In 2014, United Nations Programme on HIV/AIDS (UNAIDS) developed targets aimed at global HIV epidemic control: 90% of all people living with HIV (PLHIV) to know their status, 90% of diagnosed PLHIV to receive antiretroviral therapy (ART), and 90% of PLHIV on ART to achieve viral suppression by the year 2020 (UNAIDS, 2014a). According to UNAIDS 2014 Gap report, of the 35 million people estimated to be HIV infected by 2013 worldwide, 19 million (54.3%) were not aware of their HIV positive status (UNAIDS, 2014b), with a similar proportion (53.1%) documented in Kenya in 2012 (NASCOP, 2014). The UNAIDS agenda was driven by an understanding that lack of awareness of HIV positive status was the single biggest barrier to ART programming (Cohen et al., 2011; NASCOP, 2012; NASCOP, 2014; UNAIDS, 2014b). In the recent national survey, Kenya posted an achievement of 80% in the 1^st 90 of UNAIDS targets (NASCOP KENPHIA 2018 Report, 2020).

At the time, a growing body of research had indicated sub-optimal linkage of HIV infected people to ART services (Fox & Rosen, 2010; Genberg et al., 2015; Gerdts et al., 2014; Hatcher et al., 2012; Kranzer et al., 2010; Ng’ang’a et al., 2014; Rosen & Fox, 2011). In the intervening years, ‘test and treat’ has become an axiom for national HIV/AIDS programmes but it is worth remembering that concerted national testing campaigns are still a key foundation for the efficacy of those programmes. Kenya’s national HIV testing services (HTS) rapid results initiative (RRI) campaign in 2013 is a case in point, as we describe in this article which is based on a thorough analysis of the campaign data.

A total of 1,463,378 individuals received HTS during the national HTS campaign implemented from July to August 2013 (Table 1). This resulted to 114% achievement of the campaign HTS target for this HTS campaign as shown in Table 1. The geographical distribution of those tested was highest in Rift Valley (25.5%) and Nyanza (20.3%) regions, while the lowest was recorded in Central (7.1%) and North Eastern (5.7%) regions (Table1). The majority of those who tested positive were from Nyanza (43.3%) and Rift Valley (18.5%). HIV prevalence was highest in Nyanza region at 5.6%, followed by Central region at 2.5% and Nairobi at 2.4%. The Western and North Eastern region had the lowest prevalence at 0.9% and 0.3% respectively. Of those tested, 220,902 (15.1%) were children aged less than 15 years and 55,088 (7.2%) were individuals tested together as couples. Eight percent of tested individuals (116,126) were classified as key populations according to Kenya MOH guidelines (i.e. particularly at higher risk of acquiring or transmitting HIV): sex workers, men who have sex with men, people who inject drugs, truckers, fisher folk and prisoners (Table 2).

This national HTS campaign identified 37,630 (2.6%) undiagnosed PLHIV and most of these individuals (25,994; 69%) were diagnosed at health facility settings (Table 2). This proportion of PLHIV identified at health facility settings was significantly higher than 11,994 (31%) identified in the community-based settings (p<0.0001). Of the 853,922 persons tested at health facility settings, 25,994 (3.0%) were diagnosed with HIV, while 11,636 (1.9%) of 608,456 reached in community-based settings had HIV infection (Table 2). Therefore, those reached with HTS at health facility settings were more likely to test positive for HIV infection compared to those diagnosed in the community-based settings; OR=1.6 (95% CI: 1.58–1.65), p<0.0001. Overall, women had a higher HIV prevalence of 2.9% compared to men whose HIV burden was estimated at 2.2%. Among children (<15 years of age) who received HTS, 3,244 (1.5%) tested HIV positive. Of all adults who tested as couples, 1,763 (3.2%) spouses were identified as being HIV sero-discordant and the highest proportion of HIV discordant couples was identified at health facility settings (4.7%) compared to 2.1% who were diagnosed at community-based settings (Table 2). Key populations had the highest HIV prevalence in both settings (4.8%).

Over the campaign period, 29,851 (79.3%) of individuals who tested positive for HIV were reported to have enrolled into HIV C&T services (Table 2) within three months of HIV diagnosis. Of 11,636 PLHIV diagnosed at the community-based settings 9,908 (85%) were enrolled into HIV C&T services in this period compared to 19,943 (77%) of PLHIV that were identified at health facility settings. Therefore, PLHIV diagnosed at community-based settings had increased odds of enrolment into HIV C&T compared to those diagnosed at the health facility settings; OR=1.7 (95% CI: 1.64–1.84), p<0.0001. Women 19,314 (81%) were generally more likely to enroll into HIV C&T compared to men 10,537 (77%); OR=1.2 (95% CI: 1.18–1.31, p<0.0001). Adults (≥15 years old) were 2-fold as likely as children (<15 years old) to enroll into HIV C&T services; OR=1.9 (95% CI: 1.71–2.01), p<0.0001. A higher percentage of enrolment to HIV C&T was recorded among HIV positive adults (84.8%) and children (90.9%) identified in community-based settings compared to those diagnosed at health facility settings (78.2%) and (62.8%), respectively (Table 2). Similarly, higher proportions of enrolment to HIV C&T were observed among HIV positive men (84.4%) and women (85.6%) identified in community-based settings compared to those diagnosed at health facility settings (78.2%) and (62.8%), respectively. However, unlike among PLHIV diagnosed at community-based settings where female sex was not associated with enrolment to HIV C&T [OR=1.1 (95% CI: 0.99–1.22), p=0.077], female PLHIV identified at health facility settings were more likely to enroll into HIV C&T services compared to their male counterparts; OR=1.3 (95% CI: 1.22–1.38), p<0.0001.

The campaign data revealed a differential association between age and the setting of diagnosis. Among PLHIV diagnosed at health facility settings, adults (≥15 years old) had 2-fold increased odds of enrollment into HIV C&T compared to children (<15 years old); OR=2.1 (95% CI: 1.95–2.32), p<0.0001. On the contrary, children (<15 years old) were more likely to enroll into HIV C&T compared to adults (≥15 years old) among PLHIV diagnosed at community-based settings; OR=1.8 (95% CI: 1.38–2.35), p<0.0001.

This was a very successful campaign. All regional targets were achieved, with Rift Valley’s target being significantly surpassed 154% (p<0.0002). The number of previously undiagnosed PLHIV identified by this campaign was equivalent to 5.9% of the undiagnosed population as estimated by the 2012 population-based survey (NASCOP, 2014). The 3,750 previously undiagnosed HIV positive children under the age of 15 was equivalent to 8% of the total undiagnosed population of children in Kenya as estimated by the 2012 nation HIV survey (NASCOP, 2014). For Kenya, this was particularly a major success because it was well known the ART programme had limited coverage of children (UNAIDS, 2014a, NASCOP 2014).

There was a significant geographical variation both in HIV prevalence and yield of identified PLHIV. Three regions (Nyanza, Rift Valley and Nairobi) contributing approximately 75% of all PLHIV identified during the campaign. Nyanza region alone accounted for over 40% of all identified PLHIV. These finding were similar to those from the KAIS 2012 survey, which documented that the majority of undiagnosed PLHIV were in Nyanza province (35%), followed by Rift Valley (18%) and then Nairobi (11%) (NASCOP, 2014). Furthermore, the gap in knowledge of status was reduced by a big range; notably, HIV prevalence especially those above the age of 15 was 2.8% which was half the national HIV prevalence estimate in 2012 (NASCOP, 2014).

There could be many reasons why variations on yield and linkage were observed between the two HTS settings (Health facilities and Community settings) but these campaign results provided indicative scope for campaign planning, refining targets geographically and across different subpopulations in order to reach as many as undiagnosed individuals as possible, while providing useful information for refining estimates of HIV prevalence as subpopulation level. It therefore seems that testing during campaigns should be much more targeted and further explorative analysis should be conducted specially to assess the yield in HIV positivity in different testing sites, population subgroups and even within different departments of the health facilities. This will allow for better targeting in order to achieve a higher yield in identifying undiagnosed PLHIV. Likewise, the results regarding ART enrollment rates following HTS in different settings and amongst different sub-populations, provide information for refining concerted HTS campaigns. The campaign data showed that almost 80% of those who tested positive for HIV were subsequently enrolled in C&T services. However, individuals reached at community settings having better enrolment rates compared to those reached at health facility settings. Furthermore, the data showed that age and sex were patient-related factors significantly (p<0.0001) associated with enrollment for HIV treatment, suggesting that national RRIs provide a matrix for addressing age/sex-related unmet needs for HTS and enrolment for HIV C&T in HIV programing.

Kenya’s 2013 RRI campaign is an example of the utility of these interventions to boost the UNAIDS agenda:

The RRI provided information in support of Kenya’s commitment to achieving UNAIDS targets. Notably, increasing population-level knowledge of HIV status toward achieving the 1^st 90 of the UNIAD targets and highlighting the need for optimizing enrollment of newly identified PLHIV into HIV C&T to facilitate achievement of the 2^nd 90 in the UNAIDS target. These created a foundation for the test and treat policy that was implemented in Kenya in 2016. Indeed, the recent 2018 national survey documented a marked reduction in proportion of PLHIV who did not know they were HIV infected from 53.1% in 2012 to 20.5% in 2018.