Costing analysis of an SMS-based intervention to promote HIV self-testing amongst truckers and sex workers in Kenya

Objective: HIV testing rates in many sub-Saharan African countries have remained suboptimal, and there is an urgent need to explore strategic yet cost-effective approaches to increase the uptake of HIV testing, especially among high-risk populations. Methods: A costing analysis was conducted for a randomized controlled trial (RCT) with male truckers and female sex workers (FSWs) registered in the electronic health record system (EHRS) of the North Star Alliance, which offers healthcare services at major transit hubs in Southern and East Africa. The RCT selected a sample of truckers and FSWs who were irregular HIV testers, according to the EHRS, and evaluated the effect of SMSs promoting the availability of HIV self-testing (HIVST) kits in Kenyan clinics (intervention program) versus a general SMS reminding clients to test for HIV (enhanced and standard program) on HIV testing rates. In this paper, we calculated costs from a provider perspective using a mixed-methods approach to identify, measure, and value the resources utilized within the intervention and standard programs. The results of the analysis reflect the cost per client tested. Results: The cost of offering HIVST was calculated to be double that of routine facility-based testing (USD 10.13 versus USD 5.01 per client tested), primarily due to the high price of the self-test kit. In the two study arms that only offered provider-administered HIV testing in the clinic, only 1% of truckers and 6% of FSWs tested during the study period, while in the intervention arm, which also offered HST, approximately 4% of truckers and 11% of FSWs tested. These lower than expected outcomes resulted in relatively high cost per client estimates for all three study arms. Within the intervention arm, 65% of truckers and 72% of FSWs who tested chose the HIVST option. However, within the intervention arm, the cost per additional client tested was lower for FSWs than for truckers, at USD 0.15 per additional client tested versus USD 0.58 per additional client tested, driven primarily by the higher response rates. Conclusion: Whilst the availability of HIVST increased HIV testing among both truckers and FSWs, the cost of providing HIVST is higher than that of a routine health facility-based test, driven primarily by the price of the HIV self-test kit. Future research needs to identify strategies which increase demand for HIVST, and determine whether these strategies and the subsequent increased demand for HIVST are cost-effective in relation to the conventional facility based testing currently available.