Performance of an HIV risk screening tool to identify people living with HIV aged 15 years and above in primary care facilities in Uganda: a secondary program data analysis

BACKGROUND: Uganda is nearing its 1^st 90 target. 89% of people living with HIV (PLHIV) are aware of their status, but approximately 215,000 PLHIV are not yet on treatment. As Uganda and other countries scale treatment coverage, it is increasingly difficult and resource-intensive to identify remaining PLHIV, particularly as resources for HIV testing services (HTS) decline. To reduce testing volumes while increasing positivity, many countries are implementing risk-based screening tools, but there is very limited evidence on their impact. Uganda is screening for HTS eligibility among adults in outpatient departments (OPD) of public facilities. This abstract describes secondary analysis of programmatic data on the impact of risk-based screening tools in OPD settings in Uganda.
METHODS: We conducted a retrospective secondary data analysis of routinely collected program data from October to November 2019 in 24 facilities implementing HIV risk screening. Participants were clients over the age of 15 in OPD who were screened and then tested for HIV, regardless of eligibility. De-identified data was analyzed to calculate HIV positivity rates with and without screening, and sensitivity and specificity of the tool.
RESULTS: Of 19,704 patients screened, 12,971 (66%) were female and the median age was 27 (IQR: 21-35). Overall, 732 (yield 3.71% (95% CI: 3.06-4.50)) patients were positive. Based on risk screening, 14,879 (76%) patients were eligible for testing and 664 (yield 4.5% (95% CI: 4.1%-4.8%)) of these patients were positive. Overall sensitivity of the screening tool was 90.7% (95% CI: 88.4%, 92.7%) and specificity was 25.1% (95% CI: 24.5%-25.7%). With screening, the number needed to test (NTT) to identify one PLHIV fell from 32 to 22.
CONCLUSIONS: Implemented in OPD, the screening tool in Uganda reduced testing volumes 24.5%, but did not result in a statistically significant increase in yield. The tool identified 9.3% of PLHIV as ineligible for testing; these PLHIV would have been screened out and not offered HTS. Given the increasing difficulty and costs of identifying remaining PLHIV, it is critical that ministries of health carefully consider potential risks and tradeoffs in implementing risk-based screening tools which may miss PLHIV at facilities.