Seemingly Unrelated Regression analysis of the Cost and Health-Related Quality of Life Outcomes of the REVAMP randomised clinical trial

Abstract

Objective: To evaluate the nine-month cost and health-related quality of life outcomes of resistance versus viral load testing strategies to manage virological failure in low-middle income countries. Methods: We analysed secondary outcomes from the REVAMP clinical trial: a pragmatic, open label, parallel-arm randomised trial investigating resistance versus viral load testing for individuals failing first-line treatment in South Africa and Uganda. We collected resource data, valued according to local cost data and used the EQ-5D-3L to measure health-related quality of life at baseline and nine-months. We applied Seemingly Unrelated Regression Equations to account for the correlation between cost and health-related quality of life. We conducted intention to treat analyses with multiple imputation using chained equations for missing data and performed sensitivity analyses using complete cases. Results: For South Africa, resistance testing and opportunistic infections were associated with statistically significantly higher total costs and virological suppression was associated with lower total cost. Higher baseline utility, higher CD4 count and virological suppression were associated with better health-related quality of life. For Uganda, resistance testing and switching to second-line treatment were associated with higher total cost and higher CD4 was associated with lower total cost. Higher baseline utility, higher CD4 count and virological suppression were associated with better health-related quality of life. Sensitivity analyses of the complete-case analysis confirmed the overall results. Conclusion: Resistance testing showed no cost or health-related quality of life advantage in South Africa or Uganda over the nine-month REVAMP clinical trial