Reliability and validity of physical and mental health summary scores from the medical outcomes study HIV Health Survey

OBJECTIVES. Health-related quality of life measures are used to evaluate patient outcomes in clinical trials of new treatments for human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS). Summary index scores, rather than profiles of scale scores, would simplify data analysis and interpretation of findings from clinical trials and comparison across studies. METHODS. Baseline MOS HIV Health Survey scores from two clinical trials of new antiretroviral medications in HIV/AIDS patients (total n = 2253) and an observational study (n = 162) were used to develop physical health summary (PHS) and mental health summary (MHS) scores. Exploratory and confirmatory factor analysis were used to identify the factor structure of the summary scores based on MOS HIV Health Survey scales. Physical health summary and MHS scores were derived and the factor structure proved invariant across the two groups. RESULTS. Reliability of the PHS score was 0.90 to 0.92 and MHS score was 0.91 to 0.94. Mean PHS and MHS scores differed in patient groups defined by HIV disease stage, HIV disease severity, Karnofsky performance status scores, and global ratings of health status. Mean PHS and MHS scores in patient reporting worsening health status were significantly lower than scores of patients reporting stable or improving health status. CONCLUSIONS. The PHS and MHS were reproducible across different samples of HIV/AIDS patients and are reliable and valid measures for demonstrating treatment impact on patient functioning and well-being.