Application of random effects models and other methods to the analysis of multidimensional quality of life data in an AIDS clinical trial

BACKGROUND. Current analytic methods applied to multidimensional health-rel ated quality of life (HRQOL) data do not borrow strength across analyses an d do not produce summary estimates of effect. Objectives. TO compare a random effects modelling approach for the analysis of multidimensional HRQOL data to the following: (1) separate analyses for each dimension; (2) O'Brien's global test procedure; and (3) multivariate analysis of variance (MANOVA). RESEARCH DESIGN. Randomized clinical trial comparing 3 treatments (Trimetho prim-Sulfamethoxazole [TS], Dapsone-Trimethoprim [DT], and Clindamycin-Prim aquine [CP] for Pneumocystis carinii pneumonia [PCP]). SUBJECTS. Patients w ith PCP enrolled in AIDS clinical Trials Group Protocol 108. MEASURES. A 33-item battery assessing 7 dimensions of HRQOL: physical funct ioning, pain, energy, general health perceptions, disability, pulmonary sym ptoms, and constitutional symptoms. RESULTS. Analyses focused on changes in score from baseline to Day 7 (n = 1 45). Separate analyses for each dimension suggested a trend favoring CP ver sus TS, but using a Bonferroni correction no differences were statistically significant. O'Brien's global procedure for a test of no-treatment effect versus superiority of one treatment yielded P = 0.07. MANOVA did not reveal significant differences among treatment groups. A random effects model usi ng fixed treatment and dimension effects and separate random effects for ea ch person showed a significant overall treatment effect (P = 0.02); changes in scores for CP averaged 10 points greater than for TS. CONCLUSIONS. Random-effects models provide a flexible class of models for a nalyzing multidimensional quality of life data and estimating treatment eff ects because they borrow strength across dimensions.