Farr, Sherry L. PhD1; Jamieson, Denise J. MD, MPH1; Rivera, Hirmice Vásquez MD2; Ahmed, Yusuf BM1; Heilig, Charles M. PhD3
Puerto Rico has one of the highest documented rates of cesarean delivery in the world. In 2002, the rate of total cesarean delivery in Puerto Rico reached 45%, and the primary cesarean delivery rate was 34%, rates that are 72% to 85% higher than among Puerto Rican women delivering on the mainland.1 Between 1996 and 2002, the rate of primary cesarean delivery increased by 52% in Puerto Rico, but by 24% among Puerto Rican women delivering on the mainland.1
It is unclear why rates of cesarean delivery are higher and increasing more rapidly in Puerto Rico than on the mainland. Therefore, the aims of this investigation were 1) to determine the factors associated with primary cesarean delivery among births to Puerto Rican women; 2) to examine reasons for the increasing rate of primary cesarean delivery in Puerto Rico between 1996 and 2002; and 3) to examine reasons for the greater rate of primary cesarean delivery in 2002 in Puerto Rico compared with Puerto Rican women delivering on the U.S. mainland.
MATERIALS AND METHODS
To examine changes in rates of primary cesarean delivery over time in Puerto Rico, live births to Puerto Rican women delivering in Puerto Rico in 1996 were compared with those delivering in Puerto Rico in 2002. To examine differences between places of delivery, live births to Puerto Rican women who delivered on the mainland in 2002 were compared with live births to Puerto Rican women who delivered in Puerto Rico in 2002.
Data for this analysis are from the National Vital Statistics System birth certificates from Puerto Rico and the mainland. Information on method of delivery (vaginal, vaginal birth after cesarean delivery [VBAC], primary cesarean delivery, and repeat cesarean delivery), mothers' and infants' demographic and health characteristics, labor and delivery complications, and medical risk factors associated with the pregnancy are collected on birth certificates. Studies using deidentified, publicly available data, as in this study, are not required to be reviewed by the U.S. Centers for Disease Control and Prevention institutional review board.
Birth certificates in Puerto Rico do not include an item on whether the mother is of Hispanic origin. In Puerto Rico between 1996 and 2002, 96% of birth certificates listed the mother's birthplace as Puerto Rico or the mainland and, within Puerto Rico, characteristics of births were similar by mother's place of birth. Therefore, Puerto Rican births are defined as births occurring in Puerto Rico to women who were residents of Puerto Rico and born in Puerto Rico or on the mainland. The 4% of births to mothers born elsewhere were excluded. Puerto Rican births on the U.S. mainland were defined as births occurring on the mainland to women who were residents of the mainland and born in Puerto Rico or on the mainland with self-reported Hispanic origin as “Puerto Rican” on the infant's birth certificate.
Only births to women who were considered Puerto Rican, had singleton births, and who had never had a previous cesarean delivery were included in this analysis. In 1996 there were 63,1412 live births in Puerto Rico. There were 51,357 (81.3%) singleton births to Puerto Rican women without a previous cesarean delivery; of these births, 51,049 (99.4%) had information on all characteristics of interest and were included in the multivariable analysis. During 2002, there were 52,7473 births in Puerto Rico, and 40,489 (76.8%) met inclusion criteria. Of these births, 40,207 (99.3%) had information on all characteristics of interest and were included in the final multivariable model. On the mainland in 2002, there were more than 4 million live births. Of all mainland births in 2002, 47,800 (1.2%) were singleton births to Puerto Rican women without a previous cesarean delivery; of these, 41,593 (87.0%) had information on all characteristics of interest and were included in the final multivariable model.
To determine what may explain the increased rate of primary cesarean delivery in 2002 in Puerto Rico compared with the rate in Puerto Rico in 1996 and on the mainland in 2002, we examined two values for each risk factor for cesarean delivery (mother's age, education, parity, level of prenatal care, pregnancy weight gain, medical risk factors, labor induction, labor or delivery complications, and infant birth weight) among the three populations of births: 1) the distribution of the risk factor within the population, and 2) the risk for cesarean delivery associated with the risk factor. Pearson's χ2 tests were used to assess differences in the distribution of selected maternal and infant characteristics comparing births in Puerto Rico in 2002 with births in Puerto Rico in 1996, and separately, with Puerto Rican births on the mainland in 2002. The primary cesarean delivery rate (reported as a percentage) was computed as the number of cesarean deliveries per 100 live births to women who had not had a previous cesarean delivery.
Maternal characteristics examined included mother's age, number of previous live births, level of education, and weight gain during pregnancy. Medical risk factors (0, 1, 2 or more) that may be associated with cesarean delivery (diabetes, genital herpes, hydramnios or oligohydramnios, pregnancy-associated hypertension, eclampsia, incompetent cervix, previous infant weighing at least 4,000 g, previous preterm or low birth weight infant, and uterine bleeding) were compared between populations. For the multivariable analysis, we conducted a sensitivity analysis excluding pregnancies with any reported medical risk factors. Level of prenatal care during the current pregnancy was based on the Adequacy of Prenatal Care Utilization index,4 which uses total number and timing of prenatal care visits to categorize care into four levels: inadequate, intermediate, adequate, and adequate plus. In the multivariable analysis, inadequate and intermediate levels of care were combined and compared with adequate and adequate-plus levels of care. The distribution of induced labor (yes or no) was also assessed.
For infant weight and gestational age, a three-level nominal variable was created with mutually exclusive categories for preterm or low birth weight infants (less than 37 weeks gestation or less than 2,500 g), full-term infants of normal birth weight (37 weeks or more and 2,500 g to 3,999 g), and full-term infants with a high birth weight (37 weeks or more of gestation and 4,000 g or more).
The number of complications of labor and delivery (0, 1, 2 or more) that may be associated with cesarean delivery (fetal distress, breech presentation, cord prolapse, cephalopelvic disproportion, dysfunctional labor, premature rupture of membranes, abruptio placenta, moderate or heavy meconium, placenta previa, excessive bleeding, prolonged labor, and “other” labor complication) was examined in the univariable analysis. Due to low prevalence of cesarean deliveries among women with no reported complications of labor and delivery in certain birth cohorts, we conducted a sensitivity analysis among births to women with one labor complication, controlling for all other variables. Finally, whether a woman could be considered at low risk for a cesarean delivery was assessed. Based on the Healthy People 2010 criteria, a woman at low risk is defined as a woman giving birth for the first time with a singleton pregnancy of 37 weeks or more of gestation with an infant in vertex presentation.5
To examine whether an increased risk for cesarean delivery among certain subgroups of women explained the higher rate of cesarean delivery in 2002 in Puerto Rico, we examined modification of the relative risk of cesarean delivery for each characteristic above by year of delivery (1996 compared with 2002) in Puerto Rico and by place of delivery (mainland compared with Puerto Rico) in 2002. We implemented two log-linear models using the Genmod procedure in SAS 9.1 (SAS Institute Inc., Cary, NC) to generate adjusted relative risks and 95% confidence intervals for associations of interest. Both models included interaction terms between year of delivery (first model) or place of delivery (second model) and all covariates. To determine whether there was a statistically significant difference in relative risks for individual characteristics between birth cohorts (if the interaction term contributed to the overall fit of the multivariable model), we conducted an analysis of deviance, using a –2 log likelihood test for interaction, to compare models with and without each interaction term.6
Table 1 presents distributions and rates of primary cesarean delivery by subgroup. Within Puerto Rico, the risk profile of the population changed slightly between 1996 and 2002 (Table 1). Compared with births in 1996, in 2002 a slightly higher percentage of live births were to women of greater age, who were nulliparous, who gained more than 40 lb during pregnancy, with reported medical risk factors, and who had labor induced. All of these subgroups had higher rates of cesarean delivery. Percentages also increased for low birth weight or preterm births and those with reported complications of labor or delivery. A lower percentage of women could be considered at low risk for a cesarean delivery in 2002 (83%) compared with 1996 (87%). All changes in the distributions of examined characteristics were statistically significant (P<.01). However, strata changed by approximately 5 percentage points or less for all characteristics except amount of prenatal care received, with receipt of adequate-plus prenatal care increasing from 24.3% in 1996 to 38.8% in 2002. Together these changes suggest a slightly increased risk profile for cesarean delivery for births in 2002 compared with 1996.
Between 1996 and 2002 in Puerto Rico, rates of cesarean delivery increased for almost all demographic, maternal, and infant characteristics (Table 1). Overall, the unadjusted rate of cesarean delivery in 2002 (32.7%) increased by 52% relative to the rate in 1996 (21.6%), with the greatest relative increases in women less than 20 years of age (by 89%), with less than a high school education (by 76%), with intermediate or inadequate prenatal care (by 79% and 97%, respectively), and with no reported labor complication (by 73%).
Table 2 shows adjusted relative risks for each characteristic. In 2002, women in Puerto Rico were 2.1 (95% confidence interval [CI] 2.0–2.3) times as likely to deliver by cesarean as in 1996 (Table 2), which translates into one additional cesarean delivery in every 4.2 live births, controlled for examined risk factors. Low-risk women were 1.9 (95% CI 1.8–2.1) times as likely to deliver by cesarean in 2002 compared with 1996 (Table 2). Between 1996 and 2002, associations weakened or remained the same between the risk for primary cesarean delivery and all examined risk factors for all births (Table 2), and patterns were similar for births to women at low risk (data not shown).
Table 1 also shows the distribution of all characteristics among Puerto Rican women by place of delivery (Puerto Rico or mainland) in 2002. The distribution of individually examined risk factors did not indicate a greater risk for cesarean delivery in either place. A slightly higher percentage of women were considered low risk on the mainland (84%) than in Puerto Rico (83%).
In 2002, rates of primary cesarean delivery were higher in every strata of every characteristic in Puerto Rico compared with the mainland. The overall rate of primary cesarean delivery was almost twice as high in Puerto Rico (32.7%) as on the mainland (16.9%). On the mainland, rates were highest among the same subgroups of women as in Puerto Rico. However, rates were approximately 100% to 150% greater in Puerto Rico relative to the mainland for births to women aged younger than 30 years, with recommended maternal weight gain, with one medical risk factor, without induced labor, who had normal birth weight infants, and with one or no reported labor complications. The rate of cesarean delivery among low-risk women in Puerto Rico (44.8%) was almost twice that among the mainland population (22.6%).
In 2002, births in Puerto Rico were 2.4 (95% CI 2.2–2.6) times as likely to be delivered by cesarean as births on the mainland (Table 2), or one additional cesarean delivery in every 4.2 live births, after controlling for risk factors. Women at low risk without induced labor were 2.7 (95% CI 2.5–3.0) times as likely to deliver by cesarean as Puerto Rican women delivering on the mainland (Table 2). Compared with the mainland, associations between risk for cesarean delivery and all examined risk factors weakened or remained the same in Puerto Rico, both for all women (Table 2) and those at low risk (data not shown).
In both analyses, excluding births with any medical risk factor did not change relative risk estimates. Risk estimates also were similar when analyses were limited to births with one reported complication of labor or delivery.
Neither the distribution of examined risk factors nor the risk for cesarean delivery among women with these characteristics could explain the higher rate of primary cesarean delivery in Puerto Rico in 2002 compared with that in 1996 or on the mainland in 2002. The slight increases in prevalence of risk factors for cesarean delivery (advanced maternal age, more induced labor, more reported medical risk factors and labor or delivery complications, and more low birth weight or preterm births) in Puerto Rico between 1996 and 2002 may contribute, in part, to the increase in the primary cesarean delivery rate between those years. These may be actual changes or may reflect changes in reporting of examined characteristics on birth certificates. However, changes in maternal and infant demographics did not fully explain the 52% higher rate of primary cesarean delivery in Puerto Rico in 2002. Additionally, if a lower threshold for performing cesarean deliveries in Puerto Rico in 2002 accounted for the high rate of primary cesarean delivery, we might expect greater relative risks by maternal and infant characteristics in Puerto Rico in 2002 compared with 1996; however, relative risks weakened or remained the same for all characteristics examined. Even larger differences in rates of cesarean delivery within strata and differences in the relative associations between risk factors and cesarean delivery were seen comparing births on the mainland with those in Puerto Rico in 2002.
Increases in the rate of primary cesarean delivery between 1996 and 2002 also occurred on the mainland among Puerto Rican women and all women1,7; however rates increased much more sharply in Puerto Rico. The increase over time on the mainland may be due to changes in medical practice in the United States, as seen by increases among low-risk Puerto Rican women1 and low-risk women with no reported medical risk factors or labor complications.8
Several factors, such as changes in insurance coverage, use of new technologies, medical research findings, fears of litigation, or maternal or physician preferences, may influence changes in medical practice over time. In 1993 the government of Puerto Rico initiated the health reform process aimed at increasing access to health care for all.9 From 1996 to 2002 in Puerto Rico, approximately 15% more women received adequate or adequate plus prenatal care, as reported on birth certificates, with the greatest increases among younger women with a high school education or less (data not shown), the groups for which rates of primary cesarean delivery increased the most. However, increased access to prenatal care does not fully explain the dramatic increase in the primary cesarean delivery rate between 1996 and 2002. Rates of primary cesarean delivery in 2002 were greater for all women in this study, and, relative to 1996 rates, rates increased 18% to 43% among women who received adequate plus and adequate prenatal care in both years, respectively.
Research on women's and physicians' attitudes and experiences with regard to cesarean delivery is underway in Puerto Rico. Until ways to reduce rates safely among the entire population are found, physicians, public health practitioners, and other stakeholders may want to focus their efforts on reducing rates among women with no reported labor complications and those at low risk, groups whose rates rose 73% and 43%, respectively, over 6 years in Puerto Rico and were 98 to 130% higher than those on the mainland in 2002.
High rates of cesarean delivery are also documented in Hispanic populations both on the U.S. mainland and in other countries. In 2003, the total cesarean delivery rate for women of self-reported Cuban origin delivering on the mainland was 39.8%.10 However, rates for women of other Hispanic origins delivering on the mainland were similar to that for the mainland as a whole.10 Chile, Brazil, and Mexico have estimated rates of cesarean delivery above 30% overall, with rates of more than 50% in private hospitals.11 Generalizing research findings from other areas to Puerto Rico is difficult. Understanding the unique medical and cultural factors associated with the high rate of cesarean delivery in Puerto Rico will require additional study in collaboration with researchers and stakeholders in Puerto Rico.
Because this analysis is based on data from birth certificates, it should be considered in the context of several limitations. Certain information on birth certificates may be underreported.12 However, past research has shown that percent agreement with medical record data for most characteristics examined in this analysis is relatively high.12–14 Sensitivity and positive predictive value for individual labor or medical complications are generally low, but specificity and negative predictive values for these factors are generally greater than 97%,12,15 and trends in these items are consistent over time.
We were unable to examine certain factors not reported on birth certificates, such as reason for cesarean delivery, type of hospital, or type of insurance coverage, that may influence rates of cesarean delivery. Finally, no distinction could be made between cesarean deliveries that were elective, those that resulted from medical indications, and those that were conducted as emergency procedures, a concern highlighted by other researchers and public health practitioners.16
The dramatic rise in rates of primary cesarean deliveries in Puerto Rico was explained only partially by changes in demographic characteristics between 1996 and 2002. Additionally, in 2002 rates of primary cesarean delivery were lower among Puerto Rican women delivering on the mainland compared with women who delivered in Puerto Rico. Hypothesis-generating research, such as key-informant interviews and focus groups, with obstetricians, hospital administrators, women of reproductive age, and other stakeholders may shed more light on the reasons behind the dramatic rise in rates over time in Puerto Rico and difference in rates between Puerto Rico and the mainland.
1. Centers for Disease Control and Prevention (CDC). Rates of cesarean delivery among Puerto Rican women—Puerto Rico and the U.S. mainland, 1992-2002. MMWR Morb Mortal Wkly Rep 2006;55:68–71.
2. Ventura SJ, Martin JA, Curtin SC, Mathews TJ. Report of final natality statistics, 1996. Mon Vital Stat Rep 1998;46 suppl:1–99.
3. Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Menacker F, Munson ML. Births: final data for 2002. Natl Vital Stat Rep 2003;52:1–113.
4. Kotelchuck M. An evaluation of the Kessner Adequacy of Prenatal Care Index and a proposed Adequacy of Prenatal Care Utilization Index. Am J Public Health 1994 Sep;84:1414–20.
5. U.S. Department of Health and Human Services. Healthy people 2010. 2nd ed. Washington (DC): Government Printing Office; 2000.
6. McCullagh P, Nelder JA. Generalized linear models. 2nd ed. London (UK): Chapman and Hall; 1989.
7. Menacker F, Declercq E, Macdorman MF. Cesarean delivery: background, trends, and epidemiology. Semin Perinatol 2006 Oct;30:235–41.
8. Declercq E, Menacker F, MacDorman M. Rise in “no indicated risk” primary caesareans in the United States, 1991-2001: cross sectional analysis. BMJ 2005;330:71–2.
10. Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Menacker F, Munson ML. Births: final data for 2003. Natl Vital Stat Rep 2005;54:1–116.
11. Belizan JM, Althabe F, Barros FC, Alexander S. Rates and implications of caesarean sections in Latin America: ecological study. BMJ 1999;319:1397–400.
12. DiGiuseppe DL, Aron DC, Ranbom L, Harper DL, Rosenthal GE. Reliability of birth certificate data: a multi-hospital comparison to medical records information. Matern Child Health J 2002;6:169–79.
13. Buescher PA, Taylor KP, Davis MH, Bowling JM. The quality of the new birth certificate data: a validation study in North Carolina. Am J Public Health 1993;83:1163–5.
14. Piper JM, Mitchel EF Jr, Snowden M, Hall C, Adams M, Taylor P. Validation of 1989 Tennessee birth certificates using maternal and newborn hospital records. Am J Epidemiol 1993;137:758–68.
15. Roohan PJ, Josberger RE, Acar J, Dabir P, Feder HM, Gagliano PJ. Validation of birth certificate data in New York State. J Community Health 2003;28:335–46.
16. National Institutes of Health state-of-the-science conference statement: cesarean delivery on maternal request March 27-29, 2006. Obstet Gynecol 2006;107:1386–97.