Acta Ped ia t r Scand 79: 47-5 1, 1990

Birthweight Specific Perinatal Mortality in Greece

CHRYSSA TZOUMAKA-BAKOULA,’ VASSO LEKEA-KARANIKA,’ N. S. MATSANIOTIS,’ B. J . McCARTHY* and JEAN GOLDING3 From the ’First Department of Paediatrics, Athens University, Hospital for Sick Children “Aghia Sophia”, Athens, Greece, the ’Centers for Disease Control, Atlanta, USA and the ’Institute of Child Health, Bristol, Great Britain

ABSTRACT. Bakoula, C., Lekea, V., Matsaniotis, N., McCarthy, B. and Golding 1. (1st Department of Paediatrics, Athens University, Hospital for Sick Children “Aghia Sophia”, Athens, Greece). Birthweight specific perinatal mortality in Greece. Ada Paediatr Scand 79:47, 1990.

Data from the Greek Perinatal Study in April 1983 revealed an excessively high perinatal mortality rate of 21.6 per loo0 total births among singletons despite a low birthweight rate of only 4.5%. Comparison of perinatal mortality rates with Danish mortality rates in 1983, revealed the Greek rates to be three times higher than those in Denmark. When divided by time of death, the Greek stillbirth rates were two times higher and the eady neonatal mor- tality rates were four times higher than the corresponding Danish rates. Subdivision of the Greek perinatal deaths using the Wigglesworth cladcat ion showed that the biggest group (40%) consisted of deaths associated with intraparturn asphyxia. The incidence of such deaths was 10 times higher than that found in Denmark. We conclude that in reducing the excessively high perinatal mortality rate in Greece special attention should be made to im- prove intraparturn and resuscitation techniques. Key words: perinatal rnortalily, birth- weight, asphyxia, intraparturn care.

In studies of perinatal mortality it is of major importance to assess the way in which the peri- natal deaths vary with birthweight. Thus it will be possible to a) define the birthweight groups in which the mortality rates are particularly high, especially in relation to other coun- tries, and b) identify areas in which there appears to be under-reporting of perinatal deaths.

In a study of Swedish birthweight specific mortality rates the overall mortality was low, compared with that of England and Wales. Alberman (1) showed that the birthweight specific mortality rates were identical within each 500 g group up to 2500 g, but diverged thereafter. The implication was that in England and Wales the difference in mortality was due in part to a difference in birthweight distribution between the two countries but also to a difference in management of the larger infants. Under-reporting was found in the USA where a comparison of mortality rates in Georgia with those in New York resulted in con- fusion; the mortality rate among the very low birthweight infants in Georgia appeared to be exceptionally low. An important study by McCarthy (2) was then made to identify neonatal deaths that had not been registered.

Tzoumaka-Bakoula (3) showed that the perinatal morality rate in Greece in 1983 was ex- ceptionally high in comparison with that of the rest of Europe. In an effort to understand the reasons for such high rates we have compared the birthweight specific mortality rates in Greece with those of the Nordic countries in 1983. These were the lowest reported rates in Europe and may be regarded as the “gold standard” for other nations.

MATERIAL AND METHODS In this study, data were collected on all livebirths and stillbirths of 500 g or more which occurred throughout Greece during 30 consecutive days in April 1983. Questionnaires were filled out by obste- tricians and/or midwives at the time of delivery. The neonates were followed up to the age of seven days

48 C . Tzoumaka-Bakoula et al. Acta Paediatr Scand 79

and all deaths which had occurred by that time were identified. Information collected included the in- fant’s birthweight.

The death records were compared with death notifications registered in both the peripheral and cen- tral registration offices. These linked data constituted the data file for analyses. Further details of the methodology and contents of the questionnaires are published elsewhere (3, 4). In our analysis, only singleton births have been considered. Information on birthweight was missing in 10 infants. The major- ity of these had died (three stillbirths, five early neonatal deaths). These 10 cases have been omitted from the present analyses.

RESULTS

In Greece, only 4.5% of the singleton infants weighed less than 2500 g at birth (Table 1). Nevertheless, 66.1% of all stillborn infants, 59.6% of all who died in the early neonatal period, hence 62.8% of those who died in the perinatal period weighed less than 2500 g.

The overall singleton birthweight specific perinatal mortality rates for Greece in 1983 were virtually identical to those in the 1970 British National Birth Survey by Chamberlain et al. (5). The two surveys both took place in April and the overall perinatal mortality rates were almost identical (Table 2). Comparison of the birthweight specific mortality rates

Table 1. Birthweights of infants who died perinatally and survivors in Greece (singletons only)

Early Birthweight Stillbirths neonatal deaths Survivors Total births (g) n (%) n (Yo) n (Yo) n (YO)

< I 000 33 (28.7) 1 OW1 499 23 (20.0) 1500-1 999 9 (7.8) 2 o w 2 499 11 (9.6) 2 5 W 2 999 13 (11.3) 3 oo(F-3 499 17 (14.8)

4 o w 499 2(1.7) 4 5oo+ 2(1.7)

All known 115 (100)

3 5 w 3 999 5 (4.3)

24 (20.3) 24 (20.3) 11 (9.3) 11 (9.3) 16 (13.6) 22 (18.6) 6 (5.1) 3 (2.5)

118(100)

l(0.8)

l(O.0) 28 (0.3) 71 (0.7)

248 (2.3) 1 363 (12.8) 4 159 (39.3) 3 541 (33.4) 1 012 (9.6)

170 (1.6)

10 593 (loo)

58 (0.5) 75 (0.7) 91 (0.8)

270 (2.5) 1 392 (12.9) 4 198(38.8) 3 552 (32.8) 1 017 (9.4)

173 (1.6)

10 826 (100)

Table 2. Perinatal mortality rate per I000 total births among singletons in Greece, 1983, Britain (GB) , 1970 and Denmark, 1983 95% confidence intervals in brackets

Greece 1983 GB 1970 Denmark 1983 Birthweight (d (a) (b) (am) (c) (dc)

< I 000 983 (949.7-1 016.3) 1 00G1 499 627 (517.5-736.5)

2 o w 2 499 81 (48.5-113.5) 1 500-1 999 220 (134.9-305.1)

2 5 w 2 999 21 (13.5-28.5) 3 OW3 499 3 500+ 4 (2.2-5.8)

9 (6.1-1 1.9)

All known birthweights 21.6 (18.9-24.3)

93 1 1 613 1 26 1 1 69 1 18 1 7 1 5 1

21.4 1

- 210 117 28 8 3 2 *

* Not comparable rates as < I 000 g omitted.

Acta Paediatr Scand 79 Perinatal mortality in Greece 49

within each 500 g-group showed a very similar pattern. The only significant difference was found in the group of infants weighing under 1000 g. Here, although the study was made be- fore the era of neonatal intensive care units, the British perinatal mortality rate for infants weighing less than 1 OOO g was significantly lower than the Greek rate in 1983. No data were available on this very low birthweight group from the Scandinavian countries, however.

In Table 2 we present data from Denmark for 1983. Data from Sweden were virtually iden- tical (6). The mortality rates among all birthweight groups were markedly higher in Greece and about three times higher than those found in Scandinavia.

The Greek birthweight specific mortality rates are compared with those in Denmark sepa- rately for the late fetal deaths and the early neonatal deaths in Table 3. Although the stillbirth rate was twice as high in Greece, the difference was only statistically significant for

Table 3. Birthweight specific stillbirth rates (per 1000 total births) and early neonatal mortality rates (per 1000 livebirths) in Greece and Denmark, 1983 In brackets are the 95% confidence intervals for the Greek rates

Stillbirth rates Early neonatal death rates

Birthweight Greece Denmark Greece Denmark groups (g) (a) (b) (db) (a) (b) (&)

1000-1 499 307(202.&411.4) 115 2 462 (326.5-597.5) 111 4 1 500-1 999 99 (37.6-160.4) 81 1 134 (60.3-207.7) 40 3

3 000-3 499 4 (2.1-5.9) 2 2 5 (2.9-7.1) - 5 3 500+ 2 (0.7-3.3) 2 1 2 (0.7-3.3) - 2

2 000-2 499 41 (17.444.6) 21 2 42 (17.6-66.4) 7 6 2 500-2 999 9 (4.0-14.0) 6 1.5 12 (6.3-17.7) 2 6

Table 4. Singleton perinatal deaths classified according to Wigglesworth 5 recommendations; Greece 1983. Comparison of cause specific perinatal mortality rates with Denmark 1985

Wigglesworth group

Other Birthweight Antepartum Congenital Intrapartum specific (g) stillbirth defect Immaturity asphyxia cause All

<lo00 8 1 000-1 499 4 1500-1 999 5 2 000-2 499 4 2 50&2 999 4 3 000-3 499 1 3 500+ 2

All known 28

Mortality ratea Greece 1983 (a) 2.6 Denmark 1985 (b) 3.5

(a/b) 1

(12.0%) 48 (20.6%)

4.4 1.6

3

23 22 4 3

52 (22.3%)

4.8 1.7

3

21 11 9

10 12 22 8

93 (39.9%)

8.6 0.9

10

- 5 3 4

12 (5.2%)

1 . 1 0.4

3

57 47 20 22 29 39 19

233 (100%)

21.5 8.1

3 ~~

a 95% confidence intervals in brackets.

4-908301

50 C . Tzoumaka-Bakoula et al. Acta Paediatr Scand 79

weights t l 5 0 0 g. Regarding early neonatal deaths however, the Greek mortality rate for each birthweight group was substantially higher than the corresponding rate in Denmark, the difference being especially marked (5-6-fold) for birthweights in the range 2 000-3 499 g.

Classification of deaths In Table 4 the Greek perinatal deaths are classified according to the criteria given by Wigglesworth (7) and compared with Danish data from 1985 (8), as no such data from 1983 were available. It can be seen that although the Greek rate of fetal deaths does not differ from the Danish rate, in all other groups the mortality rate is significantly higher. With re- gard to malformation, immaturity and miscellaneous causes of death, the rates are three times as high but for intrapartum asphyxia the difference is ten-fold. Although the Danish data are classified under the broader heading “Intrapartum event”, and thus not directly comparable, this finding cannot be ignored.

Wigglesworth (7) has given criteria for preventive intervention, such as antenatal care for stillbirths, family planning to reduce congenital defects, neonatal care for the preterm in- fants and obstetric and neonatal care for asphyxia.

DISCUSSION

Our data on births in Greece are collected during one month only. It is feasible that by col- lecting data during a 12-month period, with a larger number of births and reduced bias as- sociated with seasonal variations in birth and death rates, we might have had somewhat dif- ferent results. Nevertheless, the differences in death rates compared to Denmark are so large and dramatic that it is most unlikely that the conclusions would have been changed.

The Greek Perinatal Study provides a unique opportunity for Greek health professionals to assess the factors associated with the high perinatal death rates. In spite of the relatively few low birthweight deliveries in Greece the perinatal mortality rate is exceptionally high and is similar to that of Great Britain thirteen years ago. Comparison of the Greek birth- weight specific mortality rates with the 1983 Nordic rates showed that the stillbirth rates were twice as high but that neonatal mortality rates were even higher. The Wigglesworth criteria are most useful for the identification of causes of perinatal death. We have for in- stance shown that 40% of the deaths were associated with intrapartum asphyxia. The mor- tality rate in this group was thus 8.6, some ten times higher than that found in Denmark two years later.

It is not immediately obvious how obstetric care could be improved in order to reduce deaths associated with intrapartum asphyxia. It is dangerous to suggest more obstetric inter- vention during labour since Greece has already one of the highest intervention rates in Europe. Indeed 2% only of Greek mothers in this survey did not have an obstetric interven- tion of some kind (C. Bakoula, unpublished). The caesarean section rate in Greece was 12.6%, and according to WHO (9) it should not be any higher. On the contrary it is possible that the aggressive obstetric care that is being practised. in Greece combined with inad- equate resucitation techniques immediately after birth may have contributed to the high rate of mortality due to intrapartum asphyxia. This possibility though, should be the focus of a further analysis of our data.

REFERENCES 1. Alberman E. Prospects for better perinatal health. Lancet 1980; I: 189-92. 2. McCarthy BJ, Terry J, Rochat RW, Quave S, Tyler CW Jr. The underregistration of neonatal deaths:

3. Tzoumaka-Bakoula C. The Greek National Perinatal Survey: I. Design, methodology, case ascertain- Georgia 197477. Am J Public Health 1980; 70: 977-82.

ment. Paediatr Perinatal Epidemiol 1987; 1: 4>51.

Acta Paediatr Scand 79 Perinatul mortality in Greece 51

4. Tzoumaka-Bakoula C. Assessment of the incidence of perinatal deaths in Greece using two separate sources. In: Golding J , ed. The third report of the WHO study of social and biological effects on perinatal mortality. Bristol: University of Bristol Press, 1987; in press.

5 . Chamberlain R , Chamberlain G , Howlett B, Claireaux A . British births 1970, Vol. 1. London: Heinemann Medical, 1975.

6. Nordisk Medicinal-Statistisk Kommittk: Births in Nordic countries: Registration of the outcome of pregnancy 1979-1983. Reykjavik, NOMESKO, 1987: 25-35.

7. Wigglesworth J. Monitoring perinatal mortality: a pathophysiological approach. Lancet 1980; 11: 684-86.

8. Sundhedsstyrelsen. Medicinsk fadsels- og misdannelsesstatistik 1986. Vitalstatistik 1988; 1 : 23. 9. World Health Organisation. Having a baby in Europe: Report on a study. Public Health in Europe

1985: 26: 1-157.

Submitted Aug. 1 , 1988. Accepted April 7, 1989

(C. B.) 1st Department of Paediatrics Hospital for Sick Children “Agia Sophia” Athens 11527 Greece