We defined a population for study using two local health districts: the Dniprovski (“Left Bank”) region of Kyiv and the city of Dniprodzerzhinsk, about 400 km southeast of Kyiv. All women who were identified as pregnant at local polyclinics in those districts and whose last menstrual period (LMP) occurred between December 25, 1992 and July 23, 1994, comprised the population. Lists of every identified pregnancy were prepared by the polyclinic sites. All women residing in the health districts were required to make their first prenatal visits there, even if they subsequently obtained care elsewhere, so those listings should have been virtually complete. For each pregnancy we collected dates of first prenatal visit, mother's birth, LMP, decision on termination (allowed on demand in the first 12 weeks), gestational age at the end of pregnancy, and vital status at delivery. World Health Organization definitions of live birth and fetal death4 were used. All research procedures were approved by institutional review boards in Ukraine and the United States.
The choice of sites was motivated by a collaboration that allowed us to collect additional data on pregnancies of a subset of women in the population. The multinational European Longitudinal Study of Pregnancy and Childhood5 sought to identify factors that affect human growth and development by identifying cohorts of pregnant women and observing their children over time. Each participating country defined their cohort of pregnancies by specifying a geographic area and recruiting women there who had LMPs in a specified time period. Children of Ukraine, the Ukrainian component of the multinational study, involved six different sites; our selection of two of them was based on their size and accessibility. Both areas are highly industrial, although demographic, economic, and ecologic conditions vary. In 1993, each had a population of approximately 250,000.
Women at the sites were eligible for the Children of Ukraine cohort if they had no plan to terminate the pregnancy, official permanent residency, were receiving care in the local facilities and not planning to move, and were fluent in Ukrainian. Eligible women were asked to participate in Children of Ukraine study at their first prenatal visit or as soon as possible thereafter. Data about cohort pregnancies were collected from medical records and self-completed questionnaires.
A delivery was classified as preterm if it occurred before 37 completed weeks after LMP. We conducted extensive examinations to evaluate the credibility of gestational age estimates, as described elsewhere.6 We used multiple sources of data to identify possible errors or outliers, which were examined individually and assigned gestational ages using a systematic protocol. We could not routinely compare the estimates with ultrasound because only very specialized diagnostic centers had access to that equipment.
The denominator for preterm birth rate is live singleton births of 20 or more weeks' gestation. Multiple births and stillbirths were excluded for comparability with rates from other countries. We computed rates for the entire population at the sites and also for the Children of Ukraine cohort and compared them with rates in Europe, Canada, and the United States. Children of Ukraine births were also classified by conditions surrounding onset of labor, including whether it was spontaneous or medically indicated because of complications.
A total of 17,137 pregnancies with LMPs between December 25, 1992 and July 23, 1994, were identified at the study sites (Table 1). Sixty percent were voluntarily terminated, all but 1% of them by the end of the 12th week of gestation. Preterm birth rates among the continuing pregnancies of 20 or more weeks' gestation were 6.6% among live-born singletons and 8.8% among all singletons.
Additional information about labor and delivery was collected from the subgroup that participated in the Children of Ukraine study. Among 6774 continued pregnancies, 67% were eligible for Children of Ukraine (Table 1). The primary reasons for ineligibility were that women were in temporary residences, receiving care elsewhere, or moving. Among eligible women, 71% agreed to participate and among those, preterm rates were lower than in the population as a whole, 5.2% for live-born infants and 5.9% for all singletons. Among preterm deliveries, only 12% were medically indicated (Table 2), the rest were idiopathic. Preterm premature rupture of membranes (PROM) occurred in 45%, or 2.2% of all births with known onset of labor. The cesarean rate was 8% overall and 12% among preterm births. Nearly half the preterm births were of low birth weight.
The lower preterm birth rate in the Children of Ukraine sample reflects some maternal characteristics. For example, 29% of the total group were less than 20 years of age compared with 19% of the Children of Ukraine sample. Fifty-five percent of the total group obtained prenatal care by the end of the first trimester, whereas 65% of the Children of Ukraine sample did so. The total group included many potentially high-risk women, such as transients and refugees, as well as women who were first seen when an emergency occurred and arrived too late to enroll in the study.
We compared Ukrainian preterm birth rates with those in other countries for any period from 1985 to 1994. (Official Ukrainian statistics are not comparable because they were not based on World Health Organization criteria.) Studies in Sweden, France, Belgium, and the Czech Republic reported preterm birth rates from 4.0% to 5.4% among live-born singleton births.7–9 In Canada, the preterm birth rate from 1992 to 1993 was 5.9%.10 None of those rates are as high as the rate for all enumerated pregnancies in the two sites (6.6%). The preterm birth rate for United States whites (used for comparability with Ukrainian ethnicity) was higher, 8.4%, in 1993.11 Data were not available to compare Ukrainian mothers with those in other countries.
Preterm birth rates for live-born infants depend on fetal mortality rates and maternal characteristics. Infants stillborn before 37 weeks' gestation are not included in the numerator or denominator of live-born preterm birth rate. That makes the numerator and denominator smaller than if the infant had been born alive, but the proportion of decrease in the numerator is relatively greater than in the denominator. Thus, the net effect of preterm stillbirths is to lower the live-born preterm birth rate relative to what it would have been if they had lived. The comparison of United States and Ukraine pregnancies illustrates the point. Table 1 shows 129 births at 20–27 weeks, with a fetal mortality rate of 91% in Ukraine. In the United States in 1992, the fetal mortality rate in that group was 34%.6 If Ukraine had had the same rate, there would have been 85 live preterm births at 20–27 weeks—an increase of 74 live preterm births in the study population—and a preterm birth rate of 7.7% instead of 6.6%.
Obstetric practice in Ukraine affects which births survive to enter the numerator of the preterm birth rate. The standard of care in Ukraine required resuscitation only at 28 weeks or later, because the very costly technology required to resuscitate and maintain infants born earlier is not available. Similarly, the proportion of medically indicated preterm deliveries influences the preterm birth rate by its effect on the fetal mortality rate. A high rate of early intervention in pathologic pregnancies might increase the preterm birth rate by shifting some births into the preterm range and by saving some infants who would otherwise be stillborn. The association between increased medical interventions and increased preterm birth rate has been noted by others.1 In Ukraine, the rate of 12% for medical interventions in preterm births is lower than rates of 19–29% in North America,12 reflecting the severe shortages of equipment and supplies. The low rate of cesarean deliveries (Table 2) is an example of the minimal use of procedures.
The comparison between preterm birth rates in Ukraine and other countries also reflects differences in measurement of gestational age. When ultrasound is used to estimate gestation, rates are higher than when LMP is used.13 Ultrasound was generally not used in Ukraine.
The termination rate in the two Ukrainian sites was 4.6 times the estimated United States rate14 because voluntary termination is the usual form of contraception in Ukraine. One cannot estimate its impact on the preterm birth rate, but it is reasonable to suspect that marginal pregnancies that would have gone on to be preterm births might be terminated more often with a liberal abortion policy. Conversely, repeated terminations using extensive cervical dilatation might increase the risk of preterm birth.15
The rate of preterm delivery in two urban areas of Ukraine was 6.6% of all singleton live births. In a subgroup of Ukrainian women at apparently lower risk, the rate was 5.2%. Those rates should be interpreted with caution. The adverse conditions in Ukraine have been well documented, and we hypothesized that Ukrainian preterm birth rates would be high, which was true when the rate for all enumerated pregnancies was compared with several European countries and Canada, but not when it was compared with the rate in the United States, a country of wealth and stability. Is reproductive health in Ukraine better than in the United States, or is the preterm birth rate an indicator of pregnancy outcome that does not necessarily show the true risk associated with early birth?
Rates of preterm delivery based on live singleton births are influenced by maternal characteristics. A low preterm birth rate might signify a low-risk population and a high preterm birth rate might be a sign of maternal characteristics that predispose to preterm birth. However, the preterm birth rate is also influenced by complex factors that determine which births are classified as preterm. Among those factors are systematic practices that alter fetal death rates, such as voluntary termination, medical interventions, and resuscitation policies. When those are done before 37 weeks, and an infant dies who otherwise would have lived, a live preterm birth is lost and the rate is decreased. Conversely, when an infant's life is saved, a preterm birth is gained and the rate increases. Thus, a low preterm birth rate could be a sign of high fetal mortality and poor reproductive health, while a high rate could be a sign of advanced technology and life-saving techniques.
The preterm birth rate might be paradoxical and should be taken as only one statistic among many. Fetal and infant morbidity and mortality are better indicators of reproductive outcome. They are not easy to measure, but they are not fraught with the difficulties of interpreting preterm birth.
1. Kramer MS, Platt R, Yang H, Joseph KS, Wen SW, Morin L, et al. Secular trends in preterm birth: A hospital-based cohort study. JAMA 1998;280:1849–54.
2. Hedegaard M, Henriksen TB, Sabroe S, Secher NJ. Psychological distress in pregnancy and preterm delivery. BMJ 1993;307:234–9.
3. United Nations Children's Fund International Child Development Centre. Crisis in mortality, health and nutrition. Regional monitoring report number 2. Florence, Italy: United Nations Children's Fund International Child Development Centre, 1995.
4. World Health Organization. International classification of diseases; manual of The International Statistical Classification of Diseases, Injuries, and Causes of Death. Ninth revision. Geneva, Switzerland: World Health Organization, 1977.
5. European longitudinal study of pregnancy and childhood (ELSPAC). Paediatr Perinat Epidemiol 1989;3:460–9.
6. Little RE, Monaghan SC, Gladen BC, Shkyryak-Nyzhnyk AZ, Wilcox AJ. The outcome of 17,137 pregnancies in two urban areas of Ukraine. Am J Public Health 1999;89:1832–6.
7. Buekens P, Wilcox AJ, Kiely J, Masuy-Stroobant G. Birthweight, preterm births and neonatal mortality in Belgium and the United States. Paediatr Perinat Epidemiol 1995;9:273–80.
8. Breart G, Blondel B, Tuppin P, Grandjean H, Kaminski M. Did preterm deliveries continue to decrease in France in the 1980s? Paediatr Perinat Epidemiol 1995;9:296–306.
9. Koupilova I, Vagero D, Leon DA, Pikhart H, Prikazsky V, Holcik J, et al. Social variation in size at birth and preterm delivery in the Czech Republic and Sweden, 1989–91. Paediatr Perinat Epidemiol 1998;12:7–24.
10. Joseph KS, Kramer MS, Marcoux S, Ohlsson A, Wen SW, Allen A, et al. Determinants of preterm birth rates in Canada from 1981 through 1983 and from 1992 through 1994. N Engl J Med 1998;339:1434–9.
11. National Center for Health Statistics. Vital statistics of the United States. 1993. Volume 1: natality. Washington DC: United States Public Health Service, 1996.
12. Savitz DA, Blackmore CA, Thorp JM. Epidemiologic characteristics of preterm delivery: Etiologic heterogeneity. Am J Obstet Gynecol 1991;164:467–71.
13. Kramer MS, McLean FH, Boyd ME, Usher RH. The validity of gestational age estimation by menstrual dating in term, preterm, and postterm gestations. JAMA 1988;260:3306–8.
14. United States Bureau of the Census. Statistical abstract of the United States: 1997. 117th ed. Washington DC: United States Bureau of the Census, 1997.
15. Hogue CJ, Cates W Jr, Tietze C. The effects of induced abortion on subsequent reproduction. Epidemiol Rev 1982;4:66–94.