Abstract
:
Background The prevalence of HIV in Uganda is 7.3%, and yet nearly 40% of people living with HIV are unaware of their status. The current HIV testing policy which is strictly blood-based poses several challenges including: a need for high level laboratory skills, stringent waste disposal needs, and painful sample collection. It is envisaged that introduction of a rapid, painless HIV oral fluid test as a potential alternative is likely to increase the number of people testing. The aim of this study was to determine the diagnostic accuracy and acceptability of rapid HIV oral testing among adults attending Kisenyi Health Centre IV in Kampala. Methods and findings We conducted a cross-sectional study among 440 adults recruited consecutively at Kisenyi Health Centre IV from January to March 2016. The diagnostic accuracy of the HIV oral test was assessed by comparing to the national HIV serial testing algorithm. We also assessed for acceptability among patients and health care workers (HCWs) by triangulating responses from a structured questionnaire, three focus group discussions and seven key informant interviews. Acceptability was defined as willingness to take the test at the time of the study and intention for future use of the test if it was availed. The prevalence of HIV infection among study participants was 14.8%. The HIV oral fluid test was highly accurate with sensitivity of 100% (95% CI; 94.5-100.0), specificity of 100% (95% CI; 99.0-100.0), positive predictive value (PPV) of 100% (95% CI; 94.5-100.0) and negative predictive value (NPV) of 100% (95% CI; 99.0-100.0). Acceptability of HIV oral testing was also high at 87.0% (95% CI; 83.6-89.9). Participants preferred HIV oral testing because it was: pain free (91%, n = 399) and did not require blood draw (82%, n = 360). Conclusion The HIV oral fluid test has high diagnostic accuracy and acceptability. HIV oral testing is a suitable addition to the national HIV testing strategies with the potential of increasing access to HIV testing services in Uganda.