OBJECTIVE: Our aim was to document the differential neonatal morbidity and intrapartum and neonatal mortality of subgroups of preterm delivery.
METHODS: This analysis included 38,319 singleton pregnancies, of which 3,304 (8.6%) were preterm deliveries (less than 37 completed weeks) enrolled in the World Health Organization randomized trial of a new antenatal care model. We classified them as preterm deliveries after spontaneous initiation of labor, either with or without maternal obstetric and medical complications; preterm deliveries after prelabor spontaneous rupture of amniotic membranes (PROM), either with or without obstetric and medical complications; and medically indicated preterm deliveries with maternal obstetric and medical complications. Severe neonatal morbidity and neonatal mortality were the primary outcomes.
RESULTS: Fifty-six percent of all preterm deliveries were spontaneous, without maternal complications. Small for gestational age was increased only among the medically indicated preterm delivery group (22.3%). Very early preterm delivery (less than 32 weeks of gestation) was highest among PROM with complications (37%). For intrapartum fetal death and neonatal death, after adjusting by gestational age and other confounding variables, we found that the obstetric and medical complications preceding preterm delivery predicted the different risk levels. Conversely, for severe neonatal morbidity the clinical presentation, ie, PROM or medically indicated, predicted the increased risk.
CONCLUSION: There are differential neonatal outcomes among preterm deliveries according to clinical presentation, pregnancy complications, gestational age at delivery, and its association with small for gestational age. This syndromic nature of the condition should be considered if preterm delivery is to be fully understood and thus reduced.
LEVEL OF EVIDENCE: III
Subgroups of preterm deliveries have different perinatal outcomes according to their clinical presentation and obstetric and medical complications.
From the United Nations Development Programme/United Nations Population Fund/World Health Organization/World Bank Special Program of Research, Development, and Research Training in Human Reproduction, Department of Reproductive Health and Research, World Health Organization, Geneva, Switzerland; Centro Rosarino de Estudios Perinatales, Rosario, Argentina; Khon Kaen University, Khon Kaen, Thailand; Norwegian Institute of Public Health, Oslo, Norway; National Guard King Khalid Hospital, Jeddah, Saudi Arabia; Hospital Gineco-Obstétrico “América Arias,” Havana, Cuba; Ministry of Health, Riyadh, Saudi Arabia; and Department of Pediatrics, McGill University Faculty of Medicine, Montreal, Quebec, Canada.
Received November 21, 2003. Received in revised form March 4, 2004. Accepted March 18, 2004.
* For a list of members of the World Health Organization Antenatal Care Trial Research Group, see the Appendix.
Supported by the United Nations Development Programme/United Nations Population Fund/World Health Organization/World Bank Special Program of Research, Development, and Research Training in Human Reproduction of World Health Organization. Additional support was provided by the Municipal Government, City of Rosario, Argentina; Ministry of Health, Cuba; National Institute of Public Health, Mexico; The Population Council—Regional Office for Latin America and the Caribbean; Ministry of Health, Saudi Arabia; Swedish Agency for Research Cooperation with Developing Countries, Swedish International Development Agency; Ministry of Public Health and Faculty of Medicine, Khon Kaen University, Thailand; Department for International Development of the United Kingdom; Mother Care—John Snow Inc.; National Institute for Child Health & Human Development, National Institutes of Health, United States; and the World Bank.
Address reprint requests to: Dr. José Villar, UNDP/UNFPA/WHO/World Bank Special Program of Research, Development and Research Training in Human Reproduction, Department of Reproductive Health and Research, World Health Organization, 1211 Geneva 27, Switzerland; e-mail: firstname.lastname@example.org.
Approximately 13 million preterm deliveries occur per year worldwide.1 Preterm delivery is the most important determinant of neonatal morbidity and mortality in developed countries2 and for hospital deliveries occurring in developing countries.3,4 Although many promising preventive and therapeutic measures have been put forward, little success has been achieved, with even an increase in preterm deliveries occurring in some circumstances.5–7 This rise could be associated with an increase in obstetric interventions between 34 to 36 weeks, multiple birth related to higher maternal age, wider use of assisted reproduction techniques,7 a gestational age estimation artifact,8 or other social or biologic factors. Moreover, most of the reduction in perinatal mortality is due to improvements in the survival of tiny infants rather than in the prevention of their early birth.7
Why are we unable to reduce preterm delivery when much knowledge has been gained and interventions seem to improve results in other areas of medicine? On one hand, the mechanisms involved in the initiation of labor (at term or before it) remain to be completely elucidated. On the other, preterm delivery is a multifactorial entity or syndrome, whose prevention is unlikely to be achieved with interventions such as nutrition supplementation9 or antibiotic prophylaxis.10
Epidemiologic heterogeneity,11–16 although still under debate,17 has been recently supported by a secondary analysis using biologic markers.18 It could even be expanded if risk factors, such as previous spontaneous preterm delivery,19 fetal growth impairment,20 or the interaction between subgroups of small for gestational age and subgroups of preterm delivery,21 are considered. There are few data comparing, in a comprehensive manner, morbidity and mortality among the suggested different clinical presentations and mechanisms of preterm deliveries.22–25 We aimed to explore the syndromic nature of this entity in terms of its perinatal outcomes.
MATERIALS AND METHODS
This is a secondary analysis using the populations enrolled in the WHO multicenter randomized trial for the evaluation, using a cluster randomization design, of a new antenatal care model where the clusters were the antenatal care clinics. This trial compared a new model of care versus the standard package.26,27
The trial was conducted in Rosario, Argentina; Havana, Cuba; Jeddah, Saudi Arabia; and Khon Kaen, Thailand, and enrolled 24,678 women in 53 antenatal clinics. In addition to these 53 original antenatal clinics participating in the trial, the 5 hospitals affiliated with 1 of the centers of the trial, the Centro Rosarino de Estudios Perinatales in Rosario, Argentina, followed the same trial's data collection procedures for their entire pregnant populations during the same study period (August 1996 to December 1998). A detailed description of the trial methodology and populations has been published elsewhere.26,28,29 The trial's results for preterm delivery (P = .70), low birth weight (P = .40), small for dates (P = .67), neonatal mortality (P = .52), and neonatal intensive care unit (NICU) stay more than 7 days (P = .11) did not differ among participant countries as evaluated by the χ2 test for heterogeneity across sites and strata.30
There was a total of 41,751 pregnant women; among them there were 929 deliveries with birth weight less than 500 g or gestational age less than 22 weeks, which were considered abortions and not included in the birth population. There were 848 women lost to follow-up, 334 multiple births, and 579 women without information on obstetric and medical complications during pregnancy, all of whom were excluded from the analysis population. Furthermore, we did not have all the information needed to calculate the best obstetric estimate of gestational age at delivery for 288 women, who were also excluded. Among the remaining 38,773 births, 454 newborns had clinically evident congenital malformations and were also excluded. Therefore, our final outcome analysis population included 3,304 preterm deliveries among the total of 38,319 singleton births (Figure 1).
Gestational age at birth was based for all deliveries on the Best Obstetric Estimate as standardized for the WHO Antenatal Care Trial. This includes the data on the last normal menstrual period, gestational age at time of positive pregnancy test, early ultrasound examination if requested by the attending staff for dating, uterine size during first trimester, routine uterine fundal height measures, and obstetric clinical evaluation.31 Preterm deliveries were those with a gestational age at birth of less than 37 completed weeks.
Data collection forms used in the trial were validated by the study supervisors, who completed a second data-collection form for all deliveries taking place during 1 randomly selected day. The clinic staff and data clerks were unaware of this procedure, which was repeated twice in all the study hospitals (total sample 759 women). Agreement between these 2 observers was assessed by the percentage of agreement and κ statistics adjusted for strata for qualitative variables and the intraclass correlation coefficient for quantitative variables. Kappa statistics were 0.88 (95% confidence interval 0.84–0.94) for preterm delivery status and intraclass correlation coefficient for gestational age at birth 0.93, lower (1-tailed) 95% confidence interval 0.92. The κ statistics for baseline characteristics were 0.79 or higher.
We classified the preterm deliveries according to the presence of severe obstetric and medical complications during pregnancy (yes/no) and the characteristics of initiation of labor or delivery (clinical presentation). Obstetric and medical complications were considered as the presence of at least 1 of the following conditions: chronic hypertension (diastolic blood pressure more than 90 mm Hg at booking before 20 weeks of gestation, or being under antihypertension treatment); hypertension (more than 140/90 mm Hg) in the present pregnancy after the 20th week; preeclampsia (blood pressure more than 140/90 mm Hg and 2.0 g or more in 24 hours or 2+ proteinuria by dipstick) or eclampsia; hemorrhage during pregnancy but before labor by trimester as considered clinically relevant by the attendant; any cardiopathy, renal disease or any other severe medical complications diagnosed after local protocols, severe anemia (Hb less than 90 g/L), and hemolytic disease in the present pregnancy (rhesus-isoimmunization) considered as a positive Coombs test or Rh sensitization in any pregnancy. Type 1 diabetes mellitus and gestational diabetes were considered separately because both are generally associated with increased fetal weight and preterm deliveries, whereas the other complications are associated with impaired fetal growth.
The characteristics of initiation of preterm labor or delivery were spontaneous initiation of labor, labor after prelabor spontaneous rupture of amniotic membranes (PROM; occurring at least 1 hour before the onset of sustained labor), and medically indicated (defined as induced labor or elective cesarean delivery). Therefore, 5 main subgroups of preterm deliveries were studied.
There were 2 small subgroups of medically indicated preterm deliveries that because of their special characteristics were only considered in a descriptive manner. They were medically indicated, because of suspected small for gestational age, but without maternal complications, and those medically indicated because of a previous cesarean delivery or breech presentation.
Primary outcomes were newborn length of stay at a NICU or at other special care unit (7 days or more), which was used as an indicator for severe neonatal morbidity and prehospital discharge neonatal death. Secondary outcomes were intrapartum fetal death and small for gestational age (below the 10th percentile of birthweight for gestational age chart32 previously used in this population27). For “days of stay in the newborn intensive care units,” the denominator was all newborns that did not die in the first week of life. For intrapartum fetal death, we used pregnancies as denominator. Neonatologists were not aware of the hypotheses tested and followed locally accepted clinical protocols.
The form of the WHO antenatal care trial26 was translated and adapted to the local terminology and practices. A trained female interviewer, supported by the medical staff, abstracted information, and if needed, questioned mothers during the postpartum period. A data quality–monitoring unit was established at each research center.29 All data collection staff were aware that a trial was being conducted but they were not informed of the specific hypotheses being tested.
The proportion of preterm deliveries groups was estimated as a fraction of the total population. We estimated the rates of neonatal morbidity and mortality for each of the preterm deliveries groups. We identified potential confounding variables and included them with country and gestational age in the logistic regression models. We compared, for each outcome, adjusted and unadjusted models, excluding variables describing pregnancy history. We also explored the effect of clustering and found it was negligible. After gestational age was included in models, the effect of adding birth weight to them was evaluated using the Hosmer-Lemeshow goodness-of-fit test.33
We estimated adjusted probabilities of each outcome by week of gestation using the logistic regression models and compared probability curves in the figures against the preterm deliveries “spontaneous without complications” group as the reference group, using the Wald test to contrast each curve's coefficients.33 For the outcomes “length of stay in newborn intensive care” and neonatal death, restricted cubic splines with 3 knots were used in the probability models.34 Finally, interaction terms of gestational age and preterm deliveries groups were included in the probability models and retained if statistically significant. All calculations were made with a SAS Software 8.2 (SAS Institute, Cary, NC). The project was approved by the WHO Committee for Research into Human Subjects and the institutional review boards of the participating institutions.
There were 3,304 preterm births (8.6%) among the 38,319 singleton births without congenital malformations. Fifty-six percent were spontaneous preterm delivery without evidence of obstetric and medical complications; a further 17% had PROM without obstetric and medical complications, whereas 4% were PROM with complications. “Spontaneous with complications” represented 12% and “medically indicated with complications” 11%. The most common medical complications were hypertensive disorders including chronic hypertension and preeclampsia. Spontaneous preterm deliveries without obstetric and medical complications had statistically significant heavier birth weights than the other preterm deliveries groups (P < .001); PROMs with complications were the lightest; mean gestational age at delivery was similar for spontaneous preterm deliveries without complications and medically indicated preterm deliveries; and the other 3 groups had statistically significantly shorter gestational age (P < .001; Table 1).
The proportion of small for gestational age was highest in the medically indicated preterm deliveries (22.3%), whereas all the other preterm groups had considerably lower rates (Table 2). The “PROM with complications” had the highest (37%) proportion of very early preterm deliveries (less than 32 weeks; P < .001), as compared with spontaneous preterm deliveries without complications (17%). The other 3 groups had rates between 24.4% and 26.2% of very early preterm delivery (data not shown in tables).
In general, medically indicated preterm deliveries were more strongly associated with previous stillbirth, last pregnancy with preeclampsia, and previous gynecologic surgery than the other subgroups. The PROM groups were associated with smoking during present pregnancy more often than the other subgroups. Other sociodemographic indicators in general did not differentiate among preterm delivery subgroups, although maternal education less than primary was overrepresented among mothers in the “spontaneous preterm deliveries without complications” group, and age more than 40 years as higher among mothers of medically indicated preterm deliveries. (Table 3).
Preterm deliveries with obstetric and medical complications were associated with the highest rate of intrapartum fetal death at a given gestational age, regardless of whether they were spontaneous, indicated, or PROM after adjustment for previous stillbirth and preeclampsia-eclampsia in previous pregnancy (Figure 2). Excluding these 2 variables from the models did not modify the results. The 2 groups without complications had consistently lower intrapartum fetal death. Adjusting by gestational age and country did not change these patterns. Birthweight was not included because it reduced the goodness of fit of the model. The interaction term between gestational age (linear) and preterm delivery subgroup was not statistically significant (P = .94).
For NICU stays of 7 days or more, the most important determinant factor for a differential risk in our population was the spontaneous presentation of the preterm delivery. After adjusting for the effect of previous stillbirth, previous low birth weight, parity, reproductive tract infection, diabetes, substance abuse, country, and gestational age, the spontaneous preterm delivery groups (with and without complications) had, at a given gestational age, lower risk than the other groups (Figure 3).
The 2 PROM groups and the medically indicated had similar morbidity risk up to approximately 30 weeks of gestation. After that time, we observed that the PROM without complications tended to have lower risk than these 2 other groups (statistically significant interaction P < .001; Figure 3). Birthweight was not included in the model because it reduced its goodness of fit. Excluding from the models the variables indicating pregnancy history did not modify these results.
The risk of neonatal death declined 26 weeks onwards for all groups, and after 32 weeks of gestation, the risk was less than 10% for all groups. The 3 groups with maternal complications had the highest risk at each gestational age. The lowest risk was found in the PROM without complications group. All these analyses were adjusted by urinary tract infection and country and included gestational age as cubic spline (Figure 4). There was no interaction between gestational age (both linear and spline) and preterm delivery subgroups (P = .29 and P = .59).
Finally, there were 26 preterm deliveries that were medically indicated because of a suspected small for gestational age fetus, of whom 19 were below the 10th percentile of birthweight for gestational age at birth. The mean birthweight of this group was 1,839 g (standard deviation [SD] 450 g) and the mean gestational age 35 weeks (SD 1.2 weeks). The overall neonatal death rate of this group was 3.4%, and 73% of them had 7 or more days of NICU stay.
There were also 47 preterm deliveries “medically” indicated associated with previous cesarean delivery or breech presentation but without any maternal complications. This group had a mean birthweight of 2,498 g (SD 606 g), a mean gestational age of 34.6 weeks (SD 1.9 weeks), and neonatal death rate of 7%, and 32% of them had a NICU stay of 7 days or more.
This large, prospective, observational study provides evidence, from a multiethnic population, of differential neonatal morbidity and fetal and neonatal mortality of the preterm delivery syndrome classified according to severe obstetric and medical complications and the clinical presentation of preterm deliveries. Differential rates of small for gestational age and early preterm delivery were also observed.
Our study was based on all women initiating antenatal care during a fixed period at clinics serving determinate geographic areas. Data were recorded as part of an antenatal care research project, standardized and monitored.29 Ultrasound measures were used to corroborate uncertain gestational age.21 Antepartum fetal death was considered a consequence of prematurity and not the inverse, and therefore data are presented only for intrapartum fetal death.35
The incidence of preterm deliveries in our total population, close to 9%, was remarkably similar to that in other ethnically different populations studied recently,13,17,36 supporting the concept that preterm delivery rates are relatively stable across populations.37 It was difficult to compare the distribution of the subgroups of preterm deliveries among studies because of sampling characteristics, the composition of the pregnant population, and the definitions used.
We separated spontaneous preterm deliveries with and without obstetric and medical complications. For those with maternal complications, early delivery was at least partially related to the underlying maternal conditions; for spontaneous without complications, there are no preventive strategies,7,9 and tocolytic treatments are unlikely to be effective or even desirable.7 Earlier reports had not separated these 2 entities,15,21,38,39 whereas others grouped spontaneous preterm deliveries and preterm delivery with PROM.16,20,40 In our population, spontaneous preterm deliveries without clinically evident complications represented 56% of all preterm deliveries, whereas the 11% of medically indicated was within the range of other populations.7,15,12
It had been suggested that overall preterm delivery is associated with small for gestational age births.21 However, we have observed such an increased risk only among medically indicated preterm deliveries, in agreement with a recent large European preterm delivery study.41 For these pregnancies, strategies to prevent early delivery would have to be directed to the underlying conditions.
Obstetric and medical complications, rather than clinical presentation, tended to be the dominant factor for intrapartum fetal death (Figure 2). It is unrealistic, therefore, to expect that single preventive or treatment modalities would affect intrapartum death among preterm deliveries, which is related to conditions preceding preterm labor or even pregnancy.
PROM and medically indicated preterm deliveries had the highest severe neonatal morbidity. Spontaneous preterm deliveries had a clearly different morbidity pattern at all gestational ages from the others independently of complications. We observed an interaction effect with gestational age for this outcome. However, this needs confirmation because of the small number of events at each gestational age and the limitations of the summary morbidity indicator we used.
For neonatal mortality, similar to intrapartum death, the 3 groups with obstetric and medical complications had the highest risk for mortality. It should be noted that at approximately 20 weeks of gestation, neonatal mortality was not 100% (Figure 4), because of the very early referral of some of these newborns to other hospitals, which was recorded in our data as “hospital discharge alive.” Unfortunately we did not have data on the outcome of these few infants in the other institutions.
The present analyses require confirmation in other settings. Nevertheless, the different ethnic groups we studied, the stable rate of preterm delivery, and the considerable reduction of risk after 32 weeks reproduced in our population support the idea that the heterogeneity concept could be present in other populations.
We believe that it is the understanding of the complex interaction between clinical presentations, infections and inflammations, maternal complications, and impaired fetal growth42 in relation to specific outcomes that will provide the way forward to the prevention of preterm delivery.
1. Villar J, Gurtner de la Fuente V, Ezcurra E, Campodonico L. Pre-term delivery: unmet need. In: Keirse M, editor. New perspectives for the effective treatment of pre-term labour. Royal Turnbridge Wells, Kent, UK: Wells Medical; 1994.
2. Kramer MS, Demrissie K, Yang H, Platt RW, Sauvé R, Liston R for the Fetal and Infant Health Study group of the Canadian Perinatal Surveillance System. The contribution of mild and moderate pre-term birth to infant mortality. JAMA 2000;284:843–9.
3. Ziadeh SM. Obstetrical outcomes amongst preterm singleton births. Saudi Med J 2001;22:342–6.
4. Yasmin S, Osrin D, Paul E, Costello A. Neonatal mortality of low-birth-weight infants in Bangladesh. Bull World Health Organ 2001;79:608–14.
5. Joseph KS, Kramer MS, Marcoux S, Ohsson 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.
6. Demrissie K, Rhoads G, Amanth CV, Alexander GR, Kramer MS, Kogan MD, et al. Trends in preterm birth and neonatal mortality among Blacks and Whites in the United States from 1989 to 1997. Am J Epidemiol 2001;154:307–15.
7. Goldenberg R. The management of preterm labor. Obstet Gynecol 2002;100:1020–37.
8. Bakketeig L. Ultrasound dating of pregnancies changes dramatically the observed rates of pre-term, post-term and small for gestational age birth: a commentary. Iatrogenic 1991;1:174–5.
9. Villar J, Merialdi M, Gülmezoglu M, Abalos E, Carroli G, Kulier R, et al. Nutritional interventions during pregnancy for the prevention of maternal morbidity and preterm delivery: an overview of randomized controlled trials. J Nutr 2003;133(suppl):1606S–35S.
10. Klebanoff M, Carey J-C, Hauth J-C. Failure of metronidazole to prevent preterm delivery among pregnant women with asymptomatic Trichomonas vaginalis infection. N Engl J Med 2001;345:487–93.
11. Savitz D, Blackmore C, Thorp J. Epidemiologic characteristics of pre-term delivery: etiologic heterogeneity. Am J Obstet Gynecol 1991;164:467–71.
12. Tucker JM, Goldenberg R, Davis R, Copper R, Winkler CL, Hawth JC. Etiologies of pre-term birth in an indigent population: is prevention a logical expectation? Obstet Gynecol 1991;77:343–7.
13. Berkowitz GS, Blackmore-Prince C, Lapinski RH, Savitz DA. Risk factors for preterm birth subtypes. Epidemiology 1998;9:279–85.
14. Ananth C, Savitz D, Luther E, Bowes W. Preeclampsia and preterm birth subtypes in Nova Scotia, 1986 to 1992. Am J Perinatol 1997;14:17–23.
15. Blackmore C, Savitz D, Edwards L, Harlow S, Bowes W. Racial differences in the patterns of pre-term delivery in central North Carolina, USA. Paediatr Perinat Epidemiol 1995;9:281–95.
16. Adams M, Read J, Rawlings J, Karlass F, Sarno A, Rhodes P. Pre-term delivery among black and white enlisted women in the United States Army. Obstet Gynecol 1993;81:65–71.
17. Pickett K, Abrams B, Selvin S. Defining preterm delivery—the epidemiology of clinical presentation. Paediatr Perinat Epidemiol 2000;14:305–8.
18. McGrath S, Mclean M, Smith D, Bisitis A, Giles W, Smith R. Maternal plasma corticotropin-releasing hormone trajectories vary depending on the cause of preterm delivery. Am J Obstet Gynecol 2002;186:257–60.
19. Mercer BM, Goldenberg RL, Moaward A, Meis PJ, Iams JD, Das AF, et al. The preterm prediction study: effect of gestational age and cause of preterm birth on subsequent obstetric outcome. National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. Am J Obstet Gynecol 1999;181:1216–21.
20. Zeithin J, Ancel PX, Saurel-Cubizolles M, Papiernik E. Are risk factors the same for small for gestational age versus other preterm births? Am J Obstet Gynecol 2001;185:208–15.
21. Hediger M, Scholl T, Scholl J, Miller L, Fischer R. Fetal growth and the etiology of pre-term delivery. Obstet Gynecol 1995;85:175–82.
22. Wolf E, Vintzileos AM, Rosenkrantz TS, Rodis J, Salafia C, Pezzullo J. Do survival and morbidity of very-low-birthweight infants vary according to the primary pregnancy complication that results in pre-term delivery? Am J Obstet Gynecol 1993;169:1233–9.
23. Gray PH, Hurley TM, Rogers YM, O'Callaghan MJ, Tudehope DI, Burns YR, et al. Survival and neonatal and neurodevelopmental outcome of 24–29 week gestation infants according to primary cause of preterm delivery. Aust N Z J Obstet Gynaecol. 1997;37:161–8.
24. Owen J, Baker SL, Hauth J, Goldenberg RL, Davis RO, Copper RL. Is indicated or spontaneous preterm delivery more advantageous for the fetus? Am J Obstet Gynecol 1990;163:868–72.
25. Kimberlin DF, Hauth JC, Owen J, Bottoms SF, Iams JD, Mercer BM, et al. Indicated versus spontaneous preterm delivery: an evaluation of neonatal morbidity among infants weighing </= 1000 grams at birth. Am J Obstet Gynecol 1999;180:683–9.
26. Villar J, Bakketeig L, Donner A, Al-Mazrou Y, Ba'aqeel H, Belizan JM, et al. The WHO antenatal care randomised controlled trial: rationale and study design. Paediatr Perinat Epidemiol 1998;12(suppl):27–58.
27. Villar J, Ba'aqeel H, Piaggio G, Lumbiganon P, Belizan J, Farnot U, et al; WHO Antenatal Care Trial Research Group. WHO antenatal care randomized trial for the evaluation of a new model of routine antenatal care. Lancet 2001;357:1551–64.
28. Donner A, Piaggio G, Villar J. Methodological considerations in the design of the WHO Antenatal care Randomized Controlled Trial. Paediatr Perinat Epidemiol 1998;12:59–74.
29. Pinol A, Bergel E, Chaisini K, Diaz E, Gandez M. Managing data for a randomized controlled trial: experience from the WHO Antenatal Care trial. Paediatr Perinat Epidemiol 1998;12:142–55.
30. Fleiss JL. Statistical methods for rates and proportions. New York: Wiley; 1981.
31. WHO antenatal care randomized trial: manual for the implementation of the new model. WHO/RHR/01.30. Geneva: World Health Organization; 2002.
32. Williams RK, Creasy RK, Cunningham GC, Hawes W, Norris FD, Tashiro M. Fetal growth and perinatal viability in California. Obstet Gynecol 1982;59:624–32.
33. Hosmer DW, Lemeshow S. Applied logistic regression. 2nd ed. New York: Wiley; 2000.
34. Harrell FE. Regression modeling strategies: with applications to linear models, logistic regression, and survival analysis. New York: Springer; 2001. p. 20–4.
35. Kramer MS, Liu S, Luo Z, Yuan H, Platt R, Joseph KS. Analysis of perinatal mortality and its components: time for a change? Am J Epidemiol 2002;156:493–7.
36. Leung TN, Roach VJ, Lau TK. Incidence of preterm delivery in Hong Kong Chinese. Aust N Z J Obstet Gynaecol 1998;38:133–7.
37. Villar J, Belizan JM. The relative contribution of prematurity and fetal growth retardation to low birth weight in developed and developing societies. Am J Obstet Gynecol 1982;143:793–8.
38. Main DM, Gabbe SG, Richardson D, Strong S. Can pre-term deliveries be prevented? Am J Obstet Gynecol 1985;151:892–8.
39. Meis PJ, Ernest JM, Moore ML. Causes of low weight births in public and private patients. Am J Obstet Gynecol 1987;156:1165–8.
40. De Haas I, Herlow B, Cramer D, Frigoletto F. Spontaneous pre-term birth: a case-control study. Am J Obstet Gynecol 1991;165:1290–6.
41. Zeitlin J, Ancel PY, Saurel-Cubizolles MJ, Papiernik E. The relationship between intrauterine growth restriction and preterm delivery: an empirical approach using data from a European case-control study. Br J Obstet Gynaecol 2000;17:750–8.
42. Klebanoff M. Conceptualizing categories of preterm birth. Prenat Neonat Med 1998;3:13–15.
The World Health Organization Antenatal Care Trial Research Group consists of the following additional members: Olav Meirik, Epidemiologist; Allan Donner, Consultant Epidemiologist-Statistician; Alain Pinol, Systems Analyst; Milena Vucurevic and Catherine Hazelden, Statistical Assistants; Ana Langer, Focal person, satisfaction studies; Gunilla Lindmark, Obstetrician/Gynaecologist; Miranda Mugford, Focal person, economical evaluation; Vivian Wong, Obstetrician/Gynaecologist; Julia Fox-Rushby, Health Economist; Gustavo Nigenda, Public Health Specialist; Mariana Romero, Public Health Epidemiologist; Georgina Rojas, Psychologist; and Chusri Kuchaisit, Public Health Nurse. Data Coordinators: Eduardo Bergel, Argentina; Elva Diaz, Cuba; Mohammed Gandeh, Saudi Arabia; and Yaowaret Singuakool, Thailand. Field Coordinators: Alicia del Pino, Argentina; Juan Vazquez, Cuba; Abdul-hameed Helal, Saudi Arabia; and Kamron Chaisiri, Thailand. Cited Here...