Attitudes toward pregnancy among women enrolled in prevention of mother-to-child transmission of HIV (PMTCT) services in Moshi, Tanzania

Abstract

For pregnant women living with HIV (WLWH), feelings about pregnancy may influence their emotional well-being and health seeking behaviors. This study examined attitudes toward pregnancy and associated factors among women enrolled in prevention of mother-to-child transmission of HIV (PMTCT) services in Moshi, Tanzania. 200 pregnant WLWH were enrolled during their second or third trimester of pregnancy and completed a structured survey. Univariable and multivariable regression models examined factors associated with attitudes toward pregnancy, including demographics, interpersonal factors, and emotional well-being. Attitudes toward the current pregnancy were generally positive, with 87% of participants reporting feeling happy about being pregnant. In the final multivariable model, having higher levels of partner support, being newly diagnosed with HIV, and having fewer children were significantly associated with more positive attitudes toward their pregnancy. Findings point to a need for tailored psychosocial support services in PMTCT, as well as comprehensive reproductive health care for WLWH.

Background

Despite a large body of research on maternal and child health outcomes in the context of prevention of mother-to-child transmission of HIV (PMTCT) programs, little is known about how women living with HIV (WLWH) feel about being pregnant. A woman’s attitudes toward her pregnancy may impact not only her emotional well-being, but also her health seeking behavior. Positive maternal attitudes toward pregnancy are associated with improved maternal and child health outcomes (1–3), and therefore are important to understand and address among WLWH. Pregnancy attitudes are not uniformly defined in the literature, but this manuscript uses the term to encompass both pregnancy intention and feelings about pregnancy (1,4–6).

In Africa, about one-third of all reported pregnancies are unintended (7,8). Factors that may influence pregnancy intention range from sociodemographic to behavioral (9,10); in Tanzania, women with stable relationships, high socioeconomic status, and access to family planning services were more likely to report that their pregnancy was intended (10). Studies that have examined pregnancy among WLWH in sub-Saharan Africa have found that unintended pregnancies are higher in WLWH compared to their HIV-negative counterparts (11,12). Unintended pregnancies may pose a threat to positive health outcomes, as studies have found that unintended pregnancies contribute to poor antiretroviral therapy (ART) outcomes and increased risk of mother-to-child transmission among pregnant WLWH (2,13,14).

Increased coverage of ART and greater options for safe conception contribute to changing circumstances for women living with HIV that may impact pregnancy attitudes (15,16). With evidence that ART adherence and subsequent viral suppression can virtually eliminate the chance of HIV transmission during sexual intercourse, clinical monitoring and counseling can help HIV-serodifferent couples to safely conceive a child and prevent HIV transmission during conception (17,18). With effective viral suppression through ART and engagement in PMTCT programs, it is safe to conceive and give birth to an HIV-negative child (14,18). However, misinformation about the risk of transmission and concerns about stigma remain common (19–21). As a result, a woman’s HIV status may still influence her intention to become pregnant (20,22,23), and in turn may make her feel less positive toward her pregnancy and have important implications for the health of both the mother and child. At the same time, WLWH often express a desire for children, and sometimes point to their HIV status as a motivating factor for conceiving, as they say children provide a purpose in life and allow them to maintain a sense of normalcy and uphold societal expectations (24,25).

In Tanzania, the robust implementation of PMTCT programs has increased mothers’ linkage to HIV care and access to ART (26,27), allowing for a reduction in both vertical transmission to the child and horizontal transmission to partners (20). In sub-Saharan Africa, a disproportionate burden of new HIV infection falls on women of reproductive age, who may still have fertility desires (28,29). Understanding factors that influence pregnancy attitudes among WLWH may inform services to help women achieve their pregnancy desires and contribute to improved health outcomes for a growing population of vulnerable women in their reproductive years. This study aims to examine attitudes toward a current pregnancy and associated factors among pregnant WLWH enrolled in PMTCT care in Moshi, Tanzania.

Methods

Study setting and participants

This paper presents a cross-sectional analysis of baseline data from an observational cohort study that enrolled 200 pregnant WLWH attending antenatal care at nine clinics in Moshi municipality, located in the Kilimanjaro Region of Tanzania. The nine clinics included six urban and three rural facilities. A more detailed overview of the study and setting have been published elsewhere (30). All of the study sites followed the Tanzanian national protocol for PMTCT, which includes the provision of HIV clinical services and ART free of charge (31). Women were initiating PMTCT care under Option B+ guidelines recommended by the World Health Organization, whereby all pregnant WLWH initiate ART during antenatal care (ANC) and continue using ART for life (32). Patients receive integrated ANC and HIV care in the PMTCT clinic; monthly appointments include both ANC services and ART management.

Women were eligible for the study if they were 18 years or older, at least 16 weeks pregnant, able to give informed consent, fluent in Swahili, and had been using ART for at least one month. Of 436 pregnant WLWH who attended ANC at the nine study clinics during the enrollment period, 221 were screened for study enrollment; 8 declined to participate and 13 were not eligible (30). A total of 200 women were enrolled between July 2016 and August 2017, including both women with established HIV diagnoses and women who were newly diagnosed with HIV during their current pregnancy.

Procedures

Study participants completed a structured baseline survey, which was orally administered in Swahili by trained research assistants. The survey took approximately 60 minutes to complete. Survey data were collected on a paper form which was then entered into a secure online database using double data entry to ensure data quality. Study procedures received ethical approval from the institutional review boards at Duke University, Kilimanjaro Christian Medical University College, and the Tanzanian National Institute for Medical Research (NIMR), and all participants provided informed consent.

Measures

Measures were translated from English to Swahili, back-translated to English, and reviewed by the bilingual research team for quality and cultural compatibility. All measures included in this analysis were assessed at the baseline (i.e., pregnancy) timepoint.

Outcome measure: Pregnancy attitudes

The primary outcome was attitudes toward the current pregnancy. We adapted a measure of “pregnancy desirability” that was developed by Speizer and colleagues (6). We selected items that our multi-national team felt captured local concepts of pregnancy intention (e.g., “This pregnancy is a result of a plan I made to get pregnant”), timing (e.g., “This pregnancy came at the right time of my life”), partner’s desire for pregnancy (e.g., “My partner wanted me to become pregnant”), and feelings about the pregnancy (e.g., “When I learned I was pregnant, I felt happy”). The final measure included eight items (Table 2). For each item, participants responded on a scale from 0 (Strongly Disagree) to 3 (Strongly Agree). Items were summed (0–24), with higher scores indicating more positive attitudes about the pregnancy. The measure exhibited strong internal reliability in the current sample (α=0.91).

Table 2.

Pregnancy attitudes at baseline (n = 200)

	Agree or Strongly Agree (%)	Disagree or Strongly Disagree (%)
I feel happy about being pregnant.	87.0	12.5
When I learned I was pregnant, I felt happy.	85.5	14.5
This pregnancy came at the right time of my life.	75.0	25.0
Before getting pregnant, I knew that I wanted to have a baby.	73.0	26.5
This pregnancy is a result of a plan I made to get pregnant.	71.0	29.0
My partner wanted me to become pregnant.	69.5	30.0
I looked forward to telling my friends that I was pregnant.	53.5	46.5
If I could do it all over again, I would still choose to be pregnant at this time.	49.0	51.0

^aPercentages may not add up to 100% due to participants who declined to respond

Covariates

Previous research in Tanzania suggests that mistimed and unwanted pregnancies are associated with both intrapersonal and interpersonal factors (10). In examining factors that are associated with attitudes toward pregnancy among WLWH, we therefore chose to examine co-variates at both the individual and interpersonal levels.

Demographics.

We assessed several pertinent demographic variables including age, relationship status, level of education, and number of living children. Women were classified as having a new HIV diagnosis if they first learned of their HIV status during the current/index pregnancy, and were classified as having an established diagnosis if they had known their HIV status prior to the index pregnancy.

Individual-level characteristics.

Depressive symptoms were assessed using the Edinburgh Postnatal Depression Scale (EPDS; α=0.87) (33). Self-acceptance of one’s HIV diagnosis was measured with the acceptance subscale of the Illness Cognition Questionnaire (ICQ; α=0.83) (34). Attitudes toward long-term use of ART, which included thoughts about dependency on medication and self-rated understanding of ART, were measured using the Beliefs About Medicines Questionnaire (BMQ; α=0.56) (35). Household food insecurity was assessed using an adapted version of the Household Food Insecurity Access Scale (HFIAS; α=0.92) (36).

Interpersonal-level characteristics.

The Perceived Availability of Support scale (PAS; α=0.82) was used to measure perceived general social support, which included eight items (37). Perceived general support from the father of the child was assessed using the Norbeck Social Support Questionnaire (NSSQ; α=0.96), which included seven items about emotional, financial, and practical support (38). Participants self-reported if they had ever disclosed (i.e., told any other person about) their HIV status and were asked whether or not they had experienced verbal or physical abuse perpetrated by an intimate partner or forced sex perpetrated by anyone in the past three months.

Analysis

Statistical analysis was conducted using Stata Version 16.1 (39). Scaled measures were summed; mean imputation was applied for missing items when participants completed at least 75% of scale items. To describe the sample, categorical variables were presented as frequencies and percentages, and continuous variables were summarized by means and standard deviations. For descriptive purposes, responses for the pregnancy attitudes outcome measure were dichotomized into strongly agree/agree and disagree/strongly disagree responses for each item. The dichotomized item assessing pregnancy intention (“This pregnancy is a result of a plan I made to get pregnant”) was compared between women with new and established HIV diagnoses, with a chi-squared test of independence to examine the significance of the relationship. To explore factors associated with pregnancy attitudes, we first used univariable linear regression models to examine associations between potential predictors and the summed pregnancy attitudes measure. Factors with a p-value of .10 or less in univariable analyses were included in the final multivariable regression model; additionally, education, age, and relationship status were identified a priori to be included in the multivariable model regardless of univariable significance. After running the multivariable regression, multicollinearity was examined by checking variance inflation factors (VIF), and a White test was conducted; results indicated no strong presence of multicollinearity or heteroskedasticity requiring further adjustments to the model (40,41).

Results

Description of the sample

Key characteristics of the sample are described in Table 1. Participants ranged in age from 18 to 44 years, with a mean age of 30 years (SD=6.22). The majority of participants had a primary school education or less (n=117; 58.5%); about half were married (n=98; 49.0%), and one-third (n=74; 37.0%) were in a relationship but not married. Most participants (n=156; 78.0%) had been pregnant prior to the index pregnancy. Of these, 32 (20.5%) had experienced a miscarriage, and 12 (7.7%) had experienced a stillbirth in previous pregnancies. Among women with living children (n=141; 70.5%), 11 (7.8%) had at least one child living with HIV.

Table 1.

Characteristics of the sample (n = 200)

	n (%)
Relationship status
Married	98 (49.0)
In a relationship, not married	74 (37.0)
Single	19 (9.5)
Separated or divorced	6 (3.0)
Widowed	3 (1.5)
Education
Primary or lower	117 (58.5)
Secondary or higher	83 (41.5)
HIV diagnosis
Newly diagnosed during current pregnancy	94 (47.0)
Established diagnosis	106 (53.0)
HIV disclosure
Disclosure to at least one person	159 (79.5)
No disclosure	41 (20.5)
Experience of recent interpersonal violence
Reported experiencing any abuse in last 3 months	21 (10.5)
No reported abuse in last 3 months	179 (89.5)
Number of living children
0	59 (29.5)
1	70 (35.0)
2	39 (19.5)
3	23 (11.5)
4 or more	9 (4.5)
	Mean	SD
Age (years)	30.32	6.22
Depression – EPDS [range 0–30; higher score ~ more depressive symptoms]	6.22	5.96
HIV acceptance – ICQ [range 6 – 24; higher score ~ more acceptance]	17.52	3.43
Attitudes toward long-term ART use – BMQ [range 10 – 50; higher score ~ more positive attitudes]	39.05	4.53
Food insecurity – FHI [range 0 – 12; higher score ~ more food insecurity]	1.28	2.41
Social support – PAS [range 8 – 40; higher score ~ more support]	28.86	6.62
Partner support – NSSQ [range 7 – 35; higher score ~ more support]	22.78	8.56

The sample was nearly evenly split between participants with a new HIV diagnosis during the index pregnancy (n=94; 47.0%) and participants with an established HIV diagnosis (n=106; 53.0%). Most participants (n=159; 79.5%) had disclosed their HIV status to at least one person. Interpersonal violence (e.g., verbal or physical abuse perpetrated by an intimate partner or forced sex perpetrated by anyone) in the past three months was reported by 21 participants (10.5%).

Pregnancy attitudes

Participant responses on the pregnancy attitudes scale ranged from a sum of 2 (most negative attitudes toward pregnancy) to 24 (most positive attitudes toward pregnancy and the maximum score possible), with an average score of 14.98 (SD=5.29). A description of responses by question is presented in Table 2. All but one of the eight statements were positively endorsed by a majority of participants. Most participants (n=142; 71%) reported that their pregnancy was planned. Pregnancy intention was significantly higher among women newly diagnosed with HIV during the index pregnancy (n=74; 78.7%), compared to women with an established HIV diagnosis (n=68; 64.2%) (X²(1)=5.14, p=.023).

Factors associated with pregnancy attitudes

In univariable analyses, age, relationship status, time of HIV diagnosis, number of living children, depression, food insecurity, social support, perceived support from the father of the child, and prior disclosure of one’s HIV status were significantly associated with attitudes toward pregnancy (Table 3). Being in a stable relationship (B=3.927; 95% CI 1.870, 5.984), having disclosed one’s HIV status (B=1.931; 95% CI .121, 3.740), reporting higher levels of general social support (B=.230; 95% CI .123, .337), and having higher levels of perceived partner support (B=.228; 95% CI .148, .308) were significantly associated with more positive attitudes toward pregnancy. Older age (B=−.120; 95% CI −.237, −.002), more children (B=−1.604; 95% CI −2.207, −1.001), having an established HIV diagnosis (B=−2.619; 95% CI −4.053, −1.185), having higher levels of depressive symptoms (B=−.188; 95% CI −.309, −.067), and reporting more food insecurity (B=−.482; 95% CI −.781, −.182) were significantly associated with less positive attitudes toward pregnancy.

Table 3.

Factors associated with positive attitudes toward pregnancy (n = 200)

	Univariable	Multivariable
	Regression coefficient B (95% CI)	Regression coefficient B (95% CI)
Demographics
Age	−.120 (−.237, −.002)*	.038 (−.090, .166)
Education
Secondary or higher education	1.239 (−.250, 2.728)	−.727 (−2.083, .629)
[ref: Primary or lower education]
Relationship status
Married/In a relationship	3.927 (1.870, 5.984)***	2.182 (−.043, 4.407)
[ref: Single/Divorced/Widowed]
HIV diagnosis
Established diagnosis	−2.619 (−4.053, −1.185)***	−2.145 (−3.570, −.720)**
[ref: New diagnosis]
Individual characteristics
Number of living children	−1.604 (−2.207, −1.001)***	−1.756 (−2.513, −.999)***
Depression – EPDS	−.188 (−.309, −.067)**	−.017 (−.130, .096)
[range 0 – 30; higher score ~ more depressive symptoms]
HIV acceptance – ICQ	.071 (−.145, .287)	-
[range 6 – 24; higher score ~ more acceptance]
Attitudes toward long-term ART use – BMQ	.116 (−.047, .279)	-
[range 10 – 50; higher score ~ more positive attitudes]
Food insecurity – FHI	−.482 (−.781, −.182)**	−.106 (−.380, .169)
[range 0 – 12; higher score ~ more food insecurity]
Interpersonal characteristics
Social support – PAS	.230 (.123, .337)***	.091 (−.017, .199)
[range 8 – 40; higher score ~ more support]
Partner support – NSSQ	.228 (.148, .308)***	.155 (.054, .255)**
[range 7 – 35; higher score ~ more support]
HIV status disclosure
Disclosed to at least one person	1.931 (.121, 3.740)*	1.477 (−.251, 3.204)
[ref: No disclosure]
Experience of recent interpersonal violence
Experienced any abuse in last 3 months	−.195 (−2.605, 2.214)	-
[ref: no abuse experienced]

^*p<0.05

^**p<0.01

^***p<0.001

In the final multivariable model (Table 3), attitudes toward pregnancy remained significantly associated with perceived partner support, recency of HIV diagnosis, and number of living children. Having higher levels of perceived partner support (B=.155; 95% CI .054, .255) was associated with more positive attitudes about pregnancy, and having an established HIV diagnosis (B=−2.145; 95% CI −3.570, −.720) and more children (B=−1.756; 95% CI −2.513, −.999) were associated with less positive attitudes toward pregnancy. Other variables were not significantly associated with attitudes toward pregnancy in the final model.

Discussion

This study examined attitudes toward a current pregnancy and associated factors in a cohort of women in Moshi, Tanzania who were receiving antenatal care and HIV treatment during their pregnancy. Overall attitudes toward pregnancy were positive, which is an encouraging finding, but one that also points to the importance of identifying women with more negative attitudes who may require tailored support during their pregnancies. A majority of participants (71.0%) reported that their pregnancy was the result of a plan to get pregnant; pregnancy intentions were higher among women diagnosed with HIV during the index pregnancy compared to women with an established HIV diagnosis. Being newly diagnosed with HIV during the index pregnancy, having fewer children, and having higher perceived levels of partner support were identified as predictors of positive attitudes toward the current pregnancy. Results from this study can inform counseling in the context of PMTCT and comprehensive family planning programs, inclusive of preconception care, for WLWH.

Similar to other studies among WLWH, which found lower pregnancy intention among women with more children (42–44), we found that women who had more children felt less positively toward their current pregnancy. Qualitative work in sub-Saharan Africa has shown that women’s pregnancy intentions are often thoughtful and take into consideration the number of children already in the family, birth spacing, and economic needs (25,45,46). However, our finding that more than 1 in 4 women in the study still reported unplanned or unintended pregnancies speaks to the need for continued attention to access and uptake of family planning. WLWH may require tailored family planning counseling that considers their unique circumstances, as they often feel increased economic vulnerability and relationship insecurities related to their HIV status (47–49), particularly in contexts with substantial gender inequity, which may lead to reluctance to bring another child into the household.

Perceived support from the father of the child was similarly predictive of attitudes toward the pregnancy. While there is little literature about the impact of partner support on pregnancy attitudes in sub-Saharan Africa or in the context of HIV, qualitative data among pregnant women in the United States found that a lack of partner emotional and instrumental support is associated with pregnancy unwantedness, a construct that captures negative attitudes toward the pregnancy (50). Partner support is reflective of both emotional support for the mother, as well as perceived financial and instrumental support for the child in the future. Limited partner support, therefore, has implications for both the mother’s immediate well-being, as well as the future she foresees for her child. A woman’s negative feelings resulting from limited perceived support from a partner may be compounded by the relationship insecurity that accompanies a positive HIV status for some women (51–53).

In multivariable and univariable analyses, we observed that a woman’s prior knowledge of her HIV status had a significant impact on pregnancy attitudes. Women who had an established HIV diagnosis prior to becoming pregnant reported less positive attitudes toward the current pregnancy and were more likely to report that the pregnancy was unintended. Fears about transmission of HIV to a child, anticipation of criticism from a medical provider for becoming pregnant while living with HIV, or concern about the unwanted disclosure of one’s HIV status during pregnancy or childbirth may contribute to negative feelings about a pregnancy (20,23,54–57). Women living with HIV who become pregnant may experience increased anxiety or experience stigma from others who may deem it irresponsible to become pregnant while living with HIV (24,56,58,59).

PMTCT programs should be aware of the myriad of feelings that women living with HIV might have toward their pregnancy; they should be equipped to offer social support to help women process their pregnancy and cope with negative feelings, especially as related to the intersection between pregnancy and HIV. Appropriate counseling should address patients’ worries about being pregnant and living with HIV and provide clear information for how to prevent transmission of HIV to the child, which may help alleviate negative feelings about the pregnancy and improve care engagement. PMTCT programs report retention rates below those of the general adult population in HIV care (19); although not a focus of this paper, poor retention may be influenced by attitudes toward the pregnancy. Helping women to identify intrinsic motivations to stay in care can have important implications for vertical transmission of HIV, health outcomes for the mother and child, and potential forward transmission to sexual partners.

These findings underscore the importance of family planning education and access to contraception for all individuals, including those living with HIV. For women with an established HIV diagnosis, our data suggest that a large proportion did intend to become pregnant; clear conversations about pregnancy planning must therefore be an essential component of HIV services for women and men of reproductive age to provide education on safe conception practices and pregnancy care to avoid HIV transmission (18,20,60). Counseling, coupled with access to a range of modern contraceptives, can also help women prevent unintended pregnancies or delay pregnancy until a time when the pregnancy is desired, and should not be neglected in the immediate postpartum period. Integration of family planning services into HIV care and strategies including provision of family planning vouchers to postpartum WLWH have been associated with better knowledge of family planning and higher uptake of contraceptive methods (61,62), showing promise for achieving safer pregnancies and reducing unintended pregnancies. Shared decision-making tools have shown effectiveness in facilitating patient-provider communication around contraceptive use (63). These tools could be adapted to support WLWH to make fertility decisions that are consistent with their preferences and values.

At the same time, many women continue to be newly diagnosed with HIV during pregnancy. This points to a significant need to increase HIV community prevention efforts, including a focus on HIV testing and treatment among men. HIV-serodifferent couples who are looking to conceive need tailored HIV risk reduction counseling that focuses on evidence-based strategies, including treatment as prevention, periconception pre-exposure prophylaxis (PrEP), and self-insemination (64,65). Models of safer-conception counseling among HIV-serodifferent couples in sub-Saharan Africa have been received positively among clients and have been effective at promoting safe conception (17,18,66).

Finally, our data suggest that PMTCT care should include men when possible, given the positive influence of partner support on pregnancy attitudes. In Tanzania, national guidelines encourage women to bring their partners to the first antenatal care visit for HIV counseling and testing (31), but male partner engagement receives limited attention after a woman’s initial ANC visit. PMTCT services should find opportunities to meaningfully engage male partners throughout the pregnancy and postpartum period to build and maintain positive, mutual support, while also being mindful not to undermine care for women (67). For women who do not have a supportive partner, conversations with a counselor or provider could help women to identify other sources of support in women’s lives, or connect them with clinic-based services aimed at facilitating peer support.

This study is novel in its use of a comprehensive measure of pregnancy attitudes, which was assessed during participants’ second or third trimester of pregnancy. This allowed participants to respond in the context of their current pregnancy, rather than a hypothetical pregnancy or reflecting on a previous pregnancy. While this study provides novel data on attitudes toward pregnancy among pregnant women living with HIV in Tanzania, findings should be interpreted in light of study limitations. The study enrolled only women who had been taking ART for at least one month and therefore does not include women who stopped attending antenatal care visits after the first visit or women who never engaged in antenatal care. These women may have felt more negatively about their pregnancies or differed from the study sample in ways that make the findings not generalizable to all pregnant women living with HIV. Furthermore, the cross-sectional nature of the study does not determine directionality between two factors that could influence each other (e.g., pregnancy attitudes and depression), nor does it capture changes in pregnancy attitudes over time. Participants in their second or third trimester may have had difficulty recalling their feelings when they first discovered they were pregnant or may have experienced evolving emotions over time. As surveys were interviewer-administered, social desirability bias may have led women to report more positive attitudes toward pregnancy than they actually felt, particularly given that elective abortion is illegal in Tanzania (68). The analysis may be additionally limited by overlap in constructs; notably, the pregnancy attitudes scale included an item assessing partner’s feeling about the pregnancy (“My partner wanted me to become pregnant”), which may have overlapped with the partner support measure (NSSQ). Finally, the measure of pregnancy attitudes we used differed from measures of pregnancy intention, fertility desire, or pregnancy wantedness that have been used in previous studies, making it difficult to directly compare our findings to existing literature. We adapted this measure for our study purposes because we did not identify existing measures that captured the construct of pregnancy attitudes; thus, there is a need for the development of a robust measure of pregnancy attitudes so that this construct can be examined across settings and over time.

Conclusions

This study found that women had more positive attitudes toward the current pregnancy if they were newly diagnosed with HIV, reported higher levels of perceived partner support, and had fewer children. Most participants reported that the index pregnancy was intended, yet pregnancy intention was significantly lower among women with an established HIV diagnosis as compared to women who were newly diagnosed with HIV during the index pregnancy. As timing of HIV diagnosis relative to conception may influence pregnancy attitudes, PMTCT programs should pay particular attention to the feelings of women with established HIV diagnoses, especially those who may struggle with new emotions or changing life circumstances that could impact their care engagement as a newly pregnant WLWH. Comprehensive psychosocial support during PMTCT may improve women’s attitudes toward their pregnancy, particularly if they lack partner support. Regular HIV appointments and postpartum health care are key opportunities for education and psychosocial support to assist women to make decisions about future pregnancy intentions, and, if desired, be offered contraceptives to prevent unwanted pregnancies or discuss preconception care to prevent HIV transmission, particularly with HIV-serodifferent couples. Being responsive to women’s pregnancy attitudes during HIV care may result in improvements in care engagement and health outcomes for both women and children.