Predicting Lost to Follow-Up Status Using an Adolescent HIV Psychosocial Attrition Risk Assessment Tool: Results From a Mixed Methods Prospective Cohort Study in Uganda

Abstract

Background: Low retention in care for adolescents living with HIV (ALHIV) has been a key driver of suboptimal viral load suppression rates in Uganda. The objective of this study was to develop a psychosocial risk assessment tool and evaluate its ability to predict the risk of attrition of ALHIV between the ages 15 and 19 years. Setting: The study was conducted in 20 facilities in Central and Western Uganda from August 2021 through July 2022. Methods: A mixed methods prospective cohort study was conducted in two phases. In the first phase, the Adolescent Psychosocial Attrition Risk Assessment tool was developed and revised using feedback from focus group discussions and interviews. In the second phase, the ability of the Adolescent Psychosocial Attrition Risk Assessment tool to predict attrition among ALHIV was evaluated using diagnostic accuracy tests. Results: A total of 597 adolescents between the ages 15 and 19 years were enrolled, of which 6% were lost to follow-up at the end of the study period. A 20-question tool was developed, with 12 questions being responded to affirmatively by >50% of all participants. Using a cut-off score of 6 or more affirmative answers translated to an area under the curve of 0.58 (95% CI: 0.49 to 0.66), sensitivity of 55% (95% CI: 36% to 72%), and specificity of 61% (95% CI: 56% to 65%). Conclusion: Although the Adolescent Psychosocial Attrition Risk Assessment tool was not effective at predicting lost to follow-up status among ALHIV, the tool was useful for identifying psychosocial issues experienced by ALHIV and may be appropriate to administer during routine care visits to guide action.