The disclosure dilemma: Willingness to disclose a positive HIV status among individuals preparing for HIV testing during antenatal care in Tanzania

Abstract

HIV status disclosure can reduce transmission risks and improve care engagement. Individuals may have strong feelings about HIV disclosure even prior to diagnosis. We assessed willingness to disclose a positive HIV status among pregnant women and their male partners awaiting routine HIV testing during antenatal care in Tanzania (n=939). Logistic regression models were used to examine factors associated with willingness to disclose to one’s inner circle (partner/family member) and outer circle (friend/neighbor) in the event of an HIV diagnosis. Almost all (93%) were willing to disclose to at least one person; participants were more willing to disclose to their inner circle (91%) vs. outer circle (52%). Individuals with some form of employment, more stigmatizing attitudes of social distancing of PLWH, greater anticipated HIV stigma, more perceived social support, and prior contact with someone living with HIV were more likely to disclose to their inner circles. Individuals who were older, male, and who had higher levels of perceived social support were more willing to disclose to their outer circle. These findings increase the understanding of the intra- and interpersonal factors that influence HIV disclosure decisions. Tailored pre- and post- HIV test counseling are needed to facilitate social support and overcome barriers to disclosure if they test positive for HIV.

Introduction

Among people living with HIV (PLWH), decisions related to whether, how, and when to disclose one’s positive HIV status can have far-reaching impacts for health and well-being. Across multiple studies and settings, serostatus disclosure has been associated with improved HIV care engagement, increased social support in managing the disease, and decreased risk of HIV transmission from unprotected sex [1–3]. Conversely, PLWH who have not disclosed their HIV status are more likely to report symptoms of depression and describe their nondisclosure as a considerable source of worry and stress in their lives [4,5]. Concealing a positive HIV status can lead to challenges in attending HIV clinic appointments, taking antiretroviral medication, and obtaining social support to effectively cope with the physical, emotional, and social implications of HIV [6–8].

Despite the benefits of HIV status disclosure for PLWH, the decision whether or not to disclose is complex and driven by a careful weighing of the perceived consequences and the perceived benefits of disclosure [9]. When considering disclosure, patients describe persistent feelings of shame and sadness related to their illness, as well as worry about stigma and discrimination they may face as their status becomes known to others [5,10]. Among women, and particularly pregnant women, disclosure decisions are further complicated by gender inequities and traditional gender roles that can lead to reliance on male partners and family members to meet basic needs for the mother and her children [11,12]. Women living with HIV (WLHIV) frequently express fears of abandonment, withdrawal of support, accusations of infidelity, or physical violence in response to disclosure [13]. Although rare, studies have shown that these events do occur [14,15]. Because women face specific challenges to disclosing during pregnancy, women who are diagnosed with HIV in antenatal care (ANC) are less likely to disclose than women testing in other settings [13], and disclosures are commonly limited to a small circle that may include a sexual partner, family members, or trusted friends [16,17].

When disclosure is done thoughtfully, it can have positive impacts for PLWH. Among pregnant women in Tanzania, a study showed that disclosure outcomes were generally very positive, with 73% of women reporting increased emotional support after disclosing their status, 30% reporting increased freedom to use antiretroviral medication, and 26% reporting increased financial support for HIV treatment; in contrast, only 14% reported experiencing stigma or discrimination, while 11% experienced a separation or divorce after disclosure [9]. In light of these findings, it is important to provide pregnant and postpartum WLHIV with counseling and support for their disclosure decisions.

To date, the majority of HIV status disclosure studies and interventions have focused on disclosure decisions weeks or months after an HIV diagnosis [18,19]. Universal HIV testing and counseling during pregnancy, supported by the WHO and the Option B+ protocol [20], create an opportunity to address the topic of HIV disclosure much earlier, immediately upon receiving a positive test result or even prior to HIV testing. Given that couples often present together to the first ANC visit, there is also an opportunity to facilitate supported disclosure to one’s sexual partner in the presence of a trained counselor.

In this study, we surveyed pregnant women and their male partners prior to HIV testing in ANC and assessed their willingness to disclose a hypothetical positive HIV status to various members of their social network, including their inner circle (partner or family member) and outer circle (friend or neighbor), and we examined the factors associated with willingness to disclose a positive HIV status. Our study was based on the consequences-based model of HIV disclosure [9], which suggests that individuals make decisions about disclosure based on the perceived consequences of a disclosure. We hypothesized that perceived disclosure consequences would be driven by perceptions of HIV stigma [21], but that, even controlling for HIV stigma, perceived social support would lead to greater willingness to disclose [22]. The results of this study can be used to develop tailored counseling interventions that educate and empower pregnant women and male partners to disclose their HIV status in the event of a positive test result.

Methods

Study Setting

The study was conducted in the Kilimanjaro Region of Tanzania at two urban health clinics where ANC and PMTCT services are offered. Together, the clinics see approximately 2500 pregnant women each year, with approximately 4.8% of attendees living with HIV. As part of the Tanzanian national protocol [23] and WHO Option B+ guidelines [20], all pregnant women attending their first ANC appointment are tested for HIV and, if diagnosed with HIV, initiate antiretroviral therapy on the same day [24]. In Tanzania, pregnant women are encouraged to attend the first ANC appointment with their male partners [25].

Sample and Recruitment

The sample included 498 pregnant women who attended ANC with a male partner who were enrolled in a pilot randomized clinical trial of an HIV stigma reduction intervention [26]. Out of the 498 pregnant women who enrolled in the study and were attending with male partners, 441 of the male partners enrolled in the study. Women and their partners attending a first ANC appointment between April and November 2019 were identified by clinic nurses, and, if interested, referred to a private office to meet with a member of the research team. Women were eligible to participate if they were attending a first ANC appointment for their current pregnancy. Women known to be living with HIV at the time of enrollment were enrolled in the larger trial but excluded from the current analysis. Men were eligible if they were accompanying a female partner to her first ANC appointment. All participants provided written informed consent prior to enrollment. The study received ethical approval from the institutional review boards of the Tanzanian National Institute for Medical Research and the Duke University Institutional Review Board.

Procedures

Enrolled participants completed a structured survey using audio computer-assisted self-interview (ACASI) technology on tablets running Questionnaire Development System (QDS) software (Nova Research Company, Version 5.0). Participants completed the assessment in a private study office using individual tablets. They were first guided through three practice questions by a research assistant, and then independently read and listened to the survey questions and response options in Kiswahili before indicating their response on a touch screen. Following the survey, participants were randomized to study condition and returned to the clinic for HIV testing.

Measures

Measures were selected based on previous research in East Africa, including measure validation when available, and evaluation of face validity by the Tanzanian researchers on our team. All measures were translated from English into Swahili and then back-translated and discussed to reach consensus on best translation. A full description of the survey can be found in the study’s protocol [26].

The primary outcome of interest for this study was willingness to disclose a hypothetical positive HIV status. This construct was measured using a four-item scale with questions about disclosure to one’s partner (“If I had HIV, I would tell my partner”), family member (“If I had HIV, I would tell a family member”), friend (“If I had HIV, I would tell a friend”), and neighbor (“If I had HIV I would tell a neighbor”). For each item, participants responded on a scale from 0 (Strongly Disagree) to 3 (Strongly Agree). Willingness to disclose a positive HIV status to one’s partner and/or a family member were grouped as willingness to disclose to one’s inner circle, while willingness to disclose to one’s friend and/or a neighbor were grouped as willingness to disclose to one’s outer circle. For descriptive purposes, responses for the willingness to disclose outcome were dichotomized into “willing to disclose” (response of strongly agree or agree) and “not willing to disclose” (response of disagree or strongly disagree).

We assessed demographic information including participants’ age, sex, level of education, employment status, and marital status. HIV stigmatizing attitudes were assessed using an 18-item scale, adapted for the Kiswahili language and Tanzanian context [27]. The measure yielded sub-scale scores representing two dimensions of HIV stigma: interpersonal distancing and social isolation (HSS1; α = 0.84, potential range 0–30) and blame and moral judgment (HSS2; α = 0.93, potential range 0–18). The interpersonal distancing and social isolation dimension probed participant behaviors in response to PLWH, asking for participant endorsement of statements (e.g. “I would not eat with someone I knew had HIV,” “I would not let someone with HIV to look after my child”). The second dimension asked for participant endorsement of statements measuring attitudes related to judgment of PLWH (e.g. “Getting HIV is a punishment for bad behavior,” “I would think less of someone if I found out the person has HIV”). Anticipated stigma was measured using a 12-item scale (AHAS; α = 0.90, potential range 0–36) assessing the degree to which people would expect to experience prejudice and discrimination from others if they were living with HIV and their positive status was known [28]. Social support was measured using the Perceived Availability of Support Scale (PASS; α = 0.87, potential range 0–24), an 8-item scale assessing the degree to which people perceived they had access to social support if they needed it [29]. Contact with people living with HIV was measured with a set of 6 items asking participants if they knew someone personally who was living with HIV (a family member, spouse, close friend, etc.).

Data Analysis

Descriptions of demographic variables were presented as simple frequencies. Scaled measures were summed and mean imputation was applied for missing items when participants completed at least 75% of that scale’s items, a common method when few items are missing, as was the case in these data [30,31]. Descriptions of scaled measures were presented as median and interquartile range. Three quality-check items were included throughout the survey (e.g., “Please select Strongly Agree for this item”). Of the 939 participants, 842 (90%) responded to at least two quality-check items correctly. All analyses were repeated with this 90% and compared to assess response validity; the results did not meaningfully change, so the full sample was retained in the final analysis.

Univariable logistic regression models were used to examine associations between potential predictors and participants’ hypothetical willingness to disclose a positive HIV status to one’s inner and/or outer circle. Predictors were selected based on previous literature indicating that they might influence one’s willingness to disclose [17,32–35]. As disclosure endorsements differed significantly by sex, stratified univariable analysis was conducted to explore differences by sex, and when results significantly differed, interaction terms were explored. Because no interaction terms were significant in the multivariable models, only main effects were included. Odds ratios with a 95% confidence interval were reported. Factors with a p-value of .10 or less in univariable analyses were included in the final multivariable regression model.

Results

Description of sample

Table 1 describes the demographics of the sample (498 pregnant women and 441 male partners). The average age of female participants was 25 years old (SD = 5.1), and the average age of male participants was 30 years old (SD = 6.9). About half of participants (n=597, 52.3%) had a primary school education or less. Over two-thirds of the participants reported being married (n=653, 69.5%).

Table 1.

Demographic characteristics of participants (n=939).

	n	Percentage (%)
Sex
Male	441	47.0
Female	498	53.0
Age
18 – 25 years	428	45.6
26 – 34 years	384	40.9
35 years and older	127	13.5
Education
Primary or none	497	52.3
Secondary	391	41.6
High	51	5.4
Marital status
Married	653	69.5
Not married	286	30.5
Religion
Muslim	475	50.6
Christian	462	49.2
Employment Status
No income-earning activity	214	22.8
Some income-earning work	524	55.8
Formally employed	201	21.4
	Median	Q1, Q3
HIV Stigmatizing Attitudes (HSS1) – Social distancing	7	1, 13
HIV Stigmatizing Attitudes (HSS2) – Blame and judgment	4	1, 8
Anticipated HIV Stigma (AHAS)	8	3, 13
Perceived Availability of Support (PASS)	16	13, 20

Willingness to disclose

Table 2 describes willingness to disclose in this population. The vast majority of participants were willing to disclose to at least one other person (n=870, 92.7%), with substantially greater likelihood of being willing to disclose to someone in their inner circle (n=856, 91.2%) than their outer circle (n=489, 52.1%). The highest number of participants were willing to disclose to a family member (n=799, 85.1%), with slightly fewer (n=779, 83.0%) being willing to disclose to a partner. For the outer circle, about half of participants were willing to disclose to a friend (n=469, n=49.9%), and far fewer (n=245, 26.1%) willing to disclose to a neighbor.

Table 2.

Willingness to disclose to inner and outer circles.

	Overall (%)	Pregnant women (%)	Male partners (%)
Willing to disclose to anyone	870 (92.7)	457 (92.3)	413 (95.4)
Willing to disclose to inner circle	856 (91.2)	450 (90.7)	406 (93.1)
Partner	779 (83.0)	405 (81.5)	374 (85.8)
Family member	799 (85.1)	419 (84.3)	380 (86.6)
Willing to disclose to outer circle	489 (52.1)	222 (44.8)	267 (61.7)
Friend	469 (49.9)	210 (42.3)	259 (59.5)
Neighbor	245 (26.1)	99 (20.0)	146 (33.6)

Factors associated with willingness to disclose

For willingness to disclose to the inner circle, univariable logistic regression revealed that those with higher levels of education, income-earning activities, more stigmatizing attitudes of social distancing of PLWH, greater anticipated HIV stigma, more perceived social support, and prior contact with someone living with HIV were more likely to disclose to their inner circles (Table 3). In the final multivariable model (Table 3), nearly all factors that were significant in the univariable logistic regressions remained significant. Having higher levels of education and formal employment were no longer significant (p-values <0.10 but not <0.05). Participants had greater odds of disclosing to their inner circle if they had some income-earning work (AOR=2.027; 95% CI: 1.161, 3.539) more stigmatizing attitudes of social distancing of PLWH (AOR=1.049, 95% CI: 1.009, 1.092), greater anticipated HIV stigma (AOR=1.083, 95% CI: 1.037, 1.131), greater perceived social support (AOR=1.086, 95% CI: 1.039, 1.134), and prior contact with someone living with HIV (AOR=2.023; 95% CI: 1.226, 3.337).

Table 3.

Regression analysis for willingness to disclose to the inner circle (n=939)

	Univariable AOR (95% CI)	Multivariable AOR (95% CI)
Sex
Female	REF
Male	1.383 (0.857 – 2.234)
Age
For every one-year increase	0.991 (0.957 – 1.026)
Education
None or primary	REF	REF
Secondary or higher	1.676 (1.028 – 2.732) *	1.577 (0.933 – 2.668) ‡
Marital status
Married	REF
Divorced/widowed	1.237 (0.728 – 2.104)
Employment status
No income-earning activities	REF	REF
Some income-earning work	1.865 (1.096 – 3.175) ‡	2.027 (1.161 – 3.539) *
Formally employed	1.847 (0.935 – 3.649) *	1.855 (0.903 – 3.807) ‡
HIV Stigmatizing Attitudes (HSS1) – Social distancing
For every one-point increase	1.053 (1.016 – 1.092) **	1.049 (1.009 – 1.092) *
HIV Stigmatizing Attitudes (HSS2) – Blame and judgment
For every one-point increase	1.002 (0.950 – 1.056)
Anticipated HIV Stigma (AHAS)
For every one-point increase	1.069 (1.028 – 1.112) **	1.083 (1.037 – 1.131) ***
Perceived Availability of Support (PASS)
For every one-point increase	1.075 (1.032 – 1.121) **	1.086 (1.039 – 1.134) ***
Contact with PLWH
No contact	REF	REF
Contact	1.997 (1.242 – 3.213) **	2.023 (1.226 – 3.337) **

Note.

^‡p < .10.

^*p < .05.

^**p < .01.

^***p < .001.

CI, confidence interval.

For willingness to disclose to one’s outer circle, univariable logistic regression (Table 4) revealed that being male, being older, having some form of income-earning activities, and having more perceived social support were significant predictors of being likely to endorse hypothetical HIV disclosure to their outer circles. In the final multivariable logistic regression (Table 4), sex, age, and perceived availability of support scores were significantly associated with participants’ willingness to disclose to the outer circle. Participants had greater odds of disclosing to their outer circle if they were male (AOR=1.814; 95% CI: 1.327, 2.480), were older (AOR=1.045; 95% CI: 1.020, 1.070), and had higher levels of perceived availability of support (AOR=1.057; 95% CI: 1.030, 1.085).

Table 4.

Regression analysis for willingness to disclose to the outer circle (n=939)

	Univariable AOR (95% CI)	Multivariable AOR (95% CI)
Sex
Female	REF	REF
Male	1.985 (1.527 – 2.581) ***	1.814 (1.327 – 2.480) ***
Age
For every one-year increase	1.054 (1.032 – 1.077) ***	1.045 (1.020 – 1.070) ***
Education
None or primary	REF
Secondary or higher	1.189 (0.918 – 1.539)
Marital status
Married	REF
Divorced/widowed	1.037 (0.783 – 1.373)
Employment status
No income-earning activities	REF	REF
Some income-earning work	1.310 (0.953 – 1.803) ‡	0.792 (0.550 – 1.141)
Formally employed	1.643 (1.112 – 2.427) *	0.890 (0.571 – 1.387)
HIV Stigmatizing Attitudes (HSS1) – Social distancing
For every one-point increase	0.992 (0.975 – 1.009)
HIV Stigmatizing Attitudes (HSS2) – Blame and judgment
For every one-point increase	0.995 (.967 – 1.024)
Anticipated HIV Stigma (AHAS)
For every one-point increase	0.998 (0.981 – 1.016)
Perceived Availability of Support (PASS)
For every one-point increase	1.046 (1.021 – 1.073) ***	1.057 (1.030 – 1.085) ***
Contact with PLWH
No contact	REF
Contact	0.936 (0.721 – 1.216)

Note.

^‡p < .10.

^*p < .05.

^**p < .01.

^***p < .001.

CI, confidence interval.

Discussion

For PLWH, disclosing a positive HIV status is a complex decision, and particularly so during the pregnancy and postpartum periods [5]. Since disclosure carries both reward and risk, it is not a “one size fits all” solution and it is important to consider the how individuals weigh the potential consequences and benefits as they consider whether to disclose and to whom. Although frameworks have been developed to understand the factors that influence disclosure decisions among different populations [7], few studies have assessed disclosure decision making among individuals preparing for HIV testing.

In this study, we found that willingness to disclose a positive HIV status to one’s inner circle (partner or family member) was nearly universal, while only half were willing to disclose to a member of their outer circle (friend or neighbor). In an earlier study on patterns of disclosure among postpartum WLHIV in Tanzania [3], actual rates of disclosure were substantially lower than the hypothetical rates we observed. These findings showcase a need for research to follow individuals longitudinally in order to understand discrepancies between intent to disclose and actual disclosure, which may represent an example of the “know-do gap” [36].

As we would expect, we found that participants who perceived themselves as having greater social support were more willing to endorse disclosure to individuals in their inner and outer circles. Increased social support allows participants to feel that their disclosure would be met in a supportive manner, representing a perceived benefit of disclosure [8,33,37]. It also shows that possible consequences of disclosure [9] may be mitigated by increased social capital, which improves coping with an HIV diagnosis, quality of life, care engagement, and health outcomes [38,39].

Surprisingly, and contradicting prior literature [35,43], our study also found that greater anticipated HIV stigma and stigmatizing attitudes of social distancing of PLWH were associated with an increased willingness to disclose a positive HIV status to members of one’s inner circle. A potential explanation for these unexpected findings is that participants who anticipate high levels of stigma from others and who hold stigmatizing attitudes may try to mitigate these feelings by disclosing to a few, trusted confidants in their families [43]. A fear of the manifestations of stigma such as abandonment and abuse can serve as a barrier to HIV disclosure, but our findings suggest awareness of such potential interruptions in support may motivate pregnant women and their partners to seek support from family members, who can provide a buffer from stigmatizing experience from the community. It is also notable that individuals who knew someone living with HIV were twice as willing to disclose, highlighting the impact of having role models to facilitate disclosure decisions.

In our study, males and older participants expressed more willingness to disclose a positive HIV status to their friends and/or neighbors. Slight differences among disclosure rates by sex are reported in previous literature [44] and may be attributed to gender dynamics and gender roles, which influence HIV care engagement, support, and disclosure decisions [45–47]. These roles may place women in vulnerable positions to disclose their HIV status, with fear of abandonment and loss of social and economic security as barriers to disclosure [45]. Although previous research regarding actual HIV disclosure among older and younger participants have shown mixed results [48,49], a similar study assessing intention to disclose a positive HIV status among undiagnosed participants saw that older individuals had greater intention to disclose than did younger participants [17]. Oftentimes, younger adults report more experiences of HIV stigma and discrimination than older adults and are more likely to fear job loss or other social and economic consequences upon disclosure [52,53].

This study demonstrates the need for targeted and tailored programs to support HIV disclosure among women and their partners who might be diagnosed with HIV during ANC. In particular, it emphasizes the need to understand how individuals evaluate both the potential consequences and benefits of HIV disclosure to various members of their social networks. Universal HIV testing in ANC provides the opportunity to counsel individuals in order to empower them to make informed decisions about disclosure, build the necessary skills for disclosure, and ultimately close the gap between disclosure intention and behavior [54]. Counseling prior to HIV testing may help patients to reflect on the importance and implications of HIV disclosure and commit to HIV disclosure prior to receiving an HIV diagnosis. Immediately after receiving an HIV diagnosis, counseling can reinforce prior intentions around disclosure and help individuals plan for disclosures. In order to facilitate difficult and complex conversations around disclosure, healthcare providers need on-going training in patient-centered counseling, making them attuned to and respectful of the many factors that may influence disclosure, including age, sex, education, and perceived social support as it pertains to whom an individual wishes to disclose to.

The study findings should be interpreted in light of the study’s limitations. All survey findings were self-reported by participants, which may be subject to social desirability bias, although this was likely mitigated by the use of tablet-based ACASI software. The cross-sectional nature of the study design does not allow for conclusions on causality. Finally, the study did not include pregnant women who presented to ANC without a male partner; these women may be very different in their support systems and their willingness to disclose.

Conclusions

HIV status disclosure remains a crucial element for improving HIV care engagement and the psycho-social well-being of PLWH. In the light of the results of this study, there is a need to better understand the pathways towards HIV disclosure. This is particularly important among women and younger people, whom we identified as less likely to disclose to their outer circles. These two groups constitute the majority of those living with HIV in the sub-Saharan Africa region, and supporting them to make informed decisions about HIV disclosure can help to meet global targets for HIV prevention and treatment [55]. This study highlights the opportunity to use the ANC setting not only to understand the complexities of disclosure decision making, but also to support the disclosure process in an environment where HIV testing is mandatory.