Correlates of unawareness of HIV status and recency of HIV infections among women living with HIV: findings from population-based surveys in 13 African countries

Abstract

Background

Despite efforts to increase HIV testing availability, a substantial number of women living with HIV (WLHIV) remain unaware that they have HIV. We explored the demographic, socioeconomic and clinical characteristics associated with being unaware of HIV status among WLHIV.

Methods

Secondary analysis of data from 13 population-based HIV impact assessment surveys. We used weighted χ² analysis and log-binomial regression to identify associations between awareness of living with HIV and various factors.

Results

Among 27 983 WLHIV, 7459 (26.6%) were unaware that they were living with HIV. Women at the extremes of age; 15–24 years (adjusted prevalence ratio (aPR): 1.84; 95% CI 1.67 to 2.03, p value <0.01), those >60 years; and those living in rural areas (aPR: 1. 09; 95% CI 1.02 to 1.18, p value 0.02) were more likely to be unaware of that they were living with HIV. Of the 7459 women who were unaware that they were living with HIV, 7071 (94.8%) had long-term HIV infection. Factors associated with long-term HIV infection included: older age 35–44 years (aPR: 1.03; 95% CI 1.02 to 1.08, p value <0.01), 45–59 years (aPR: 1.05; 95% CI 1.01 to 1.09, p value 0.02) and having no sexual partner in the past 12 months (aPR: 1.04; 95% CI 1.00 to 1.09, p value 0.04).

Conclusion

A high proportion of women who were unaware that they were living with HIV had long-term HIV infections. HIV testing interventions should be targeted towards these women to improve early access to HIV treatment.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• HIV testing is a critical entry point for treatment and prevention, yet a substantial number of women living with HIV (WLHIV) in sub-Saharan Africa remain unaware of their status. Existing literature has identified barriers such as stigma, limited healthcare access and low perceived risk, particularly among young women.

WHAT THIS STUDY ADDS

• This study reveals that about one in four WLHIV in 13 sub-Saharan African countries is unaware of their HIV status, with about 94.8% of these women having long-term infections. It identifies key factors associated with unawareness, including extremes of age i.e younger or older age, rural residence, not being engaged in any recent sexual activity, non-condom use and never visiting a Tuberculosis (TB) clinic.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Findings highlight the need for targeted HIV testing interventions, especially for women at the extremes of age, those not actively engaged in sexual relationships, those in rural settings and those in higher income settings. Policies promoting integrated, routine and stigma-free HIV testing, especially outside traditional healthcare settings, could help close this gap and improve early diagnosis and treatment.

Introduction

The UNAIDS 2025 AIDS targets state that 95% of people living with HIV should know their status, 95% of these should be initiated on antiretroviral therapy (ART) and 95% of those on ART should be virally suppressed by 2030.¹ However, by the end of 2022, out of 39.0 million PLHIV globally, only 76% were receiving ART, and 71% achieved virologic suppression.^{1 2} Only five sub-Saharan countries, that is, Botswana, Eswatini, Rwanda, Tanzania and Zimbabwe, had achieved all three 95-95-95 targets by the end of 2022.²

In the same year, there were 1.3 million new HIV infections globally with 46% of these being women. In sub-Saharan Africa, adolescent girls and young women accounted for more than 77% of new infections among young people aged 15–24 years in 2022.³ This means that young women were two times as likely to get newly infected with HIV as young men.² Various factors contribute to the risk of acquiring HIV among young women, including early sexual debut, engaging in multiple concurrent sexual relationships, intergenerational sex and inability to negotiate for consistent condom use in both long-term relationships and casual partnerships. Additionally, women on the fringes of society, eg, commercial sex workers, are more likely to acquire HIV due to unfavourable policies and frequent gender-based violence.^{4 5}

HIV testing is a critical entry point for both treatment and prevention initiatives. Testing early during HIV infection (within the first 12 months of infection) has several benefits for both the patient and the population, including early initiation of ART, leading to optimum viral load suppression, which in turn leads to reduced transmission. Although various interventions have been implemented to increase access to HIV testing services for young women, for example, community-based outreaches, targeted community-based testing for close contacts of persons newly diagnosed with HIV, and more recently, increased access to oral self-testing, several barriers to access to HIV testing still exist. These may be patient-related, for example, low perceived risk of having HIV infection,^{6 7} or health system-related, for example, lack of dedicated staff to conduct tests, and logistical difficulties in carrying out community outreaches.^{6 7} To further improve access to HIV testing, these barriers should be overcome.

Utilising data from population-based surveys, we aimed to describe the proportion of women in sub-Saharan Africa who are unaware that they are living with HIV and to determine the socio-demographic and clinical factors associated with this. We also aimed to establish how long women who were unaware that they were living with HIV had had the infection and the factors associated with long-term HIV infection among these women. Findings from this study could help refine future interventions to improve access to HIV testing for women in sub-Saharan Africa.⁸

Methods

We carried out a secondary analysis of data from population-based HIV impact assessments (PHIAs) conducted in sub-Saharan Africa from October 2015 to March 2019. The methods used in these surveys have been detailed elsewhere.⁹ In summary, the PHIAs utilised two-stage cluster sampling to obtain nationally representative samples of households in each country, stratified by designated geographical zones. Individuals were eligible to participate if they slept in the household the night before the interview and were at least 15 years old. Face-to-face interviews were conducted to capture demographic, behavioural and clinical information, including self-reported knowledge of HIV status and testing history. In each household, home-based HIV counselling and testing were carried out according to each nation’s testing algorithm. Participants who tested positive for HIV received a point-of-care CD4 count, lab-based HIV viral load and had a dry blood spot specimen taken to screen for detectable concentrations of antiretroviral drugs (ARVs). Participants were also tested for recency of their HIV infection following an algorithm that uses the HIV-1 LAg Avidity test, HIV-1 viral load and the detection of HIV antiretrovirals in blood (figure 1).

Figure 1. Classification of recent versus long-term HIV-1 infection using LAg avidity assay, viral load and ARV detection.²⁹ ODn, normalised optical density.

Study outcomes

• HIV positive status: patients were determined to be HIV positive if they had a positive rapid test at the household, followed by laboratory confirmation using the Geenius HIV 1/2 Supplemental Assay (Bio-Rad, Hercules, California).¹⁰

• Viral load suppression (VLS): viral load suppression was defined as <1000 HIV RNA copies per millilitre of blood.¹¹

• Awareness of living with HIV: participants who tested positive for HIV during the survey were considered aware that they were living with HIV if they (a) reported in the interview being HIV positive or (b) had ARVs detected in their blood.

• Unawareness of living with HIV: participants who tested positive for HIV during the survey were considered unaware that they were living with HIV if they reported having a negative or unknown HIV status, and ARVs were not detected in their blood.

• Long-term HIV infection: participants were classified as having long-term HIV infection based on the PHIA recent infection testing algorithm (RITA), which combines HIV-1 LAg Avidity assay results with clinical and laboratory indicators to improve accuracy. A participant was considered to have a long-term infection if:

1. The LAg Avidity test result directly indicated a long-term infection (normalised optical density ≥1.5); OR

2. The LAg result suggested a recent infection but one or more of the following criteria applied, indicating the infection was actually long-standing:

   1. CD4 count <200 cells/mm³, suggestive of advanced immunosuppression.

   2. HIV viral load <1000 copies/mL, which may reflect long-term infection with viral suppression (either due to ART or natural control)

   3. Detection of ART in the participant’s blood, or self-reported ART use.

   4. Presence of AIDS-defining illnesses or clinical symptoms consistent with advanced HIV disease.

These additional criteria were applied to avoid misclassifying long-term infections as recent, particularly among individuals on treatment or with disease progression. This classification approach is consistent with PHIA protocols and international guidelines for cross-sectional HIV incidence estimation.

Data management and analysis

Data were downloaded from the PHIA website in a STATA-readable format and imported into STATA V.18 for cleaning and analysis. The cleaning process involved several steps to ensure consistency and readiness for pooled analysis across countries. We began by retaining only the variables relevant to our study, including age, residence (urban/rural), wealth index, marital status, region, number of sexual partners, condom use at last sex, Tuberculosis (TB) diagnosis, Sexually Transmitted Diseases (STD) diagnosis and Sexually Transmitted Infections (STI) symptoms in the past 12 months. Variables with differing coding schemes across countries such as marital status or regional classifications were harmonised for uniformity. The wealth index, which in PHIA is derived from household asset data and categorised into quintiles (poorest, poor, middle, richer, richest), was recoded into three categories for our analysis: low (poorest and poor), middle (middle) and high (richer and richest). Observations with missing or implausible values for key variables (eg, missing age or wealth index) were excluded. We also assessed the extent of missing data and found it to be minimal and unlikely to bias the results. All variables were properly labelled, and categorical variables were formatted appropriately for analysis. Once cleaning was complete, we appended datasets from 13 countries—Cameroon (2017–2018), Côte d’Ivoire (2017–2018), Eswatini (2016–2017), Ethiopia (2017–2018), Kenya (2018), Lesotho (2016–2017), Malawi (2015–2016), Namibia (2017), Rwanda (2018–2019), Tanzania (2016–2017), Uganda (2016–2017), Zambia (2016) and Zimbabwe (2015–2016)—into a single dataset. We also created a new variable to indicate the country of origin, to account for potential country-level clustering in subsequent analyses. All data management procedures were conducted using STATA V.18.

Countries were grouped into Western Africa, Eastern Africa, South-Eastern Africa and Southern Africa using standard UN regional definitions. Data were weighted to account for sample selection probabilities and to adjust for non-response and non-coverage. Normalised weights in Rwanda and Kenya were rescaled to population weights for consistency with the rest of the countries. We used Taylor series expansion to obtain robust variance estimators for the complex survey data. Weighted χ² analysis was conducted to identify associations between awareness of HIV+status, sociodemographics, clinical characteristics and risky sexual behaviours among women living with HIV. Variables with a p value of less than 0.2 on the χ² statistic were then entered into a bivariable and multivariable log-binomial regression model to identify associations between awareness of HIV positive status and various sociodemographic and clinical characteristics. We also fit a log-binomial model to identify factors associated with the long-term HIV infection among WLHIV who were unaware of their HIV+status. We calculated weighted crude and adjusted-prevalence ratios, examining the variables of interest and accompanying 95% CIs. The level of significance was set to α=0.05 in the fitted models. All analyses were conducted in STATA V.18.12,14

Implementation

Implementation of the surveys that contributed data to this study was carried out by the respective ministries of health, supported by the International Center for AIDS Care and Treatment Programs at Columbia University, and the University of California San Francisco. The initial surveys in each country obtained the relevant approvals from the research and ethics review boards of the respective countries and the U.S. Centers for Disease Control and Prevention. We requested permission from the PHIAs website to access data for our countries of interest. Secondary data that were deidentified were used for this study.

This study involved secondary analysis of deidentified, publicly available data from the PHIA surveys. All PHIA surveys obtained written informed consent from participants at the time of primary data collection, including parental or guardian consent for minors and assent from those aged below the age of majority, as per each country’s ethical guidelines. Since this analysis used secondary, anonymised data with no direct participant contact, no additional consent was required. Permission to access and analyse the data was obtained through a formal request to the PHIA data custodians. This study does not directly involve human participants.

Patient and public involvement statement

None.

Results

Demographic and clinical characteristics

A total of 27 983 women living with HIV from 13 countries were included in this study. About 26.6% of these women were not aware that they were living with HIV. Most women who were unaware that they were living with HIV were from the East African region (47.8%), were between 25 and 34 years old (33%) and were residing in urban areas (50.4%) (table 1). The majority of the women who were unaware that they were living with HIV also had long-term infections, as was evidenced by low CD4 cell counts (61.1% had CD4 cell counts<500 cells/µL) (table 1).

Table 1. Demographic and clinical characteristics among women living with HIV who were unaware and aware that they were living with HIV.

Variables	Aware N=22 543 (%) 73.4	Unaware N=5440 (%) 26.6	P value
Age groups
15–24	2097 (9.2)	1334 (21.89)	<0. 01
25–34	6718 (30.0)	1832 (33.0)
35–44	7487 (34.0)	1214 (24.3)
45–59	5514 (24.0)	879 (17.0)
≥60	737 (3.0)	163 (3.9)
Residence
Urban	9489 (50.4)	2519 (50.4)	01
Rural	13 054 (49.6)	2921 (49.6)
Marital status
Never married or never lived together	4640 (13.2)	1208 (18.5)	<0.01
Married or living together	10 365 (47.5)	2552 (47.4)
Divorced/separated/widowed	7453 (39.3)	1668 (34.1)
Wealth index
Low	9219 (34.6)	2021 (34.7)	0.54
Middle	4745 (22.0)	1159 (20.8)
High	8554 (43.4)	2258 (44.6)
Number of sexual partners in the past 12 months
None	5440 (33.1)	992 (23.7)	<0.01
One	11 989 (59.3)	2886 (63.9
Two or more	1323 (7.6)	514 (12.5)
Condom use in the past 12 months
No sexual intercourse	6061 (33.0)	1189 (25.2)	<0.01
Used condom	7249 (25.7)	862 (12.1)
Did not use a condom	7416 (41.3)	2987 (62.6)
STD diagnosis in the past 12 months
No STD diagnosis	5820 (94.6)	1579 (95.8)	0.20
STD diagnosis	303 (5.4)	67 (4.2)
STI symptoms in the past 12 months
No STI symptoms	7669 (89.3)	2364 (87.7)	0.17
STI symptoms	902 (10.7)	291 (12.3)
TB diagnosis
Not visited a TB clinic	14 820 (70.3)	5051 (94.9)	<0.01
Visited TB clinic (did not have TB)	4196 (16.0)	239 (3.2)
Visited TB clinic (had TB)	2456 (13.7)	124 (1.8)
Ever tested for HIV
No	333 (2.3)	1365 (28.7)
Yes	22 210 (97.7)	4075 (71.3)
Region
Western Africa	1094 (10.7)	894 (23.0)	<0.01
Eastern Africa	5355 (48.3)	1664 (47.8)
Southeastern Africa	5608 (29.4)	1566 (24.7)
Southern Africa	10 486 (11.6	1316 (4.5)
CD4 count
<200	1497 (8.5)	729 (15.7)	<0.01
≥200 and <500	7502 (40.3)	2377 (45.4)
≥500	11 516 (51.2)	1900 (38.9)
Viral load
Suppressed (<1000 copies/mL)	18 638 (81.4)	610 (10.9)	<0.01
Not suppressed	3855 (18.6)	4801 (89.1)
Recency of HIV infection
Recent infection	30 (0.1)	326 (5.2)	<0.01
Long-term infection	22 499 (99.9)	5104 (94.8)

The frequencies are not weighted, and the percentages are weighted. Zambia did not ask questions about the number of sexual partners in the past 12 months. Uganda, Cameroon, Ivory Coast, Eswatini, Lesotho, Namibia, Rwanda and Tanzania did not ask questions about STD diagnosis in the past 12 months. Uganda, Cameroon, Ivory Coast, Eswatini, Lesotho, Namibia and Rwanda did not ask questions about STI symptoms in the past 12 months. Rwanda and Kenya did not measure CD4 count.

%, weighted percentage; N, unweighted frequency; STD, Sexually Transmitted Diseases; STI, Sexually Transmitted Infections; TB, Tuberculosis.

Factors associated with being unaware that one is living with HIV

Compared with women of 35–44 years, women at the extremes of age 15–24 years (aPR; 1.84; 95% CI 1.67 to 2.03, p≤0.01) and 60 years and above (aPR; 1.63; 95% CI 1.37 to 1.94, p≤0.01) were more likely to be unaware that they were living with HIV (table 2).

Table 2. Factors associated with being unaware that one is living with HIV.

Variables	Unadjusted PR (95% CI)	P value	Adjusted PR (95% CI)	P value
Age groups
35–44	Ref		Ref
15–24	2.79 (2.59 to 3.02)	<0.01	1.84 (1.67 to 2.03)	< 0.01
25–34	1.54 (1.43 to 1.65)	<0.01	1.22 (1.12 to 1.34)	< 0.01
45–59	1.0 (0.92 to 1.09)	0.95	1.04 (0.93 to 1.17)	< 0.51
60 or more	1.3 (1.10 to 1.53)	<0.01	1.63 (1.37 to 1.94	< 0.01
Residence
Urban	Ref		Ref
Rural	1.00 (0.94 to 1.06)	1.00	1.09 (1.02 to 1.18)	0.02
Marital status
Married or living together	Ref		Ref
Never married	1.27 (1.17 to 1.38)	<0.01	1.17 (1.06 to 1.29)	<0.01
Divorced/separated/widowed	0.90 (0.84 to 0.97)	<0.01	1.09 (1.00 to 1.19)	0.04
Wealth index
Low	Ref		Ref
Middle	0.95 (0.88 to 1.04)	0.30	1.02 (0.93 to 1.11)	0.717
High	1.02 (0.95 to 1.09)	0.64	1.20 (1.11 to 1.31)	< 0.01
Number of sexual partners in the past 12 months
One	Ref		Ref
None	1.36 (1.25 to 1.49)	< 0.01	1.38 (1.21 to 1.57)	< 0.01
Two or more	1.82 (1.60 to 2.06)	<0.01	1.06 (0.94 to 1.19)	0.34
Condom use in the past 12 months (among those who are sexually active)
Used condom	Ref		Ref
Did not use a condom	2.44 (2.20 to 2.70)	<0.01	1.95 (1.75 to 2.17)	< 0.01
TB diagnosis
Not visited a TB clinic	Ref		Ref
Visited TB clinic (did not have TB)	4.78 (4.02 to 5.68)	<0.01	3.84 (3.17 to 4.65)	< 0.01
Visited TB clinic (Had TB)	0.67 (0.50 to 0.89)	0.01	0.65 (0.47 to 0.92)	0.01
Region
Western Africa	Ref		Ref
Eastern Africa	0.60 (0.55 to 0.65)	<0.01	0.65 (0.60 to 0.71)	< 0.01
Southeastern Africa	0.53 (0.49 to 0.58)	<0.01	0.53 (0.48 to 0.59)	< 0.01
Southern Africa	0.28 (0.26 to 0.31)	<0.01	0.42 (0.39 to 0.46)	< 0.01

PR, Prevalence Ratio; TB, Tuberculosis.

Women living in rural areas (aPR; 1.09; 95% (1.02 to 1.18, p=0.017), those who were never married (aPR: 1.16; CI 1.06 to 1.28, p=0.002) and those who were divorced/separated/widowed (aPR: 1.10; 95% CI 1.00 to 1.19, p=0.040) were more likely to be unaware that they were living with HIV. Women in the high wealth index category (aPR: 1.20; 95% CI 1.10 to 1.30, p≤0.001), those who did not have any sexual partners in the past 12 months (aPR: 1.37; 95% CI 1.20 to 1.56, p≤0.001), women who had never visited a TB clinic (aPR: 3.91; 95% CI 3.22 to 4.73, p≤0.001) were more likely to be unaware that they were living with HIV. Compared with those who had used a condom, women who did not use a condom in the past 12 months (aPR: 1.96; 95% CI 1.75 to 2.18, p≤0.001) were more likely to be unaware that they were living with HIV (table 2).

Factors associated with long-term HIV infections among women unaware that they are living with HIV

Among women who were unaware that they were living with HIV, 94.8% had long-term infections. Being 35–44 years (aPR: 1.03; 95% CI 1.02 to 1.08, p value <0.01), 45–59 years (aPR: 1.05; 95% CI 1.01 to 1.09, p value 0.02), women who had no sexual partner (aPR: 1.04; 95% CI 1.00 to 1.09, p value=0.04), did not use a condom at last sexual intercourse (aPR: 1.04; 95% CI 1.00 to 1.08, p value 0.04) were more likely to have long-term HIV infections (table 3).

Table 3. Factors associated with long-term HIV infections among women unaware that they are living with HIV.

Variables	Unadjusted PR (95% CI)	P value	Adjusted PR (95% CI)	P value
Age groups
15–24	Ref		Ref
25–34	1.03 (1.01 to 1.06)	0.02	1.03 (1.00 to 1.06)	0.08
35–44	1.06 (1.04 to 1.09)	<0.01	1.05 (1.02 to 1.08)	<0.01
45–59	1.06 (1.04 to 1.10)	<0.01	1.05 (1.01 to 1.09)	<0.01
60 or more	1.07 (1.03 to 1.10)	<0.01	1.04 (0.99 to 1.09)	0.11
Residence
Urban	Ref		Ref
Rural	0.99 (0.98 to 1.01)	0.27	1.00 (0.98 to 1.01)	0.58
Marital status
Married or living together	Ref		Ref
Never married or lived with a man	1.00 (0.98 to 1.03)	0.80	1.01 (0.98 to 1.03)	0.72
Divorced/separated/widowed	1.02 (1.00 to 1.04)	0.04	1.01 (0.98 to 1.03)	0.55
Wealth index
Low	Ref		Ref
Middle	0.99 (0.97 to 1.01)	0.38	1.00 (0.97 to 1.03)	0.97
High	1.01 (0.99 to 1.03)	0.27	1.01 (1.00 to 1.04)	0.22
Number of sexual partners in the past 12 months
One	Ref		Ref
None	0.97 (0.96 to 0.99)	0.01	1.04 (1.00 to 1.09)	0.04
Two or more	0.94 (0.90 to 0.98)	<0.01	0.96 (0.91 to 1.02)	0.17
Condom use in the past 12 months
Used condom	Ref		Ref
Did not use a condom	1.03 (1.02 to 1.06)	0.06	1.04 (1.00 to 1.08)	0.04
TB diagnosis
Not visited a TB clinic	Ref		Ref
Visited TB clinic (did not have TB)	0.99 (0.95 to 1.04)	0.77	0.99 (0.94 to 1.04)	0.71
Visited TB clinic (had TB)	1.01 (0.95 to 1.07)	0.80	0.98 (0.90 to 1.06)	0.61
Region
Western Africa	Ref		Ref
Eastern Africa	0.98 (0.96 to 1.00)	0.05	0.97 (0.95 to 0.99)	0.01
Southeastern Africa	0.98 (0.96 to 0.997)	0.02	0.98 (0.96 to 1.01)	0.20
Southern Africa	0.96 (0.94 to 0.98)	<0.01	0.98 (0.95 to 1.01)	0.12

PR, Prevalence Ratio; TB, Tuberculosis.

Discussion

This study reveals that a significant proportion of women in sub-Saharan Africa is unaware that they are living with HIV, with the majority of these women having long-term HIV infections. This lack of awareness was particularly pronounced among women at the extremes of age, younger women aged 15–24 and those aged 60 and above, highlighting an age-related disparity in HIV awareness. Young women face numerous barriers to HIV testing, including stigma, limited access to healthcare services and a lack of perceived risk.¹⁵ Older women, on the other hand, might not perceive themselves at risk or might have missed earlier waves of HIV testing campaigns.¹⁶

The analysis shows that women living in rural areas and those with a high wealth index are more likely to be unaware of their HIV status.¹⁷ Rural women often face additional challenges, such as limited healthcare infrastructure and long distances to testing centres, which impede regular testing and follow-up.¹⁸ Interestingly, women in the high wealth index category also show a higher likelihood of unawareness, which may be attributed to lower engagement with public health interventions typically targeted at lower income populations.¹⁹

Women who did not use condoms were more likely to be unaware that they were living with HIV. Additionally, women who had never visited a TB clinic were more likely to be unaware of their HIV status. This is most likely because the current practice in most TB clinics for HIV high-burden countries is to test all patients with presumptive TB for HIV, as guided by the WHO.^{20 21} This highlights the potential role of integrated healthcare delivery in improving HIV awareness and testing uptake.

From this study, women who did not have a stable sexual partner (widowed or divorced) were more likely to be unaware of their HIV-positive status compared with married women. These findings may also indicate that the absence of a stable partner can contribute to lower perceived risk and, therefore, lower motivation to seek healthcare services. However, this could also be because women who are in stable relationships or marriages have more chances of interacting with HIV testing and counselling opportunities, for example, through antenatal care services.^{22 23}

These findings underscore the need for more inclusive and targeted HIV testing strategies.²⁴ For younger women, interventions could include integrating HIV testing with reproductive health services and increasing community-based testing opportunities.²⁵ For older women, campaigns should focus on routine testing and awareness programmes that specifically address older age groups.¹⁶ Additionally, enhancing access to testing in rural areas through mobile clinics and community health workers could bridge the gap for rural women,¹⁷ and designing workplace programmes that target women in formal employment may help reach these categories of women.

A striking 94.8% of women who were unaware of their HIV-positive status had long-term HIV infections.²⁶ Long-term infection was evidenced by the absence of recent infection markers and low CD4 counts. This finding indicates that these women have been living with the virus for several years without diagnosis, thereby missing the benefits of early treatment, such as improved health outcomes and reduced transmission risk.²⁷ Some of the factors associated with long-term HIV infections, for example, not having a sexual partner in the past 12 months, and not using condoms during the last sexual encounter, overlapped with the factors associated with being unaware of one’s HIV status, suggesting an interplay of social and behavioural dynamics that delay HIV testing and diagnosis.²⁸

To address the high proportion of long-term HIV infections, it is essential to promote regular HIV testing among women who do not perceive themselves at risk, such as those with no stable sexual partners, and those who have been divorced or widowed.¹⁸ Strategies could include routine screening during general health visits and leveraging social networks to disseminate information and reduce stigma associated with HIV testing.¹⁵ Among these women, condom use should also be emphasised.

Limitations

This study has several limitations that should be considered when interpreting the findings. First, the RITA used to classify HIV infections as recent or long-term inherently recategorises anyone on ART or who is virally suppressed as having a long-term infection, regardless of their actual time since infection. While this approach is appropriate at the population level, it may underestimate the number of recent infections among individuals who became aware of their status and initiated treatment soon after infection. As such, some women classified as having long-term infections may have been infected within the past year but were misclassified due to the algorithm’s design. Second, while the analysis of CD4 counts provides helpful context in assessing the chronicity of infection, the absence of a detailed HIV testing history limits our ability to validate or refine these classifications further. Access to information on prior HIV testing or date of diagnosis would enhance the interpretation of infection timing and treatment engagement. Additionally, standard limitations of self-reported survey data apply. Responses to questions about HIV testing history, sexual behaviour and healthcare utilisation are subjected to recall bias and social desirability bias, which may affect the accuracy of the reported information. This was a robust study that utilised population-representative data from 13 countries. In addition, the study connected biomarkers to demographic and behavioural information. However, since this was an analysis of secondary data that was already collected, some variables like the number of sexual partners in Zambia and CD4 count were missing for certain countries and therefore could not be analysed.

In conclusion, there is a significant proportion of WLHIV who are unaware of their status, and the majority of these have long-term infections. These findings highlight critical gaps in current HIV testing and awareness programmes emphasising the need for targeted interventions.

Data availability statement

Data are available upon reasonable request.