Analysis of Perinatal Mortality and Its Components: Time for a Change?

American Journal of EPIDEMIOLOGY Volume 156 Number 6 September 15, 2002 Copyright © 2002 by The Johns Hopkins Bloomberg School of Public Health Sponsored by the Society for Epidemiologic Research Published by Oxford University Press COMMENTARY Analysis of Perinatal Mortality and Its Components: Time for a Change? 1 Department of Pediatrics, McGill University Faculty of Medicine, Montreal, Quebec, Canada. of Epidemiology and Biostatistics, McGill University Faculty of Medicine, Montreal, Quebec, Canada. 3 Health Surveillance and Epidemiology Division, Centre for Healthy Human Development, Health Canada, Ottawa, Ontario, Canada. 4 Perinatal Epidemiology Research Unit, Departments of Pediatrics and of Obstetrics and Gynecology, Dalhousie University, Halifax, Nova Scotia, Canada. 2 Department Received for publication June 5, 2002; accepted for publication July 9, 2002. Since the midtwentieth century, stillbirths (late fetal deaths) and early neonatal deaths have often been combined into a single category of “perinatal” deaths. In the past, such a combination was justified by the fact that asphyxia was a common cause of death during labor (intrapartum stillbirth) and shortly after birth and by geographic and temporal differences in classification of livebirths versus stillbirths. In more recent years, however, the etiologic determinants have diverged sharply, with many fewer early neonatal deaths caused by asphyxia and relatively many more caused by congenital anomalies. Moreover, the increasingly common stratification of pregnancy outcome measures by gestational age or birth weight leads to the use of an inappropriate denominator (total livebirths plus stillbirths within each gestational age or birth weight category) for denoting risk for the stillbirth component, because all unborn fetuses (including the majority of those not born within the specified gestational age or birth weight range) are at risk of being stillborn in that range. The authors suggest that, whenever possible, stillbirths and early neonatal deaths should be reported separately, with gestational age-specific risks of stillbirth based on all fetuses at risk, and that antepartum and intrapartum stillbirths be reported separately. Am J Epidemiol 2002;156:493–7. abnormalities; fetal death; infant, newborn RATIONALE FOR COMBINING STILLBIRTHS AND EARLY NEONATAL DEATHS Anyone who has witnessed a large number of deliveries is aware of the occasional difficulty in distinguishing stillborn infants from those liveborn infants who are extremely preterm (near the limit of viability), asphyxiated, or neurologically depressed and who die soon after birth. Because of this difficulty, the World Health Organization developed and promulgated what is currently the universally adopted definition of a livebirth: “. . . the complete expulsion or extrac- Correspondence to Dr. Michael S. Kramer, Department of Epidemiology and Biostatistics, McGill University, 2300 Tupper Street, Room F-265, Montreal, Quebec, H3H 1P3, Canada (e-mail: ). 493 Michael S. Kramer1,2, Shiliang Liu3, Zhongcheng Luo1,2, Hongbo Yuan1,2, Robert W. Platt1,2, and K. S. Joseph4 for the Fetal and Infant Health Study Group of the Canadian Perinatal Surveillance System 494 Kramer et al. In other settings, cultural, religious, and economic factors may exert an opposite influence. Some families may prefer to register, name, and bury a stillborn or a liveborn infant near the borderline of viability who later dies as a way of “healing” their loss, and in some jurisdictions, maternity leave benefits may depend on registration. ETIOLOGIC DIFFERENCES BETWEEN STILLBIRTHS AND EARLY NEONATAL DEATHS Stillbirths and early neonatal deaths differ substantially with respect to their principal causes. As discussed in the previous section, this may have been less true several decades ago, when many more term infants died of asphyxia during labor or shortly after birth (4). Even today, conditions such as abruptio placentae and fetal growth restriction can cause either stillbirth or early neonatal death. However, in most developed countries at the present time, the etiologic differences are far more striking than the similarities. Fetuses with congenital anomalies incompatible with fetal growth and development are often aborted early in gestation (first trimester) (10, 11). Other anomalies, however, do not become life threatening until birth. For example, an in utero existence protects fetuses with severe congenital heart defects that lead to physiologic compromise only after birth and the relative shift from pulmonary to systemic blood flow. Similarly, congenital abnormalities of the gastrointestinal tract (e.g., tracheoesophageal fistula) or lungs (e.g., pulmonary hypoplasia) only become life threatening after the enteral nutrition and respiratory function, respectively, required by extrauterine life. Unpublished data from the linked birth, stillbirth, and infant death file of the Canadian Perinatal Surveillance System support these physiologic arguments. Among singleton births at 25 weeks or more in Canada in 1991–1997, for example, the risk of stillbirth due to congenital anomalies was only 3.8 per 10,000 total births, whereas the risk of early neonatal death due to congenital anomalies was 9.3 per 10,000 livebirths. The corresponding risks for death due to asphyxia were reversed: 17.1 per 10,000 for stillbirths and 2.9 per 10,000 for early neonatal deaths. Congenital anomalies caused 45.8 percent of early neonatal deaths versus 9.4 percent of stillbirths, while the reverse trend was observed for deaths due to asphyxia: 14.4 percent of early neonatal deaths versus 42.5 percent of stillbirths. Moreover, etiologic determinants differ widely according to whether the stillbirth occurs antepartum or intrapartum, that is, before or during labor. Antepartum stillbirths often occur with severe maternal, placental, or fetal abnormalities, including umbilical cord complications (12–14), preeclampsia (14, 15), intrauterine growth restriction (12, 13, 16– 19), abruptio placentae (14, 20), and infection (21, 22). Maternal smoking, advanced maternal age, grand multiparity, and obesity are also widely recognized determinants of antepartum stillbirth (12–16, 23–25), while one fourth occur without known cause (13, 15). Intrapartum fetal deaths are usually the result of fetal distress and/or obstructed labor and often reflect poor access to or quality of clinical care during delivery (26). In developed countries, the vast majority (85–90 percent) of stillbirths occur antepartum (12, 26), whereas this proportion is Am J Epidemiol Vol. 156, No. 6, 2002 tion from its mother of a product of conception, irrespective of the duration of pregnancy, which, after such separation, breathes or shows any evidence of life, such as beating of the heart, pulsation of the umbilical cord, or definite movement of voluntary muscles, whether or not the umbilical cord has been cut, or the pla (...truncated)