OBJECTIVES To identify a parsimonious factor structure for the 43-item PROQOL-HIV questionnaire as an easy-to-use instrument for measuring health-related quality of life in clinical studies.

METHODS Using a representative sample of 3,022 HIV-infected patients from the Vespa2 French national survey, HRQL was summarized using the 8-dimension PROQOL-HIV questionnaire. Exploratory Factor Analysis (FA) was used to isolate 4 correlated factors. Dimensionality and internal consistency were checked using parallel analysis and Cronbach’s alpha. A confirmatory FA was applied on the international validation sample (N=791, 8 countries). Summated scale scores were compared to individual scores computed from FA and partial credit models.

RESULTS The four dimensions down to: physical health and symptoms (11 items, α=0.931), health concerns and mental distress (10 items, α=0.908), social and intimate relationships (7 items, α=0.898), and treatment impact (10 items, α=0.936). Four items exhibited PROMAX rotated loadings < 0.4, three of which were found to cross-load on two factors. This four-factor solution suggests acceptable fit to the international validation sample (RMSEA=0.082, 90% CI [0.079;0.084]; NNFI=0.880). Correlations between unweighted sum scale scores and factor or IRT scores were above 0.9 in most cases.

CONCLUSIONS PROQOL-HIV is the only specific scale taking into account patient’s experience with treatment. This simplified scoring version will allow researchers and clinicians to better monitor and to correlate HRQL improvements in patients undergoing different treatment strategies.