The rate of cesarean deliveries has continued to increase in the United States, and, in parallel, an increasing number of pregnant women with human immunodeficiency virus (HIV) infection are undergoing cesarean deliveries.1,2 It is estimated that 8,500 pregnant U.S. women with HIV infection deliver annually.3 In 1999, data indicated a lower risk of mother-to-child transmission of HIV with cesarean delivery,4,5 and the rate of cesarean deliveries soon approached 50%.1,2,6,7 Since then, well-tolerated and better studied highly active antiretroviral therapy (HAART), which consists of three or more antiretroviral agents to suppress HIV viral load, has become a cornerstone of HIV management in pregnancy.8 In concert, the rate of mother-to-child transmission of HIV in the United States has decreased to less than 2%.9,10
Human immunodeficiency virus-in-pregnancy guidelines have evolved over the past two decades.1 In 1998, new data demonstrated a 50–80% lower risk of mother-to-child transmission of HIV with cesarean delivery.4,5 These results led the American College of Obstetricians and Gynecologists in 1999 to recommend that pregnant women with HIV infection be offered a scheduled cesarean delivery at 38 weeks of gestation11 (the time period is hereafter referred to as “1998–1999”). Emerging data in the HAART era suggested that effective treatment markedly reduced the risk of mother-to-child transmission through maternal viral suppression, regardless of mode of delivery (hereafter referred to as “2000–2008”).12,13 In 2009, the U.S. Public Health Service recommended HAART for all pregnant women with HIV infection (hereafter referred to as “2009–2013”).14 The U.S. Department of Health and Human Services and the American College of Obstetricians and Gynecologists currently recommend cesarean delivery before labor and before ruptured membranes to prevent mother-to-child transmission for pregnant women with HIV infection with greater than 1,000 copies/mL or unknown HIV plasma viral load.1,2,15–17
We examined changing patterns and factors associated with mode of delivery among pregnant women with HIV infection in the United States in relation to evolving HIV-in-pregnancy guidelines from 1998 to 2013.
MATERIALS AND METHODS
This is a secondary analysis of two large observational studies of pregnant women with HIV infection, Pediatric AIDS Clinical Trials Group 367 (1998–2004) and International Maternal Pediatric Adolescent AIDS Clinical Trials Protocol P1025 (late 2002–2013). Pediatric AIDS Clinical Trials Group 367 was a prospective chart abstraction study of pregnant women with HIV infection at 72 centers in the United States and Puerto Rico designed to assess maternal HIV infection, mother-to-child transmission, antiretroviral therapy use, mode of delivery, and plasma viral load (Shapiro D, Tuomala R, Pollack H, Burchett S, Read J, Cababasay M, et al. Mother-to-child HIV transmission risk according to antiretroviral therapy, mode of delivery, and viral load in 2,895 US women (PACTG 367). Eleventh Conference on Retroviruses and Opportunistic Infections; 2004; San Francisco, California). The subsequent International Maternal Pediatric Adolescent AIDS Clinical Trials Protocol P1025 was a prospective cohort study of pregnant women with HIV infection at 56 centers in the United States and Puerto Rico designed to assess maternal health as well as safety and efficacy of antiretroviral therapy and interventions for the prevention of mother-to-child transmission.18 All women with singleton or multiple gestations with a live birth at greater than 20 weeks of gestation were eligible for enrollment. Institutional review boards approved the protocols at all clinical sites, and written informed consent was obtained from all enrolled women.
Primary study outcomes included 1) cesarean delivery compared with vaginal delivery; then, among women who underwent cesarean delivery: 2) elective compared with nonelective; and 3) indications for cesarean delivery, grouped as HIV-indicated cesarean delivery compared with other clinical indications. Consistent with the originally defined modes of delivery, an “elective” cesarean delivery was defined as cesarean delivery performed before the onset of labor and before ruptured membranes, and a “nonelective” cesarean delivery was defined as cesarean delivery performed after onset of labor, ruptured membranes, or both.19 All indications for cesarean delivery were reviewed by two obstetricians (K.K.V. and R.E.T.) and were classified into the following five categories: 1) HIV infection (ie, for prevention of mother-to-child transmission), 2) repeat cesarean delivery, 3) arrest disorder, 4) fetal indication, and 5) maternal indication. In analyses, we dichotomized this outcome as an “HIV-indicated” cesarean delivery (the primary clinical indication for delivery was HIV infection) compared with a “non–HIV-indicated” cesarean delivery (ie, the other four categories of indications). Data were not available to identify the indication for a woman's first or initial cesarean delivery.
Maternal demographic and clinical covariates of interest included the following: study (Pediatric AIDS Clinical Trials Group 367 vs International Maternal Pediatric Adolescent AIDS Clinical Trials Protocol P1025), estimated date of delivery year, maternal age at delivery, timing of HIV diagnosis (before or during pregnancy), trimester of prenatal care initiation, initial trimester of antiretroviral exposure, first and last antiretroviral regimen in pregnancy, total number of trimesters of antiretroviral exposure, first and last CD4 cell count in pregnancy, first and last HIV RNA in pregnancy, ruptured membranes while in labor, intrapartum and neonatal antiretroviral use, gestational age at delivery, neonatal birth weight, and HIV treatment guideline period, namely 1998–1999, 2000–2008, and 2009–2013 (as described previously).
Descriptive statistics for maternal demographic and clinical characteristics were calculated, both overall and by mode of delivery. Logistic regression models with odds ratios (ORs) and 95% CIs were used to estimate the association between predictors and each of the three study outcomes: vaginal delivery, elective cesarean delivery, and HIV-indicated cesarean delivery. Multivariable logistic regression models adjusted for the following variables, namely maternal age at delivery, timing of HIV diagnosis, initial trimester of antiretroviral use, first CD4 cell count in pregnancy, last HIV viral load before delivery, gestational age at delivery, and HIV treatment guideline period; and in the initial model, which included women who delivered vaginally, ruptured membranes in labor. Separate multivariable logistic regression models for each of the three predefined HIV treatment guideline periods were constructed to assess whether predictors of an HIV-indicated cesarean delivery varied by time among 1998–1999, 2000–2008, and 2009–2013 (ie, effect modification).
The multivariable models included only complete cases (ie, only women who had complete data for all covariates included in the model). Multiple imputation was used to assess the effect of missing data on the previously described conclusions. For each study outcome, 30 imputations of missing covariate values were generated for the same study population and the same covariates as the previously described models assuming that variables were missing at random. To obtain a monotone missing data pattern, continuous covariates were first imputed using Markov Chain Monte Carlo methods, and then categorized, and categorical variables were imputed using logistic regression. For each of the 30 imputations, a logistic regression model was fit with the same covariates and same study population as the complete case analysis plus the women who had been excluded from the complete case analysis using values from the imputation model. The adjusted ORs and standard errors from each of the 30 logistic regressions were then pooled, and final estimates of the adjusted ORs and 95% CI were calculated using SAS Proc MIANALYZE. All analyses used SAS 9.4. Statistical significance was defined as a two-sided P value <.05.
A total of 7,846 women were enrolled in Pediatric AIDS Clinical Trials Group 367 (4,758 women) and International Maternal Pediatric Adolescent AIDS Clinical Trials Protocol P1025 (3,088 women). Of these, 1,402 (17.8%) were excluded from this analysis for the following reasons: ineligible for study enrollment, no prenatal care at a study site, no obstetric baseline data, no estimated date of delivery, or incomplete information on mode of delivery, antiretroviral use, or both. Therefore, this analysis included 6,444 pregnant women with HIV infection. In 1998, the overall rate of mother-to-child transmission was 4% and decreased over time: 4% from 1998 to 1999, 1% in 2000–2008, and 0.3% in 2009–2013 (P<.001).
Twenty-one percent of women delivered between 1998 and 1999, 58% between 2000 and 2008, and 21% between 2009 and 2013 (Table 1). More than half of the women (56%) had been enrolled in the earlier Pediatric AIDS Clinical Trials Group 367 cohort. Most (3,419 [53%]) delivered vaginally. More than two thirds (68%) had been diagnosed with HIV infection before pregnancy, and 57% accessed prenatal care in the first trimester. Overall, only 22% of women had a plasma viral load greater than 1,000 copies/mL at delivery. This changed by time period: 41% from 1999 to 2000, 21% 2001–2008, and 8% 2009–2013 (P<.001).
In 1998, 30% of pregnant women with HIV infection underwent cesarean delivery compared with 21% of U.S. women overall; by 2013, 48% of pregnant women with HIV infection underwent cesarean delivery compared with 30% of U.S. women overall (Fig. 1). Since 2008, approximately 50% of pregnant women with HIV infection underwent cesarean delivery annually, of whom 62% underwent elective cesarean delivery. Among 774 women undergoing repeat cesarean delivery, 74% (n=574) were elective.
Of 3,025 women who underwent cesarean delivery, data on the indication were available for 2,886 (95.4%). The 139 women with an unknown indication for cesarean delivery did not differ by time period, viral load before delivery, or gestational age at delivery. Overall, 977 (34%) of cesarean deliveries had an indication of HIV infection, as did 774 (27%) repeat cesarean deliveries (Table 2). From 1998 to 2013, the primary indication for cesarean delivery changed from HIV infection to repeat cesarean delivery (Fig. 2). The rate of HIV-indicated cesarean deliveries peaked at 48% in 2004 and then dropped to 12% by 2013. Between 1998 and 2013, additional changes occurred: increase in repeat cesarean deliveries (16–42%), decrease in arrest disorder (25–12%), decrease in fetal indications (39–28%), and increase in maternal indications (4–7%). Among the subset of women in International Maternal Pediatric Adolescent AIDS Clinical Trials Protocol P1025 in whom parity and indication were both collected, 54.5% (263/483) of nulliparous women underwent primary cesarean delivery, of whom 35.9% (82/228) were HIV-indicated; 80% of these women had a suppressed HIV viral load (ie, less than 1,000 copies/mL) at delivery.
In multivariable analyses of cesarean delivery compared with vaginal delivery (Table 1), women who delivered between 2000 and 2008 and between 2009 and 2013 (adjusted ORs 1.62–1.99) were more likely to undergo cesarean delivery compared with women who delivered between 1998 and 1999. Factors associated with an increased risk of cesarean delivery included maternal age older than 30 years, HIV viral load 500 copies/mL or greater before delivery, and gestational age at delivery between 32 and 40 weeks. Women with ruptured membranes in labor and who did not undergo antiretroviral therapy during pregnancy were less likely to undergo cesarean delivery. Multiple imputation analyses were used to investigate the effect of missing data, and the results were similar (Appendix 1, available online at http://links.lww.com/AOG/B81).
In multivariable analyses of elective compared with nonelective cesarean delivery, women were more likely to undergo elective cesarean delivery between 2000–2008 and 2009–2013 (adjusted ORs 1.49–1.59) in comparison with 1998–1999, as were women who delivered between 37 and 40 weeks of gestation and who had an HIV viral load greater than 1,000 copies/mL before delivery. Among the 774 women who underwent repeat cesarean delivery, there were no statistically significant predictors (not shown).
In multivariable analyses of HIV-indicated compared with non–HIV-indicated cesarean delivery (Table 3), women who delivered between 2000 and 2008 were more likely to undergo HIV-indicated cesarean delivery compared with those who delivered between 1998 and 1999. Conversely, those who delivered between 2009 and 2013 were less likely to undergo HIV-indicated cesarean delivery. Women who were diagnosed with HIV infection during pregnancy, who initiated antiretroviral therapy in the third trimester, who had a last HIV viral load of 500 copies/mL or greater before delivery, and who delivered between 37 and 40 weeks of gestation had significantly higher odds of undergoing HIV-indicated cesarean delivery. In multiple imputation analyses, the results were similar (Appendix 2, available online at http://links.lww.com/AOG/B81).
We performed subgroup analyses to examine how the predictors of an HIV-indicated compared with non–HIV-indicated cesarean delivery changed over time by fitting separate multivariable logistic regression models for each guideline time period (ie, 1998–1999, 2000–2008, and 2009–2013) (Table 4). Women older than 30 years were progressively less likely to undergo HIV-indicated cesarean delivery in 2000–2008 and 2009–2013 than in 1998–1999. Women who were diagnosed with HIV during pregnancy were more likely to undergo an HIV-indicated cesarean delivery in both 2000–2008 and 2009–2013, but not in 1998–1999. Women who initiated antiretroviral therapy in the third trimester were more likely to undergo HIV-indicated cesarean delivery in 2009–2013, but not in 1998–1999 or 2000–2008. Women with an HIV viral load of 500 copies/mL or greater before delivery were more likely to undergo HIV-indicated cesarean delivery in 2000–2008 and 2009–2013, but not in 1998–1999. In multiple imputation analyses for each guideline period, the results were similar (not shown; Appendix 3, available online at http://links.lww.com/AOG/B81).
We analyzed data for more than 6,000 pregnant women with HIV infection across multiple sites in the United States and Puerto Rico spanning 1998–2013, a time period of evolving obstetric guidelines for HIV infection and access to more effective HAART. Repeat cesarean delivery replaced HIV infection as the primary indication for cesarean delivery among pregnant women with HIV infection. Nevertheless, despite improvements in treatment with mother-to-child transmission at the lowest rate and guideline changes, the prevalence of cesarean delivery remained high among pregnant women with HIV infection in the HAART era. It is likely that the prevalence of cesarean delivery is multifactorial, influenced by HIV infection as well as a history of prior cesarean delivery. Additionally, pregnant women with HIV infection undergoing HIV-indicated cesarean delivery were increasingly more likely over time to be diagnosed with HIV infection during pregnancy, had initiated antiretroviral therapy later in pregnancy, and had a higher HIV viral load at delivery. These findings highlight that HIV-indicated cesarean deliveries were increasingly reserved for women at high risk for mother-to-child transmission, suggesting that obstetric care providers and patients at tertiary care sites where these two studies were conducted were changing HIV-in-pregnancy clinical care as new data emerged.
Human immunodeficiency virus infection was less likely to be the primary indication for cesarean delivery among pregnant women with HIV infection in the United States over time, similar to recent European data.20,21 Although the rate of HIV-indicated cesarean delivery quickly rose from 1998 to 2000 after guidelines recommending that all pregnant women with HIV infection be offered a scheduled cesarean delivery at 38 weeks of gestation, the rate then declined until 2013. This change in indication for cesarean delivery was temporally consistent with guideline changes recommending HAART for all pregnant women with HIV infection, regardless of disease severity, and high rates of viral suppression at the time of delivery.12,22
We found that, since 2008, almost half of pregnant women with HIV infection have undergone cesarean delivery annually. The main indication for cesarean delivery in pregnant women with HIV infection from 2009 onward was a history of prior cesarean delivery, not HIV infection. The rate of elective cesarean delivery was close to 60% of all cesarean deliveries throughout the study. During the current study, increasing numbers of pregnant American women without HIV infection were also undergoing cesarean delivery, although the cesarean delivery rate increased less markedly from 21% to 30%.23 A prior analysis from this cohort showed that morbidity was highest with nonelective cesarean delivery and lowest with vaginal delivery, and rates of morbidity in the HAART era were lower than in earlier historical cohorts of pregnant women with HIV infection (19% vs 29–49%).19,24 Our findings suggest that, similar to pregnant women without HIV infection,25 pregnant women with HIV infection, particularly those on HAART with effective viral suppression, should be offered and should consider trial of labor after cesarean delivery (TOLAC) when clinically appropriate.
Women who were first diagnosed with HIV infection during pregnancy, who initiated HAART later in pregnancy, and who had an unsuppressed HIV viral load at delivery were more likely to undergo HIV-indicated cesarean delivery, and these associations were more pronounced over time. These results highlight the importance of continued screening for HIV before pregnancy.26,27 Additionally, once women with HIV infection are pregnant, prompt enrollment in prenatal care and early initiation of antiretroviral therapy are necessary to maximize the likelihood of viral suppression, decrease the risk of mother-to-child transmission, and increase the chances of vaginal delivery, including TOLAC.28 These evidence-based interventions have the promise to further decrease the high rate of cesarean delivery among pregnant women with HIV infection in the United States.
This analysis has limitations. This retrospective observational study across greater than 60 clinical sites does not provide information about how pregnant women with HIV infection were counseled about their options for delivery, including TOLAC, compared with pregnant women without HIV infection. The current study was conducted at tertiary care centers with expertise in HIV and obstetric care and hence these findings may not apply to other clinical settings with fewer resources. The current study did not uniformly collect parity data across the two studies; however, in the subset of women in International Maternal Pediatric Adolescent AIDS Clinical Trials Protocol P1025 in whom parity was collected, the frequency of cesarean delivery, an HIV-associated cesarean delivery, and a suppressed plasma viral load (ie, less than 1,000 copies/mL) at delivery were similar between nulliparous women and the overall cohort. Among women with a history of prior cesarean delivery, we do not have data on indications for the first cesarean delivery. It is possible that some repeat cesarean deliveries resulted after an initial cesarean delivery for HIV transmission. It may be difficult to distinguish these cases as distinct from an HIV-indicated cesarean delivery. Some women may prefer repeat cesarean delivery to decrease perinatal HIV transmission in light of previous counseling in addition to the risks associated with TOLAC. However, it is likely that TOLAC counseling along with counseling about current recommendations for prevention of perinatal HIV transmission could result in vaginal delivery for many of these women.
Maternal HIV infection continues to influence mode of delivery, although to a smaller extent than before use of HAART. For a pregnant women with HIV infection and her obstetric care provider, many factors, both HIV-related and patient preference, may influence the final choice of mode of delivery.24 As in pregnant women without HIV infection, interventions aimed at reducing the cesarean delivery rate among pregnant women with HIV infection will need to target obstetric management as well as HIV-specific clinical scenarios.
1. Jamieson D, Read JS, Kourtis AP, Durant TM, Lampe MA, Dominguez KL. Cesarean delivery for HIV-infected women: recommendations and controversies. Am J Obstet Gynecol 2007;197(suppl):S96–100.
2. Dominguez K, Lindegren ML, D'Almada PJ, Peters VB, Frederick T, Rakusan TA, et al. Increasing trend of cesarean deliveries in HIV-infected women in the United States from 1994 to 2000. J Acquir Immune Defic Syndr 2003;33:232–8.
4. International Perinatal HIV Group, Andiman W, Bryson Y, de Martino M, Fowler M, Harris D, et al. The mode of delivery and the risk of vertical transmission of human immunodeficiency virus type 1—a meta-analysis of 15 prospective cohort studies. The International Perinatal HIV Group. N Engl J Med 1999;340:977–87.
5. European Mode of Delivery Collaboration. Elective caesarean-section versus vaginal delivery in prevention of vertical HIV-1 transmission: a randomised clinical trial [published erratum appears in Lancet 1999;353:1714]. Lancet 1999;353:1035–9.
6. Kind C, Rudin C, Siegrist CA, Wyler CA, Biedermann K, Lauper U, et al. Prevention of vertical HIV transmission: additive protective effect of elective cesarean section and zidovudine prophylaxis. Swiss Neonatal HIV Study Group. AIDS 1998;12:205–10.
7. Kuhn L, Bobat R, Coutsoudis A, Moodley D, Coovadia HM, Tsai WY, et al. Cesarean deliveries and maternal-infant HIV transmission: results from a prospective study in South Africa. J Acquir Immune Defic Syndr Hum Retrovirol 1996;11:478–83.
8. Anderson B, Cu-Uvin S. Pregnancy and optimal care of HIV-infected patients. Clin Infect Dis 2009;48:449–55.
9. Garcia P, Kalish LA, Pitt J, Minkoff H, Quinn TC, Burchett SK, et al. Maternal levels of plasma human immunodeficiency virus type 1 RNA and the risk of perinatal transmission. Women and Infants Transmission Study Group. N Engl J Med 1999;341:394–402.
10. Chappell CA, Cohn SE. Prevention of perinatal transmission of human immunodeficiency virus. Infect Dis Clin North Am 2014;28:529–47.
11. American College of Obstetricians and Gynecologists. Scheduled cesarean delivery and the prevention of vertical transmission of HIV infection. ACOG Committee Opinion No. 219. Washington, DC: ACOG; 1999.
12. Stoszek S, Duarte G, Hance LF, Pinto J, Gouvea MI, Cohen RA, et al. Trends in the management and outcome of HIV-1-infected women and their infants in the NISDI Perinatal and LILAC cohorts, 2002-2009. Int J Gynaecol Obstet 2013;122:37–43.
13. Cooper E, Charurat M, Mofenson L, Hanson IC, Pitt J, Diaz C, et al. Combination antiretroviral strategies for the treatment of pregnant HIV-1-infected women and prevention of perinatal HIV-1 transmission. J Acquir Immune Defic Syndr 2002;29:484–94.
14. Perinatal HIV Guidelines Working Group. Public Health Service Task Force recommendations for use of antiretroviral drugs in pregnant HIV-infected women for maternal health and interventions to reduce perinatal HIV transmission in the United States. 2009. Available at: http://aidsinfo.nih.gov/ContentFiles/PerinatalGL.pdf
. Retrieved November 12, 2017.
15. American College of Obstetricians and Gynecologists. Scheduled cesarean delivery and the prevention of vertical transmission of HIV infection. Committee Opinion No. 234. Washington, DC; 2000.
16. US Department of Health and Human Services. Recommendations for use of antiretroviral drugs in pregnant HIV-1-infected women for maternal health and interventions to reduce perinatal HIV transmission in the United States. 2016. Available at: https://aidsinfo.nih.gov/guidelines/html/3/perinatal-guidelines/0
. Retrieved November 12, 2017.
17. Read J, Tuomala R, Kpamegan E, Zorrilla C, Landesman S, Brown G, et al. Mode of delivery and postpartum morbidity among HIV-infected women: the women and infants transmission study. J Acquir Immune Defic Syndr 2001;26:236–45.
18. Katz I, Leister E, Kacanek D, Hughes MD, Bardeguez A, Livingston E, et al. Factors associated with lack of viral suppression at delivery among highly active antiretroviral therapy-naive women with HIV: a cohort study. Ann Intern Med 2015;162:90–9.
19. Livingston E, Huo Y, Patel K, Tuomala RE, Scott GB, Stek A, et al. Complications and route of delivery in a large cohort study of HIV-1-infected women—IMPAACT P1025. J Acquir Immune Defic Syndr 2016;73:74–82.
20. Aebi-Popp K, Mulcahy F, Glass TR, Rudin C, Martinez de Tejada B, Bertisch B, et al. Missed opportunities among HIV-positive women to control viral replication during pregnancy and to have a vaginal delivery. J Acquir Immune Defic Syndr 2013;64:58–65.
21. Briand N, Jasseron C, Sibiude J, Azria E, Pollet J, Hammou Y, et al. Cesarean section for HIV-infected women in the combination antiretroviral therapies era, 2000-2010. Am J Obstet Gynecol 2013;209:335.e1–12.
22. Baroncelli S, Tamburrini E, Ravizza M, Dalzero S, Tibaldi C, Ferrazzi E, et al. Antiretroviral treatment in pregnancy: a six-year perspective on recent trends in prescription patterns, viral load suppression, and pregnancy outcomes. AIDS Patient Care STDS 2009;23:513–20.
23. Menacker F, Declercq E, Macdorman MF. Cesarean delivery: background, trends, and epidemiology. Semin Perinatol 2006;30:235–41.
24. Read JS, Newell MK. Efficacy and safety of cesarean delivery for prevention of mother-to-child transmission of HIV-1. The Cochrane Database of Systematic Reviews 2005, Issue 4. Art. No.: CD005479. DOI: .
25. Tilden E, Cheyney M, Guise JM, Emeis C, Lapidus J, Biel FM, et al. Vaginal birth after cesarean: neonatal outcomes and United States birth setting. Am J Obstet Gynecol 2017;216:403.e1–8.
26. Rahangdale L, Sarnquist C, Maldonado Y, Cohan D. Patient acceptance of and satisfaction with rapid HIV testing in a labor and delivery setting. J Womens Health (Larchmt) 2008;17:465–71.
27. Prenatal and perinatal human immunodeficiency virus testing: expanded recommendations. Committee Opinion No. 635. American College of Obstetricians and Gynecologists. Obstet Gynecol 2015;125:1544–7.
28. Taylor A, Nesheim SR, Zhang X, Song R, FitzHarris LF, Lampe MA, et al. Estimated perinatal HIV infection among infants born in the United States, 2002-2013. JAMA Pediatr 2017;171:435–42.
Supplemental Digital Content
© 2018 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.