Uptake of HIV testing

HIVST increased the uptake of HTS. Three (35–37) of the five randomized controlled trials reported on uptake of HIV testing. A meta-analysis of these results indicated that HIVST doubled the uptake of HIV testing compared to standard HTS (Relative Risk (RR) = 2.12; 95% CI: 1.51, 2.98) (see Fig. 2.2). (See Annex 16 for further details.)

Fig. 2.2

Meta-analysis of HIV testing uptake at three and six months. M-H: Mantel-Haenszel; CI: Confidence interval.

The two studies, in Kenya where women distributed HIVST kits to their male partners, also reported an increased uptake of couples testing when compared with the approach of giving male partners letters or vouchers inviting them to test at a clinic (35,36).

Among men who have sex with men in Hong Kong SAR, one study (37) reported that uptake of HIV testing was higher in the HIVST group compared to the facility-based HIV testing group, both in the sub-groups of recent testers (>1–3 tests in 3 years) (RR = 1.75; 95% CI: 1.46, 2.08) and non-recent testers (0 tests in 3 years) (RR = 2.22; 95% CI: 1.61; 3.08). Analysis of uptake in this study also showed that the men who have sex with men who reported condomless anal intercourse at enrolment were also more likely to test if they were in the HIVST group compared to the standard HTS group (RR = 1.75; 95% CI: 1.26, 1.81) (37). These results suggests that higher risk men who have sex with men may be more likely to take up HIVST than standard facility-based HTS.

These findings are consistent with results from other systematic reviews (4,6,38), trials and observational studies from Kenya (39,40), Lesotho (41), Malawi (10,42) and Zimbabwe (16), which report similar increases in the uptake of HIV testing following the offer of HIVST. For instance, a two-year cluster randomized trial in Malawi reported uptake of directly assisted community-based HIVST among the general population at approximately 76.5% (crude uptake: 84%, 14 004/16 660) when adjusted to account for population turnover (10). Over the two-year period, the study reported that 44% were first-time testers. Uptake was consistently high in years one and two among adolescents (16–19 years of age) (95%; 2374/2502 and 2405/2502), young people (16–29 years of age) (90%; 8333/9315 and 8503/9315), women (85%; 6835/7802 and 6445/7802) and men (68%; 5902/8643 and 5924/8643) (10). Likewise, studies among key populations, primarily men who have sex with men and female sex workers, also report high uptake of HIV testing when HIVST is offered (4,5).

Frequency of HIV testing

HIVST increased the frequency of HIV testing among men who have sex with men. Two of the five randomized controlled trials reported on frequency of HIV testing among men who have sex with men (33,34). Across both studies, men in the HIVST group had a mean of two more tests in a 12–15 month period than those in the facility-based HTS group (mean difference = 2.13; 95% CI: 1.59, 2.66) (33,34) (see Fig. 2.3).

Fig. 2.3

Meta-analysis of mean number of tests among men who have sex with men in 12–15 month period. M-H: Mantel-Haenszel; SD: Standard deviation; CI: Confidence interval; IV: Independent variable.

In Australia, among non-recent testers (0 tests in >2 years or never tested) HIVST substantially increased the frequency of HIV testing compared to standard testing (33,34). The rate ratio for this sub-group showed that men in the HIVST group had a testing rate 5.54 times higher in a 12-month period than men in the standard HIV testing group (Rate Ratio = 5.54; 95% CI: 3.15, 9.74).

In the United States, HIVST was also shown to increase the uptake of quarterly HIV testing among men who have sex with men (76%; 74/98) compared to those receiving standard HTS (54%; 53/99) (33).

Although the results considered above are from a small number of studies, the reported increases in the frequency of HIV testing could have important public health implications, particularly when reaching people with undiagnosed HIV infections and those at high ongoing risk who are not accessing existing services. Sustained increases in frequency of HIV testing among higher risk populations of this magnitude or more, facilitated by HIVST, could identify a greater number of infections and result in a reduction in HIV incidence if individuals are effectively linked to prevention and treatment services (43,44).

HIV risk behaviour following HIVST

There was no increase in HIV risk behaviour identified following HIVST. One of the five randomized controlled trials reported on risk behaviour following HIVST (33). In this study, men who have sex with men in the HIVST group did not increase condomless anal intercourse compared to those receiving standard HTS, as reported at a nine-month follow-up (RR = 0.94: 95% CI: 0.55, 1.61). In this same study, men in the HIVST group acquired fewer bacterial sexually transmitted infections (STIs) than those in the standard HIV testing group (RR = 0.41; 95% CI: 0.15, 1.13).

Additionally, in Australia, HIVST was found to neither decrease nor increase the frequency with which men who have sex with men tested for STIs (34).

Although not reported in the reviewed randomized controlled trials, other observational studies suggest some users may be interested in using HIVST to screen potential sex partners (11,18,45,46). Studies have shown this could prevent users with a reactive self-test result from engaging in condomless sex (18,45). This evidence also suggests HIVST could be used to serosort (select a sex partner based on their self-reported HIV status) and to inform decisions about behaviour, including on using HIV prevention, such as condoms (18,45). In many of these studies, however, users were unaware that current HIVST technologies have a three-month window period and do not identify acute or early HIV infection.

While the decision to not use condoms may be appropriate in longer term relationships, where both partners have low ongoing risk, serosorting following HIVST is not recommended in population groups with high HIV incidence (for example, sex workers and men who have sex with men) (47). Clear messages about the potential risks of using HIVST to make decisions about risk behaviour and use of HIV prevention need to be communicated, particularly to key populations who may be at high ongoing risk of HIV.

Social harm

Only one instance of harm, which was not directly related to HIVST, was identified in the review (35). In that randomized controlled trial, there was one report of intimate partner violence (IPV) in both the HIVST group and the standard HIV testing group (1/297 in the HIVST group; 1/303 in the control group) (35). In the HIVST group, the harm was not directly related to HIVST: the female participant reported that the occurred because she took part in the study without consulting her husband (35). The participant reported leaving home for approximately three weeks, before returning. At a two-month follow-up visit, she informed researchers that she and her husband had reconciled. In the standard HIV testing group, one female participant also reported IPV (35).

These results suggest that HIVST may not directly influence the risk of IPV, but that these risks largely depend on the setting, context and relationship dynamics of couples and partners. Such findings are consistent with those reported by systematic reviews assessing harm across all forms of HTS (48), couples HTS (49), and self-testing for HIV and other conditions and diseases (50), as well as other observational studies. In an urban setting in Malawi, results from a two-year cluster randomized trial reported no cases of IPV, self-harm or suicide resulting from HIVST or other HIV testing in the community (10). In the United States, a study reported that, out of 124 events where men offered HIVST to their male sexual partners, 7 resulted in a verbal confrontation and none resulted in physical violence (45).Several studies have also shown that some initial reports of “coercion” to test were rather “persuasion” or “encouragement” to test (10,20,51). For example, in Malawi, it was primarily men who self-tested with their female partner who reported being coerced to test; of whom 94.4% (252/267) also said they would recommend HIVST to friends and family, and 92.2% (130/141) said they were highly satisfied with HIVST (10).

Nevertheless, while most users consider HIVST to be empowering (20,51), a study in Malawi on couples who received HIVST kits reported that two of 17 couples felt pressured by their partner to self-test and found dealing with serodiscordant results challenging (20). Likewise, in Kenya, four of 265 HIV-negative pregnant and postpartum women and female sex workers who distributed HIVST kits to their male partners and clients reported instances of IPV; two were among postpartum women and two were among female sex workers (18). It was unclear if these cases were directly linked to HIVST since prior to the intervention 41% of all female study participants said they had experienced IPV in the preceding 12 months. The two women in postpartum care who reported experiencing verbal abuse in a confrontation with their husbands said this was regarding a reactive test result (18). Both women left their homes but later returned and reconciled with their husbands. One woman reported that her husband (who was diagnosed HIV-positive) has since enrolled in care and that they are now using condoms during sex (18). Of concern are the two cases among female sex workers who distributed HIVST to their clients and experienced physical violence; one reported forced condomless sex (18), suggesting that not all testing approaches are appropriate for all contexts, and that caution may still be needed among vulnerable populations.

While the overall results are encouraging, it is critical for programmes to recognize the importance and complexity of monitoring, reporting, evaluating and assessing social harm in relation to HIVST. As recommended with all HTS, programmes need to consider context-specific approaches to implementing HIVST in ways that are ethical, safe and acceptable. In addition, risk mitigation in relation to social harm and the establishment of active monitoring and reporting systems are important.

HIVST positivity

A major aim of any HTS approach is to be as efficient and effective as possible in reaching people with undiagnosed HIV. One of the five randomized controlled trials reported on HIV positivity (33), and indicated that men who have sex with men in the HIVST group were twice as likely to have an HIV-positive test than those in the standard care group (RR = 1.97; 95% CI: 0.37, 10.52).

No other randomized controlled trials included in the GRADE analysis reported on HIV positivity, that is, the proportion of people with a reactive and confirmed positive HIV self-test result. However, several other studies on HIVST, which do not make a direct comparison with standard HTS, report HIV positivity ranging from 3–14% among the general population in sub-Saharan Africa and from 1–30% among key populations (see Table 2.1).

Table 2.1

Summary of HIV positivity studies implementing HIVST.

Despite the lack of a comparison with standard HTS, these studies describe HIV positivity on a par with that reported in many other HTS approaches in similar populations and settings (38).

It is important to note that, as HTS coverage increases, the proportion of HIV-positive tests and new diagnoses will likely decrease across all approaches. Thus, more strategic and focused methods will be required in order to continue to achieve similar, or greater, levels of HIV positivity for all HTS approaches, including HIVST.

Linkage to further HIV testing and HIV prevention, treatment and care services

One of the five randomized controlled trials reported on linkage to further testing following HIVST (36). In that study, 72% (n=396) of male partners of pregnant women who received an HIVST kit reported attending a facility for further HTS (36). This outcome could not be compared with standard facility-based HTS because only HIVST requires linkage to further testing.

No other studies included in the GRADE analysis reported on linkage to further HTS or HIV prevention, treatment and care services. Therefore, the findings from additional trials and observational studies, which did report on outcomes related to linkage to care, were summarized. An overview of these findings is included here.

Across general populations in sub-Saharan Africa, linkage to care was 50–56%. Among the general population in Malawi, community-based HIVST together with home-based assessment and initiation of treatment led to a threefold increase in linkage to care compared with standard HTS and facility-based care (181/8194, 2.2% vs. 63/8466, 0.7%; RR = 2.94; 95% CI, 2.10, 4.12; p < 0.001) (60). Also in Malawi, a two-year cluster randomized trial with the same assisted community-based HIVST approach but which utilized clinic referral cards to facilitate linkage estimated that 56.3% (524/930) of people diagnosed as HIV-positive after HIVST were linked to care (10). In a cohort study in Kenya, linkage to care was similar, with 50% (2/4) of women in antenatal care or postpartum care reporting that their male partner linked to care within three months of self-testing and being diagnosed HIV-positive (18).

Among key populations, linkage to care was approximately 20–100%. Two observational studies in the United States reported 100% linkage to care among four men who have sex with men, two in each study, who self-tested and were confirmed HIV-positive (59,61). In Hong Kong SAR, 20% (2/10) of men who have sex with men who were uncertain of their self-test result, or had a reactive self-test result, linked to further HIV testing or sought medical advice, while the majority (8/10) used another HIVST kit to retest (62). In Viet Nam, a pilot study of community-based HIVST among men who have sex with men, transgender people and people who inject drugs reported that 100% of those self-testing for HIV received confirmatory testing and were linked to care (55). In Kenya, 88% (23/26) of female sex workers reported that their male partners or clients who self-tested HIV-positive enrolled in HIV care (18). In Zimbabwe, 99% (97/98) of female sex workers with a reactive self-test result who were confirmed HIV-positive linked to care (56).

Although these findings are based on a small number of cases, they suggest that, as with all HTS approaches, linkage rates are often suboptimal unless an evidence-based linkage intervention is utilized. However, results also suggest that HIVST plus home-based assessment or treatment may be a particularly effective approach, as well as other assisted community-based strategies, couples and partners HIVST, and facility-based HIVST approaches.

Implementing these linkage services requires trained personnel to provide follow-up services. Further evaluation of strategies and approaches that can facilitate linkage to care, as well as prevention, following HIVST are needed, particularly for key populations, who may be less likely to link to services, especially in settings with restrictive laws and policies.

(See section 2.3.1 on planning for service delivery for further information on linkage to care.)

Key populations

Acceptability and willingness to use HIVST is generally high among key populations (4,18,56,112,131,132) despite some reported concerns about the potential lack of support, possible social harm, the level of accuracy of test results, and the related costs which could hinder access (4,72,80). Overall, the benefits of HIVST most commonly cited by key populations were convenience and privacy, followed by HIVST being an easy and painless testing option (4,80) as well as not needing a facility visit (112).

Some studies among men who have sex with men and female sex workers reported preferences for oral fluid-based RDTs for HIVST because this is a painless testing option. In contrast, other users said they preferred fingerstick/whole blood-based RDTs for HIVST as they considered them to be more accurate (11,55). For example, a study in Viet Nam reported that people who inject drugs preferred HIVST using a fingerstick/whole blood-based RDT, whereas more men who have sex with men and female sex workers said they favoured oral fluid-based RDTs (55).

Although there are many possible ways and channels for distributing HIVST, some studies suggest that men who have sex with men, transgender women and female sex workers prefer HIVST to be available over-the-counter at pharmacies and other locations or through the Internet (4). Several studies in Australia, China, Brazil, Peru, the United Kingdom and the United States reported that a high proportion of men who have sex with men and transgender women were comfortable accessing HIVST through the Internet and online gay dating sites, which mail a kit to their home or a location of their choice (4,9,57,66,100,154). Other options, such as vending machines and distribution at events, have also been shown to be acceptable among men who have sex with men and transgender people (98,99,101). In Kenya, female sex workers reported preferences for accessing HIVST kits at pharmacies or private clinics rather than obtaining them at public sector clinics (112), as well as a willingness to distribute HIVST to their peers, social network, primary partners and clients (18).

General populations

In the studies considered, HIVST was viewed to be highly acceptable by the general adult population, where there was strong support to promote and use HIVST.

In many African countries, interest in HIVST was consistently high. In Zimbabwe, a cross-sectional study among 289 adults found 80% would self-test, and nearly 90% would self-test if the cost was low (124). Likewise, in Zambia, 76% (1216/1600) said if HIVST were available they would definitely use it (122). In Kenya, a synthesis of several studies reported that privacy and personal empowerment were key motivations to self-test for HIV and that users would like to access HIVST kits at clinics or pharmacies (111). In-depth qualitative interviews in South Africa also indicated high acceptability of HIVST, as it was perceived to be a way to overcome barriers to existing HTS, such as lack of trust in health workers and the health system (117).

In Europe and the region of the Americas, similar levels of high acceptability were identified among general populations (76,150,153,154). In the United States, in 2006, prior to the availability of HIVST, 56.2% (95% CI: 54.7, 57.7) of adults who participated in a telephone survey indicated HIVST as acceptable (76), and two thirds of at-risk populations expressed a desire to use HIVST (76). In Spain, 80% (2699/3373) of individuals attending a street-based HTS site were in favour of making HIVST available (153). Additionally, 84% (174/207) of those accessing unassisted HIVST using fingerstick/whole blood-based RDTs in Spain said they were motivated to self-test for HIV in the future (150). In the United Kingdom, among 555 users who self-tested using a fingerstick/whole blood-based RDT, 98% said they would use it again and reported it was easy to perform (154).

Potential barriers to HIVST among the general population included concerns about the risk of violence, self-harm or suicide; misuse of HIVST kits; accuracy; lack of support; and potentially high costs. For example, in Kenya, 61% (n=1133) of people said HIVST kits might be misused, and there were also concerns about suicide or non-disclosure of self-test results (110). In Zambia, similar concerns were reported, albeit on a smaller scale: although 35% (n=1617) of respondents said they had concerns about HIVST, 98% of them said these concerns were minor and could be addressed (122). Concerns about the accuracy of HIVST were generally in relation to how to perform the test correctly and whether fingerstick/whole blood-based RDTs or oral fluid-based RDTs were more accurate (122). In one qualitative study in South Africa, a proportion of male respondents indicated they would prefer HIVST using fingerstick/whole blood-based RDTs because they perceived them to be more accurate (117).

Although concerns were raised about social harm in several of the studies, it is important to note that nearly all of those studies were conducted among individuals who had never self-tested for HIV. In contrast, in the studies in which HIVST was implemented, reports of social harm were rare (see section 2.2.1). Furthermore, in the studies considered, despite the various concerns raised by users, there was a consistently high level of interest and desire for HIVST to be readily available.

Couples and partners

In the studies considered, couples and partners from the general population and key population groups in Kenya (14,18,40,105), Malawi (20,51) and the United States (45,58,92,94,95,102,162) reported a high level of acceptability and interest in using HIVST.

Men who have sex with men in the United States, particularly those with casual partners and who do not use condoms, reported that they were interested in using HIVST as a form of harm reduction by screening potential sex partners (45,92,95,102,162). In two related studies, high-risk men who have sex with men suggested HIVST could improve honesty in disclosure of HIV status (102) and reported that, after their partner had a reactive self-test result, they provided emotional support and linked their partner to care (92). Service providers also report that couples-based HIVST could be a health promotion approach for men who have sex with men (102,162).

Among heterosexual cohabiting couples, HIVST may be a preferred approach to both determine one's HIV status and strengthen a relationship. In Malawi, HIVST was seen as an enabling, innovative way to test with a partner, strengthen a relationship bond and address concerns about suspected infidelity (20,51). Among women, in Malawi, the motivation to self-test as a couple was reportedly driven by long-term goals of health and “togetherness”, while men reported that they needed persuasion to self-test (although this was perceived as beneficial) and viewed HIVST as more flexible and less intimidating than HTS at a facility (20). HIVST also provided a way for couples to disclose a previously concealed HIV-positive status, alleviating internal conflict (20). Studies in Malawi and Kenya also demonstrated that distribution of an HIVST kit to a partner was perceived to be both safe and acceptable (19,105).

In Kenya, HIVST was highly acceptable among serodiscordant couples who were aware of their partner's status and were using PrEP (40). According to a cross-sectional study among 120 couples, 92.5% of couples found HIVST easy to use and nearly 40% self-tested with their partner (40).

As with all HTS, it is important to consider the potential risks, as well as benefits, to couples and partners. While studies to date, and the evidence summarized above, indicate that there are many benefits and that the risk of harm is minimal, it remains true that coercion or IPV are possible, depending on the context, setting and relationship dynamics. Coping with serodiscordant results can be challenging, whichever way HIV testing is performed or delivered. Some partners who self-test together and have non-reactive test results may forgo condoms, which could lead to the acquisition of HIV or other STIs if they have additional sexual partners. However, other couples who self-test together and have a reactive or serodiscordant result could be more likely to use condoms and support each other in linking to further HIV testing and HIV prevention and treatment services – bringing multiple benefits, including the prevention of future HIV transmission. Like other forms of couples and partner HTS, HIVST may not be appropriate for individuals or couples who report IPV in their current relationship (49). Thus, clear messages and information need to be delivered to couples and partners to mitigate potential risks and maximize benefits. (See section 2.2.1 above and Chapter 4 for additional information, as well as section 5.4 in Chapter 5 of the Consolidated guidelines on HIV testing services.)

Young people (15–24 years of age)

Although the number of existing studies that look at young people's (15–24 years of age) willingness to use HIVST kits are few, they report considerable interest in accessing HIVST in Canada (68), France (143), South Africa (116) and the United States (12,80,103,104).

In Canada, a survey of university students reported that 81% thought unassisted HIVST using an oral fluid-based RDT was acceptable, despite some concerns about accuracy and linkage to care (68). Similar findings were identified among South African university students (116), who also suggested HIVST was empowering, could “normalize” HIV, and that subsidized or free test kits may help people access HIVST.

In the United States, three studies reported high willingness to use HIVST kits among high-risk young people (12,103,104). One of these studies suggested the key motivations to self-test for HIV were ease of access, no need for a clinic visit, quick results, and the fact that HIVST kits could be used in non-monogamous relationships (104). However, the study noted that young people lacked information about the window period (104). Another study among young African Americans also indicated HIVST as preferable to facility-based HTS because it provides privacy and increases convenience, whilst reducing stigma and normalizing HTS (12,103). However, young people expressed concerns about accessing confirmatory testing, coping with a reactive self-test result, and whether people with a low socio-economic status would be able to understand the instruction materials provided (12,103). Similar preferences were reported among young people in France, especially if HIVST were available free of charge or came with the option of having assistance (143).

Health workers and other key stakeholders

HIVST is highly acceptable to health workers in high HIV prevalence settings for testing themselves, particularly those who have never tested before. The rate of informal HIVST among health workers is high in Ethiopia, Kenya, Malawi, Mozambique and Zimbabwe (7), where health workers view HIVST as a way to reduce stigma and discrimination around HIV testing and as a way for their family members to test (7). In Ethiopia, 70% (n=307) of health workers reported they had self-tested for HIV informally (22). Of those who self-tested, 82% said they did so for greater confidentiality and 14% said they self-tested because they lacked the time to receive standard HTS (22). In Kenya, a feasibility study reported that, following distribution of HIVST kits and provision of an information session, 89% (680/765) of survey respondents said they would recommend HIVST to fellow health workers (14).

Health workers and key stakeholders who provide or support the delivery of HTS report HIVST to be highly acceptable when implemented appropriately, using accurate and easy-to-use RDTs, and when there is linkage to care (69,157,158). In two studies in Canada, key stakeholders and health workers said HIVST should be made available nationally (67,69). One of the studies in Canada showed that one third of providers wanted further education on HIVST and preferred a more community-based approach for distributing HIVST kits (67). In Zimbabwe, focus groups revealed that, while health workers perceived HIVST to be a way to reach people with undiagnosed HIV, particularly men and those living in areas where there is limited access to HTS, they had concerns about how HIVST could impact their jobs and the jobs of other health workers (125).

(See Annex 17 for the full GRADE systematic review, including values and preferences in relation to whether HIVST should be offered as an additional HTS approach.)

Performance of HIV RDTs for self-testing

An HIV RDT used and interpreted by a self-tester performs as well as an HIV RDT used and interpreted by a trained health worker. Across 16 studies using directly assisted or unassisted HIVST, and one study reporting on both, kappa agreement between self-testers and trained health workers was almost perfect (see Fig. 2.4). No differences in concordance between self-testers and trained health workers were identified by type of approach, observation, type of specimen or HIV positivity (directly assisted: 0.98, 95% CI: 0.96–0.99 vs. unassisted: 0.98, 95% CI: 0.96–0.99; I² 33%, 95% CI: 19.5–98.1).

Fig. 2.4

Concordance of test results performed by self-tester compared to trained health worker (measured by Cohen's kappa) (n=16). Source: Figueroa et al, 2016 (181).

HIV RDTs used by self-testers can achieve high sensitivity and specificity. While there was a wide range in reported sensitivity, only two of 20 studies reported sensitivity of less than 80%; one of these studies did not provide sufficient information on how to interpret faint positive lines (175), and the other study suggested lengthy instructions were a barrier among participants in the rural arm, where literacy levels were low (52). Excluding these two studies with sensitivity less than 80%, sensitivity and specificity estimates were higher for fingerstick/whole blood-based RDTs (84,142,169,177) compared to oral fluid-based RDTs (10,17,42,108,131,167,168,170,171,173–175) (sensitivity 96.2–100% vs. 80–100%; specificity 99.5–100% vs. 95.1–100%). However, overall, there was no difference in sensitivity or specificity between approaches that offered direct assistance compared to those that did not (181) (see Table 2.2).

Table 2.2

Sensitivity and specificity of RDTs used for self-testing by type of approach (n=16).

Although sensitivity and specificity were typically high, common errors in performing the self-test or misinterpreting the self-test results were identified, such as errors in collecting the specimen (finger-prick or oral-swab) (42,131,168,178) and incorrect use or spilling of the buffer (42,168,170,171,174). Some of these errors resulted in invalid self-test results, such as errors in collecting the specimen, and others resulted in reduced test sensitivity. Studies using fingerstick/whole blood-based RDTs (84,85,150,169,178,180) reported a greater proportion of invalid results compared to studies using oral fluid-based RDTs (0.4–9.5% vs. 0.2–4.5%) (17,42,84,85,131,167,168,170,171,173,174,179,180).

In Zimbabwe, a study among a rural population with low literacy levels, which reported poor test sensitivity (66.7% 95% CI 9.4, 99.2), suggested user errors performing the self-test and interpreting the results were likely due to difficulties in reading and comprehending the instructions for use (168). Also, several studies that included people with known HIV-positive status reported that these users may have been more likely to make errors in performing a self-test or interpreting the self-test result (171,177).

HIV RDTs for self-testing among individuals on antiretroviral (ARV) drugs

Three of the 25 studies included in the review reported on whether or not participants were taking ARV drugs for treatment or prevention, including studies among participants with known HIV-positive status. These studies used an oral fluid-based RDT; in two of them, participants taking ARV drugs had a non-reactive self-test result and were later confirmed HIV-positive using a validated testing algorithm (10,168), whilst in the third study, both the HIVST and confirmatory test results were non-reactive because the reference standard was the same oral fluid-based RDT (179). While there is limited evidence on the impact of ARV drugs on the performance of HIV RDTs, several studies suggest they may be more likely to cause false non-reactive test results when using serology (antibody) tests because HIV antibodies are suppressed and remain undetected (2,182). This may be a risk for all serology tests, particularly oral fluid-based RDTs. (See Chapter 7 of the Consolidated guidelines on HIV testing services.) Therefore, it cannot be ruled out that studies reporting low sensitivity may have included populations taking ARV drugs.

Although HIVST can be highly accurate – as high as 100% sensitivity and 100% specificity, particularly with validated tests and instructions for use (181) – as with all HIV testing, a single reactive HIV RDT result is not sufficient to provide an HIV-positive diagnosis. All reactive self-test results must be confirmed by a trained tester starting from the beginning of a validated national testing algorithm (2). Currently, it is not recommended for individuals to attempt their own confirmatory testing through HIVST.

(For more detailed information on this systematic review on the performance of HIV RDTs for self-testing, please refer to Annex 19.)

Directly assisted and unassisted HIVST approaches

Depending on the population, interventions may vary in terms of the intensity and technologies used to support self-testers. The support that is offered can be either directly assisted or completely unassisted (see Table 2.3). For example, populations with disabilities, low literacy levels, and some rural communities may require direct assistance in the form of in-person demonstrations and explanations before, during and after self-testing. However, no matter the type of approach, users with access to the Internet and social media, as well as frequent and repeat testers, may be able to conduct the test using the instructions alone or accessing further assistance through telephone hotlines, text message services, videos and other support tools provided with the test kit.

Table 2.3

Support tools for directly assisted and unassisted HIVST approaches.

Linkage to further HTS and HIV prevention, treatment and care

In addition to identifying the best approach for delivering HIVST, programmes need to consider how to facilitate linkage to prevention, treatment and care following HIVST. Further research, monitoring and evaluation of strategies that facilitate linkage to care following HIVST are needed. Box 2.5 summarizes the approaches to linkage currently being implemented that can be considered.

Monitoring and reporting systems

Monitoring and reporting systems are critical for all approaches to HTS, including HIVST. Because of the discreet, private nature of HIVST, there may be particular challenges with collecting information on how effective an HIVST programme is, or monitoring the experiences of users and tracking possible social harm. Although the instances of harm reported to date have been few, it is essential for programmes to utilize or adapt existing systems to monitor and report on social harm or other adverse events, as well as corrective action and follow-up, in order to address harm if and when it occurs.

In order to track programmatic outcomes, monitoring and evaluation of HTS indicators may need to be adapted to include HIVST. Many HTS programmes are starting to use a combination of the tools described in Box 2.6 to improve monitoring of HIVST.

Further reading

• Technical specifications for WHO prequalification - Human immunodeficiency virus (HIV) rapid diagnostic tests for professional and/or self-testing. Geneva: World Health Organization; 2016. (http://www.who.int/diagnostics_laboratory/guidance/en/).
• Operational research note on HIV self-testing. Geneva: Global Fund; 2016. (http://www.theglobalfund.org/documents/core/infonotes/Core_OpResearchImplementationHIVSelfTesting_BriefingNote_en/).
• Consolidated guidelines on HIV testing services. Geneva: World Health Organization; 2015. (http://www.who.int/hiv/pub/guidelines/hiv-testing-services/en/).
• March 2014 supplement to the consolidated HIV guidelines on the use of antiretroviral therapy – a public health approach. Geneva: World Health Organization; 2014. (http://www.who.int/hiv/pub/guidelines/arv2013/arvs2013upplement_march2014/en/).
• A short technical update on HIV self-testing. Geneva: Joint United Nations Programme on HIV/AIDS; 2014. (http://www.unaids.org/sites/default/files/media_asset/JC2603_self-testing_en_0.pdf).
• WHO/UNITAID HIV self-testing technology landscape – 2nd edition. Geneva: UNITAID; 2016. (http://unitaid.org/images/marketdynamics/publications/UNITAID_HIV_rapid_diagnostic_tests_for_self-testing.pdf).