Pediatr Blood Cancer

Adherence to Transcranial Doppler Screening Guidelines Among ChildrenWith Sickle Cell Disease

Michael J. Eckrich, MD, MPH,1 Winfred C. Wang, MD,2 Elizabeth Yang, MD, PhD,3 Patrick G. Arbogast, PhD,4

Anthony Morrow, BS,5 Judith A. Dudley, BS,5 Wayne A. Ray, PhD,5 and William O. Cooper, MD, MPH3,5*

INTRODUCTION

Sickle cell disease (SCD) is a multisystem disease process that

affects more than 80,000 individuals in the United States alone

[1]. SCD is characterized by several different types of hemoglo-

binopathies, with the most severe disease occurring in persons

with hemoglobin SS (HbSS) or sickle-b0-thalassemia (HbSb0)

[2]. Children with the severe forms of SCD are at risk for a

number of life-threatening complications, including cerebrovascu-

lar accident (CVA). The estimated incidence of CVA in patients

with severe SCD is 0.61 per 100 patient years [2], and approxi-

mately eight percent will experience a stroke before the age of

fourteen years [3]. Overt neurovascular events such as CVA can

be devastating, both in terms of morbidity and overall health care

costs [4,5].

The use of transcranial Doppler (TCD) ultrasound provides a

practical and safe mechanism for the detection of patients at

highest risk for CVA [6]. In 1997, the Stroke Prevention Trial

in Sickle Cell Anemia (STOP) demonstrated a profound benefit of

transfusion therapy for children with HbSS or HbSb0 whose

screening TCD indicated a high risk for stroke [7]. Because of

the strong predictive ability of TCD in the STOP trial [8], current

recommendations from the National Heart Lung and Blood Insti-

tute (NHLBI) include screening all patients with SCA starting at

age 2 and continuing yearly until age 16 [7]. The recommendation

for annual TCD screening was reiterated in the 2002 Fourth

edition on the Management of Sickle Cell Disease [9] and the

AAP Health Supervision for Children with Sickle Cell Disease

guidelines [10].

Yet, despite the availability of a stroke-preventing interven-

tion, children with HbSS or HbSb0 at high risk for stroke often do

not receive recommended screening [11]. Previous experience

with preventive care recommendations for children with SCA

has demonstrated poor adherence rates with penicillin prophylaxis

[12,13]. The goal of this study was to assess adherence to

TCD screening guidelines in a cohort of at risk children in

the year before and the years following the original NHLBI

recommendations, and to determine factors that predict adherence

to recommendations.

METHODS

We conducted a retrospective cohort study utilizing data from

the Tennessee Medicaid program (TennCare) and the Tennessee

Department of Health. These data have been used to perform over

300 studies of health care outcomes including several studies of

SCD [13–15] and critical database elements have been validated

[16].

We first identified the population of all persons in the state

who had claims-based evidence of SCD using the International

Classification of Diseases, Ninth revision (ICD-9). Enrollees who

had two outpatient visits separated by 30 days, or a single inpa-

tient admission for ICD-9 codes (282.6, 282.60, 282.61, 282.62,

282.63, 282.69, 282.41, 282.42) were identified [12,13]. Children

with SCD eligible for TCD’s were further identified as those who

were enrolled in TennCare by their 2nd birthday, maintained

Background. Little is known about adherence to guidelines rec-ommending yearly screening with transcranial Doppler (TCD) ul-trasonography to detect stroke risk for children with severe sicklecell disease. The objective was to determine the proportion ofchildren with hemoglobin SS (HbSS) or sickle-b0-thalassemia(HbSb0) aged 2–16 years who received recommended TCD screen-ing from 1997 to 2008, and to identify factors associated withadherence. Procedure. A retrospective cohort study includedpatients enrolled in Tennessee Medicaid with HbSS or HbSb0

who received care at the two largest sickle cell centers in Tennes-see. The outcome of interest was adherence with guidelines forannual screening TCD’s, identified from computer claims and vali-dated through medical record review. The cumulative rate of chil-dren who received a TCD per year was calculated using theKaplan–Meier method. Cox proportional hazards regression was

used to examine the association of child, family, and health careuse characteristics with receiving a TCD. Results. Among 338 TCDeligible at-risk children, 232 (68.6%) had at least one TCD duringthe study period. The yearly cumulative incidence of annual TCD’sincreased from 2.5% in 1997 to 68.3% in 2008. In multivariatemodels, calendar year, maternal education, and increased numberof sickle cell related outpatient visits were associated with an in-creased rate of receiving a TCD. Conclusions. Publicly insuredchildren with HbSS or HbSb0 had increasing adherence with TCDscreening guidelines between 1997 and 2008, though 31% had noTCD at all during follow-up. Increasing number of sickle cell relatedoutpatient visits was associated with increasing adherence toscreening guidelines. Pediatr Blood Cancer� 2012 Wiley Periodicals, Inc.

Key words: sickle cell disease; stroke prevention; transcranial Doppler ultrasound (TCD)

1Center for International Blood and Marrow Transplant Research

(CIBMTR), Milwaukee, Wisconsin; 2Department of Hematology,

St Jude Children’s Hospital, Memphis, Tennessee; 3Department of

Pediatrics, Vanderbilt University, Nashville, Tennessee; 4Department

of Biostatistics, Vanderbilt University, Nashville, Tennessee; 5Depart-

ment of Preventive Medicine, Vanderbilt University, Nashville,

Tennessee

Grant sponsor: Agency for Healthcare Research and Quality CERT

Program; Grant number: HS10384.

Conflict of interest: Nothing to declare.

*Correspondence to: William O. Cooper, MD, MPH, Suite 313

Oxford House 1313 21st Avenue South Nashville, TN 37232-4313.

E-mail: william.cooper@vanderbilt.edu

Received 1 February 2012; Accepted 29 May 2012

� 2012 Wiley Periodicals, Inc.DOI 10.1002/pbc.24240Published online in Wiley Online Library(wileyonlinelibrary.com).

continuous enrollment for 365 days prior to the 2nd birthday

(allowing for brief administrative gaps), and had no evidence of

a prior stroke as determined by the presence of an ICD-9 code

(430, 431, 433-437.7 excluding 435-Transient ischemic attack).

The study began in 1997, the year before the STOP trial was

initially published [8]. The first day of study follow-up for an

individual cohort member was defined as the first day when a

person met study entry requirements. Follow-up continued until

the end of study, which occurred with loss of TennCare enroll-

ment for >60 days, occurrence of stroke, age >16 years, death, or

the end of the study period in 2008.

Children were considered to be assigned to a designated sickle

cell center based on the distance from the child’s residence to the

center on each birth date, consistent with the practice for center

assignment in the state. We selected children who were assigned

to the two largest sickle cell centers in the state, accounting for

84.9% of children in the state with SCD. One of the centers

implemented a clinical program for TCD screening including

placement of a TCD machine in clinic, the other center encour-

aged adherence but did not implement a separate clinical pro-

gram. For children at the two centers, medical records and

hemoglobin identification patterns were reviewed to confirm the

presence of SCD and to identify the disease subtype. Only chil-

dren with confirmed HbSS or HbSb0 were included in the cohort.

For potential cases, for whom we were unable to obtain medical

records (14.5%), case status was determined through manual re-

view of hemoglobin identification patterns and hospital claims

indicating multiple hospitalizations for sickle cell related compli-

cations. Maternal age at birth of the child, and years of education

were collected from birth certificates linked to Medicaid enrol-

ment files. Presence of chronic health conditions in the mother

was identified from the mother’s claims. Neighborhood per capita

income was estimated from United States Census data based on

the child’s residence on each birthday [12,13].

We identified all TCD’s occurring during the study period

from claims files using current procedural terminology (CPT)

codes 76506, 76536, 93886, 93888, 93880, 93882. We validated

receipt of TCD’s by reviewing medical records, demonstrating a

positive predictive value of 100% and a sensitivity of 90.5% for

computerized procedure claims.

We considered two outcomes: (1) having any TCD during

study follow-up; and, (2) annual adherence to recommended

TCD screening defined as having a TCD between one birthday

and the day prior to the next birthday. The cumulative rate of

children who received a TCD per year was calculated using the

Kaplan–Meier method to account for partial year follow-up. Cox

proportional hazards regression was used to examine the associa-

tion of calendar year, sociodemographic factors, outpatient visits

for sickle cell care (identified from ICD-9 codes) and care unre-

lated to SCD with receiving a TCD while accounting for variable

follow-up periods. Every year of follow-up for a child was treated

as a separate observation period since our goal was to assess

children receiving a TCD for each year of follow-up. Robust

standard errors were used to account for children having multiple

years of follow-up [17]. Measures of association were expressed

as hazard ratios and 95% confidence intervals. Final multivariate

models included calendar year, gender, number of siblings, ma-

ternal age, the presence of a chronic health condition in the

mother, maternal education, neighborhood income, distance to a

designated sickle cell center, and SCD and non-SCD outpatient

visits. Potential covariates were selected on the basis of their

potential relationship to the study outcome [18–22].

Study data were obtained from the Bureau of TennCare and

the Tennessee Department of Health. The study was reviewed and

approved by the local Institutional Review Boards and was con-

sidered not to represent human subjects research as individuals

not involved in the conduct of the analysis performed linkages

among data sets and analyses were performed only on data sets

that included no personal identifying information and could not be

linked back to individuals.

RESULTS

We identified 764 children with evidence of SCD at the two

centers based on health care claims, which represented 84.9% of

all children in the state with SCD diagnoses. We obtained records

for 653 children (85.5%), and identified the following hemoglobin

electrophoresis patterns: 315 with HbSS (48.2%), 14 HbSb0

(2.1%), 59 HbSbþ-thalassemia (9.0%), 184 HbSC (28.2%),

8 S/HPFH (1.2%) [23]. Twenty-two children identified from

claims had HbSTrait (3.4%). Fifty one children (7.8%) had no

evidence of SCD; most of these had only a rule out diagnosis or

had another non-sickle hemoglobinopathy. The final study cohort

included 338 children (315 children with HbSS, 14 children with

HbSb0, and 9 children for whom no records were found added to

the cohort through review of hemoglobin identification patterns

and claims).

Among children in the cohort, slightly more than one third had

mothers of age <21 years, 43.2% had mothers with less than a

high school education, and 10.1% had mothers with chronic

health conditions (Table I). Most of the children lived in neigh-

borhoods with low per capita incomes. Less than 20% of the

cohort lived <5 miles from the designated sickle cell center.

More than 90% had at least one sickle cell visit (excluding visits

associated with a TCD) and all had health care encounters during

study follow-up.

Among the 338 children in the cohort, 232 (68.6%) received at

least one TCD during the study period. The annual cumulative

incidence of children who received a TCD per year increased

from 2.5% in 1997 to 68.3% in 2008 (Fig. 1). In multivariate

models, calendar year, maternal education, and number of SCD

outpatient visits were associated with an increased rate of receiv-

ing a TCD (Table II). Children whose mothers had a high school

education or higher were 24% more likely to have an annual TCD

(adjusted hazard ratio 1.24, 95% confidence interval 1.03–1.49).

Children with one or more outpatient SCD visits were more than

twice as likely to receive annual TCD screens as those with no

outpatient SCD visits. In addition, increasing number of SCD

outpatient visits was associated with an increasing likelihood of

receiving an annual TCD screen.

DISCUSSION

Between 1997 and 2008, there was a significant increase in the

rate of TCD adherence among children with SCD in Tennessee.

Yearly cumulative incidence rates of annual TCD increased from

2.5% in 1997 to 68.3% in 2008. More years of maternal education

and receipt of outpatient sickle cell services were associated with

an increased rate of receiving a TCD. Of note, while rates of

screening increased significantly during the study period, 31%

2 Eckrich et al.

Pediatr Blood Cancer DOI 10.1002/pbc

of children in the cohort received no TCD’s at all during study

follow-up.

The finding that increasing utilization of health care is related

to increased adherence to guidelines is consistent with prior pop-

ulation-based studies in children with SCD [13]. Visits to pro-

viders for SCD provide opportunities for education about the risk

for stroke and the importance of preventive care and screening to

reduce stroke risk. Thus, interventions to increase regular sickle

cell healthcare maintenance visits might be an appropriate mech-

anism to improve overall quality of care for children with SCD.

Availability of outpatient care and TCD screening was not likely

to contribute to our findings as all children in the cohort were

enrolled in Tennessee Medicaid. Furthermore, distance from the

child’s residence to their designated sickle cell center was not a

significant predictor of adherence in our analysis. In addition,

SCD visits temporally associated with a TCD were not included

in our definition of outpatient SCD care.

While annual incidence of TCD screening did increase

throughout the study, nearly 1/3 of the children had no TCD at

all during study follow-up. It is well known that patients with

SCD are at risk for non-adherence to guideline recommended care

[15]. In the setting of stroke prevention, commonly cited reasons

for non-adherence to TCD screening include the lack of ordering

by the treating physician and failure to keep appointments for

TCD [24]. Because we determined non-adherence to guidelines as

absence of medical claims for TCDs, it is not possible to deter-

mine the reasons for non-adherence in this study. It is possible

that TCD appointments were not made, or patients were not

seen for their SCD care exclusively within these institutions.

Further study to elucidate process factors related to use of

TCD would be important. Additional studies could target the

impact of TCD adherence on SCD-related stroke, as has been

demonstrated in other preventive interventions for SCD compli-

cations [25].

A prior single institution study indicated a higher rate of TCD

adherence than was observed in our study [11], whereas our data

represent SCD patients from a large geographic area of the state

which includes two sickle cell centers. The current study may

reflect uptake of a guideline in actual community-based experi-

ence. The rate of TCD adherence seen in our study demonstrated

improvement in the decade since the NHLBI guidelines were

released.

Our study has limitations, including the lack of information

from non-publicly insured patients. Although the majority of

sickle cell patients are publicly insured, utilization of medical

care for patients with commercial or private insurance differs

from Medicaid-insured patients, and is not represented in our

cohort [26]. Further, claims databases may result in misclassifica-

tion of children with sickle cell and receipt of TCD’s. However,

we reviewed medical records of all cohort members to confirm

SCD type and receipt of TCD’s to reduce misclassifications.

In conclusion, rates of TCD screening to identify stroke risk in

children with severe SCD increased between the 1997 release of

guidelines recommending annual screening and 2008. Use of

outpatient services for SCD increased the likelihood of receiving

an annual TCD. Despite this encouraging increase, nearly one

TABLE I. Characteristics of Children With Sickle Cell Anemia

and Sickle-b0-Thalassemia During the Study Period

n %

Total 338

Gender

Male 177 52.4

Female 161 47.6

Number of siblings

None 105 31.1

1þ siblings 233 68.9

Maternal agea

< 21 years old 124 37.4

21þ years old 208 62.6

Maternal chronic condition

No 304 89.9

Yes 34 10.1

Maternal education

Less than high school 146 43.2

High school and above 192 56.8

Median neighborhood income ($)

<9,398 65 19.2

9,399–11,528 77 22.8

11,529–14,027 69 20.4

14,028–18,156 64 18.9

18,157þ 63 18.6

Distance to sickle cell disease center

<5 miles 62 18.3

5þ miles 276 81.7

Sickle cell outpatient visits (per year)b

0 26 7.7

1–3 46 13.6

4–7 48 14.2

8þ 218 64.5

Non-sickle cell outpatient visits (per year)b

0 13 3.9

1–3 17 5.0

4–7 38 11.2

8þ 270 79.9

aSix children (1.8%) had missing maternal age in birth certificate file;bOutpatient visits occurring on the same date as a transcranial Doppler

examination were not included.

Fig. 1. Cumulative incidence per year of children adherent

with transcranial Doppler stroke screening recommendations among

children with sickle cell anemia and sickle-b0-thalassemia in Tennes-

see. Vertical bars denote 95% confidence intervals. [Color figure can

be seen in the online version of this article, available at http://

wileyonlinelibrary.com/journal/pbc]

TCD Screening Adherence in Children With SCD 3

Pediatr Blood Cancer DOI 10.1002/pbc

third of children received no screens. These children represent an

important target for further interventions to improve the quality of

care for children with this devastating disease.

ACKNOWLEDGMENT

The authors would like to thank Bertha Davis and Gail

Fortner, RN at St Jude Sickle Cell Center for their gracious

assistance in assembling data for this manuscript. The study

was funded by the Agency for Healthcare Research and Quality

CERT Program (grant # HS10384).

REFERENCES

1. Steinberg MH. Management of sickle cell disease. N Engl J Med 1999;340:1021–1030.

2. Ohene-Frempong K, Weiner SJ, Sleeper LA, et al. Cerebrovascular accidents in sickle cell disease:

Rates and risk factors. Blood 1998;91:288–294.

3. Balkaran B, Char G, Morris JS, et al. Stroke in a cohort of patients with homozygous sickle cell

disease. J Pediatr 1992;120:360–366.

4. Adams RJ. Big strokes in small persons. Arch Neurol 2007;64:1567–1574.

5. Lo W, Zamel K, Ponnappa K, et al. The cost of pediatric stroke care and rehabilitation. Stroke

2008;39:161–165.

6. Adams R, McKie V, Nichols F, et al. The use of transcranial ultrasonography to predict stroke in sickle

cell disease. N Engl J Med 1992;326:605–610.

7. Clinical Alert: Periodic Transfusions Lower Stroke Risk in Children with Sickle Cell Anemia. National

Heart Lung and Blood Institute (NHLBI). 1997. September 18. Available from: URL: http://www.nlm.

nih.gov/databases/alerts/sickle97.html. Accessed on: 13 June 2012.

8. Adams RJ, McKie VC, Hsu L, et al. Prevention of a first stroke by transfusions in children with sickle

cell anemia and abnormal results on transcranial Doppler ultrasonography. N Engl J Med 1998;339:

5–11.

9. The National Institutes of Health HL. Management of Sickle Cell Disease, NIH Publication

No. 02-2117. 2002.

10. Health supervision for children with sickle cell disease. Pediatrics 2002;109:526–535.

11. Raphael JL, Shetty PB, Liu H, et al. A critical assessment of transcranial doppler screening rates in a

large pediatric sickle cell center: Opportunities to improve healthcare quality. Pediatr Blood Cancer

2008;51:647–651.

12. Warren MD, Arbogast PG, Dudley JA, et al. Adherence to prophylactic antibiotic guidelines among

Medicaid infants with sickle cell disease. Arch Pediatr Adolesc Med 2010;164:298–299.

13. Sox CM, Cooper WO, Koepsell TD, et al. Provision of pneumococcal prophylaxis for publicly insured

children with sickle cell disease. JAMA 2003;290:1057–1061.

14. Shankar SM, Arbogast PG, Mitchel E, et al. Impact of proximity to comprehensive sickle cell center on

utilization of healthcare services among children with sickle cell disease. Pediatr Blood Cancer

2008;50:66–71.

15. Shankar SM, Arbogast PG, Mitchel E, et al. Medical care utilization and mortality in sickle cell

disease: A population-based study. Am J Hematol 2005;80:262–270.

16. Ray WA, Griffin MR. Use of Medicaid data for pharmacoepidemiology. Am J Epidemiol

1989;129:837–849.

17. Dupont William D. Statistical modeling for biomedical researchers: A simple introduction to the

analysis of complex data. Cambridge UK: Cambridge University Press; 2008.

18. Barakat LP, Lutz M, Smith-Whitley K, et al. Is treatment adherence associated with better quality of

life in children with sickle cell disease? Qual Life Res 2005;14:407–414.

19. Driscoll MC, Hurlet A, Styles L, et al. Stroke risk in siblings with sickle cell anemia. Blood

2003;101:2401–2404.

TABLE II. Factors Associated With Transcranial Doppler (TCD) Guideline Adherence for Children With Sickle Cell Anemia and

Sickle-b0-Thalassemia in Tennessee

Number of children Cumulative incidence Adjusted hazard ratio 95% Confidence interval

Gender

Male 177 0.31 1.00 Reference

Female 161 0.34 1.11 (0.94, 1.32)

Number of siblings

None 105 0.33 1.00 Reference

1þ siblings 233 0.33 1.08 (0.89, 1.31)

Maternal agea

<21 years old 124 0.33 1.00 Reference

21þ years old 208 0.32 0.89 (0.73, 1.08)

Maternal chronic condition

No 304 0.34 1.00 Reference

Yes 34 0.34 1.07 (0.81, 1.42)

Maternal education

Less than high school 146 0.30 1.00 Reference

High school or above 192 0.35 1.24 (1.03, 1.49)

Median neighborhood income ($)

<9,399 65 0.27 1.00 Reference

9,399–11,528 77 0.33 1.21 (0.92, 1.59)

11,529–14,027 69 0.32 1.04 (0.78, 1.38)

14,028–18,156 64 0.34 1.13 (0.85, 1.51)

18,157þ 63 0.36 1.03 (0.77, 1.36)

Distance to SCD center

<5 miles 62 0.29 1.00 Reference

5þ miles 276 0.34 1.11 (0.94, 1.32)

Sickle cell outpatient visits

0 26 0.17 1.00 Reference

1–3 46 0.32 2.29 (1.70, 3.09)

4–7 48 0.37 2.78 (2.02, 3.84)

8þ 218 0.45 2.80 (2.05, 3.82)

Non-sickle cell outpatient visits

0 13 0.25 1.00 Reference

1–3 17 0.32 1.05 (0.79, 1.40)

4–7 38 0.34 1.05 (0.79, 1.40)

8þ 270 0.37 1.17 (0.87, 1.58)

aSix children (1.8%) had missing maternal age in birth certificate file.

4 Eckrich et al.

Pediatr Blood Cancer DOI 10.1002/pbc

20. Haque A, Telfair J. Socioeconomic distress and health status: The urban-rural dichotomy of services

utilization for people with sickle cell disorder in North Carolina. J Rural Health 2000;16:43–55.

21. Elliott V, Morgan S, Day S, et al. Parental health beliefs and compliance with prophylactic penicillin

administration in children with sickle cell disease. J Pediatr Hematol Oncol 2001;23:112–116.

22. Witherspoon D, Drotar D. Correlates of adherence to prophylactic penicillin therapy in children with

sickle cell disease. Children’s Health Care 2010;35:281–296.

23. Lorey FW, Arnopp J, Cunningham GC. Distribution of hemoglobinopathy variants by ethnicity in a

multiethnic state. Genet Epidemiol 1996;13:501–512.

24. Fullerton HJ, Gardner M, Adams RJ, et al. Obstacles to primary stroke prevention in children with

sickle cell disease. Neurology 2006;67:1098–1099.

25. Halasa NB, Shankar SM, Talbot TR, et al. Incidence of invasive pneumococcal disease among

individuals with sickle cell disease before and after the introduction of the pneumococcal conjugate

vaccine. Clin Infect Dis 2007;44:1428–1433.

26. Mvundura M, Amendah D, Kavanagh PL, et al. Health care utilization and expenditures for privately

and publicly insured children with sickle cell disease in the United States. Pediatr Blood Cancer

2009;53:642–646.

TCD Screening Adherence in Children With SCD 5

Pediatr Blood Cancer DOI 10.1002/pbc