Insights from a trend analysis of HIV indicators using Population-based HIV Impact Assessment (PHIA) surveys in Zimbabwe

Abstract

Introduction

To fight HIV, the Universal Test and Treat (UTT) HIV program was introduced and has been praised for improving HIV outcomes in many settings but also faces criticism for abbreviated counselling impacting adherence, and implementation challenges in resource-limited settings. Therefore, we aimed to assess the changes in ART enrollment, HIV viral load suppression, and adherence to ART between 2015 and 2020, a period that coincided with the implementation of the UTT program in Zimbabwe.

Methods

We used data from nationally representative surveys before (2015, n = 2,792) and after (2020, n = 2,617) the national launch of the UTT program in 2016 in Zimbabwe. By applying survey weights, the prevalence of the HIV indicators was estimated and compared between the two survey years and within subgroups after stratification. Adjusted Odds ratios were also calculated to assess differences in the HIV indicators between the surveys. Importantly, our analyses were observational and did not establish a causal relationship between the UTT program and the observed changes in HIV indicators.

Results

HIV-positive status awareness of HIV-positive adults (15 years and older) was 76.8% (95% CI: 75.0-78.5) in 2015 and 86.8% (95% CI: 85.2–88.3) in 2020. Of those who were aware of their HIV-positive status, ART enrollment improved in all subgroups, with overall enrollment increasing from 88.7% (95% CI: 87.3–89.9) in 2015 to 97.0% (95% CI: 96.1–97.7) in 2020, adjusted odds ratio (aOR), 3.7 (p < 0.001). Overall viral load suppression also improved significantly from 85.4% (95% CI: 83.6–87.0) to 90.3% (95% CI: 88.9–91.6), aOR,1.5 (p < 0.001). However, young people didn’t show an increase in viral load suppression. Also, men had lower rates than women in all HIV indicators, and there was no significant change in adherence to ART between the survey years.

Conclusion

In general, there were significant improvements in ART enrollment rates and viral load suppression rates, with no change in ART adherence. These improvements might be attributed to several factors, including the 2016 UTT program. The existing gender and age disparities can be addressed by expanding ART services during weekends and unconventional hours, as well as by utilizing mobile and community-based HIV services for young individuals and men.

Summary box

What is already known on this topic

• Many countries, including Zimbabwe, have adopted the Universal Test and Treat (UTT) HIV program based on its success in improving HIV outcomes and potentially reducing HIV infections.

• However, the program has faced criticism for its shortened counseling sessions, which may impact retention in care and adherence to ART, as well as its implementation challenges in resource-limited settings.

• Evidence from nationally representative data on the program’s progress in a resource-limited country, particularly when considering age and gender stratification, is lacking.

What this study adds

• There was a significant increase in ART enrollment and viral load suppression following the launch of the UTT program in a resource-limited country.

• There were no changes in adherence to ART rates following the introduction of the UTT in Zimbabwe.

• The increase in ART enrollment and viral load suppression was lesser in young people and men compared to other groups.

How this study might affect research, practice, or policy

• This study addresses a crucial research gap and offers valuable insights into the plausible effectiveness of the UTT program and the challenges it faces.

• It informs policymakers, healthcare providers, and stakeholders in Zimbabwe and beyond, and facilitates evidence-based decision-making and resource allocation for more effective HIV interventions to improve HIV outcomes in the most affected subpopulations.

Introduction

HIV/AIDS remains a significant public health challenge globally, with sub-Saharan Africa being the most heavily affected region [1]. Among the countries in this region, Zimbabwe has experienced a significant burden of HIV infection. According to the latest available data, the national HIV prevalence in Zimbabwe stands at an alarming 12.9% [2]. For many years, Zimbabwe has encountered substantial difficulties in achieving optimal rates of antiretroviral therapy (ART) enrollment and viral suppression among individuals living with HIV [3]. Despite efforts to expand treatment access, the country has grappled with low ART coverage and suboptimal outcomes [3]. These challenges have indicated the urgent need for effective and comprehensive HIV programs.

Over the years, Zimbabwe has launched several programs to address these challenges, including the major recent Universal Test and Treat (UTT) initiative, which was launched in December 2016 following the recommendation of the World Health Organization [4]. This program replaced the previous CD4-based approach, where the initiation of ART was determined by the CD4 cell count falling below a certain threshold [4]. In contrast, the UTT program aims to provide immediate antiretroviral treatment to all individuals diagnosed with HIV, regardless of their CD4 cell count or clinical stage [4]. Projections and mathematical models simulate that the UTT approach can effectively reduce HIV infections and potentially pave the way for an HIV-free era [5, 6]. The UTT program has also undergone evaluation in the sub-Saharan region. For instance, a retrospective evaluation of the UTT program in South Ethiopia demonstrated improved survival rates among patients enrolled in the UTT program compared to those in the CD4-based program [7]. Similarly, a clinical trial conducted in Eswatini revealed that participants in the UTT program exhibited better retention in care and viral load suppression than those in the CD4-based program [8].

On the other hand, the UTT program has been subject to criticism due to shorter counselling sessions, as patients are initiated on ART on the same day, unlike the CD4-based program, which involved 2 to 3 counselling sessions before ART initiation. Additionally, initiating ART without HIV clinical symptoms in most UTT patients might impact adherence and retention in HIV care, as some may perceive themselves as not being at risk of death from HIV [9]. Furthermore, implementation challenges for the UTT program have been reported in large observational studies, clinical trials, and cohort studies, particularly in resource-limited settings [10–12].

Although the UTT program holds promise, the changes following its national implementation in developing countries, including Zimbabwe, remain relatively uncertain. The scarcity of resources, particularly in Zimbabwe, has significantly hampered the delivery of HIV services. Inadequate healthcare infrastructure and a shortage of healthcare professionals have collectively contributed to the ongoing struggle to effectively combat the HIV epidemic [13]. Low staffing levels and inadequate infrastructure for storing medical equipment in rural healthcare centers in Zimbabwe pose a challenge for ART programs and the plausible effectiveness of UTT programs [14]. Consequently, conducting a large-scale national trends analysis in resource-limited countries, particularly Zimbabwe, is crucial for gaining insights following the implementation of the 2016 UTT program. Recent studies, including that of Chowdhury et al. [15], have documented provincial spatial heterogeneity in HIV prevalence, ART coverage, and viral load suppression in Zimbabwe, offering important insights into the plausible regional effectiveness of UTT programs. Furthermore, although one clinical trial conducted in Eswatini included retention in care and viral load suppression as study endpoints, the small sample size limited more detailed subgroup analyses [8]. Larger-scale evaluations are needed to assess changes in HIV indicators following the implementation of the 2016 UTT program in key subpopulations across sub-Saharan Countries, as previous studies suggest significant geospatial heterogeneity in HIV outcomes that may also vary by age and gender [15–17]. Hence, our study used nationally representative surveys to assess the changes in ART enrollment, HIV viral load suppression, and adherence to ART between 2015 and 2020, a period that coincided with the implementation of the UTT program in Zimbabwe. Furthermore, we sought to assess the overall impact on these HIV indicators as well as the subgroup-specific impact after stratifying the data by age and gender. This study addresses a critical gap in the literature by offering novel insights into the potential benefits and associated challenges post-implementation of the 2016 UTT program.

Methods

Data from two Population-based HIV Impact Assessment (PHIA) surveys conducted in 2015 and 2020 in Zimbabwe were utilized for our research. The PHIA surveys, which are implemented in collaboration with the US President’s Emergency Plan for AIDS Relief (PEPFAR), aim to provide a nationally representative assessment of the HIV epidemic in Sub-Saharan African countries. The surveys involved a two-stage clustered sampling method. Initially, enumeration areas were selected using a probability-proportional-to-size sampling approach, followed by the random selection of households within each area. Trained staff members conducted interviews with household members aged 15 and above, including overnight visitors, after obtaining written informed consent. Computer-assisted interviews were employed to ensure accuracy. Furthermore, blood samples were collected, and rapid HIV testing was conducted on participants who consented by trained staff in accordance with national HIV guidelines. The results were shared with all participants, and those who tested positive were referred for care and treatment. Participants who declined to receive their results were considered as refusing in the biomarker aspect of the study. Ethical approval for these surveys was obtained from the Zimbabwe Health and Human Research Review Board, with additional approval granted by the Institutional Review Board (IRB) at the Columbia University Medical Center and by the US Centers for Disease Control and Prevention (CDC).

Study sample

The data were from surveys conducted in 2015 and 2020, with a total of 22,875 adult participants (aged 15 years and above) who underwent rapid HIV testing based on national HIV guidelines and received valid results in the 2015 survey. Following the same methodology, the number of participants in the 2020 survey was 19,535. In the 2015 survey, 19,312 participants (84.4% of the sample) tested HIV-negative and were excluded from the study. Similarly, in the 2020 survey, 16,577 participants (84.9%) who tested HIV-negative were excluded. We further eliminated those who were not aware of their HIV-positive status from 3563 HIV-positive participants in 2015 and 2958 in 2020. Our final sample for ART enrollment analysis consisted of 5,409 HIV-positive adult participants, with 2,792 from 2015 to 2,617 from 2020. In the analysis of ART adherence, 296 individuals (5.9%) were excluded due to missing data points related to their medication-taking behaviors. However, among individuals receiving ART, one participant was excluded from the viral load suppression analysis due to missing data.

Study variables

To establish the trend for HIV indicators between 2015 and 2020, four key HIV indicators were used as dependent variables. The first indicator focused on the prevalence of ART enrollment among individuals aware of their HIV-positive status. Participants who had previously tested for HIV and reported to have received a positive diagnosis were considered aware of their HIV-positive status. Those reporting current use of antiretroviral (ARV) drugs and having detectable drug levels in their blood were categorized as currently enrolled on ART. To account for the detection of antiretrovirals, adjustments were made to self-reported HIV-positive awareness and ART status. The prevalence of viral load suppression was then calculated among adults enrolled on ART. Viral load suppression was defined as having HIV viral load levels less than 1000 copies/ml. Individuals with viral load levels of 1000 copies/ml and above were classified as having non-suppressed viral load. Furthermore, to measure viral load suppression among those on ART for at least 12 months, the prevalence of viral load suppression was also calculated for this subgroup. Finally, adherence to ART was assessed based on self-reported missed doses in the past 30 days. Participants who reported missing any ARV pills at least one day in the past 30 days were classified as non-adherent, while those who did not miss any pills were considered adherent. Gender (men and women) and age groups (15–25, 26–35, 36–45, 46–55, and 56+) were used as strata and control variables in the study, and the year of the surveys was used as an independent variable.

Statistical analysis

Sample characteristics were analyzed using frequencies and unweighted proportions as part of the descriptive analysis. To estimate the prevalence of HIV indicators in both 2015 and 2020, survey weights were incorporated into the analysis. Age and gender stratification were employed to compare differences between the two survey years. Additionally, adjusted odds ratios of the HIV indicators were calculated to assess their trends between 2015 and 2020, with the statistical significance threshold set at p < 0.05. The regression analysis controlled for age and gender, and the results were adjusted using survey weights to account for the complex sampling design. The statistical analyses were conducted utilizing Stata software (version 15.1; Stata Corp, College Station, TX, USA).

Ethics statement

The Population-based HIV Impact Assessment (PHIA) surveys are carried out in partnership with the US President’s Emergency Plan for AIDS Relief (PEPFAR). Ethical approval for these surveys was obtained from the Zimbabwe Health and Human Research Review Board, with additional approval granted by the Institutional Review Board (IRB) at the Columbia University Medical Center and by the US Centers for Disease Control and Prevention (CDC). Additionally, while PHIA data is publicly available, access to the data is only granted through an application process to verify its proper usage. The authors were authorized to access and use the data by the PHIA Project. Informed consent was obtained in writing from every recruited participant, or from their parent/guardian in the case of participants who were under 18 years old. The participants were informed that their provided information would be confidential and anonymous.

Results

Table 1 displays our study’s frequency and unweighted proportions of sample characteristics. ART enrollment proportions were calculated from 2,792 adult participants in 2015 and 2,617 in 2020 who were HIV-positive and aware of their HIV status. We then excluded participants who were not on antiretroviral therapy (ART) and those with missing viral load results to calculate viral load suppression proportions. The final sample sizes for calculating viral load suppression proportions were n = 2,497 in 2015 and n = 2,547 in 2020. The proportions of viral load suppression were also calculated for participants who had been on ART for more than 12 months, with n = 2,038 in 2015 and n = 2,217 in 2020. Adherence to ART in the past 30 days was determined for participants who were on ART and had no missing values in adherence, with n = 2,337 in 2015 and n = 2,412 in 2020.

Table 1.

Sample characteristics

Variable	2015	%	2020	%
Gender	n = 2792		n = 2617
Women	1877	67	1840	70
Men	915	33	777	30
Age
15–25	245	9	217	8
26–35	689	25	488	19
36–45	956	34	795	30
46–55	537	19	652	25
56+	365	13	465	18
Antiretroviral therapy (ART)
On ART	2498	89	2547	97
Not on ART	294	11	70	3
Viral Load Suppression (VLS)	n = 2497		n = 2547
Suppressed	2166	87	2313	91
Not suppressed	331	13	234	9
12 + months ART VLS	n = 2038		n = 2217
Suppressed	1772	87	2010	91
Not Suppressed	266	13	207	9
Adherence to ART	n = 2337		n = 2412
Adherent	2058	88	2153	89
Non-adherent	279	12	259	11

Table 2 presents the prevalence of various HIV indicators in 2015 compared to 2020. The study sample consisted of adults who were aware of their HIV-positive status. In 2015, 76.8% (CI 95% 75.0-78.5) of individuals who tested HIV-positive during the survey were aware of their status, and this prevalence significantly increased to 86.8% (CI 95% 85.2–88.3) in 2020 (p-value < 0.001, results not shown in the table). Overall, there was a significant increase in all HIV indicators, except for adherence to antiretroviral therapy (ART), when comparing the survey before the UTT program (2015) to the survey after the program (2020). Stratifying the data by age and gender, ART enrollment among individuals aware of their HIV-positive status significantly increased in all subgroups, except for the 55-year-old and older age group, which already had a high enrollment rate, above 95% in 2015. Viral load suppression and viral load suppression 12 or more months after ART initiation showed similar patterns in the subgroup analysis. Both genders experienced a significant increase in both viral load suppression variables. However, among the age subgroups, only the 26 to 45 years age group showed a significant increase in both viral load suppression variables. Similar to ART enrollment rates, older people already had higher viral load suppression rates in 2015. Overall, there was no significant change in adherence to ART in the subgroup analysis between the two surveys, except for a slight increase among women (88.3% to 90.6%, p-value = 0.049) and the 46–55 age group (87.4% to 92.6%, p-value = 0.014).

Table 2.

The weighted prevalence of different HIV indicators in 2015 and 2020

HIV Indicator	2015	Weighted		2020	Weighted		p-ValueC
		%	95% CI		%	95% CI
Antiretroviral therapy (ART) enrollment	n = 2792			n = 2617
Total	2498	88.7	87.3–89.9	2547	97	96.1–97.7	< 0.001
Women	1680	88.8	87.1–90.4	1799	97.6	96.6–98.3	< 0.001
Men	818	88.4	85.8–90.6	748	95.9	93.9–97.3	< 0.001
Age
15–25	217	86.9	81.3–91	208	95.6	91.1–97.9	0.016
26–35	569	81.6	78.1–84.6	466	95.2	92.7–96.8	< 0.001
36–45	858	89.5	87-91.5	774	97.3	95.5–98.4	< 0.001
46–55	505	94.2	91.7–96	639	98.0	96.2–98.9	0.011
56+	349	96.0	93.1–97.7	460	97.3	95.2–99.2	0.294
Viral Load Suppression (VLS)	n = 2497			n = 2547
Total	2166	85.4	83.6–87.0	2313	90.3	88.9–91.6	< 0.001
Women	1476	87.0	85.1–88.7	1642	91.0	89.4–92.4	< 0.001
Men	690	82.7	79.3–85.6	671	89.0	86.1–91.4	0.002
Age
15–25	174	81.3	75.1–86.3	167	80.7	72.2–87.0	0.886
26–35	469	81.1	76.9–84.8	403	86.3	82.2–89.6	0.05
36–45	735	84.0	80.6–86.9	702	90.3	87.2–92.7	0.003
46–55	465	91.1	87.7–93.7	602	94.3	91.5–96.2	0.089
56+	323	92.3	87.6–95.3	439	95.5	92.5–97.4	0.154
12 + months on ART VLS	n = 2038			n = 2217
Total	1772	85.5	83.6–87.3	2010	90.0	88.5–91.3	< 0.001
Women	1200	86.1	84.0-88.1	1442	90.6	88.6–92.2	0.001
Men	572	84.5	81.0-87.5	568	89.0	86.0-91.4	0.037
Age
15–25	112	81.2	72.9–87.4	121	78.8	68.3–86.5	0.674
26–35	340	78.5	74.0-82.4	324	85.1	80.6–88.7	0.028
36–45	629	84.6	80.9–87.7	609	89.5	86.2–92.1	0.027
46–55	408	92.0	88.8–94.3	545	94.2	91.1–96.2	0.251
56+	283	92.7	87.6–95	411	96.1	93.1–97.8	0.115
Adherence to Antiretroviral therapy (ART)	n = 2337			n = 2412
Total	2058	87.6	86.1–89.0	2153	88.7	87.1–90.2	0.291
Women	1402	88.3	86.4–89.9	1560	90.6	88.9–92.1	0.049
Men	656	86.6	83.7–89.0	593	85.2	81.5–88.3	0.528
Age
15–25	143	80.2	72.7–85.9	158	86.1	76.6–92.1	0.269
26–35	456	86.1	81.8–89.5	377	86.2	81.6–89.8	0.970
36–45	727	89.0	86.2–91.2	635	86.3	83.0-89.1	0.187
46–55	422	87.4	83.7–90.3	564	92.6	89.8–94.7	0.014
56+	310	92.6	87.8–95.7	419	92.7	89.5–94.9	0.991

^cChi-Square test p-values, Bold meansp < 0.05

Table 3 displays the adjusted odds ratios (aOR) for the HIV indicators after controlling for gender and age. Adults who were aware of their HIV-positive status were 3.7 times more likely to be enrolled on ART in 2020 compared to 2015 (aOR = 3.70, 95% [2.80–4.99], p < 0.001). The odds of achieving viral load suppression were 50% higher among adults on ART in 2020 compared to 2015 (aOR = 1.50, 95% [1.22–1.84], p < 0.001). Similarly, the likelihood of achieving viral load suppression 12 or more months after ART initiation was 43% higher among adults in 2020 compared to 2015 (aOR = 1.43, 95% [1.15–1.73], p < 0.001). However, there was no significant change in adherence to ART in 2020 compared to 2015.

Table 3.

The odds ratios for different HIV indicators comparing 2020 to 2015

Variables	aOR	95% CI	p-Value
ART enrollment
2015	1
2020	3.70	2.80–4.99	< 0.001
Viral Load Suppression
2015	1
2020	1.50	1.22–1.84	< 0.001
12 + Months viral load suppression
2015	1
2020	1.43	1.15–1.78	0.001
Adherence to ART
2015	1
2020	1.06	0.86–1.30	0.591

Model adjusted for age and gender (aOR: Adjusted Odds Ratio; 95% CI: 95% Confidence Interval; Bold means p-value <0.05)

Discussion

This research study stands out as a notable contribution by utilizing data from two surveys conducted during both pre- and post-program launch periods to determine if there were changes in relevant HIV indicators following the implementation of the UTT program. The primary objectives of the study were to assess significant differences in the prevalence of ART enrollment, viral load suppression rates, and adherence to ART before and after the program’s implementation. Consistently, the results demonstrated a substantial improvement in ART enrollment and viral load suppression rates, which might plausibly be attributed to the implementation of the 2016 UTT HIV program. Importantly, there were no statistically significant changes in adherence to ART after the program’s implementation. These findings highlight significant improvements in ART accessibility and viral replication control while also affirming the absence of a negative influence on ART adherence. These changes occurred during a period that coincided with the implementation of the UTT program; however, other factors such as government commitment and policies, international funding and support, and health system strengthening could also explain the observed changes.

A considerable increase in ART enrollment was observed, with the prevalence rising significantly from 88.7% (87.3–89.9) to 97.0% (96.1–97.7) between the surveys. This corresponds to 3.7 times higher odds of being enrolled in ART in 2020 compared to 2015 for individuals aware of their HIV-positive status. Clinical trials [10, 12] for the UTT program further support these findings, demonstrating a significant increase in ART enrollment by reducing the rate of individuals who were not linked to ART or were lost to follow-up before initiation. Subgroup analysis by age and gender revealed a noteworthy improvement in ART enrollment across all groups, except for those aged 56 and above, who already had an enrollment rate exceeding 95% in 2015. Additionally, it is important to highlight that all subgroups achieved the 95% ART enrollment goal for individuals aware of their HIV status, as set by the United Nations [1].

Furthermore, an overall increase in viral load suppression was observed when comparing the two time points. Both overall viral load suppression and viral load suppression 12 or more months after ART initiation exhibited significant improvements between the two surveys, with adjusted odds ratios of 1.5 (p-value < 0.001) and 1.43 (p-value < 0.001) respectively. Viral load suppression serves as a proxy measure for retention in care [8], and our results align with other studies indicating higher retention rates among participants eligible for early ART initiation compared to those eligible at a later stage after HIV diagnosis in the CD4-based approach [8]. However, the age group of 45 years and above did not observe an improvement in viral load suppression between the surveys, possibly due to already high rates (over 90%) in 2015. Additionally, the younger age group (primarily 15–25) did not witness improvement in viral load suppression, despite having a prevalence rate below 90%. Consistent with this observation, a systematic review by Reif et al. [18] affirms that in low- and middle-income countries (LMICs), ART adherence and viral suppression outcomes remain poorer among young adults. The low prevalence may be attributed to low utilization of healthcare services among young people in general [19]. Recent innovative approaches targeting the needs of adolescents and youth are therefore warranted. Moreover, even though viral load suppression significantly increased, in all subgroups, it did not fulfill the United Nations’ recommended 95% target for those initiated on ART [1], and men had lower rates than women. Importantly, no changes were observed in adherence to ART between the two surveys, although a slight increase was noted in certain subgroups. Contrary to the belief that UTT programs might reduce ART adherence [9], our findings, which examined the time points before and after the 2016 UTT program implementation, did not reveal such an effect, nor the opposite. However, contrary to what’s reported by Hayes et al. [9] and the findings in our study, a clinical trial conducted in Eswatini reported a positive association between the implementation of the UTT program and retention in care among people enrolled in ART [8].

The study benefits from several strengths, including the use of nationally representative data from both surveys and the objective measurement of most HIV indicators, rather than relying on self-reporting. However, it is important to acknowledge certain limitations inherent in our study methodology. Our investigation relied on two cross-sectional surveys to explore the associations between variables, and it is essential to recognize that cross-sectional surveys may pose challenges in establishing causality. The absence of a control group in our study design makes it vulnerable to potential confounding factors. Additionally, potential confounders related to temporal trends in the healthcare landscape, such as advancements in healthcare quality over time, must be considered when interpreting our findings. It is also difficult to isolate a single factor responsible for the changes in HIV indicators observed in the study. Additionally, the minor overlap between Zimbabwe’s first PHIA data collection, completed in August 2016, and the rollout of the UTT program, initiated in June 2016 at pilot sites and expanded nationally in December 2016 [4], likely means the impact of UTT on the 2015 PHIA data is minimal, with early effects not fully captured. The concurrent implementation of other programs, momentum from previous years, and health systems strengthening may have influenced our outcomes.

Conclusion

Even in a resource-limited country like Zimbabwe, we demonstrated improved rates of ART enrollment, surpassing the 95% standard set by the United Nations. Additionally, viral load suppression rates increased overall, though they still fell short of the 95% target set by the United Nations. Among age groups, young people did not experience an increase in viral suppression, highlighting the need for programs that specifically target youth involvement in HIV intervention efforts. Furthermore, although viral load suppression rates increased among men, they still lagged behind those of women. Therefore, we recommend offering ART services during weekends and unconventional hours to accommodate young people in school and working men. Additionally, mobile and community-based HIV services have proven successful in increasing the utilization of HIV services among young people and men [20].

Contrary to the hypothesis suggesting a potential negative influence of UTT programs on ART adherence [9], the findings of this study could not be interpreted as supporting that notion, even considering the plausible effects of the 2016 UTT program. There is, however, room for improvement in ART adherence through strategies such as increasing patient education and counselling, implementing reminder systems like text messages and phone calls, and incorporating adherence support tools, such as pill reminder mobile applications. The findings from this research will not only contribute to the existing literature on HIV programs but also inform policymakers, healthcare providers, and stakeholders in Zimbabwe and beyond, facilitating evidence-based decision-making and resource allocation for more effective HIV prevention and treatment strategies.

Authors’ contributions

BSN, YVAP and GS: intellectual content, concept, design, data acquisition, data analysis, manuscript preparation, writing and manuscript review. SK, HZA, NGY, KA, INS and WA: manuscript editing, conceptualization, literature search and data acquisition. HY, BM, DP and BC: concept development, manuscript preparation, reviews, literature search and design.

Funding

None.

Data availability

Permission was obtained from the Population-based HIV Impact Assessment (PHIA) Project to acquire the data for this study. Upon request and registration through the website ( http://phia.icap.columbia.edu), the data is made accessible to the public.

Declarations

Ethics approval and consent to participate

The Population-based HIV Impact Assessment (PHIA) is conducted in partnership with the US President's Emergency Plan for AIDS Relief (PEPFAR). The surveys received ethical approval from the Zimbabwe Health and Human Research Review Board, as well as additional approval from the Institutional Review Board (IRB) at the Columbia University Medical Center. Participants were included in the study after providing their informed consent, either personally or through their parents/guardians if they were under 18 years old. The consent form clearly outlined the measures taken to protect their privacy, and participants were assured of their right to withdraw from the study without facing any negative consequences. This study strictly followed ethical guidelines and the research protocols approved by the Institutional Review Boards.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.