Côte d’Ivoire has a population of 22.6 million and Abidjan, the economic capital of Côte d’Ivoire, has a population of 4.7 million. Adult HIV prevalence in the country is estimated at 2.8%, the highest among countries in the West Africa region [1]. UNAIDS estimates suggest that 500,000 adults and children are living with HIV in Côte d’Ivoire, of which 225,839 are receiving ART [1]. Since 2004, Côte d’Ivoire offers its population free healthcare for the management of HIV infection including HIV testing, consultations, ARV treatment and laboratory tests. A recent population-based HIV impact assessment (PHIA) survey in Côte d’Ivoire suggests that 37% of adults (15–64) are aware of their positive HIV status, highlighting substantial gaps towards achieving HIV epidemic control [2]. Given resource constraints, targeted HIV case identification strategies are needed to reach the remaining individuals with undiagnosed HIV infection.

This study evaluates the outcomes of the Family Approach to Counseling and Testing (FACT) program in Côte d’Ivoire. The FACT program, which began in 2015, provides facility-based HIV testing to the main partner, children and other household family member of HIV-positive index cases. Family-centered HIV care models have been implemented and evaluated, mainly in the context of treatment for children living with HIV [3]. Also, family-centered prevention models have been evaluated and tend to focus on assessing outcomes for mothers and children in family planning clinics and maternal health clinics [4–6].

The WHO’s guidelines on HIV partner services, also known as index contact testing or partner notification, were issued in 2016 [7], subsequent to the start of the FACT program. HIV partner services consist in offering voluntary HIV testing services to sexual and/or drug injecting partners of people with HIV. Several studies have demonstrated the effectiveness of HIV index contact testing for identifying new cases of HIV among adults, with positivity rates ranging from 35 to 62% [8–12]. Studies also show new case identification rates among children of HIV positive index cases ranging from 4 to 18% [13–16]. Index contact testing may be a relatively cost-effective approach to HIV case finding, particularly among those who would not otherwise access health care services. Economic studies of HIV index contact testing in resource constrained settings suggest that such interventions are cost-effective [17, 18]. To gauge the effectiveness of the FACT program, we examined the number of family members enumerated, tested, and linked to care and treatment initiation through the program across 36 sites, in Abidjan, between June 2015 and December 2016. We also evaluated the association between the characteristics of the index case and the family member testing positive.

This study demonstrates that the FACT program for HIV testing in Côte d’Ivoire is a feasible and effective approach to case finding, particularly among partners of HIV index cases.

Studies of partner testing reported HIV testing uptake rates ranging from 2 to 65% [11, 12, 21]. The FACT data show that 75% of sexual partners, 89% of children, and 78% of “others” came to the facility testing (Fig. 1). The higher uptake rate obtained through the FACT program may result from the family-focused approach, where trained social workers and HIV counselors assist a family in the management of HIV and offer benefits to the family as a whole.

Strikingly, women index cases were 47 times more likely (p < 0.0001) than men index cases to refer a positive partner, perhaps owing to the near universal coverage of Côte d’Ivoire’s Prevention of Mother to Child Transmission (PMTCT) program, which started in 2004, and where women are routinely screened for HIV in antenatal clinics. HIV positive women ought to be prioritized for partner enumeration interviews to increase HIV case identification among men, for whom the gap in knowledge of HIV positivity is highest at 76% [22].

A majority of index cases were men (n = 1326; 81%), maybe because, as heads of household, they felt more comfortable consenting to participate in the FACT program. A majority of these men (62%) did not refer a sexual partner, and among the 302 women index cases, 32% did not refer a sexual partner. Given the high rates of new diagnoses among partners, maximizing the number of index cases who refer one or more partners is key to increasing case identification. Further research may inform this this gender difference.

From 2015 to 2018, the average positivity rate of PEPFAR-supported HIV testing, among those less than 15 years age in Côte d’Ivoire, was 0.7% [2], by contrast, children of index cases tested through the FACT program had a 5% positivity rate. Median life expectancy for children with perinatally-acquired HIV is 2 years, without ART [23], and in Côte d’Ivoire, an estimated 38,000 children living with HIV are unaware of their status [1], thereby reinforcing the need to test the children of index cases and getting those HIV-positive children on life-saving treatment. In their 2019 brief, the WHO strengthened their recommendation for HIV partner services, with emphasis on the importance of extending HIV partner services to offer HIV testing for untested children of HIV positive clients [24].

From 2015 to 2018, PEPFAR-supported facility-based testing programs reported nearly 6 million HIV tests and with a 3% positivity rate [2]. The FACT program showed HIV positivity rates of 21% and 11%, among the partners and other family member of index cases. Relative to provider-initiated facility-based testing, the FACT approach requires additional resources to conduct the enumeration interview with the index case, and follow-up on the HIV testing status of their family members. While we did not collect any cost data, these additional resources are unlikely to outweigh the incremental HIV case finding benefit of the FACT program. And, economic evaluations of HIV index testing strategies in resource constrained setting indicates that these approaches are cost-effective [17, 18].

Among the “others” family members, a focus on testing the parents of index cases is warranted given their high positivity rate, 41% for mothers and 36% for fathers, particularly mothers of index cases under the age of 15 of whom 11 out of 12 tested positive. Focusing the FACT approach on immediate family members, rather than extended ones (e.g. aunts, uncles, and cousins) may free up resources to scale-up the approach.

A majority (76%) of index cases referred at least one child, however, most (57%) index cases over the age of 15 did not enumerate any sexual partners (Table 2). Given that the highest yield of testing was observed among partners of index cases, increasing the number of sexual partners named is key to increasing new HIV case identification. Improving the interview skills of counselors, offering support for HIV disclosure and offering masked couples testing and/or enabling a provider to contact the partner to refer them for testing may help reduce barriers faced by the index case in terms of naming and referring their sexual partners for HIV testing.

Subsequent to our study period, the FACT program expanded its services to offer several referral approaches to index contact testing including assisted approaches where the provider is engaged in referring the contacts for testing, and, an analysis of those program outcomes categorized by referral approach is available [25].

Our analysis has several limitations. First, data were abstracted from index patient charts and variables relating to family members were limited. More information about family members may increase our ability to establish association between characteristic of the family member and testing result. Also, characteristics of HIV counselors (e.g., gender, age, experience with HIV/AIDS) were not captured and may influence index patients’ decision to refer their family members. The amount of time between the enumeration interview and testing of the family member was not known. Such data would allow to improve follow-up and contact tracing. Third, marital status was missing in nearly 10% of all index cases overall, and was therefore excluded from the multivariable model. Lastly, 158 (9%) of index cases were removed due to illogical progression of events from referral to testing, to ART initiation. The majority of these errors are likely unbiased data abstraction errors and therefore their removal is unlikely to alter the significance of our findings.

The FACT program is undoubtedly a valuable addition to HIV testing efforts in Côte d’Ivoire. Scale-up of HIV testing among the family contacts of index cases is warranted to maximize HIV case finding and curtail the epidemic.