Value in Health Volume 22, Supplement 3, Pages S638-S639
OBJECTIVES: Ceiling effects limit the discriminative ability of measures in detecting clinically relevant change. The HIV Treatment Satisfaction Questionnaire: change version (HIVTSQc) is a condition-specific measure designed to overcome ceiling effects found with the status version (HIVTSQs) of the questionnaire. Both HIVTSQs and HIVTSQc include the same 12 item stems (e.g. side-effects, demands). However, HIVTSQc response options measure relative change in satisfaction rather than absolute satisfaction. Here we compare the HIVTSQs and c for their ability to demonstrate change.
METHODS: LATTE-2 (NCT02120352), a Phase IIb, randomized, open-label trial, evaluated antiviral activity, tolerability, and safety of intramuscular injections of cabotegravir LA plus rilpivirine LA at 4 or 8-week intervals, relative to cabotegravir plus abacavir/lamivudine once-daily tablet, in HIV-1 infected patients new to antiretrovirals. HIVTSQc endpoint scores (32 weeks) were compared with endpoint-minus-baseline (week 16 of tablet induction period) scores from the HIVTSQs (HIVTSQsDiff).
RESULTS: Analyses included 250 patients, including 47 in the tablet control group and 199 in the injection groups (four missing). At baseline 203 patients were categorized ‘at ceiling’ (AC) and 47 categorized ‘not at ceiling’ (Not AC). A three-way (questionnaire, treatment, ceiling) analysis of variance revealed a significant questionnaire-by-treatment-type interaction effect (p<0.001). Differences in treatment satisfaction between the tablet (M=19.32, SD=12.89) and injection groups (M=27.80, SD=7.28) were found only when using the HIVTSQc (p<0.001). Furthermore, effect sizes were significantly stronger for the HIVTSQc (r=0.59-0.88) than for the HIVTSQsDiff (r=0.03-0.58) particularly for AC patients in the injection groups (HIVTSQc: AC r=0.88, HIVTSQsDiff: AC r=0.16, z=10.86, p<0.001).
CONCLUSIONS: The HIVTSQc demonstrated superior responsiveness to change compared to a standard difference score and was particularly effective in patients AC at baseline. The HIVTSQc, used in conjunction with the HIVTSQs, overcomes the problem of ceiling effects when a status measure is used alone and provides greater validity in determining benefits of new treatments for HIV.
