The impact of health care providers on female sterilization among HIV-positive women in Brazil

Abstract

This paper explores the reproductive preferences and outcomes of HIV-positive women in two cities in Brazil. We used three types of data, all drawn from women who delivered in public sector hospitals: (1) clinical records of 427 HIV-positive women; (2) pre- and postpartum in-depth interviews with 60 HIV-positive women; and (3) a prospective survey carried out among 363 women drawn from the general population. The HIV-positive samples were collected on women who had prenatal care between July 1999 and June 2000, and the general population survey was conducted with women who started prenatal care between April 1998 and June 1999. Among the women in the clinic sample, we found dramatic differences in the proportion sterilized postpartum: 51% in São Paulo vs. 4% in Porto Alegre, compared to 3.4% and 1.1%, respectively, of women in the general population. Our qualitative data suggest that HIV-positive women in this study had strong preferences to have no more future children and that female sterilization was the preferred way to achieve this end. Therefore, we conclude that the large difference in rates is mainly due to HIV-positive women's differential access to sterilization in the two settings. In-depth interviews revealed that women in São Paulo were often encouraged by clinic staff to be sterilized postpartum. In contrast, HIV-positive women in Porto Alegre clinics were not offered sterilization as an option and those who requested it were repeatedly put off. The striking difference found in the frequency with which doctors provide postpartum sterilization to seropositive women in our study sites deserves attention and discussion in the respective medical communities. At the higher level of national policy on reproductive rights, there may be grounds for reopening discussion about the norms regarding postpartum procedures, and for devoting far more resources to expanding contraceptive options.

Introduction

The Brazilian constitution accords women and men the right to freely decide on the number and spacing of their children as well as the means to implement those reproductive decisions. But it is one thing to legislate the means to exercise reproductive rights and another entirely for citizens to realize those rights in practice. This is particularly true for vulnerable groups such as HIV-positive women. On the one hand, HIV-positive women might have an especially strong desire for highly effective contraception that imposes a minimum of risk and inconvenience. Either because they have reached their ideal number of children, they are carriers of HIV, or some combination of these and other reasons, women with HIV may be strongly motivated to avoid future pregnancies. Accordingly, they should be provided with a full range of contraceptive options, especially in a country like Brazil where therapeutic abortion is illegal except in cases of rape or to save the life of the woman. On the other hand, health care providers and institutions might be inclined to dissuade such women from future childbearing. Moreover, because seropositive women have frequent and extended contact with health care providers, it is important to ensure that they are given the opportunity to make informed choices regarding contraception and childbirth rather than simply following whatever the “standard” practices or policies may be in the places where they receive care.

What are the contraceptive options available to seropositive women after giving birth? If they are more concerned with effectiveness than reversibility, and also with avoiding additional surgical procedures, then postpartum female sterilization might be an attractive option. On the other hand, sterilization, although widespread, has a checkered history in Brazil, and has recently been regulated by legislation that places strict limits on the use of postpartum procedures. Also, in Brazil as elsewhere, there is regional variation in the use of sterilization as a contraceptive method and in the fraction of procedures that are performed postpartum (BEMFAM, 1997; EngenderHealth, 2002; Westoff and Davis, 2000), in part due to differences in physicians’ attitudes about sterilization or local rules and practices that may either limit or encourage sterilization for women who want no more children (Bumpass, Thomson, & Godecker, 2000; Caetano and Potter, 2004; Potter, 1999).

In this paper, we assess the prevalence of postpartum female sterilization, the influence that health care providers and the HIV diagnosis have on the use of this procedure and the extent to which demand for postpartum sterilization is met among seropositive women in São Paulo and Porto Alegre, Brazil. Before turning to our methods and results, we briefly review trends and differentials in female sterilization in Brazil as well as the recent legislation affecting the legal status of sterilization. We then summarize the evolution of the HIV epidemic among Brazilian women, as well as some findings regarding future childbearing intentions and contraceptive practice among seropositive women in other countries.

As the demand for fertility limitation increased in Brazil during the past three decades, female sterilization became the most widely used contraceptive method. The rapid spread of female sterilization has raised a number of concerns. Until recently, sterilization had a dubious legal status at best, and was not reimbursed by either public or private insurance. The surgery was paid for either by the patient or her family, or sometimes doctors absorbed their costs, and it was most often performed simultaneously along with another surgical procedure, often a Caesarean section. One fear was that the demand for female sterilization was leading to a large number of unnecessary Caesarean deliveries (Faundes and Cecatti, 1993). Questions were also raised about possible efforts to control the growth of the poor population by this means (Brasil Congreso Nacional, 1993; Folha de São Paulo, 1992; Pernambuco Assembléia Legislativa 1992; Rodrigues Filho, 1994), as well as the use of sterilization for electoral purposes (Caetano, 2000; Caetano and Potter, 2004).

Among married women, sterilization prevalence went from 26.9% in 1986 to 40.1% in 1996 according to the nationally representative surveys conducted in those years (BEMFAM (1987), BEMFAM Sociedade Civil Bem-Estar Familiar (1997)). In addition to the usual differentials according to age and parity, there are two other gradients that mark the prevalence of sterilization in Brazil. It is higher among the approximately 25% of women who deliver in private hospitals, in part due the extremely high Caesarean rate in this sector (BEMFAM, 1997, Potter, Berquó, & Perpétuo, 2001; Hopkins, 2000). Second, there is a pronounced regional variation with the highest rates found in the Northeast and Central West regions, and the lowest found in the three states comprising the South region. While the factors underlying these differentials are not well understood, in the Northeast, sterilization is more likely to be provided free of charge in public sector hospitals than it is in the South and Southeast where a larger fraction of these procedures are performed for a fee or in conjunction with a Caesarean delivery in private hospitals (Caetano and Potter 2004). In a previous paper (Potter et al., 2003), strong demand for postpartum sterilization was found among both private and public patients who said they wanted no more children (38% and 46%, respectively, when asked in the first-half of pregnancy). Also, a much higher proportion of women who delivered in the private sector were able to realize their preference for a postpartum sterilization compared to women in the public sector (69% vs. 33%). Nearly half the women who delivered in public hospitals said they wished they had been sterilized before leaving the hospital after the birth of their last child.

In August 1997, Law 9263, for the first time, legalized the practice of female (and male) sterilization, but also placed explicit regulations upon its use. The law and the rules for its implementation issued by the Ministry of Health (in 1997 and 1999) (Brasil Ministério da Saúde, 1997) specified that to request a voluntary surgical sterilization men and women had to be at least 25 years old or have two living children and, if married, had obtained spousal consent. In addition, the law required a minimum 60 day waiting period between the time of the request and the surgery and prohibited sterilization procedures during delivery, following an abortion or in the 41 days following a birth. The law made it possible, also for the first time, for clinics and hospitals to be reimbursed by the public health insurance system for voluntary sterilization procedures, but only after they were licensed for this purpose by a state or municipal level authority. Licensing was in turn contingent on the ability to provide both counseling and services for alternative methods of contraception. The law also prescribed penalties for those who performed sterilizations that did not meet the stated criteria.

The law does, however, define two conditions of “proved necessity” that allow a woman to have a voluntary sterilization performed during delivery, following abortion or in the immediate postpartum period. As during the pre-1997 sterilization regime, women who have a history of successive previous Caesareans can continue to use this exception under the new law. In addition, postpartum sterilizations are allowed for women who are carriers of a previously existing illness and for whom a second surgery or additional exposure to anesthesia would represent a major health risk. In these cases, the medical indication must be attested to in writing by two doctors (Brasil Ministério da Saúde, 1999).

Both these exceptions are open to interpretation by individual doctors; interpretations which, in turn, are framed by the institutional context in which they work. In the absence of clear guidelines and with contradictory rulings by the federal and state-level medical councils, there would appear to be considerable ambiguity regarding what exactly constitutes an exception. For instance, it is unclear how many previous Caesarean sections are considered “too many” and which diseases qualify and what constitutes a “major risk”. Similarly, there are no official guidelines from the Ministry of Health regarding a woman's HIV/AIDS status and eligibility for postpartum sterilization. It is not clear whether a woman has to have AIDS and a compromised immune system in order to be excepted from the prohibition on postpartum procedures, if seropositivity is a sufficient cause, or whether either condition matters. Apart from our study, to date, there is no information on HIV-positive women's ability to obtain postpartum sterilization or how the Brazilian health care services and professionals are addressing this issue.

Brazil has approximately 89,527 officially documented cases of AIDS among women, the vast majority of whom are of reproductive age (Brasil Ministério da Saúde, 2003). Szwarcwald and Carvalho (2001) estimate that about 218,000 women 14 years and older are infected with HIV. In addition, Brazil has seen a dramatic drop in the male to female ratio of the disease, which fell from 28:1 in 1985 to 2:1 in 1997. Since 1994, the ratio has been 1:1 among 15–19 year-olds (Vermelho, Silva, & Costa, 1999). Prevalence rates vary dramatically across Brazil and higher rates are found in the South and Southeast regions of the country. In addition, women become infected with HIV primarily through their male sexual partners, confirming that heterosexual transmission is now the primary engine of the epidemic among women (Szwarcwald and Bastos, 1998). In 2000, for example, heterosexual transmission was responsible for over 93% of HIV cases among women 13 and over compared to 62% of the cases in 1991 (Brasil Ministério da Saúde, 2003). As is true in a large number of contexts (Barbosa, 2003), the HIV/AIDS epidemic disproportionately strikes the poor and less-educated and this pattern is even more pronounced among Brazilian women. In 1991, 51% of women with AIDS had not completed their primary school education compared to 39% of seropositive men who had less than a primary education. By 2000, this proportion had increased for women and men to 67% and 60%, respectively (Brasil Ministério da Saúde, 2001). At the same time, antiretrovirals are available free of charge to all HIV-infected Brazilians (World Health Organization, 2000).

The publicly funded Brazilian health care system provides HIV-infected women in urban Brazil with two main types of specialized prenatal care: hospital-based outpatient clinics and free-standing clinics. The hospital-based clinics, which are either within or nearby the hospital, may or may not be affiliated with medical school and vary from a highly integrated and organized HIV-specific services with specially trained doctors, nurses and social workers to a loose collection of a small number of physicians who oversee the care of the HIV-infected pregnant women. The free-standing clinics also vary in the complexity of their operations: they can range from one obstetrician and nurse to a more integrated team of health professionals that include doctors, nurses, psychologists and the like. What sets the hospital outpatient clinic apart is its clear referral system for childbirth. In large cities, such as São Paulo and Porto Alegre, a woman who receives prenatal care at a hospital clinic is guaranteed a bed in that hospital at the time of her delivery. The free-standing clinics, on the other hand, are not affiliated with any particular hospital. This means that, at the time of delivery, HIV-positive women who went for prenatal care at the free-standing clinics must search for a hospital that has a bed available to admit them. In this study, we draw from both types of specialized publicly funded clinics available to HIV-positive women in Brazil.

What little is known about the reproductive intentions of HIV-positive women in Brazil and how HIV status influences their goals comes from two recent studies done in São Paulo. Paiva et al. (2002) surveyed 1068 HIV infected women and found that 87% no longer wanted to have children but these figures were lower among younger women: 75% of 25 to 29 year-olds and 61% of 18 to 24 year-olds. The majority of women who still wanted to have children related their desire for future childbearing to a sense of “fulfillment as a woman,” to the desire to constitute a family or to not yet having the desired number of children (Paiva et al., 2002). A second study conducted in 1997 among 148 women of an outpatient clinic in a reference center for STD/AIDS in São Paulo found that 79% wanted no more children in the future. HIV status, however, was not the reason given for wanting an end to childbearing. Instead, “having many children” was the factor associated with not wanting more children (Santos et al., 2002).

In 1994, the Pediatric AIDS Clinical Trials Group Protocol 076 (known as ACTG 076), in a placebo-controlled trial, showed that treatment with zidovudine reduced mother-to-child transmission of HIV by nearly 70% (Connor et al., 1994). Treatment in the ACTG 076 trial included administration of zidovudine (AZT) to the pregnant woman five times daily during pregnancy, intraveneously during labor and delivery and to the infant orally every 6 h for six weeks following birth. Zidovudine treatment was effective for a wide variety of viral loads and was confirmed in subsequent analyses with more mother-child pairs 18 months after birth (Sperling et al., 1996). Zidovudine treatment was therefore recommended for use for all HIV-positive pregnant women. Current drug regimens vary, depending on when the HIV is diagnosed in the mother, and whether her treatment regimen includes other antiretroviral medications. Moreover, a recent study found that a single dose of nevirapine to the mother, in addition to the oral zidovudine treatment begun at 28 weeks’ gestation, was highly effective in reducing perinatal transmission of HIV (Lallemant et al., 2004).

Before the implementation of the ACTG 076 protocol, however, most US studies showed no significant differences in pregnancy incidence, abortion rates or contraceptive use among HIV-positive women compared to their HIV negative counterparts (Williams, Watkins, & Risby, 1996; Lindsay et al., 1995). Moreover, studies that measured the impact of knowledge of the benefits of prenatal antiretroviral treatment found that this additional information had no effect on fertility rates, contraceptive use or abortion rates among HIV-positive women in the US (Healton, Taylor, Messeri, Weinberg, & Bamji, 1999; Smits et al., 1999). Among the US studies, only Bedimo, Bessinger, & Kissinger (1998) found significant differences between HIV-positive and HIV-negative women across multiple reproductive outcomes, which, in addition to higher rates of abortion, included a decline in fertility rates and an increase in sterilization rates.

In contrast to the US studies, European and Australian studies done prior to the use of the ACTG 076 protocol consistently found that, in contrast to their HIV-negative counterparts, HIV-positive women had lower fertility rates, increased sterilization rates, and increased incidence of voluntary termination of pregnancy (Thackway et al., 1997; Stephenson & Griffioen, 1996; Lindgren et al., 1998; De Vincenzi et al., 1997). After the introduction of ACTG 076, Canadian and European cohorts of HIV-positive women continue to have lower fertility rates (Lindgren et al., 1998; Van Benthem et al., 2000; Hankins, Tran, & Lapointe, 1998). Van Benthem et al. (2000) found that rates of induced abortion were somewhat higher more than four years after HIV diagnosis compared to the four-year period before diagnosis (45% and 42%, respectively). Hankins et al. (1998), on the other hand, found that rates of induced abortion fell after the ACTG 076 protocol was introduced. Significantly, Lindsay et al. (1995) found that HIV-positive women were nearly three times more likely than HIV-negative women to be sterilized postpartum, even after controlling for parity, age, race, and marital status. On the other hand, in countries where the ACTG 076 protocol is not available, such as Rwanda and Burkina Faso, studies found no difference in pregnancy incidence or contraceptive use between seropositive and seronegative women (Allen et al., 1993; Nebie et al., 2001).

Focusing exclusively on the impact of health care providers and policies on sterilization rates among HIV-positive women, Pinchun (1994) found hospital policies to prevent vertical transmission of HIV highly influenced outcomes in Chonburi, Thailand: 100% of seropositive women who went to the hospital prior to 24 weeks gestational age had their pregnancy terminated, 44.4% of whom subsequently had a tubal ligation. Figueroa-Damian and Villagrana-Zesati (2001) found in Mexico, where postpartum sterilization is widely available, that 49 of 72 (68%) HIV-positive women were sterilized postpartum. Sterilization acceptance was associated with already having a child, prenatal care from 1995 forward and having at least one child infected with HIV. The authors speculate that prenatal care physicians may have influenced women's decisions to become sterilized (Figueroa-Damian & Villagrana-Zesati, 2001).

While the evidence is still not entirely clear that HIV-positive women differ from other women in childbirth preferences or practices, in some countries they appear to be less likely to continue childbearing. Unfortunately, it is not possible to know whether these observed differences in existing studies are due to differences between seropositive and seronegative women in their childbearing preferences or because HIV-positive women have better (or more frequent) access to services, or are more easily convinced to follow the clinic/hospital policies regarding pregnancy termination or sterilization.

This study adds to the current literature of the reproductive experiences of HIV-positive women because it provides, for the first time, an assessment of how Brazil's recently enacted law regulating sterilization is implemented among women who are HIV-positive. Moreover, it calls into question the merits of the law as currently written. We build on prior research which showed that certain groups of Brazilian women are more likely to become sterilized than others, despite similar levels of preference for the procedure. And, to our knowledge, this is the first prospective study to explore sterilization preferences and outcomes among different groups of seropositive women in Brazil. On a more global scale, this paper adds to what we know about how HIV-positive women negotiate different medical contexts and how that negotiation is mitigated or influenced by the accepted medical practices in which she seeks care.