Postpartum contraceptive preferences of HIV-infected women in the era of highly active antiretroviral therapy (HAART) and scheduled cesarean deliveries

Contraception 84 (2011) 150 – 154

Original research article

Postpartum contraceptive preferences of HIV-infected women in the era of highly active antiretroviral therapy (HAART) and scheduled cesarean deliveries☆ Methodius G. Tuulia,⁎, Thinh H. Duongb , Nicole P. Yostb , Jane Ellisb , Charlotte V. Burkeb , Patrice L. Basanta-Henryb , Michael Lindsayb a

Department of Obstetrics of Gynecology, Washington University in St Louis School of Medicine, St Louis, MO 63110, USA b Department of Gynecology and Obstetrics, Emory University School of Medicine, Atlanta, GA 30303, USA Received 22 July 2009; revised 29 October 2010; accepted 1 November 2010

Abstract Background: We sought to determine if postpartum tubal ligation among HIV-infected women changed with the introduction of highly active antiretroviral therapy (HAART) and scheduled cesarean delivery. Methods: Retrospective cohort study of HIV-infected women delivered before (Pre-HAART) and after (Post-HAART) the introduction of HAART and scheduled cesarean delivery. Rates of the primary outcome, postpartum tubal ligation (PPTL), were compared by univariable and multivariable analyses. Results: We found that 34.5% (60/174) of women in the Post-HAART period chose PPTL, compared to 22.0% (18/82) in the Pre-HAART period [unadjusted OR=1.87 (95% CI 1.02–3.44), p=.04]. When stratified by mode of delivery, rates of PPTL were not significantly different between the two periods. Similarly, in multivariable analysis controlling for confounders, rates of PPTL were not different between the two periods [adjusted OR=1.40 (95% CI=0.66–2.99), p=.39]. Conclusions: HIV-infected women on HAART are overall more likely to have PPTL, but cesarean delivery appears to be the facilitator of this choice. © 2011 Elsevier Inc. All rights reserved. Keywords: Postpartum; Contraceptive preferences; HIV-infected; HAART; Cesarean

1. Introduction The number of HIV-infected women, especially of reproductive age, continues to increase [1]. In the care of pregnant women infected with HIV, prevention of vertical transmission from mother to fetus or newborn is a major goal. In 1994, the benefit of maternal antepartum and intrapartum zidovudine therapy followed by newborn

☆

Presented as Poster at the 56th Annual Clinical Meeting of the American College of Obstetricians and Gynecologists, New Orleans, LA, 2008. ⁎ Corresponding author. Department of Obstetrics and Gynecology, Washington University School of Medicine, Campus Box 8064, St. Louis, MO 63110, USA. Tel.: +1 314 747 2362, 314 253 2982 (Pager). E-mail address: [email protected] (M.G. Tuuli). 0010-7824/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.contraception.2010.11.001

zidovudine therapy for the first 6 weeks after birth was published [2]. Specifically, the risk of maternal-infant HIV transmission was reduced by approximately two thirds with this therapy (from 25 to 8%). Subsequently, the vertical transmission rate was further reduced to about 2% with the antepartum use of antiretroviral therapy and scheduled cesarean delivery in select HIV-infected women [3]. The use of highly active antiretroviral therapy (HAART) during pregnancy was demonstrated to reduce the rate of vertical transmission to below 2% [4]. The advent of HAART has also had a remarkable impact on the long-term prognosis of HIV-infected individuals including women. It has been suggested that confidence in vertical transmission reduction strategies and the improvement in long-term maternal prognosis significantly influence the reproductive intentions and, consequently, the contraceptive preferences of HIVinfected women [5,6].

M.G. Tuuli et al. / Contraception 84 (2011) 150–154

Studies prior to the use of antiretroviral therapy showed that HIV infected women were less likely than HIV-uninfected women to report pregnancy intentions and actual pregnancies [7–10]. A study in our facility from 1987 through 1989, prior to the introduction of any interventions for the prevention of vertical transmission, revealed that seropositive women were significantly more likely than seronegative women to undergo tubal sterilization (27% versus 15%) [11]. A subsequent study from 1993 through 2002 at Parkland Memorial Hospital demonstrated a 39% reduction in the choice of postpartum tubal sterilization following the introduction of Pediatric AIDS Clinical Trials Group (PACTG) Protocol 076, which involved the use of zidovudine for the prevention of vertical transmission of HIV [5]. Since then, the use of HAART and scheduled cesarean delivery for select HIV-infected women has become the standard of care in the United States for the prevention of vertical transmission [12]. To date, few studies have reported on the postpartum contraceptive choices of HIV-infected women in the era of HAART. The objective of this study was to evaluate postpartum contraceptive preferences among HIV-infected women and in particular, determine whether the choice of postpartum tubal ligation (PPTL) has changed since the introduction of HAART and scheduled cesarean delivery.

2. Materials and methods This was a retrospective cohort study comparing postpartum contraceptive choices among HIV-infected women delivered at Grady Memorial Hospital (GMH), a 1000-bed inner-city hospital, from January 1, 2002, to December 31, 2006, to those delivered from January 1, 1992, to December 31, 1993. Patients who were delivered between January 1, 1992 and December 31, 1993, were designated the Pre-HAART group. Patients delivering during this period did not benefit from prenatal anti-retroviral therapy and elective cesarean deliveries for prevention of vertical transmission of HIV. Those who were delivered between January 1, 2002, and December 31, 2006, were the PostHAART group. These patients had the opportunity to benefit from antiretroviral therapy and elective cesarean delivery. Pregnant women initiating prenatal care at GMH are offered HIV screening. Those found to be positive are managed in a high-risk obstetrics clinic overseen by a dedicated Maternal Fetal Medicine faculty. In the PostHAART period, all HIV-infected women attending the clinic are offered antiretroviral therapy starting in the second trimester. Patients who are already on therapy prior to pregnancy are generally continued on their regimen. Viral loads are measured in each trimester. In the Pre-HAART period, cesarean sections were reserved for obstetric indications. During the Post-HAART period the mode of delivery offered to HIV-infected women is based on the viral load. Generally, when the viral load is N1000 copies/ml HIVinfected women are offered elective cesarean section at 38

151

weeks. Those with viral loads b1000 copies per milliliter are allowed a trial of vaginal delivery unless there is an obstetric indication which dictates cesarean delivery. During prenatal visits, the attending physician and family planning counselor provide contraceptive counseling, and women are encouraged to select a method. Those who desire permanent sterilization sign consent forms during the antepartum period. Requirements for tubal ligation in the state of Georgia stipulate that women are at least 21 years old and sign consent forms at least 30 days prior to the time of the procedure. For women delivering prior to 40 weeks' gestation, tubal ligation may be performed if consents were signed at least 72 h prior to the procedure. Tubal ligation is performed at the time of cesarean delivery or on the first or second postpartum day following vaginal delivery. For those women who choose other methods of contraception and those who are undecided, the family planning nurse counsels them in the postpartum period. The contraceptive preferences of patients are recorded in their hospital charts and included in the electronic discharge summary. All HIV-infected women who delivered at GMH during the study period were eligible for the study. Women were excluded if they delivered before the gestational age of viability (b24 weeks) or had incomplete records. We reviewed medical records and abstracted demographic information (age, race, education, marital status, employment), delivery information (parity, mode of delivery), CD4 count, viral load and postpartum contraceptive choice (tubal sterilization, oral contraceptives, patch, vaginal ring, intrauterine device (IUD), depo medroxyprogesterone acetate, condoms, abstinence). All data entries were double-checked for accuracy. Since women were not eligible for tubal ligation if they were younger than 21 years old and could not have signed consent forms if they did not have prenatal care, we limited analysis of the choice of tubal ligation to women who were 21 years or older and received prenatal care. Univariable analysis was performed to identify factors that were associated with the choice of tubal ligation as a contraceptive method. Continuous variables were compared using the Student's t test while categorical variables were compared using chi-square or Fisher exact tests as appropriate. Stratification was used to control for individual confounders while multivariable analysis was used to control for multiple confounders simultaneously. The factors to be included in the multivariate analysis were selected from results of the univariate analyses. Backward selection was used to reduce the number of variables in the model. Differences in the hierarchical explanatory models were tested using the likelihood ratio test. All variables that were statistically significant, as well as those of known clinical importance as predictors of tubal ligation were included in the final model. All statistical analyses were completed using STATA software package, version 10, Special Edition (College Station, TX, USA). Tests with p values b.05 were considered statistically significant.

152

M.G. Tuuli et al. / Contraception 84 (2011) 150–154

The study was approved by the Institutional Review Board of Emory University.

3. Results For the study period January 1, 1992, to December 31, 1993, and January 1, 2002, to December 31, 2006, a total of 340 HIV-infected women were delivered at GMH. Of these, 113 were in the Pre-HAART group (January 1, 1992, to December 31, 1993), while 227 were in the Post-HAART group (January 1, 2002, to December 31, 2006). Table 1 depicts selected patient characteristics. The mean age did not differ significantly between the two groups. However, when subdivided into age categories, the two groups differed significantly, with more patients in the Pre-HAART group in the 21–34-year age category. The two groups also differed significantly with respect to race, marital status, educational level, proportion receiving prenatal care and mode of delivery. The rate of cesarean delivery was significantly higher in the Post-HAART group. Rates of the primary outcome variable, PPTL, in the two groups are shown in Fig. 1. A total of 84 women were excluded from comparison of choice of tubal ligation

Table 1 Selected characteristics of HIV-infected women delivered at Grady Memorial Hospital, January 1, 1992 to December 31, 1993 and January 1, 2002 to December 31, 2006

Age, years (mean±S.D.) Age (y) b21 21–34 N34 Race African-American Caucasian Hispanic Other Marital status Single Married Divorced Parity Primiparous Received prenatal care Education Kindergarten Middle Junior high High General equivalency diploma College Received HAART CD4 count b200 Viral load b1000 copies/ml Cesarean delivery

Pre-HAART (n=113)

Post-HAART (n=227)

p value

26.9 ± 5.1

27.5 ± 6.1

.40

10 (8.9%) 95 (84.8%) 7 (6.3%)

30 (13.3%) 158 (69.9%) 38 (16.8)

.01

109 (96.5%) 2 (1.8%) 2 (1.8%) 0 (0%)

189(85.1%) 16 (7.2%) 16 (7.2%) 1(0.5%)

.02

89 (92.7%) 5 (5.2%) 2 (2.1%)

145 (65%) 48 (21.5%) 30 (13.5%)

b.01

27 (23.9%) 89 ((78.8%)

59 (26.0%) 201(88.9%)

.68 .01

0 (0%) 3 (3.3%) 7(7.8%) 79 (87.8%) 0 (0%) 1 (1.1%) 0 (0%) 18 (17.8%) 9 (8 %)

11 (5.6%) 37(18.7% 13 (6.6%) 120 (60.6%) 7 (3.5%) 10 (5.1%) 200 (90.9%) 52 (24.2%) 126 (65%) 140 (66.7%)

b.01

b.01 .37 b.01

because they were younger than 21 years old (n=34), had no prenatal care (n=44) or both (n=6). In univariable analysis, women delivered during the Post-HAART period were significantly more likely to choose PPTL compared to those delivered during the Pre-HARRT period [34.5% (60/174) vs. 22.0% (18/82), unadjusted OR 1.87 (95% CI 1.02– 3.44), p=.04]. In addition, older age, multiparity and cesarean delivery were significantly associated with the choice of PPTL (Table 2). When stratified by mode of delivery, there was no evidence of interaction between mode of delivery and the Pre-HAART or Post-HAART period on the choice of PPTL (χ2 test for heterogeneity, p=.65). The choice of PPTL was not significantly different between women delivered in the Post-HAART and Pre-HAART periods for either mode of delivery [cesarean: 34.0% vs. 22.0%, OR 1.83 (95% CI 0.37-9.13), p=.46; vaginal: 24.0% vs. 17.0%, OR 1.52 [0.71-3.25], p=.53) (Table 3). The pooled OR suggested no association between choice of PPTL and the period of delivery after controlling for mode of delivery (Mantel-Haenszel pooled OR 1.49 [95% CI=0.72 – 3.07, p=.28]). Finally, in multivariable analysis controlling for age, parity, CD4 count and mode of delivery, choice of PPTL was not significantly different between the Post-HAART and Pre-HAART groups [adjusted OR=1.40 (95% CI=0.66–2.99), p=.39]. Table 4 shows the reversible contraceptive choices among HIV-infected women. To test the hypothesis that increased confidence in vertical transmission strategies results in increased reproductive intentions among HIV-infected women, we used the choice of a long-term method of contraception as a surrogate marker for absence of further reproductive intention that is not solely dependent on a surgical procedure. For purposes of this analysis, tubal ligation, Norplant and IUD were considered long-term methods, while condoms only, depomedroxyprogesterone acetate, oral contraceptives, ring and patch were considered short term methods. From this analysis, the choice of a longterm contraceptive method was not significantly different between the Post-HAART and Pre-HAART groups [35% vs. 31 %, OR 1.00 (0.71–1.34), p=.98].

4. Discussion Our results indicate that HIV-infected women in our facility were significantly more likely to choose PPTL during the period when HAART and scheduled cesarean sections were used for prevention of vertical transmission. After controlling for age, parity CD4 count and mode of delivery, this association was no longer significant. These findings differ from those of previous studies on the choice of tubal ligation as a contraceptive method among HIV-infected women. In a study by Lindsay et al. [11], HIVinfected women were more likely to select tubal ligation when compared to HIV-uninfected controls during a period of time when no interventions were available for the

M.G. Tuuli et al. / Contraception 84 (2011) 150–154

153

Fig. 1. Tubal ligation choices of HIV-infected women delivered at Grady Memorial Hospital, January 1, 1992, to December 31, 1993, and January 1, 2002, to December 31, 2006.

prevention of vertical transmission. That finding cannot be directly compared to ours, since different controls were used in the two studies (HIV-uninfected vs. HIV-infected with neither HAART nor cesarean delivery). In a subsequent study by Stuart et al. [5], the choice of postpartum tubal ligation as a method of contraception was reduced significantly with the introduction of the PACTG Protocol 076, which involved the use of zidovudine for prevention of vertical transmission. The authors attributed this to confidence in the proven efficacy of this protocol in the prevention of vertical transmission. Following the above premise, we expected to observe a further reduction in the choice of tubal ligation in the era of HAART and scheduled cesarean delivery, a more effective strategy for the prevention of perinatal HIV transmission. Contrary to this expectation, our results showed a higher rate of tubal ligation in the Post-HAART group. This unanticipated finding is likely the result of confounding, since the increase was not significant after controlling for age, parity, CD4 count and mode of delivery. It is noteworthy that after controlling for only mode of delivery by stratification, the association was also no longer significant. This suggests that Table 2 Factors associated with choice of tubal ligation in univariable analysis among HIV-infected women delivered at Grady Memorial Hospital, January 1, 1992 to December 31, 1993 and January 1, 2002 to December 31, 2006a

Delivery in Post-HAART period Age per year Cesarean delivery Multiparity a

n

(OR) 95% CI

p-value

174 256 129 196

1.87 (1.02-3.44) 1.06 (1.01-1.13) 2.09 (1.21-3.62) 5.23 (2.14-12.75)

.04 .03 b.01 b.01

Analysis limited to women 21 years or older who had prenatal care.

the use of cesarean delivery in the current protocol probably facilitates patients' decision to undergo tubal ligation, since it can be performed at the same time without the need for an additional procedure. Furthermore, the finding that choice of long-term contraceptive methods was not significantly different between the two groups suggests that reproductive intentions have probably not changed with the use of HAART and scheduled cesarean delivery. Rather, women who have no further reproductive intentions chose tubal ligation if they were undergoing cesarean delivery and another long-term method if they were not. However, the true reason for the choice of tubal ligation can only be accurately ascertained in a prospective study. Our study has a number of strengths. The large sample size and comprehensive data enabled us to control for several potential confounders. In addition, the staff at the study site has been remarkably stable, minimizing the chance that bias in contraceptive counseling during the two periods influenced patients' decisions on PPTL. However, the possibility Table 3 Choice of tubal ligation in the Pre-HAART and Post-HAART groups by mode of delivery among HIV-infected women delivered at Grady Memorial Hospital, January 1, 1992-December 31, 1993 and January 1, 2002December 31, 2006a Tubal ligation, n (%) Pre-HAART Post-HAART (n=113) (n=227) Vaginal (n=149) 17 (17%) Cesarean (n=174) 2 (22%) Pooled OR -

16 (24%) 48 (34%) -

OR (95% CI)

p value

1.52 (0.71-3.25) .29 1.83 (0.37-9.13) .46 1.49 (0.72-3.07) .28

Mantel-Haenszel χ2 test for heterogeneity, p=.65. a Analysis limited to women 21 years or older who had prenatal care.

154

M.G. Tuuli et al. / Contraception 84 (2011) 150–154

Table 4 Reversible contraceptive choices among HIV-infected women delivered at Grady Memorial Hospital, January 1, 1992 to December 31, 1993 and January 1, 2002 to December 31, 2006

rather than a change of confidence in the effectiveness of vertical transmission prevention strategies.

Method

Pre-HAART (n=94)

Post-HAART (n=163)

References

Depo medroxyprogesterone acetate, n (%) Oral contraceptives, n (%) Norplant, n (%) Condoms only Patch, n (%) IUD, n (%) Ring, n (%) Abstinence, n (%)

52 (55.3) 21 (22.3) 21 (22.3) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

64 (39.3) 37 (22.7) 0 (0) 23 (14.1) 29 (17.8) 6 (3.7) 2 (1.2) 2 (1.2)

[1] Schneider E, Glynn MK, Kajese T, McKenna MT. Epidemiology of HIV/AIDS — United States, 1981–2005. MMWR Morb Mortal Wkly Rep 2006;55:589–92. [2] Connor EM, Sperling RS, Gelber R, et al. Reduction of maternal-infant transmission of human immunodeficiency virus Type 1 with zidovudine treatment. Pediatric AIDS Clinical Trials Group Protocol 076 Study Group. N Engl J Med 1994;331:1173–80. [3] The International Perinatal HIV Group. The mode of delivery and the risk of vertical transmission of human immunodeficiency virus type 1 — a meta-analysis of 15 prospective cohort studies. N Engl J Med 1999; 340:977–87. [4] Cooper ER, Charurat M, Mofenson L, 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. [5] Stuart GS, Castano PM, Sheffield JS, et al. Postpartum sterilization choices made by HIV-infected women. Infect Dis Obstet Gynecol 2005;13:217–22. [6] Kirshenbaum SB, Hirky AE, Correale J, et al. “Throwing the dice”: pregnancy decision-making among HIV-positive women in four U.S. cities. Perspect Sex Reprod Health 2004;36:106–13. [7] Lester P, Partridge JC, Chesney MA, et al. The consequences of a positive prenatal HIV antibody test for women. J Acquir Immune Defic Syndr Hum Retrovirol 1995;10:341–9. [8] Kline A, Strickler J, Kempf J. Factors associated with pregnancy and pregnancy resolution in HIV seropositive women. Soc Sci Med 1995;40:1539–47. [9] Sunderland A, Minkoff HL, Handte J, et al. The impact of human immunodeficiency virus serostatus on reproductive decisions of women. Obstet Gynecol 1992;79:1027–31. [10] Levin L, Henry-Reid L, Murphy DA, et al. Incident pregnancy rates in HIV infected and HIV uninfected at-risk adolescents. J Adolesc Health 2001;29(3 Suppl):101–8. [11] Lindsay MK, Grant J, Peterson HB, et al. The impact of knowledge of human immunodeficiency virus serostatus on contraceptive choice and repeat pregnancy. Obstet Gynecol 1995;85(5 Pt 1):675–9. [12] American College of Obstetricians and Gynecologists (2000; reaffirmed 2003). Scheduled cesarean delivery and the prevention of vertical transmission of HIV infection. ACOG Committee Opinion No. 234. Washington, DC: American College of Obstetricians and Gynecologists.

that the counseling offered could have changed over the years cannot be entirely ruled out. The availability of both paper and electronic medical records enabled us to doublecheck patient information for accuracy. Finally, the choice of a robust primary outcome (tubal ligation) further reduced ascertainment bias. Along with these strengths, it is important to consider some limitations when interpreting the results. Retrospective cohort studies by design have the potential for confounding and bias. While the comprehensive data allowed us to adjust for relevant confounders, there is the potential for residual confounders, which we did not collect data on and therefore could not control for. Also, we included only confirmed tubal ligations in the primary outcome. While this ensured that we captured only those who really wanted to have PPTL, it resulted in the exclusion of patients who desired the procedure but could not have it for one reason or another. This may have resulted in an underestimation of patients' preference for PPTL. Despite these limitations, our results significantly add to the existing literature on contraceptive preferences of HIV-infected women, therefore assisting physicians in counseling these patients. In conclusion, our results show that pregnant HIV-infected women on HAART are overall more likely to choose PPTL. Cesarean delivery appears to be the facilitator of this choice