=8,27

8 p=0.55

7Adolescent (10-19) 875 (10.8, 102-11.5) 5 (11.6, 3.9-

25.1)880 (10.8, 10.1-

11.5 )26 (12.0, 8.0-

17.1)869 (10.8, 10.1-

11.5)895 (10.8, 10.2-

11.5)Adult (> 19) 7,202 (89.2, 88.5-

89.8)38 (88.4, 74.9-

96.1)7,240 (89.2, 88.5-89.8)

190 (88.0, 82.9-92.0)

7,193 (89.2, 88.5-98.9)

7,383 (89.2, 87.3-88.7)

     Maternal HIV status, n (%, CI)

12,218 (99.5, 99.4-99.6)

59 (0.5, 0.3-0.6) 12,277

n=12,277

p<0.001

264 (2.1, 1.9-2.4)

12,168 (97.9, 96.6-98.1) 12,432

n=12,432

p=0.05

Negative 6,372 (52.2, 51.3-53.0)

19 (32.2, 20.6-45.6)

6,391 (52.1, 51.2-52.9 )

120 (45.5, 39.3-51.7)

6,362 (52.3, 51.4-53.2)

6,482 (52.1, 51.3-53.0)

Positive 2,911 (23.8, 23.1-24.6)

13 (22.0, 12.3-34.7)

2,924 (23.8, 23.4-24.6)

66 (25.0, 19.9-30.7)

2,896 (23.8, 23.0-24.6)

2,962 (23.8, 23.1-24.6)

Unknown 2,935 (24.0, 23.3-24.8)

27 (45.8, 32.7-59.2)

2,962 (24.1, 23.4-24.9)

78 (29.5, 24.1-35.4)

2,910 (23.9, 23.2-24.9)

2,988 (24.0, 23.3-24.8)

122

Viability and duration of pregnancy at deliveryThere was a significant association between viability and duration of pregnancy at delivery (p<0.001).

There were 12,218 (99.5%) alive deliveries of whom 11,988 (98.1%) were term deliveries, 212 (1.7%) were pre-term deliveries and 18 (0.1%) were post-term. Out of 59 total deliveries recorded as dead, 39 (66.1%) were term deliveries and 20 (33.9%) were pre-term deliveries.

Pre-term deliveries reported a higher proportion of infants recorded as dead compared to term deliveries.

Pregnancy outcome and age Results from our study indicated no difference between perinatal survival or preterm delivery and age category (adolescents and adults) (p=0.867). Out of all 8,077 (99.5%) alive deliveries with recorded age, 875 (10.8%) were adolescents and 7,202 (89.2%) were adults. Out of all 43 (0.5%) deliveries recorded as dead, five (11.6%) were adolescents and 38 (88.4%) were adults. There were 216 (2.6%) pre-term deliveries with recorded of whom 26 (12.0%) were adolescents and 190 (88.0%). Out of all 8,062 term deliveries with recorded age, 869 (10.8%) were adolescents and 7,193 (89.2%) were adults.

Pregnancy outcome and maternal HIV statusHIV status of a mother has been identified as a risk factor for poor birth outcomes of infants (4). There was a significant association between unknown maternal HIV status and viability (p<0.001).

Findings showed that out of all 12,218 (99.5%) alive deliveries, 6,372 (52.2%) were from HIV negative women, 2,911 (23.8%) were from HIV positive women and 2,935 (24.0%) women had unknown HIV status. Out of all 59 (0.5%) deliveries recorded as dead, 19 (32.2%) were from HIV negative women, 13 (22.0%) were from HIV positive women and 27 (45.8%) women had unknown HIV status which was higher than HIV negative and positive women (p<0.001).

Findings also showed a significant association between HIV status of women and duration of pregnancy at delivery (p=0.05). HIV status demonstrated a similar

123

trend in terms of term deliveries while there was slight variation in terms of pre-term deliveries. Out of 264 (2.1%) pre-term deliveries, 120 (45.5%) were from HIV negative women, 66 (25.0%) were from HIV positive women and 78 (29.5%) had unknown HIV status. From the total 12,168 (97.9%) term deliveries, 6,362 (52.3%) were from HIV negative women, 2,896 (23.8%) were from HIV positive women and 2,910 (23.9%) had unknown HIV status (p=0.05).

Table 3: Logistic regression assessing risk factors for mortality Logistic regression

Risk of mortalityCharacteristics N (%) OR [CI] p-valueBooking status 57Booked 37 (64.9%) 1 -Unbooked 20 (35.1%) 3.06[1.77-5.28] p<0.001

HIV status of mother 59HIV negative 19 (32.2%) 1 -HIV positive 13 (22.0%) 1.50[0.74-3.04] p=0.261HIV unknown 29 (49.2%) 3.11[1.73-5.60] p<0.001

Age 43Adolescent 5 (11.6%) 1 -Adult 38 (88.4%) 0.92[0.36-2.35] p=0.864

Adolescents 5Early-mid adolescent (10-16 years) 1 (20.0%) 1 -Late adolescent (17-19 years) 4 (80.0%) 0.51[0.056-4.56] p=0.543

Duration of pregnancy at delivery 59Term 39 (66.1%) 1 -Pre term 20 (33.9%) 25.49[14.65-

44.45] p<0.001

Table 4: Logistics regression assessing factors associated with booking status

Logistic regressionRisk of being unbooked

Characteristics N OR [CI] p-value

HIV status of mother 1,817HIV negative 63 (3.5%) 1 -HIV positive 101 (5.6%) 3.60[2.62-4.94] p<0.001HIV unknown 1,653 (91.0%) 126.38[97.60-

163.64] p<0.001

Age 1,767Adolescent 217 (12.3%) 1 -Adult 1,550 (87.8%) 0.83[0.71-0.98] p=0.026

Adolescents 217

124

Early-mid adolescent (10-16 years) 23 (10.6%) 1 -Late adolescent (17-19 years) 194 (89.4%) 1.09[0.67-1.79] p=0.724

Our study showed that there were three times greater odds of mortality in the unbooked group as compared to the booked group of women [OR=3.06, (1.77-5.28), p<0.001]. HIV positive women had 50% greater odds of mortality compared to HIV negative women, though this was not significant [OR=1.50, (0.74-3.04), p=0.261].

There was a three times greater odds of mortality in women who had unknown HIV status [OR=3.11, (1.73-5.60), p<0.001]. Our study showed that there was no significant difference in risk of mortality between adolescent and adult pregnancies [OR=0.92, (0.36-2.35), p=0.864]. Furthermore our study showed that there was a 0.51 lesser odds of mortality in late adolescents compared to early-mid adolescents though this was not significant [OR=0.51, (0.056-4.56), p=0.543]. There was a 25 times greater odds of mortality in pre-term deliveries as compared to term deliveries [OR=25.49, (14.65-44.35), p<0.001].

Our study showed that there was a 3.6 times greater odds of HIV positive women to be unbooked [OR=3.60, (2.62-4.94), p<0.001]. Results from our study further showed that there was a 126 times greater odds of women with unknown HIV status to be unbooked [OR=126.38, (97.60-163.64), p<0.001]. Our study further showed that there was a 0.83 less odds of adults to be unbooked as compared to adolescents [OR=0.83, (0.71-0.98), p=0.026]. Lastly, there was no difference between early-mid adolescents and late adolescents in terms of booking status [OR=1.09, (0.67-1.79), p=0.724].

Discussion

This study showed an unbooked rate of 15% within region F. This proportion of those being unbooked was higher than expected. South African Millennium Development Goals country report showed that more 95% of women attend antenatal care at least once and 73% women attend antenatal care four times or more (9,10), yet in the inner city only 85% of pregnant women attend antenatal care. The reasons for this were not explored in this study, however the inner city is known for its highly migrant population and high number of foreigners. Some studies have suggested that xenophobia (real or perceived) within the health system deters foreigners from attending health care facilities (11).

125

Booking status and birth outcomes as key measures of antenatal care

Our study showed an association between booking status and birth outcomes (infant born alive or dead) which was consistent with evidence from previous studies (6, 12-14).

Our study showed no association between booking status and duration of pregnancy at delivery. Though there have not been many studies showing an association between the two variables it is important to note that women attending antenatal clinics have a greater chance of being assessed for any possible risks that could cause any adverse event during pregnancy which includes pre-term birth. Unbooked women do not have such advantages.

Antenatal booking and birth outcomes between adults and adolescents

Our study showed that there was an association between booking status and age (adults and adolescents) suggesting the differences in antenatal care seeking behaviour between adolescents and adults. This study showed a higher unbooking rate within adolescents as compared to adults which supported previous literature and existing knowledge which has indicated that adolescents perceive health services as non-friendly because of stigma and discrimination (15). Our study showed that adolescents and HIV positive women were more likely to be unbooked. Both being pregnant as an adolescent and being HIV positive are associated with stigma and discrimination and this may contribute to their higher unbooked rate.

As part of recommendations, it is important to encourage antenatal booking and ensure all antenatal services are accessible to all pregnant women. This could be achieved through community involvement and awareness around antenatal services. Research studies that will focus on barriers to accessing antenatal care should be considered in order to provide adequate health services. Continuous education of all women on the importance of antenatal care services and sensitization on antenatal booking would improve antenatal attendance. Government should continue strengthening the implementation of MomConnect at all health facilities (mHealth). It is important to assess level of knowledge around antenatal services and access from adolescents. Government should

126

promote adolescent friendly services while discouraging stigma and discrimination directed to adolescents. More studies around perceptions and birth outcomes in adolescents should be conducted. Maternal HIV testing during both antenatal visits as well as at delivery where the status is unknown (if the mother was unbooked) should be emphasized and facilitated as much as possible. As a component of health systems strengthening within antenatal care, policies and guidelines should be constantly and systematically reviewed and amended using evidence based outputs and accurate data.

Limitations of the studyOwing to data being collected retrospectively from birth registers, poor data quality was the main challenge i.e. incomplete recording of data within the birth register. There was also a possibility that amongst the three facilities that were studied; information collected was not collected in a standardized and systematic manner. From Wits RHI Health Systems Strengthening project, there has been evidence from baseline assessments conducted by Wits RHI that health facilities do not use standardized registers and are collecting different information using different tools. Another limitation was missing information or data from women who delivered at the defined facilities but were not recorded in the birth register. Data was collected at a health care facility level and consequently do not reflect women who delivered at home and do not subsequently attend a health care facility.

Conclusion

Our study of registers of women who delivered in the inner city of Johannesburg found an unbooking prevalence of 15% which is significantly higher than the country average of 5% of pregnant women not attending antenatal care prior to delivery. Our study found that being an adolescent or HIV positive was associated with a lower booking rate. Furthermore, unbooked women had significantly worse birth outcomes than booked women. Our study also showed that the risk of perinatal mortality was associated with an HIV unknown status.

Reference1.Bilano VL, Ota E, Ganchimeg T, Mori R, Souza JP. Risk Factors of Pre-Eclampsia/Eclampsia and Its Adverse Outcomes in Low- and Middle-Income Countries: A WHO Secondary Analysis. PLoS One. 2014;9:e91198.

127

2. Osungbade KO, Ayinde OO. Birth outcomes among booked and unbooked women at a secondary health facility in southwest Nigeria: implications for strengthening perinatal health services. J Child Health Care. 2011;15:320-8.3. Finlayson K, Downe S. Why do women not use antenatal services in low- and middle-income countries? A meta-synthesis of qualitative studies. PLoS Med. 2013;10:e10013734. Wabiri N, Chersich M, Zuma K, et al. Equity in maternal health in South Africa: analysis of health service access and health status in a national household survey. PLoS One. 2013;8:e73864.5. Joshi C, Torvaldsen S, Hodgson R, Hayen A. Factors associated with the use and quality of antenatal care in Nepal: a population-based study using the demographic and health survey data. BMC Pregnancy Childbirth. 2014;14:94.6. Osungbade KO, Ayinde OO. Birth outcomes among booked and unbooked women at a secondary health facility in southwest Nigeria: implications for strengthening perinatal health services. J Child Health Care. 2011;15:320-8.7. Rispel LC, Peltzer K, Phaswana-Mafuya N, Metcalf CA, Treger L. Assessing missed opportunities for the prevention of mother-to-child HIV transmission in an Eastern Cape local service area. S Afr Med J. 2009;99:174-9.8. National Department of Health.The 2012 National Antenatal Sentinel HIV & Herpes Simplex Type-2 Prevalence Survey in South Africa. Pretoria: National Department of Health, 2012:1-88.9. Solarin I, Black V. "They Told Me to Come Back": Women's Antenatal Care Booking Experience in Inner-City Johannesburg. Matern Child Health J. 2013;17:359-6710. Statistics South Africa.Millennium Development Goals country report 2010. Republic of South Africa: Government of South Africa. The Civil Society Organisations (CSOs). United Nations Resident Representative in South Africa, 2010:1-139.11. Steinberg J. Mozambican and Congolese Refugees in South Africa: A mixed reception. Institute for Security Studies. 2008.12. Ekwempu CC. The influence of antenatal care on pregnancy outcome. Trop J Obstet Gynaecol. 1988;1:67-71.13. Smith D. Saving mothers and babies. Nurs Times. 2013;109:27.14. Chopra M, Daviaud E, Pattinson R, Fonn S, Lawn JE. Saving the lives of South Africa's mothers, babies, and children: can the health system deliver? Lancet. 2009;374:835-46.15. Christiansen CS, Gibbs S, Chandra-Mouli V. Preventing early pregnancy and pregnancy-related mortality and morbidity in adolescents in developing countries: the place of interventions in the prepregnancy period. J Pregnancy. 2013.

128

PRACTICE OF OPERATIVE VAGINAL DELIVERY: A QUESTIONNAIRE BASED STUDY

JM Devjee, *J MoodleyDepartment of Obstetrics and Gynaecology, King Dinizulu Central Hospital, *Women’s Health and HIV Research Group, Department of Obstetrics and Gynaecology, University of KwaZulu-Natal, Durban

BackgroundCaesarean sections are increasing while operative vaginal deliveries rates seem to be in decline in South Africa. The actual reasons for the decline in the use of operative vaginal delivery is not known, despite the fact there is evidence to show that both forceps and the ventouse are useful methods to expediate vaginal delivery safely. We therefore embarked on a questionnaire based study to determine the current practice operative of vaginal delivery in public hospitals in the Durban Metropolitan areas. MethodsQuestionnaires were distributed to doctors in the public sector. Doctors working at all levels of care consisted of interns, medical officers, registrars in training, specialists were requested to fill in a questionnaire regarding operative vaginal deliveries. The main questions in the questionnaire included1. Current use, training and supervision in relation to the use of ventouse

and forceps deliveries2. The number of operative deliveries and their views on the use of

operative vaginal deliveries

ResultsTwo hundred and twenty seven questionnaires were distributed doctors of varying experience and responses to questionnaires were received from all the doctors giving a response rate of 100 %. Of the 227 responses, 197 (86.8%) responses were complete and eligible for analysis. The rest were incomplete and excluded from the analysis.

One hundred and sixty two (82.2%) of the respondents indicated that there was a role for operative vaginal delivery. Of these 78 (48%) had reservations of the use of operative vaginal delivery because of the lack of confidence, fear of litigation and fear of complications. All the respondents indicated that they taught the procedure to their colleagues. All respondents learnt operative vaginal delivery via mannequins, advanced midwives, registrars and specialists. Seventeen (8.6%) of the doctors preferred vaginal deliveries (ventouse n=13 and forceps n=4). One hundred and eighty two (92.4%) preferred the use C/S, 15 (7.6%) would perform both operative vaginal delivery and CS. Only 2 (1%) doctors performed operative vaginal delivery in the last 6 months. Thirty one (15.7%) were trained to use “Kiwi cup”.

DiscussionThe main finding in this questionnaire based study was that doctors were reluctant to perform operative vaginal delivery as a result the frequency of the use of operative vaginal delivery (forceps and vacuum) was low. Pattinson et al similarly showed that the use operative vaginal delivery were only 0.5% in Pretoria, South Africa

Conclusion

129

The practice of operative vaginal delivery is low in the Durban, Metropolitan area.

130

AN AUDIT OF VACUUM DELIVERIES AT ZITHULELE HOSPITAL

Dr C. Benjamin Gaunt Zithulele Hospital, Eastern Cape, South Africa; ben@zithulele.org

Abstract“There is a skill about vacuum extraction which tests the mind and hands of the operator in a finer way than do forceps deliveries – for this is a gentleman’s way of delivering a baby…” (Turnbull’s Obstetrics, 1989)

An audit of all the vacuum deliveries (including failed attempts that ended in Caesarean section) performed at Zithulele Hospital during the period November 2012 to December 2014 was performed as a quality improvement exercise. A total of 319 successful and 21 failed vacuums were reviewed to assess neonatal outcomes as well as maternal complications. The vacuum delivery rate was 7.4% and outcomes within the rather wide ranges reported in other audits internationally. Five minute Apgar scores were ≥8 in 91% of babies, but significantly worse in failed vacuums. There were four early neonatal deaths. The primary maternal complication was perineal injury, with 11.3% experiencing third or fourth degree tears, with incidence varying between practitioners. Although subject to limitations, such as accuracy of data recording, this audit is the first of its kind at our hospital and provides a useful baseline as well as helpful feedback to individual practitioners. It may also provide a helpful comparison for other institutions working in a similar context.

IntroductionZithulele Hospital is a deeply rural district hospital in the Eastern Cape province of South Africa. Over the past decade it has seen a substantial increase in the number of deliveries performed (up 275% from 745 in 2005 to 2’087 in 2014) and a corresponding drop in perinatal mortality rate (from 44.2 in 2005 to 19.4 per 1000 >1000g in 2014). Obstetric care is provided by midwives (only two with an advanced qualification) and generalist hospital doctors, ranging from community service doctors to

more experienced non-specialists, two of whom have diplomas in obstetrics. One feature of the obstetric service is that vacuum assisted deliveries are still a strong component and all new staff are instructed how to perform them. We felt it appropriate to audit all vacuum deliveries to ensure that our outcomes are consistent with other centres and to find points we need to pay careful attention to.

2005/6

2006/7

2007/8

2008/9

2009/10

2010/11

2011/12

2012/13

2013/14

2014/15

7271004 1001

1166 11831344

1778 1871 19232087

Total deliveries

Figure 1: Total deliveries over time

2006/7

2007/8

2008/9

2009/10

2010/11

2011/12

2012/13

2013/14

2014/15*

44.235.4

28.5 27.4 26.921 21.6 24.6 22.2

Perinatal mortality rate (>1000g)

Figure 2: Perinatal mortality rate over time

MethodThe audit was conducted as a retrospective review of the delivery register where all deliveries are recorded by the midwives. We wanted a cohort of about 300 deliveries to audit, so chose the period 1 November 2012 to 31 December 2014, providing a cohort of 319 for analysis. This also allowed a wider spread of doctors, as there is usually quite significant staff turnover in January each year, while still keeping it current. Details for every vacuum assisted delivery – including attempts which failed – were

132

captured onto an Excel spreadsheet and then analysed. Variables assessed included maternal parity, weight of baby, Apgar scores at 1 and 5 minutes, perinatal deaths, tears and any major complications. Cases in which the 5 minute Apgar score was less than 5, which included all the cases where the baby died, had their file drawn for a more detailed analysis and deeper learning. Outcomes were also analysed in order to detect variations between attending practitioners.

ResultsGeneralA total of 319 successful vacuum assisted deliveries were analysed. This represented 7.4% of the 4292 deliveries done during this period. The caesarean section rate for the corresponding period was 17.8% Forceps are not used at Zithulele. A total of 20 different generalist medical practitioners were involved in the deliveries and six midwives performed at least one vacuum delivery.

317874%

3197%

76318%

311%

NVD Vacuums Caesarean section Vaginal breech

Figure 3 – Mode of delivery

The majority of vacuums were performed in nulliparous women: 71.8%. Another 12.9% were having their second child and the remaining 15% were para 3 or above.

Guidelines indicate that the fetus should not be premature if vacuum is to be attempted. The generally accepted weight cut off for this is 2500g. A

133

total of 18 babies (5.6%) were <2500g. This was presumably due to misjudgement of the size of the baby (half were ≥2200g) but this aspect was not analysed further, although it provides an important learning point.

The indication for vacuum was not well recorded in the delivery register (only ¼ of the time), but of those, recorded indications were for prolonged second stage, fetal distress or poor maternal effort in a 6:1:1 ratio.

Fetal outcomesInitial Apgar scores were ≥8 in 65.1%, 5-7 in 24.1% and <5 in 10.5% of successful deliveries, but this improved markedly to 90.8% ≥8 by 5 minutes, while 7.6% were 5-7 and only 1.3% <5. Of the four cases who had 5-minute Apgars of <5, three resulted in ENNDs. (Three babies who were IUDs but delivered by vacuum were excluded from the Apgar analysis.)

The spread of Apgar scores at 1 and 5 minutes differed between practitioners, but all except one had an 80% rate of Apgars ≥8 by 5 minutes. (There were no 5-minute Apgars <5 for this doctor, however.) Further details of this sub-analysis are available but not included here due to space.

8-10 5-7 <50

50

100

150

200

250208, 65%

76, 24%33, 10.5%

Apgar score at 1 min

Figure 4: Apgar scores at 1 minute

134

8-10 5-7 <50

50100150200250300350

286, 90.8%

24, 7.6% 4, 1.3%

Apgar score at 5 min

Figure 5: Apgar score at 5 minutesComplicationsPerineal injury was the most commonly noted complication. Episiotomy was done in 25.4% of cases. The practice varied significantly between different practitioners. A total of 19.8% of women did not sustain a tear, but 18.8% had first degree tears, 24.5% second degree and 11.3% third or fourth degree tears.

No tear Epis 1st 2nd 3rd0

102030405060708090

Perineal injury

Figure 6: Incidence of perineal tears

Although this was an audit and not a study designed to assess the possible protective effect of episiotomy in vacuum deliveries, it is interesting to note that neither the incidence of third degree tear nor the absence of perineal injury appeared related to the rate of episiotomy use by individual practitioner.

The next most common complication was failed vacuum. 21 failed attempts at vacuum delivery which proceeded to caesarean section were analysed. Apgars at 1 min were 8-10 in 55% of cases, 5-7 in 30% and <5

135

in 15%. At 5 min they were 75%, 15% and 10% respectively. In two of these cases the 5 minute Apgar was <5 and one of these babies died. As expected, these outcomes are significantly worse than those described for successful vacuums.

One of the cases recorded as failed vacuum suffered a ruptured uterus but case review suggested this diagnosis was missed and occurred prior to the vacuum attempt. It was therefore not regarded as a complication and was excluded from the analysis of the Apgar scores.

There were 25 caesarean sections recorded as for delayed second stage where vacuum delivery was not attempted. The outcomes were universally good: only 3 had Apgar scores between 5 and 7 at 1 minute, and all were above 8 by 5 minutes.

9 other significant complications were recorded in the register as having occurred in successful vacuum deliveries: 5 PPH, 3 shoulder dystocia and 1 cephalohaematoma. All had apparently good outcomes.

DiscussionAlthough vacuums are by definition high risk deliveries, the overall outcomes of this cohort of patients, delivered by mostly junior generalists in a deeply rural district hospital, are reasonably good and within the range of outcomes reported from elsewhere, including developed countries.

The rate of vacuum assisted deliveries is within ranges reported from elsewhere, but is one of the highest rates in South Africa. Comparison with the 2012/2013 Saving Babies report suggests that 7.7% of all vacuums performed at district hospital in South Africa are done at Zithulele, but only 0.6% of the deliveries.

136

With respect to neonatal outcomes, there were no stillbirths. The neonatal mortality rate, NMR, (including failed vacuums but excluding known IUDs) was 11.9 per 1000. As expected this is a bit higher than the overall NMR of 9.3 per 1000 for all babies >1000g in our unit, but significantly more than the 4.3 per 1000 for babies over 2500g in the same period. The NMR is higher than we’d like, but in our setting where relatively junior staff manage challenging labours, we propose that our vacuum rate is a good thing as it saves caesarean sections and their concomitant risks, while maximising fetal outcomes. The large number of practitioners involved – most of whom had performed few or no vacuum deliveries prior to working at Zithulele – shows that skills can be passed on between generalist doctors even in a rural environment. Because this was a register review it is possible that other neonatal complications, such as conjunctival haemorrhage were under diagnosed and or under reported.

Maternal outcomes were good, with generally few recorded complications. There is a possibility that this may partly be due to recording omissions. The main complication was maternal perineal injury. There was a relatively high rate of third degree tears (although comparable to rates reported in a number of other audits). There was a trend towards more failed vacuums in the latter part of the period, which requires careful attention.

Study limitations include data quality with the possibility that not all complications, especially, were recorded. The study was retrospective, but appropriate as each case was managed on its merits to the best judgement of the doctor involved. This therefore reflects a real-world environment.

137

SEVERE MATERNAL OUTCOME FROM CAESAREAN SECTION RELATED HAEMORRHAGE IN SOUTHERN GAUTENG, SOUTH AFRICA

Maswime TS, Buchmann EJDepartment of Obstetrics and Gynaecology, University of the Witwatersrand, South Africa

Caesarean section has become a relatively safe operation in highly resourced countries (1), and is now the most widely performed surgical operation in certain countries (2). In South Africa, the safety of caesarean sections has recently come into question. Maternal deaths resulting from bleeding during and after caesarean section have increased dramatically since 2011, according to the National Confidential Enquiries into Maternal Deaths (3). A near-miss is defined as a women who experienced and survived a severe health condition during pregnancy, childbirth or postpartum. Severe maternal outcome, is the number of near-misses and maternal deaths in a population, up to 42 days post-delivery (4).

Objectives1. To determine the risk factors for caesarean section related

haemorrhage.2. To determine causes of caesarean section related haemorrhage.3. To determine interventions used to reduce morbidity and mortality

from haemorrhage, during and after caesarean section.

MethodsThis is a cross-sectional prospective study that was conducted in 13 public hospitals in Southern Gauteng province from July to December 2014. The data were collected by the researcher using ongoing surveillance. Ethics approval and permission was obtained from the University of the Witwatersrand, and from the Gauteng department of health to conduct the study. Inclusion criteria: any woman with a gestational age ≥24 weeks, or who has delivered a 500 g or more baby, undergoing caesarean section with a

combined blood loss of ≥1000 mL, and at least one of the following: blood transfusion ≥3 units, a caesarean section hysterectomy, a repeat laparotomy, transfer to a higher level of care, ventilation, the use of inotropic drugs, acute dialysis, cardio-pulmonary resuscitation, admission to an intensive care unit, and death as a result of excessive haemorrhage. Women with a, ruptured uterus, extra-uterine pregnancy, and where another cause of haemorrhage was the primary reason for bleeding, were excluded.

ResultsA total of 100 women were included in the study, 93 near-misses and 7 maternal deaths. The number of deliveries was 46 775, and the number of live births was 20527 in the 6 months. The caesarean section rate was 43%, the near-miss ratio was 2.1/1000 live births, the mortality ratio was 16.1/100 000 live births from caesarean section related haemorrhage. The mortality index was 7%. The mean age was 29 (SD 6.0) for the near-misses and 30 (SD 5.3) for the maternal deaths. The median parity was 1, for near-misses and 2 for the maternal deaths. Majority of the women were booked. Previous caesarean section was common, amongst near-misses, refer table 1.

In the near-miss audit 76 (81%) caesarean sections were assessed as difficult, by the operating surgeon, and 3 (42%) in the maternal death group. The near-miss group had more interventions done at caesarean section, compared to the maternal death group, refer to the table 2.

The maternal death group had more organ supporting interventions, and the near-miss group had more interventions to arrest the bleeding. Massive transfusion was similar in both groups, refer table 3.

Placental abnormalities accounted for almost a third of the women. Previous caesarean section was reported in almost half of the near-miss group, Anaemia and thrombocytopaenia prior to caesarean section were

139

also identified, even though they were not routinely tested, refer table 6.

Atonic uterus (33%) was the most common primary cause of bleeding, followed by caesarean section trauma (29%) in the near-miss group. Placental site bleeding was the primary cause in 14 (15%) women. Three (42%) of the women in the maternal death group died without the cause of haemorrhage being established, two (29%) were caused by atonic uterus, one (14%) by trauma and one (14%) by placental site bleeding.

DiscussionThe national caesarean section rate for 2012 was 23.2%, according to the health systems trust (5). The caesareans section rate in this study was higher (43%). The maternal death: near-miss ratio was higher than expected. The near-miss ratio in low income countries is estimated to be 1:5 in low income countries by WHO (6), however the ratio was 1:13 in this study. The mortality index for obstetric haemorrhage in a South African study that was published in 1998 was 4, in a study that was published in 1998 (7), the mortality index for caesarean section related haemorrhage in this study was 7%.

ConclusionPrevious caesarean section and placental abnormalities are an important risk factor for caesarean section related haemorrhage. Atonic uterus and trauma at caesarean section are the main primary causes for haemorrhage. Maternal deaths from caesarean section related haemorrhage are associated with fewer interventions to arrest haemorrhage.

Table 1: Pre- operative factorsNear-misses % Deaths %

Booked 83 89 6 85Previous CS 44

(Exclude Para 4760

2 28

140

0)HIV Positive 36 38 3 42

141

Table 2: Interventions at caesarean sectionNear-miss (n=93)

% Death(n=7)

%

Blood transfusion 53 57 2 28Ergometrine 19 20 0 -B-lynch 15 16 0 -Hysterectomy 20 21 1 14Condom/balloon 4 4 0 -

Table 3: Near-miss defining criteriaNear-miss (n=93)

% Death(n=7)

%

Relook laparotomy 43 46 3 42Dialysis 6 6 0 -Ventilation 39 41 5 71Inotropes 11 11 6 85ICU 30 32 4 57

Massive transfusion 1st 24hrs 64 68 5 71

CPR 0 - 4 57Transfer after Cs 10 10 1 14Hysterectomy 38 40 2 28

Table 4: Risk factors for caesarean section related haemorrhageNear-missesn=93

% Deathsn=7

%

Abruption 18 19 1 14Praevia 6 6 0 -Accreta 7 7 1 14Praevia accreta 4 4 0 -Previous caesarean 44 47 2 28Anaemia (Hb<10) 23/60Thrombocytopaenia 10/34

References1. Moran N. Avoiding unnecessary caesarean sections: a key strategy to reducing obstetric hemorrhage deaths. O&G Forum 2013; 23: 21-26.2. Fawcus S, Moodley J. Alert for maternal deaths due to obstetric hemorrhage. O&G Forum 2013; 23: 33-34.

142

3. Pattinson R, Fawcus S, Moodley J. Tenth interim report on confidential enquiries into maternal deaths in south Africa 2011 and 2012. Pretoria: Department of Health, 2013. 4. Say L, Souza J, Pattinson R. Maternal near miss – towards a standard tool for monitoring quality of maternal health care. Best Pract Res Clin Obstet Gynaecol 2009; 23: 287–296.5. Health systems trust. Caesarean section rate http://indicators.hst.org.za/healthstats/76/data (accessed 25/05/15)6. Mustafa R, Hashmi H. Near-miss obstetrical events and maternal deaths. J. Coll. Physicians Surg. Pak. 2009; 19(12): 781-7857. Mantel GD, Buchmann EJ, Rees J etal. Severe acute maternal morbidity: a pilot study of a definition of a near-miss. BJOG 1998; 105: 985-990

Acknowledgement: Carnegie Corporation of New York

143

EVALUATION OF HEALTHCARE SYSTEMS IN THE DECISION TO INCISION INTERVAL FOR CAESAREAN SECTIONS AT CHBAH

Williams M, Adam Y, Maswime SChris Hani Baragwanath Academic Hospital, Department of Obstetrics and Gynaecology

BackgroundThe recommended Decision-to-incision interval for caesarean sections is 30 minutes according to the ACOG. The recommendation from the Confidential Enquiries to maternal deaths is a decision-to-incision interval of 1 hour. The decision to incision time at CHBAH is not known.

Objectives1. To determine the median decision to incision interval for caesarean

sections from May 2014 to November 2014.2. To review logistical factors that affect the number of caesarean

sections that are performed.

MethodsThis is a cross-sectional retrospective study at CHBAH, using caesarean section booking lists, the caesarean section register, and an electronic database of registrar reports.

ResultsThe decision-to-incision interval at CHBAH is longer than the recommended interval. The average number of caesarean sections done at CHBAH, is 1 per hour in two Obstetric theatres. The median number of elective caesarean sections handed over at the start of a call, is 6 (IQR-2.5-11) with a range of 0 - 38, and the median number of emergency caesarean sections handed over is 8 (IQR-4-11) with a range of 0 – 27. Theatre delays were reported for 54% of the calls. The main reason for theatre delays is complicated obstetric cases (23.12%). Other reasons for delays are Nursing shortages, and equipment shortages.

ConclusionThe decision to incision interval is affected by many factors, but the most common factor at CHBAH is complicated obstetric cases.

SESSION 5: PAPER 6

NEONATAL DEATHS IN CHILDREN’S WARDS: THREE TRIENNIA OF CHILD PIP DATA 2005-2013Dr Cindy Stephen, Red Cross Children’s Hospital, Western Cape (cindy.stephen@uct.ac.za), Dr ME Patrick, Grey’s Hospital, Kwazulu-Natal IntroductionThe Child Healthcare Problem Identification Programme (Child PIP), now well established and widely used as a mortality review tool in South African hospitals, offers a unique view of neonatal mortality that is not otherwise systematically captured. Child PIP uses the hospital-based mortality review process to assess the health profile of and quality of care received by children in the South African health system, including neonates (aged 0-28 days) admitted to children’s wards rather than to nurseries. This population of neonates is reported on in this presentation, comparing overall data from three triennia, 2005-2007, 2008-2010 and 2011-2013, as well as describing early and late neonatal death data. The latter two triennial reporting periods describe the data presented in the Triennial Ministerial Report s. MethodsIn the triennial period ending 2013 there were 198 hospitals using Child PIP and contributing data to the national database, compared with 72 at the end of the first triennial period. This covered 49 of the 52 South African health districts, an increase of 15 districts from 2007.ResultsNeonatal deaths in children’s wards (Table 1)During the three triennial periods there were 77 744 admissions of neonates to children’s wards, which consistently accounted for 7% of all admissions. There were 3 111 audited neonatal deaths, which increased from 5% in the first period to 9% of all deaths in the third period. The in-hospital mortality rate for neonates was substantially higher in all three triennial periods than for children, notwithstanding the reductions observed. There was only a 26% reduction in the IHMR for neonates from the first to the third period, as opposed to the impressive 52% for children under five years of age during the same period.Table 1: Core Data

Period 1st 2nd 3rd Hospitals 72 121 177Districts 34 40 49

All Admissions 129564 418189 626505All Deaths (Tallied) 7329 17880 16299

All In-Hospital Mortality Rate/100 admissions

5.7 4.4 2.6Neonatal Admissions 8784 27066 41894

Neonatal Deaths (Tallied) 437 1447 1542Neonatal In-Hospital Mortality Rate 5.0 5.4 3.7

All Deaths (Audited) 8212 17996 16776Neonatal Deaths (Audited) 358 1221 1532

Early Neonatal Deaths 1.5% 3.0% 3.8%Late Neonatal Deaths 2.8% 3.8% 5.4%

Under-5 Deaths 62.7% 61.8% 59.9%Neonatal Deaths/All deaths (Tallied) 6.0% 8.1% 9.5%Neonatal Admissions/All admissions 6.8% 6.5% 6.7%

Neonatal Deaths/All deaths (Audited) 4.4% 6.8% 9.1%

Profile of the neonates who die in children’s wards (Table 2)Almost 60% of neonates under 7 days of age died within 24 hours of arrival in hospital, compared to 40% of late neonates and 32% of children under 5 years. Of concern is that, despite significant overall improvements in the HIV Prevention of Mother to Child Transmission (PMTCT) programme, the HIV status of virtually half the neonates who died under 7 days of age was unknown, reflecting either PMTCT programme gaps or incomplete history taking. Septicaemia was recorded as a cause of death in almost 40% of late neonates whereas other respiratory diagnoses were the top cause of death in early neonates. Overall, septicaemia accounted for substantially more deaths in neonates (34%) than in children (25%).For two-thirds of the neonates who died, the death was regarded as either avoidable, or there was uncertainty about its avoidability.

Table 2: Profile comparing early and late neonatal deaths in children's wards

Period 2005-2013 Early Neonatal Deaths

Late Neonatal Deaths

Under-5 Length of Stay Numbe

rPerce

ntNumbe

rPerce

ntPercen

tDeath within 24 hours 746 57.8 722 39.6 31.9HIV

Exposed 348 27.0 795 43.7 25.0Infected 40 3.1 55 3.0 21.0

Unknown 630 48.8 616 33.8 30.6PMTCT

Mother negative at 363 28.1 462 25.4 22.3PMTCT guideline not 244 10.8 613 8.0 10.9

Unknown 544 42.2 601 33.0 42.8Burden of disease

Septicaemia 335 26.0 714 39.2 24.9Pneumonia 216 16.7 454 24.9 30.9

Acute diarrhoea (121) (9.4) 351 19.3 26.7Other Respiratory

Diagnosis470 36.4 (177) (9.7) (4.7)

Was death avoidable?Yes 261 20.2 475 26.1 29.5

Not Sure 621 48.1 705 38.7 36.3Total (Yes and unsure) 882 68.4 1180 64.8 65.9

ConclusionThe overall reduction in the neonatal in-hospital mortality between the three reporting periods is pleasing.

146

However the vulnerability of the neonatal population is again emphasised by:

The in-hospital mortality rate for neonates being higher than that for children, and reducing at a far slower rate

Neonates being more likely to die within 24 hours of arrival in hospital than children, especially early neonates

The HIV pandemic remaining significantly present with HIV testing gaps observed in the PMTCT programme

Sepsis being a major contributor to neonatal mortality particularly for the late neonatal group

Simple strategies built around an understanding of the vulnerability and special needs of the neonatal population may help to improve quality of care for neonates in South African hospitals, and thereby reduce neonatal in-hospital mortality.

147

SAVING MOTHERS 2011-2013: SIXTH REPORT ON CONFIDENTIAL ENQUIRIES INTO MATERNAL DEATHS. OVERVIEW

NCCEMD

MethodThe report covers the maternal deaths that were reported to the NCCEMD secretariat by 15th May 2014, and that occurred in 2011-2013. The same definitions used in previous Saving Mothers reports were used in this report.

ResultsData was entered on 4452 deaths in pregnancy and the puerperium for the period 2011-2013. The institutional maternal mortality ratio (iMMR) has decreased from 176.22/100000 live births in 2008-2010 to 154.06/100000 live births in 2011-2013. The iMMR decreased in district and tertiary hospitals but there was a slight increase in regional hospitals.In 2011-2013, the “big 5” causes of maternal deaths were non-pregnancy related infections (NPRI) (34.7%, mainly deaths due to HIV infection complicated by Tuberculosis (TB), PCP and pneumonia), obstetric haemorrhage (15.8%), complications of hypertension in pregnancy (14.8%), medical and surgical disorders (11.4%) and pregnancy related sepsis (9.5%, includes septic miscarriage and puerperal sepsis) . These five account for 86.2% of maternal deaths. Bleeding at or after caesarean section was responsible for a third of obstetric haemorrhage deaths. TB was the most common cause of deaths due non-pregnancy related infections and was probably underdiagnosed in a number of other women. The HIV status was known for 87% of women who died; 65% of were HIV positive, a small decrease from 70% in 2008-2010. Almost 90% of women who died from NPRI were HIV positive. Of these, 55% were on HAART, compared to 36% in 2008-2010. It is not known how many women were not virologically suppressed on HAART. This may be due to recent initiation of HAART, adherence problems, or virological failure. TB remains the single most common cause of mortality amongst HIV positive women, and the most common respiratory cause. There has been a significant reduction in deaths due to complications of antiretroviral therapy. This has followed the change in ART guidelines that nevirapine should not be routinely prescribed for women of reproductive age, including pregnant women, and efavirenz used instead.The iMMR of deaths due to complications of hypertension in pregnancy have declined 18% from 2002-2004 till 2011-2013, but deaths due to obstetric haemorrhage have increased 25% from 2002-2004 till 2011-2013. The iMMR causally related to mode of delivery was three times higher for operative delivery; 66.6 per 100000 live births for vaginal birth and 185.8 per 100000 live births for caesarean section. Poor clinical assessment, delays in referral, not following standard protocols and not responding to abnormalities in monitoring of patients were the most common health care provider avoidable factors. Lack of appropriately trained doctors and nurses has emerged as a significant contributory factor in maternal deaths being recorded in 15.6% and 8.8% for doctors and nurses respectively.

ConclusionsThere has been a significant reduction in maternal deaths in the 2011-2013 triennium and this reduction is mostly due to a decrease in deaths due to NPRI; however to maintain this fall and obtain an exponential fall much more still needs to be done. Assessors classified 60% of maternal deaths to be possibly or

probably preventable indicating mostly poor quality of care during the antenatal, intrapartum and postnatal periods.

149

DEATHS DUE TO ECTOPIC PREGNANCY AND MISCARRIAGE IN SOUTH AFRICA: FINDINGS AND RECOMMENDATIONS FROM SAVING MOTHERS 2011-13

N.Moran 1 ,2, J.Kwet1,3 (1National Committee for Confidential Enquiries into Maternal Deaths; 2Department of Health, KZN; 3Department of Health, North-West Province)

IntroductionOn a triennial basis, the National Committee for Confidential Enquiries into Maternal Deaths in South Africa produces a Saving Mothers report, which provides data on maternal deaths and gives recommendations on how the number of maternal deaths can be reduced. The latest report (2011-13) is due for release in 2015. This paper presents the key findings and recommendations from the chapter on early pregnancy deaths (ectopic pregnancies and miscarriages). For complete data and discussion, readers are referred to the Comprehensive Saving Mothers Report 2011-13.

MethodsStatistics were obtained from the National MaMMAS database, which contains details of all notified cases of maternal death. In addition individual case notes of maternal deaths assessed as being due to either miscarriage or ectopic pregnancy were scrutinised to identify important recurrent issues which need to be addressed if these deaths are to be prevented in the future. One such case example is presented below.

ResultsThere were 287 early pregnancy deaths in 2011-2013, an 10% increase compared to 2008-2010. This increase is greater than the overall increase in maternal deaths (23%) compared to 2005-7. One hundred and two (102) deaths were caused by ectopic pregnancy (36% increase), and 185 by complications of miscarriage, which was about the same number as in 2008-10 (186). Sixty-five percent (65%) of ectopic pregnancy deaths and

150

35% of miscarriage deaths were assessed as being clearly avoidable within the health system.

151

Ectopic pregnancyDeaths from ectopic pregnancy occurred most frequently at regional hospitals (47%), followed by district hospitals (41%) and tertiary hospitals (12%). Eight (8%) were classified as extrauterine pregnancy beyond 20 weeks’ gestation. Although the majority of women who died from ectopic pregnancy had unknown HIV status (56%), of those who were tested, 71% were HIV infected. The final cause of death was hypovolaemic shock in 76%. The most frequent patient behaviour-related avoidable factors were lack of antenatal care and delay in accessing medical help. The most frequent administrative avoidable factor was lack of appropriately trained doctors. The most frequent health care provider-related avoidable factors were failure to make the diagnosis, and therefore incorrect management, substandard care despite making the right diagnosis, and substandard resuscitation of hypovolaemic shock.

MiscarriageKZN was the only Province where miscarriage featured in the top five causes of maternal death. Deaths from miscarriage occurred most frequently at regional hospitals (52%), followed by district hospitals (30%) and tertiary hospitals (14%). Of the subcategories of miscarriage, 62% were classified as septic miscarriage, 26% as deaths from haemorrhage (non-traumatic), 5% followed legal termination of pregnancy, 4% were classified as gestational trophoblastic disease, and 3% as miscarriage deaths related to uterine trauma. Although 41% of women who died from miscarriage had unknown HIV status, of those who were tested, 81% were HIV infected. The final cause of death was septic shock in 55% and hypovolaemic shock in 30%. The most frequent patient behaviour-related avoidable factors were delay in accessing medical help, no antenatal care, and unsafe (illegal) termination of pregnancy. The most frequent administrative avoidable factors were Lack of appropriately trained doctors and delay in initiating critical care due to an overburdened service. The most frequent health care provider-related avoidable factors were substandard care despite making the right diagnosis, failure to make

152

the diagnosis or recognise the severity of the condition, and substandard resuscitation of circulatory shock, whether due to sepsis or hypovolaemia.

Case exampleA 29 year-old gravida 2 para 1 booked early for antenatal care at 10 weeks’ gestation. She was a known HIV positive woman who had started on highly active antiretroviral therapy the year before when her CD4 count had been 188. She attended antenatal care on three occasions. There was no screening for TB done, and no new CD4 count check. At 20 weeks’ gestation she presented with shortness of breath to the casualty at a district hospital. Her BP was 96/52, her pulse rate 126 and her temperature 37,6oC. The medical officer (MO) on duty made a diagnosis of respiratory tract infection and admitted the patient to the medical ward on antibiotics, without evaluating the pregnancy. In the ward the nurses heard from the patient that she had ruptured her membranes while in casualty. She started bleeding per vagina. Her condition was noted to be “very weak and dull”. This was reported by the nurses to the on-call doctor in the evening. He did not come to see the patient, but advised telephonically as follows: “Let her abort. I will review her tomorrow”. She died undelivered the same night without further resuscitation, or recording of vital signs. The doctor came to certify the death.

Points This case was assessed as a miscarriage death, but it was not

clear to what extent the miscarriage was responsible for the death as opposed to an underlying respiratory illness

There were missed opportunities during antenatal care to evaluate the patient’s HIV clinical stage and CD4 count, and to screen for TB. Taking these opportunities might have led to earlier detection of the respiratory problem, and a better outcome.

153

The care by the doctor(s) at the hospital goes beyond being incompetent. It was unprofessional and negligent, and clearly showed a lack of commitment to saving the woman’s life.

This highlights the need for a prompt meeting at the institutional level to investigate each maternal death. Such meetings should be attended by the relevant managers so that appropriate corrective action can to be taken in response to the findings.

Although disciplinary action is necessary in response to unprofessional conduct, this cannot be the primary solution to improving the commitment to quality care amongst the health care workers in an institution. What is required is for strong clinical leadership to set the example of what is expected.

Recurrent themes from analysis of individual cases

Themes relevant to both ectopic pregnancy and miscarriage deaths Need for improved family planning / contraception services Need to maintain the intensity in the fight against HIV/AIDS Need for community education about booking early in pregnancy,

reporting early to hospital when early pregnancy complications arise, and about how to access TOP services

Need for improved resuscitative management of the shocked gynae patient, irrespective of the type of shock

Need for shocked gynae patients to be prioritised in casualty departments. There must be a shared responsibility of care between casualty staff and the gynae team

Need for facility managers to ensure that doctors and nurses are aware of their professional and ethical responsibilities on duty and to hold them accountable when they neglect these responsibilities

Themes specific to ectopic pregnancy deaths

154

Need for training on recognition of ectopic pregnancy, and management, particularly the need for prompt surgery if the patient is shocked

Themes specific to miscarriage deaths Need for access to safe legal termination of pregnancy (TOP) for all

women Need for training at all levels on recognition and management of

different types of miscarriage

ConclusionSince the previous triennium, there has been an increase in maternal deaths related to ectopic pregnancy and no change in the number of deaths due to complications of miscarriage. Many of these women bled to death in health institutions due to sub-standard management by the health care team. Unsafe and illegal terminations of pregnancy remain a major contributor to miscarriage deaths. HIV infection is an important underlying risk factor for both categories of early pregnancy deaths. In order to reduce deaths from these causes, more intensive and committed management of these patients by health care workers is needed when they present to health facilities. In addition, new approaches to giving the community access to family planning and safe TOP services will be required.

Recommendations The following are areas where intervention is needed if early pregnancy deaths are to be reduced: 1) promoting and increasing access to family planning services 2) preventing and managing HIV infection 3) educating communities about early pregnancy booking, recognising and acting on danger signs in early pregnancy, and how to access safe TOP 4) training in the recognition and management of circulatory shock 5) prioritising the care of shocked gynaecological patients in casualty units 6) training on the recognition of ectopic pregnancy and its management 7) training on

155

the recognition and management of different types of miscarriage 8) ensuring that all hospitals can offer at least medical TOP services 9) ensuring that all doctors and nurses are aware of their professional and ethical responsibilities when on duty, and holding them accountable when these responsibilities are neglected.

156

AUDIT OF MATERNAL DEATH AT NATALSPRUIT HOSPITAL

Dr B Uzabakiriho, Maswime SDepartment of Obstetrics and Gynecology; Natalspruit Hospital

BackgroundThe lifetime risk of a woman dying from a pregnancy related cause in a developing country is 25 times higher than in a developed country. Ekurhuleni district was one of 25 poorly performing districts in South Africa.

Objectives 1) To determine the causes of maternal deaths at Natalspruit hospital; and the characteristics of women who die2) To describe factors related to the timing of death.

MethodsThis is a retrospective descriptive audit of all maternal deaths at Natalspruit from January 2013 to December 2014, using a period sample.

ResultsThere were 79 maternal deaths in a period of study. During that period, the number of live births was 20676. The institutional maternal mortality rate is 382.08/100000. Majority of maternal deaths occurred after hours and during weekends. HIV positive women were 44 (55.6%). Mothers who died of post partum haemorrhage were 10 (12.6%); hypertension related cases were 12 (15.1%); ectopic pregnancies and miscarriage related cases were 8 (10.1%). Nineteen (24%) women demised in ICU.

ConclusionMaternal mortality rate is considerably high at Natalspruit hospital. Majority of the women who died were HIV positive. More women die at night and during weekends.

157

REDUCTION IN INDIRECT OBSTETRIC DEATHS WITH INCREASED HIV TREATMENT IN SOUTH AFRICA: A 5-YEAR AUDIT

Dr Vivian Black, Dr Andrew Black, Prof. Helen Rees, Prof. Franco Guidozzi, Dr Fiona Scorgie (PhD), Prof. Matthew F. Chersich (PhD)

OBJECTIVE: Review facility-based maternal deaths at a tertiary-level centre in Johannesburg, South Africa over 5-years, and document coverage of HIV services and effects of antiretroviral treatment on maternal mortality.

METHODS: Patient files, birth registers and death certificates were reviewed. Cause of death was assigned independently, by two reviewers. Maternal mortality ratios (MMR) were disaggregated by HIV status and compared to 2003-2007 figures.

RESULTS: From 2008-2012, 124 deaths occurred, with 49,296 births (facility-based MMR=251.5/100,000 live births; 95% confidence interval=209.3-299.8). Overall MMR and that among HIV-negative women were similar to 2003-2007 levels. Since 2008, MMR from direct obstetric causes rose from 111.7 to 162.3/100,000. Conversely, MMR in HIV-infected women reduced compared to 2003-2007 (456.5/100,000 versus 775.8/100,000; P=0.004), and declined annually between 2008 and 2012. Overall, population attributable risk of HIV is 30.3%. HIV testing reached 94.1% in 2012 (11,937/12,691), and numbers receiving antiretroviral treatment rose several fold. Of maternal deaths, 36.4% had not attended antenatal care (40/110). TB remains the commonest cause of indirect deaths, but TB-related deaths reduced 2.96-fold compared to 2003-2007 (P=0.003). Death from pneumonia and HIV-related malignancy declined in 2008-2012, while hemorrhage mortality rose. Direct obstetric causes constituted 64.5% of deaths in 2008-2012, but only 38.7% in 2003-2007 (P<0.001).

CONCLUSION: Higher coverage of antiretroviral treatment likely accounts for reduced MMR among HIV-infected women. Increases in direct obstetric deaths are concerning. Thus, in addition to addressing the remaining deaths attributable to HIV, lowering overall MMR requires focused efforts to increase antenatal care coverage, and avert direct obstetric deaths, especially from haemorrhage.

158

MATERNAL NEAR MISS AND MATERNAL DEATH IN THE PRETORIA ACADEMIC COMPLEX – A POPULATION-BASED STUDY.

P Soma-Pillay, RC Pattinson, L Langa-Mlambo, BSS Nkosi, AP Macdonald

IntroductionThere were 4 452 maternal deaths in South Africa for the period 2011-2013.1 There has been a decrease in the institutional maternal mortality ratio (iMMR) in South Africa for the period 2011-2013 compared to the 2008- 2010 triennium, however further work needs to be done to meet the fifth Millennium Development Goal. In order to reduce maternal mortality, it is important to understand the process of obstetric care, identify weaknesses within the system and finally implement interventions for improving care.2

A woman who experiences and survives a severe health condition during pregnancy, childbirth or postpartum is classified as a maternal near miss.3

By studying cases of maternal deaths and near misses important information can be obtained about the processes that take place within health care systems responsible for the care of pregnant women. Near miss cases share many pathological and circumstantial characteristics with maternal deaths, however near miss cases have the advantage of providing additional information about obstacles that have to be overcome after the onset of an acute complication. Although a maternal near miss case can only be identified retrospectively, it is clinically useful to prospectively identify women with potentially life-threatening conditions. A woman who develops a life-threatening condition will either become a maternal near miss case or maternal death. The purpose of this study was to determine the spectrum of severe maternal morbidity and mortality in the Pretoria Academic Complex and compare the data with previous surveys and the World Health Organisation (WHO) Multicountry Survey on Maternal and Newborn Health.4

Methods

159

This was a descriptive population-based study which took place from 1 August 2013 to 31 July 2014 at 9 delivery facilities in central, south-western and eastern Tshwane. The following delivery units were included in the study: Steve Biko Academic Hospital (SBAH) (level 3), Kalafong Hospital (KAH) (level 3), Mamelodi Hospital (level 2), Tshwane District Hospital (level 1), Pretoria West Hospital (level 1), Laudium Community Health centre (CHC) with Midwife Obstetric Unit (MOU), (CHC), Eersterust MOU (CHC), Stanza Bopape and Dark City Clinics (CHC). The Steve Biko and Kalafong Hospitals are tertiary referral hospitals which receive referrals from outside the Gauteng province. The data was only analysed for women living in the Tshwane region; those living outside our complex were excluded. Cases of abortions and ectopic pregnancy were excluded from the study. Delivery data was recorded on a daily basis at all the health facilities and daily audit meetings were held at SBAH and KAH to identify women with life-threatening conditions and organ dysfunction in pregnancy. The following World Health Organisation (WHO) indicators were used to quantify women with severe complications in pregnancy2,5:Maternal near miss – a woman who nearly died but survived a complication that occurred during pregnancy, childbirth or within 42 days of termination of pregnancy. The WHO near miss criteria are listed in table 1.Maternal death – A maternal death is the death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and the site of pregnancy from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes.Life-threatening conditions/severe maternal outcome (SMO) – refers to all women who either qualified as having maternal near miss or who died. It is the sum of maternal near misses and maternal deaths.Potentially life-threatening condition – the 5 described by the WHO are: severe postpartum haemorrhage, severe pre-eclampsia, eclampsia, sepsis/severe systemic infection and ruptured uterus. The operational definitions of the 5 potentially life-threatening conditions are:

160

Severe postpartum haemorrhage – genital bleeding after delivery, with at least one of the following: perceived abnormal bleeding (1000ml or more) or any bleeding with hypotension or blood transfusion

Severe pre-eclampsia – persistent systolic blood pressure of 160mmHg or more or a diastolic blood pressure of 110mmHg; proteinuria of 5g or more in 24 hours; oliguria of < 400ml in 24 hours; and HELLP syndrome or pulmonary oedema. Excludes eclampsia.

Eclampsia – generalised fits in a patient without previous history of epilepsy. Includes coma in pre-eclampsia.

Severe sepsis/ systemic infection – presence of fever (body temperature > 380C), a confirmed or suspected infection (eg chorioamnionitis, septic abortion, endometritis, pneumonia), and at least one of the following: heart rate > 90, respiratory rate > 20, leukopenia (white blood cells < 4000), leucocytosis (white blood cells > 12 000).

Uterine rupture – rupture of uterus during labour confirmed by laparotomy.

Severe Maternal Outcome Ratio (SMOR) – refers to the number of women with life-threatening conditions per 1000 live births. This indicator gives an estimation of the amount of care that would be needed in an area.Mortality index – the number of maternal deaths divided by the number of women with life-threatening conditions, expressed as a percentage.

Table 1. The WHO near miss criteria5

Clinical criteriaAcute cyanosis Breathing rate > 40 < 6/minOliguria unresponsive to fluids or diuretics Loss of consciousness, no pulse/heartbeatJaundice concomitantly with pre-eclampsia

Gasping

Shock Coagulation disordersCerebrovascular accident Total paralysisLaboratory criteriaOxygen saturation < 90% for > 60 minutes

Acute thrombocytopenia (< 50 000 platelets)

Creatinine > 300umol/l or > 3.5mg/dl Bilirubin > 100umol/l or > 6.0mg/dl

161

Unconscious, presence of glucose and ketoacidosis in urine

Lactate > 5

PaO2/FiO2<200mmHg pH < 7.1Management criteriaUse of continuous vasoactive drug Dialysis for treatment of acute kidney

failurePuerperal hysterectomy due to infection or haemorrhage

Cardiopulmonary resuscitation

Transfusion > 5 units of red cell concentrate

Intubation and ventilation for a period > 60 minutes, unrelated to anaesthesia

Descriptive statistics in the form of means and standard deviations in the case of continuous data and frequencies and percentages in the case of categorical data was calculated. Ethical approval was obtained from the University of Pretoria Ethics committee (no: 125/2013).ResultsThere were 26 614 deliveries in the Pretoria Academic Complex during the study period. One-hundred and thirty six women developed life-threatening conditions and there were 19 maternal deaths. The severe maternal outcome ratio (SMOR) was 5.1/1000 births and the mortality index was 14.0%. The overall caesarean section rate was 25.2% and that for women with life-threatening conditions 61.02%. The HIV infection rate for the general population was 19.9%, 23.1% for near-misses and 36.8% for mothers who died. The HIV-disease status was unknown in 2.7% of patients. The spectrum of morbidity from uncomplicated pregnancies to maternal death is illustrated in figure 1.

Figure 1 (not drawn to scale)

162

Most of the patients with potentially life threatening conditions and life threatening conditions were treated at the two tertiary level hospitals. Forty-six (39.3%) women who were classified as near misses and 7 (36.8%) women who demised had to be transferred to the two tertiary level hospitals after initially presenting at a lower level of care. The most frequent indications for emergency transfer of women with life-threatening conditions to the tertiary hospitals were: severe preeclampsia (15.4%, n=21), obstetric haemorrhage (13.2% n=18) and organ dysfunction in women with underlying medical disease (6.6%, n=9) (Table 2). The mortality index for SBAH was 18.6%, 10.2% for KAH and 12.5% for Mamelodi Hospital. The distribution of patients with potentially life-threatening conditions in relation to the different levels of care is shown in figure 2.

Figure 2. Distribution of potentially life-threatening conditions, near misses and maternal deaths for the different levels of care

163

Num

ber o

f pat

ient

s

Table 2 Acute life-threatening conditions requiring tertiary carePatients referred to tertiary centre from lower levels of care

Patients already in tertiary care

n % n %Obstetric haemorrhage 18 13.2 26 19,1Pre-eclampsia 21 15.4 22 16.2Sepsis 3 2.2 11 8.1Medical/surgical disorders

9 6.6 9 6.6

Non-pregnancy related infections

3 2.2 8 5.9

Anaesthetic disorders 0 0 4 2.9Other 2 1.5Total 54 39.7 82 60.3

The frequency of potentially life-threatening disorders is shown in table 3. (Ante-partum haemorrhage and non-pregnancy related infections which are not part of the WHO definition for potentially life-threatening conditions have been included.)

164

Table 3 Frequency of potentially life-threatening disordersAll women (n=26 614)

Women with SMO (n=136)

HIV infection in women with SMO (%)

Severe haemorrhage 660 (2.5%) 51 (37.5%) 7 (13.7%)Ante-partum haemorrhage

301 (1.1%) 17 (12.5%) 1 (5.9%)

Postpartum haemorrhage 336 (1.3%) 31 (22.7%) 4 (12.9%)Ruptured uterus 23 (0.1%) 3 (2.2%) 2 (66.7%)Severe hypertensive disorders

682 (2.6%) 44 (32.4%) 4 (9.1%)

Preeclampsia 457 (1.7%) 40 (29.4%) 4 (10.0%)Eclampsia 225 (0.8%) 4 (2.9%) 0Other complicationsPuerperal sepsis 35 (0.1%) 14 (10.3%) 2 (14.3%)Non-pregnancy related infections

20 (14.7%) 20 (100%)

The mortality index for non-pregnancy related infections was 30.0%, obstetric haemorrhage 2.0%, 13.6% for hypertension and 19.0% for medical and surgical disorders (Table 4)

Table 4 Mortality index for different disease conditionsUnderlying condition

Maternal near miss (n)

Maternal Death (n)

Mortality Index (%)

Obstetric haemorrhage

50 1 2.0%

Ante-partum haemorrhage

17 0 0

Ruptured uterus 3 0 0Post-partum haemorrhage

30 1 3.2%

Hypertension 38 6 13.6%Chronic 1 0 0Preeclampsia 35 4 10.0%Eclampsia 2 2 50.0%Puerperal sepsis 14 0 0Non-pregnancy related infections

14 6 30.0%

Medical/surgical disorders

17 4 19.0%

The near-miss markers and distribution of organ dysfunction are shown in Tables 5 and 6. There were 6 maternal deaths related to HIV and AIDS. Four patients had respiratory failure secondary to TB pneumonia, one patient had bacterial meningitis and one patient demised after presenting with multi-organ failure and milliary tuberculosis. Six mothers demised after having complications related to pre-eclampsia. One patient had a liver rupture, 2 patients had intra-cranial bleeds, 2 patients had

165

respiratory failure due to pulmonary oedema and one patient had a cardiac arrest. The patient who demised as a result of postpartum haemorrhage had a placenta praevia and 2 previous caesarean sections. Although the patient had an ante-natal ultrasound confirming placental location, the diagnosis of placenta accreta was missed. Surgeons encountered a major bleed at caesarean section and the patient also had 2 re-look laparotomies but the bleeding could not be controlled. Four patients died due to underlying medical disease; one each due to breast cancer, acute on chronic pancreatitis, an anaesthetic complication in a diabetic patient and one patient with a prosthetic heart valve in pregnancy.There were no maternal deaths at the level 1 hospitals or community health centres and just 1 death at a level 2 hospital. This was a patient with advanced stage of breast cancer who was unable to get transport to a tertiary level facility. The mortality index for the two tertiary hospitals was 18.6% (SBAH) and 10.15% (KAH) and 12.5% for the level 2 hospital.

Table 5 Markers for classification of a maternal near-miss.Near-miss marker Number (%)Cerebrovascular accident 2 (1.70)Total paralysis 1 (0.85)Oxygen saturation <90% for > 60 minutes 6 (5.13)Acute thrombocytopenia (<50 000 platelets) 26 (2.22)Creatinine > 300umol/l or > 3.5mg/dl 4 (3.42)Bilirubin > 100umol/l or > 6.0 mg/dl 1 (0.85)Ketoacids in urine 4 (3.42)Use of continuous vasoactive drug 3 (2.56)Dialysis for acute renal failure 2 (1.70)Hysterectomy following infection or haemorrhage 35 (29.91)

(infection=14, haemorrhage = 21)Cardio-pulmonary resuscitation 3 (2.56)Transfusion of > 5units red cell transfusion 31 (26.50)Intubation and ventilation for > 60 minutes 18 (15.38)

Table 6 Organ system dysfunction in women with life-threatening conditions

number PercentageVascular dysfunction (hypovolemia)

54 40.0%

Uterine dysfunction 35 25.74%Coagulation dysfunction 27 19.85%Respiratory dysfunction 24 17.65%Cardiovascular dysfunction 9 6.61%Immunological dysfunction 8 5.89%Renal dysfunction 8 5.89%

166

Cerebral dysfunction 7 5.15%Hepatic dysfunction 5 3.68%Metabolic dysfunction 5 3.68%

There can be more than one organ dysfunctionTable 7 Comparison of the indices of severe acute morbidity rates at the Pretoria Academic Complex for the periods 1997-1998, 2002-2004 and 2013-2014.

1997-1998 2002-2004 2013-2014

SMOR iMMR MI SMOR iMMR MI SMOR iMMR MI

Antepartum haemorrhage

1.04 0 0 0.91 1.9 2.1 0.6 0 0

Postpartum haemorrhage

1.41 7.4 5.3 2.06 15.5 7.5 1.2 3.8 3.2

Hypertension 1.48 33.3 22.5 1.57 19.42 12.3

1.6 22.5 13.6

Puerperal sepsis 0.37 7.4 20.0 0.54 5.8 10.7

0.5 0 0

Non-pregnancy related infections

0.33 22.2 66.7 0.41 19.4 47.6

0.8 22.5 30.0

Medical and surgical disorders

0.78 11.1 14.3 0.82 11.7 14.3

0.8 15.0 19.0

Total (excluding early pregnancy losses)

5.8 96.2 16.6 7.0 85.5 12.2

5.1 71.4 14.0

Although the SMOR for the general population has remained the same since 1997-1998, both the iMMR and the MI have decreased. These findings are consistent for postpartum haemorrhage and hypertension. The SMOR for puerperal sepsis has remained constant despite the HIV epidemic with a decrease in mortality index. The SMOR and mortality index for medical and surgical conditions remain unchanged.Figure 3 illustrates the perinatal mortality related to maternal morbidity. The women with severe maternal morbidity and mortality had a much high perinatal mortality rate (PNMR), however for every woman with a complicated pregnancy almost 5 women had no life threatening condition. This explains the relatively small difference between the total PNMR and the PNMR of the non-life threatening conditions.

167

Figure 3 Perinatal mortality rate (for babies > 500g)

The primary obstetric causes of perinatal death were unexplained intra-uterine death (30.3%), spontaneous preterm labour (25.5%), ante-partum haemorrhage (12.3%), intra-partum asphyxia (9.3%), hypertensive disorders (7.4%), fetal abnormality (6.9%) and maternal disease (3.7%).

DiscussionTo our knowledge this is the first study in South Africa assessing the spectrum of morbidity for a specific region. There were 26 614 deliveries over a 12-month period (2013-2014). This is almost a doubling of deliveries since 1997-1998 when the total number of births for the biennium was 27 025 and a 35% increase since 2002-2004 (51 469 births for the triennium 2002-2004).6 Just over 4% of women developed a potentially life-threatening condition and 0.5% developed a life-threatening condition. This is lower than the WHO Multi-country Survey on

168

Near miss196.6 / 1 000

(n = 23)

Maternal death210 / 1 000

(n = 4)

Non near-missSevere morbidity

102.64 / 1 000(n = 101)

Life threatening conditions

198.0 / 1 000(n = 27)

Potentially life-threatening conditions

114.26 / 1 000 (n = 128)

Non –life threatening conditions

23.13 / 1 000(n = 589)

n

TOTAL DELIVERIES26.94 / 1 000

(n=717)

maternal and newborn health which reported an incidence of 7% for potentially life-threatening conditions and 1% for life-threatening conditions.4 However the difference between the 2 studies was that our was population-based while the WHO study was hospital-based. About 40% of women with acute life-threatening conditions did not present directly to the two tertiary level hospitals during the acute stage of disease. These patients were booked at a level 1 or 2 facility and then developed an acute condition requiring urgent transfer. Severe pre-eclampsia, obstetric haemorrhage and organ dysfunction due to an underlying medical condition were the most important reasons for emergency transfer. This indicates the necessity of having all health care professionals involved in care of pregnant women trained in the initial stabilisation and management of obstetric and neonatal emergencies. The antenatal care protocol used in our complex is based on the WHO recommendation of four antenatal visits for low risk patients.7 The low frequency of visits possibly means that cases of preeclampsia in the early stages of the disease process were missed leading to patients presenting at a later stage with acute complications requiring tertiary care. This might mean revision of our current antenatal care protocol is required. This is supported by the recent Cochrane review on patterns of routine antenatal care for low-risk pregnancy.8

There has been a decrease in the iMMR and MI at the PAC since 1997. This has been associated with decreases in MI for postpartum haemorrhage, hypertension, puerperal sepsis and non-pregnancy related infections. The mortality index for non-pregnancy related infections in the Pretoria Academic Complex was 66.7% in 1997-1999, 75% in 2000, 47.6% from 2002-2004 and 30% during this study period.6,9 The decrease reflects the implementation of the anti-retroviral program by the Department of Health and better handling of respiratory complications. Of significance is the low mortality index for postpartum haemorrhage (3.2%) which is less than half of the rate (7.5%) reported in 2002-2004 and significantly lower than of the rate (5.3%) reported in for the period 1997-1998.9.10 The decrease in MI for severe postpartum haemorrhage and puerperal sepsis

169

is most likely as a result of the introduction of strict protocols. The frequency of postpartum haemorrhage in women with life threatening conditions (22.7%) was similar to that of the WHO study (26.7%).4

However the rates of preeclampsia (29.41%) and non-pregnancy related infections (14.7%) was greater in our study (WHO 16.3% and 1.6% respectively). The rate of preeclampsia in women with life-threatening conditions was consistent with reports from Nigeria (32.5%) and Mozambique (32.9%).11,12

Vascular (hypovolemia), uterine (hysterectomy) and coagulation (low platelets) dysfunction were the most frequent organ system dysfunctions in women with life-threatening conditions (table 6). Many women had multiple complications. The disease profile in our complex has changed since the year 2000 when vascular, cardiac, immunological and coagulation dysfunction were the most important organ systems causing obstetric morbidity.13 Obstetric haemorrhage was the potentially life-threatening condition most frequently encountered in our complex (37.5%) and vascular dysfunction as a result of hypovolemia was the most common organ system dysfunction seen. The low mortality index for postpartum haemorrhage suggests that although postpartum haemorrhage is an important problem, the condition is well managed by our clinicians. Of the five potentially life-threatening conditions, hypertensive disorders contributed to 7.4% of perinatal deaths. Ninety-two point six percent of perinatal deaths were not related to antepartum and intrapartum maternal life-threatening conditions, and if postpartum maternal life-threatening conditions are included then 80% of the women with perinatal deaths did not have severe morbidity. These findings are consistent with that Allanson et al describing perinatal mortality in the Mpumalanga Province and Vogel et al in the WHO Multi-Country Survey who found that a significant proportion of women have no recognisable obstetric or medical condition at the time of perinatal death.14,15 The WHO Multi-Country Survey found a maternal complication rate of 22.9%, 27.7% and

170

21.2% in late macerated stillbirths, late fresh stillbirths and early neonatal deaths respectively and Allanson et al found a rate of maternal complications in macerated stillbirths, fresh stillbirths and early neonatal deaths of 50.4%, 50.7% and 25.8% respectively. Current early antenatal identification of both severe maternal morbidity and perinatal mortality is inadequate.The strength of this paper is the robust method of data collection. The new national birth register records maternal complications facilitating the collection of data. The Pretoria Academic Complex has been collecting and reviewing data on life-threatening conditions for more than 15 years and all the doctors are familiar with the WHO near miss criteria. Women who were classified as a near miss were interviewed about barriers encountered in accessing healthcare. This information will be presented in a separate paper. The limitations of this study are the exclusion of cases of early pregnancy loss (abortions and ectopic pregnancies) and some cases of sepsis may have been missed if patients presented late in the postpartum period. Further, maternal infections like pneumonia, tuberculosis, meningitis were not on the list of potentially life-threatening conditions, and thus the SMOR could not be calculated for this. The list of potentially life-threatening should be expanded to include medical conditions and non-pregnancy related infections. This is supported by Lumbiganon et al who demonstrated that indirect causes of maternal deaths are increasingly important in developing countries with indirect causes being responsible for about one-fifth of severe maternal outcomes.16

Recommendations The World Health Organisation has identified 5 potentially life

threatening conditions: severe postpartum haemorrhage, severe preeclampsia, eclampsia, sepsis/severe infection and ruptured uterus.2 Our study has shown that conditions such as abruptio placenta, non-pregnancy related infections and medical and surgical disorders are also important causes of obstetric morbidity and the

171

WHO should therefore consider expanding their categories of potentially life-threatening conditions.

Forty-percent of patients with life-threatening conditions presented to a level 1 or 2 facility before being transferred for tertiary care. Cases of postpartum haemorrhage and severe pre-eclampsia could not be predicted antenatally. In addition, no recognisable obstetric condition was present in majority pregnancies that ended in a perinatal death. Health workers in level 1 and 2 centres must therefore be able to recognise, stabilise and transfer pregnant women and neonates presenting with an acute obstetric emergency.

Strategies to prevent and screen for preeclampsia and improvement of emergency transport for women are essential in order to reduce obstetric morbidity and mortality.

Review of the reduced visits protocol put forward by the WHO should be considered as increasing the frequency of ante-natal visits for low risk women may increase detection of pre-eclampsia at an earlier stage of the disease process.7 However, this would require considerable increase in resources.

ConclusionThis study was able to identify the proportion of pregnancy-related morbidity in our population and compare it to other studies. The mortality index and prevalence of potentially life-threatening conditions were similar to the WHO Multi-country Survey. Although there has been a decrease in the mortality index for non-pregnancy related infection, further interventions need to be implemented to reduce morbidity and mortality associated with HIV-disease and tuberculosis. A significant proportion of women who developed severe maternal conditions were not identified during the ante-natal period indicating the necessity of ensuring all levels of care can manage the initial steps in obstetric and neonatal emergencies and an efficient emergency transport system is available.

References

172

1. Saving Mothers: Sixth Report on Confidential Enquiries into Maternal Deaths in South Africa. 2011-2013. Pretoria: Department of Health, 2015.

2. Evaluating the quality of care for severe pregnancy complications. The WHO near-miss approach for maternal health. Geneva: WHO Press, 2011.

3. Pattinson RC and Hall MH. Near Misses: a useful adjunct to maternal death enquiries. Br Med Bull 2003; 67: 231-243.

4. Souza JP, Gulmezoglu AM, Vogel J, Carroli G, Lumbiganon P, Qureshi Z et al. Moving beyond essential interventions for reduction of maternal morbidity (the WHO Multicountry Survey on maternal and Newborn Health): a cross-sectional study. The Lancet 2013; 381: 1747-1755.

5. Say L, Souza JP, Pattinson RC. Maternal near-miss – towards a standard tool for monitoring quality of maternal care. Best Practice and research Clinical obstetrics anf Gyneacology 2009; 23: 287-296.

6. Pattinson RC, Macdonald AP, Backer F, Kleynhans M. Effect of audit on critically ill pregnant women. Clin Gov Int J 2006; 11(4): 278-288.

7. Villar J and Bergsjo P. 2003. WHO Antenatal Care Randomised Trial: Manual for the Implementation of the New Model. WHO/RHR/01.30. WHO: Geneva

8. Dowswell T, Carroli G, Duley L et al. Alternative versus standard packages of antenatal care for low-risk pregnancy (Review) 2010, Issue 10. Art. No.: CD000934. DOI:10.1002/14651858.CD000934.pub2.

9. Vandecruys HIB, Pattinson RC, Macdonald AP, Mantel GD. Severe acute maternal morbidity and mortality in the Pretoria Academic Complex: changing patterns over 4 years. European Journal of Obstetrics and Gynecology and Reproductive Biology 2002; 102: 6-10.

10.Lombaard H, Pattinson RC. Common errors and remedies in managing postpartum haemorrhage. Best Practice and Clinical Obstetrics and Gynecology 2009; 23: 317-326.

11.Daru PH, MU J, Achara P et al. Near miss maternal mortality in Jos University Teaching Hospital, Jos, Plateau State Nigeria. Ibom Medical Journal 2008; 3.

12.David E, Machungo F et al. Maternal near miss and maternal deaths in Mozambique: a cross sectional, region-wide study of 635 consecutive cases assisted in health facilities of Maputo Province. BMC Pregnancy and Childbirth 2014; 14:401.

13.Mantel GD, Buchmann E, Rees H and Pattinson RC. Severe acute maternal morbidity: a pilot study of a definition for a near-miss. BJOG 1998; 105: 985-990.

14.Allanson ER, Muller M, Pattinson RC. Causes of perinatal mortality and associated maternal complications in a South African province: challenges in predicting poor outcomes. BMC Pregnancy and Childbirth 2015;

15.Vogel JP, Souza JP, Mori R et al. Maternal complications and perinatal mortality: findings of the World health Organisation Multicountry Survey on Maternal and Newborn Health. BJOG 2014; (Suppl. 1): 76-88.

16.Lumbiganon P, Laopaiboon M, Intarut N et al. Indirect causes of severe adverse maternal outcomes: a secondary analysis of the WHO

173

Multicountry Survey on Maternal and Newborn Health. BJOG 2014;121 (Suppl. 1): 32-39.

174

RETHINKING IUCD USE AS A CONTRACEPTIVE OPTION IN KWAZULU-NATAL, SOUTH AFRICA

Valerie Makatini, Pretty Harrison, Pinky Phungula

Background and RationaleMore than 1/3 of South African women deliver a child by 19.5 years of age: most of those women are unmarried and did not use a contraceptive. A high proportion of pregnancies among teenage and young adult South African women are characterized as unplanned with the highest levels among unmarried, younger women.1 Family planning (FP) can help save women’s and children’s lives, preventing as many as one in every three maternal deaths by encouraging women to delay motherhood, space births, avoid unintended pregnancies and abortions, and stop childbearing when they have reached their desired family size.

Programme approachGuided by the Millennium Development Goals, (MDGs) and the South African National Contraception and Fertility Planning Policy and Service Delivery Guidelines, the KwaZulu-Natal (KZN) Provincial Department of Health, (PDOH) developed a Provincial 5-Point Contraceptive Strategy to revitalize interest in the utilization of the FP Programme and to increase use of contraceptive prevalence. The strategy’s 5 Key Priorities are:

Promote Healthy Timing and Spacing of Pregnancies by Improving Contraceptive Awareness and Access at Health Facilities and in the Community

Improve Contraceptive Method Mix Promote Integration of Contraceptive Services with other Services Improve Health Care Provider Training and Mentoring on Contraception Improve Record Keeping and Monitoring and Evaluation

MethodUthungulu District is a largely rural district in KwaZulu-Natal, with a population of 972,856 and comprises of six local authority areas. It was identified as one of the districts with high maternal mortality in the province, and therefore prioritized for FP interventions. In order to ensure appropriate implementation of the KZN FP Strategy, a baseline assessment was conducted, in March 2012, including a rapid assessment of community awareness of FP services and availability and evaluation of key FP indicators from the District Health Information System (DHIS) data. Following on the assessment, an operational FP plan was developed for UThungulu District, and disseminated to all sub-districts. The URC ASSIST Project and KZN PDOH staff communicated the KZN 5 Point Contraceptive strategy to all the health facilities in the district and created awareness amongst Community Leaders. Key to the success of this initiative was the integration of FP at Operation Sukuma Sakhe meetings with government structures where community health issues and other community matters were disseminated. Staff also spent time on onsite visits to war room meetings and campaigns. Involvement of Traditional Health Practitioners and healers through workshops ensured cultural buy-in, while mainstream training of Health Care worker (HCWs) on Intra-uterine Contraceptive Devices (IUCD) insertions ensured transfer of skills to HCWs. Staff were also involved in community dialogues and mobilization. Key FP indicators from the DHIS data were used to evaluate the success of the intervention.

Key Results:

175

Since implementation of the FP Strategy in Uthungulu District, over 2000 people have been provided with information on FP strategies. Family planning uptake has increased with more than 1000 people accepting FP per quarter. The distribution of female condoms to clients has more than doubled from 35,753 in January – March 2012 to 74,475 in January – March 2013 ( 52% increase). While distribution of male condoms increased from 114, 6585 in January – March 2012 to 161, 4551 in January – March 2013 (29% increase). Insertion of IUCDs increased from 22 per quarter in January – March 2012 to 464 per quarter in January – March 2013. Most importantly, we have found that promotion of IUCD as a contraceptive has proven to strengthen integration of Programmes and provide opportunities for:

•HIV counseling and testing•Early identification and management of STIs.•Early identification and management of obstetric injuries and cervical abnormalities, when inserted post-partum •Opportunities for conducting pap smears•Opportunities to discuss, instruct and distribute male and female condoms; Promote “Dual Protection”

Lessons learned:Through this initiative, important lessons learnt include the fact that political, management and community leadership involvement contribute greatly to the success of health Programmes. Involvement of development partners also contributes significantly to dissemination, implementation and success of Programmes. The huge burden placed on the South African health system by the HIV epidemic has not only overshadowed FP services, but also requires attention to be paid to the contraceptive needs of HIV infected people and those at risk of infection. We learnt that it was essential to re-train, and capacitate staff on FP strategies, insertion and removal of IUCDs and create awareness in the community. Expanding the contraceptive method mix, and re-introducing the IUCD, including for young women and HIV infected women, has provided access to a long acting contraceptive method , that had become less available within the FP service in South Africa.

176

THE PAEDIATRIC ART DATA MANAGEMENT TOOL (PADMT): USING POINT-OF-CARE TECHNOLOGY IN THE FORM OF AN ELECTRONIC MEDICAL RECORD TO ENHANCE CLINICAL QUALITY SHORTFALLS IN RESOURCE-POOR PAEDIATRIC ARV SETTINGS

Dr Kim Harper [MBChB (UCT), DCH (SA), FCPaed (SA)] Paediatrician, Head of Clinical Unit, Department of Paediatrics & Child Health, Frere Hospital, East London Hospital Complex.

Introduction: Electronic medical records have been in use in many developed countries since the early 2000s. In African settings, uptake of electronic solutions to managing health care at the point of care has been slow. This is commonly attributed to the lack of technology infrastructure; however, in 2015 we are capable of advanced touch screen technology, 3G and more.

With respects to HIV management, only a quarter of the number of children eligible for ARVs in 2013 received them as opposed to half of all adults. Having a digital system in place to track and trace those children on ARVs and those at risk for loss to follow up is critical to our success in meeting ARV treatment goals.

Method:We undertook a project to digitize all paper-based records of children (<19 yrs) attending the East London Hospital Complex (ELHC) ARV clinics from 2005, for future migration into an electronic Medical Record (eMR).

We borrowed from the TREAT ASIA cohorts and modified their simple ACCESS tool. We then systematically entered all patient information that was available in the paper-based chart. In parallel, we looked at modifying an existing eMR currently in use in parts of Nigeria, Kenya, Uganda, Ethiopia, Tanzania and Rwanda known as IQCare. This system is open source, which means it is free to use but does require site-specific modifications. Between 2012 – 2013, we modified the IQCare software platform, resulting in our PADMT software point of care system.

The PADMT has been deployed along with all captured data onto the DoH computer servers at both Frere and Cecelia Makiwane Hospitals of the ELHC. This now provides access to over 2 600 patient records and more than 60 000 visits, one of the largest such cohorts in the world. The PADMT is now used as the point of care eMR by health professionals in the paediatric ARV clinics in East London.

The system is a web-based, touch screen and keyboard capable, meaning it can be used on tablets or smartphones. The program has summary sliders, overdue test reminders, weight and BMI tracking, CD4 & viral load trending, reports, lab results tracking and electronic scripting facilities etc.

The capacity of the tool to generate reports is also important. Currently, the tool can be used to build and generate aggregate level reporting

177

(useful for DoH indicator summary data), the patient level (full case history) and the disaggregate level (useful for specific searches). Patient summaries can be printed to facilitate down-referrals.

In terms of functionality, PADMT allows one to do both high and low level review of patient cases, which allows for example, MMEDs candidates to conduct research and clinicians to investigate more complex cases. Finally and importantly, the system has the ability to interface with global HIV cohorts via the use of SQL (Structured Query Language).

178

SEX DIFFERENCES IN EEG ACTIVITY BETWEEN HEALTHY NEWBORNS DURING THE FIRST DAYS OF LIFE

Fifer, WP, Burtchen, N, Myers, MM Piraquive, J.Departments of Psychiatry and Pediatrics, Columbia University Medical Center. New York

IntroductionElectroencephalogram (EEG) power, a measure of local neural synchrony, has been associated with neurodevelopmental outcomes. High frequency power is positively correlated with cognitive processes including attention, perception, and memory. Girls show advanced development of language and fine motor skills and display earlier cortical maturation than boys in early childhood and boys are known to be at higher risk for developmental and learning disabilities. It is currently not known, however, if EEG power measures are different between female and male as early as the newborn period and if these differences in early brain activity underlie sex differences in later neurodevelopmental outcomes.

Methods128-lead EEG was obtained in 112 supine sleeping infants at 18-72 hours postpartum. Inclusion criteria: gestational age = 35 0/7- 40 6/7 weeks, no birth or medical complications; no maternal psychotropic medication, diabetes, or hypertension.

ResultsFemale newborns showed significantly greater high-frequency frontal EEG power in both active and quiet sleep, as compared to male newborns. This finding was most pronounced during quiet sleep, with similar patterns in the left and right hemispheres (left: 32% increase of frontal EEG power for females p<0.01). In active sleep, differences in EEG patterns between female and male newborns were comparable to findings in quiet sleep if slightly decreased in magnitude. Sex differences in EEG power were independent of gestational age.

ConclusionsSex differences in brain function emerge early and are detectable within the first days of life. Ongoing studies are investigating if electrocortical differences in the newborn period predict electrocortical activity and neurodevelopmental outcomes at 15 months of age.

This research was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development: R37- HD32774

179

EXPERIENCE WITH INDUCED HYPOTHERMIA AT THE CHRIS HANI BARAGWANATH ACADEMIC HOSPITAL: PRELIMINARY RESULTS

Firdose Nakwa, Sithembiso VelaphiFaculty of Health Sciences, Department of PaediatricsChris Hani Baragwanath Academic Hospital and the University of the Witwatersrand

INTRODUCTIONPerinatal asphyxia is the leading cause of cerebral palsy. Incidence of hypoxic ischaemic encephalopathy (HIE) is reported to be 1.5/1000 live births in first world countries1,2. The incidence is highest in the developing world. The perinatal asphyxia rate was 8.5/1000 livebirths at the Chris Hani Baragwanath Academic Hospital (CHBAH) in 2011 and 13/1000 livebirths in 2013. The care for HIE is largely supportive however, neuroprotective strategies have shown to be effective in reducing morbidity and mortality associated with HIE. A recent meta-analysis has shown that induced hypothermia reduces mortality and morbidity at 18 months – 24 months by 24%3. The number need to treat (NNT) is 7. There are two methods of cooling selective head cooling (SHC) and total body cooling (TBC) and there has been no difference with regards to the effectiveness of the two methods4,5,6. International Liaison Committee on Resuscitation (ILCOR) has recommended as the standard of care for asphyxiated patients with moderate to severe encephalopathy. However, it is to be used with reserve in developing countries. Induced hypothermia has been introduced at CHBAH in August 2011.

OBJECTIVESThe objectives are to describe the criteria used to select asphyxiated infants eligible for cooling, to determine the age at which cooling was started, to determine the heart rate and temperature of infants that have been cooled; before, during and after cooling, and to determine the outcomes at hospital discharge and at follow-up at 18-24 months

METHODS

180

This was a retrospective record review conducted in patients with a diagnosis off asphyxia and managed with induced hypothermia at the Chris Hani Baragwanath Academic Hospital (CHBAH) between August 2011 and November 2013. The outcome variables were the fulfillment of criteria for cooling, complications related to cooling, mortality and the neurological outcome. The protocol to select patients for cooling were: 1. gestational age of ≥34 weeks and or weight ≥2000g, 2. apgar score ≤5 at 10 minutes or required resuscitation for more than 10 minutes, 3. pH<7.00 or base deficit ≥16 mmol/L based on an arterial gas done within 60 minutes of birth, 4. encephalopathy (defined as lethargy, stupor, coma) and at least one of the following:- hypotonia, abnormal reflexes, absent or weak suck and clinical seizures and/or 5. An abnormal amplitude-integrated electroencephalography (aEEG) for at least 30 minutes. Patients had to be cooled within 6 hours of life. During cooling patients were monitored for vital signs (temperature, heart rate and blood pressure) and had daily clinical examinations. Re-warming was started after 72 hours cooling at a rate of 0.5°C/hour until 36.5°C. Demographic data, complications related to cooling and the outcome (death or discharge) and Griffiths scores at 18 months; were collected.

RESULTS:There was a total of 98 patients cooled during the period. There were 55 (56%) males and 43 (44%) females. The mean maternal age and parity was 24.3 years and 1 respectively. Fifty seven percent of the patients were delivered normally and 18% were HIV positive. The mean gestational age and birth weight was 39.1 and 3182.7 respectively.

181

Table 1: Blood Gas a parameters within 1 hour of birthBlood Gases Number (%) - BMV - BMV and CPR - BMV, CPR and Adrenalin

84 (86)8 (8)2 (2)

pH pH <7.0 pH 7.0-7.25 pH >7.25

7.0 (0.15)*65 (66)16 (16)11 (11)

Base deficit (mmol/l) - < 16 (mmol/l) - > 16 (mmol/l)

21.0 (4.8)*11 (11)81 (83)

*Mean (SD) Bag mask ventilation was required by 86% of patients. Eight percent (8%) went on to require CPR and only 2% of patients required adrenalin during resuscitation. Regarding the blood gas parameters, the mean pH was 7.0 and base deficit of 21.0 which fulfilled the criteria for an asphyxiated baby. Two thirds of the patients had a pH < 7.0 and 83% of patients had a base deficit of >16 mmol/l. (Table 1)

Half of the patients had a sentinel event which included, fetal distress 22 (46%), prolonged second stage 10 (21%), maternal collapse 4 (8%), chorioamnionitis 3 (6%), cord prolapse 2 (4.25%), cord around the neck 2 (4.25%), failed instrumental delivery 2 (4.25%), MSL 2 (4,25%), and abruption placentae 1 (2%). A quarter had an apgar of 5 at 5 minutes. A total of 60 % were resuscitated at 10 minutes and beyond. Ninety six percent had moderate to severe encephalopathy and 2 infants with mild HIE were cooled. Brain monitoring occurred in 58% of patients. Of those 45 % had moderate suppression and 48% had a severe suppression on the aEEG tracing (Table 2). Induced hypothermia was commenced at a mean time of 4 hr 42min. Seven patients were ventilated, 5 were on nCPAP for apneas and 2 on Conventional Mechanical Ventilation (CMV) for MAS and PPHN. The median temperature and heart rate during cooling and rewarming was maintained (Fig 1 and 2).

182

Table 2: Number of patients fulfilling criteria for coolingCriteria Number (%)Sentinel event 48 (49)Apgars <5 at 10 min 25 (26)Resuscitation at 10 min 59 (60)Ph <7.0Base Deficit >16 mmol/l

65 (66)81 (83)

Encephalopathy - HIE 1 - HIE 2 - HIE 3

2 (2)81 (83)15 (15)

aEEG pattern Continuous normal voltage Discontinuous normal voltage Burst Suppression Continuous Low Voltage Flat Trace

584 (7)

26 (45)10 (17)16 (28)2 (3)

Complications during cooling included thrombocytopenia and coagulopathy. Thrombocytopenia defined as a platelet count <150 X109/l, was reported in 22 patients that were cooled and only 3 patients had an INR > 1.5 IU/l. Sepsis was identified in 37% of patients; 30% of which had clinically presumed sepsis with a CRP > 10mg/dl. There were 6

183

Figure 1. Temperature chart during cooling and rewarming

Figure 2. Heart rate during cooling and rewarming.

positive cultures – 4 infants had streptococcus agalactiae, 1 Pseudomonas aeruginosa and 1 Enterobacter cloacae. Cranial sonars were performed on 65 % of infants. The reported findings were 7 (11%) with a normal sonar, 45 (69%) had white matter involvement, 4 (6%) had basal ganglia and thalami involvement (deep gray matter involvement) and 9 (14%) had diffuse involvement. Regarding outcomes 15 patients died - 11 were Sarnat Stage 2 (HIE 2) and 4 had severe encephalopathy (HIE 3). A Griffiths score < 85 is associated with developmental delay. A Griffiths score was performed in 20 infants in the cohort. Only 5 (25%) patients had a score < 85, 12 (60%) between 85 and 100 and 3 (15%) > 100.

CONCLUSION:The patients were selected according to the protocol. We were able to cool infants within 6 hours. The temperature during cooling was maintained and the complications during cooling were minimal. The mortality was at 15% and patients that demised had intractable seizures. There were few patients assessed at 12-18 months with a Griffiths motor developmental assessment scale (GMDS). Therapeutic Hypothermia is feasible within a South African setting outside of an NICU setting. However, resources remain a major factor in managing these patients according to the protocol. Long-term follow-up is important to fully assess the effectiveness of induced hypothermia at 18 months in infants that have been cooled.

ACKNOWLEDGEMENTS:Registrars, Nurses, Consultants that care for the babies. Dr K. Thandrayen for her technical support with statistics and analysis and Dr S. Lowick for assistance at the clinic with the Griffiths motor developmental assessment.REFERENCES:

184

1. Kurinczuk JJ1, White-Koning M, Badawi N. Epidemiology of neonatal encephalopathy and hypoxic–ischaemic encephalopathy. Early Hum Dev. 2010 Jun;86(6):329-382. Vannucci RC. Current and potentially new management strategies for perinatal hypoxic-ischemic encephalopathy. Pediatrics 1990; 85: 961–8.3. 3. Tagin MA, Woolcott CG, Vincer MJ, Whyte RK, Stinson DA. Hypothermia for neonatal hypoxic ischemic encephalopathy: an updated systematic review and meta-analysis. Arch Pediatr Adolesc Med. 2012;166(6):558–566

4. Gunn AJ, Gluckman PD, Gunn TR. Selective head cooling in newborn infants after perinatal asphyxia: a safety study. Pediatrics 1998;102(4 Pt 1):885–92

5. Azzopardi DV, Strohm B, Edwards AD, et al. Moderate hypothermia to treat perinatal asphyxial encephalopathy. NEngl J Med 2009;361:1349-58;

6. Jacobs SE, Berg M, Hunt R, Tarnow-Mordi WO, Inder TE, Davis PG. Cooling for newborns with hypoxic ischaemic encephalopathy. The Cochrane Library 2013; Issue 1. Art. No.: CD003311. DOI: 0.1002/14651858.CD003311.pub3.

185

THE LIFELONG EFFECTS OF EARLY NEONATAL AND CHILDHOOD ADVERSITY AND TOXIC STRESS. - MATERNAL INFANT SEPARATION AND THE BUSY NICU ENVIRONMENT

Vanessa Booysen

A vital and productive society with a prosperous and sustainable future is built on a foundation of healthy child development. Health in the earliest years—beginning with the future mother’s well-being before she becomes pregnant—lays the groundwork for a lifetime of the physical and mental vitality that is necessary for a strong workforce and responsible participation in community life. When developing biological systems are strengthened by positive early experiences, children are more likely to thrive and grow up to be healthy, contributing adults. Sound health in early childhood provides a foundation for the construction of sturdy brain architecture and the achievement of a broad range of skills and learning capacities. Together these constitute the building blocks for a vital and sustainable society that invests in its human capital and values the lives of its children.Early experiences and environmental influences can leave a lasting signature on the genetic predispositions that affect emerging brain architecture and long-term health. Many adult diseases should be viewed as developmental disorders that begin early in life and that persistent health disparities associated with poverty, discrimination, or maltreatment could be reduced by the alleviation of toxic stress in childhood.The lifelong costs of childhood toxic stress are enormous.Toxic stress I.E: Strong and prolonged activation of the body’s stress management systems in the absence of the buffering protection of adult support. Toxic stress Disrupts brain architecture and leads to stress management systems that respond at relatively lower thresholds, thereby increasing the risk of stress-related physical and mental illness. Toxic stress can lead to potentially permanent changes in learning (linguistic, cognitive, and social emotional skills), behaviour (adaptive versus maladaptive responses to future adversity), and physiology (a hyper responsive or chronically activated stress response) and can cause physiologic disruptions that result in higher levels of stress related chronic diseases and increase the prevalence of unhealthy lifestyles that lead to widening health disparities as well as impairments in learning, behaviour, and both physical and mental well-beingTherefore the consequences of significant adversity early in life prompt an urgent call for innovative strategies to reduce toxic stress within the context of a coordinated system of policies and services guided by an integrated science of early childhood and early brain development.Effective early neonatal interventions as BSSC, Birth Skin to Skin Care and KMC, Kangaroo Mother Care provide critical opportunities to prevent these undesirable outcomes The future of paediatrics lies in its unique leadership position as a credible and respected voice on behalf of children, which provides a powerful platform for translating scientific advances into more effective strategies and creative interventions to reduce the early childhood adversities that lead to lifelong impairments in learning, behaviour, and health.Paediatricians and Neonatal Nurses need to serve as both front-line guardians of healthy child development and are strategically positioned to build strong foundations for lifelong emotional and mental well-being.

186

A DESCRIPTIVE STUDY OF THE PROFILE OF SUSPECTED INFLICTED CHILD DEATHS IN JOHANNESBURG OVER A FOUR YEAR PERIOD.

Luke Lamprecht*, Lorna Jacklin, Sheri Errington

The reduction of infant mortality is a key millennium development goal. The South African government has taken on the challenge to reduce infant mortality in line with the targets set by the World Health Organisation. According to the Medical Research Council of the primary causes of childhood death are violence are accidents, which are primarily preventable and a major cause or morbidity if the injuries are not fatal.The paper will describe all of the deaths of children who presented to the Johannesburg Forensic Pathology services over a four year period. These cases will be described by mechanism of death, cause of death, history obtained by the South African Police Service and information obtained from the crime scene.Cases that are determined to be child murder will then describe in more detail with regard to high risk areas in Johannesburg as well as a review of the investigations held into these murders.

187

LONG TERM DEVELOPMENTAL OUTCOMES OF CHILD VICTIMS OF INFLICTED INFANT HEAD INJURIES OVER A 10 YEAR PERIOD IN A JOHANNESBURG CHILD ABUSE CLINIC

Lorna Jacklin*, Luke Lamprecht, Sheri Errington

Inflicted Infant Head Injuries (IIHI) are a well described form of child abuse worldwide. The outcomes of these assaults to infants vary from death to severe brain damage and disability to more mild developmental complications. The mechanism of the injury often causes damage to the brain from focal bleeds and injuries to diffuse axonal injury, swelling and atrophy. In addition there is often injury to the eyes and in certain cases breaks and fractures to the infant’s bones.This paper will present cases seen in a Johannesburg child abuse clinic over a 10 year period. The initial injuries will be presented, treatment received over time and then the most current developmental status of the children seen for follow-up in the neurodevelopmental clinic. The aim it to highlight the possible morbidity that comes with this form of infant abuse in order to generate awareness to attempt to prevent this abuse occurring or to identify it early so as to minimise the long term consequences.

188