Comparison of the measurement properties of the AQoL and SF-6D in critical illness.

OBJECTIVE Multiattribute utility (MAU) instruments are short instruments measuring quality of life, health utility scores and treatment cost-effectiveness. Many studies have compared MAU instruments, but few have involved intensive care unit patients. Our aim was to compare the measurement properties of two MAUs, the assessment of quality of life (AQoL) and Medical Outcomes Study Short Form 6D (SF-6D), in a sample of patients with critical illness. METHODS Prospective observational study conducted in an 18-bed mixed tertiary Australian ICU. Eligibility criteria were: admitted to the ICU > 48 hours, aged > 18 years, and not imminently at risk of death. Participants completed the AQoL and SF-6D on admission to the ICU as a "then-test" of pre-ICU status, and 6 months after ICU discharge. We assessed the reliability, validity, sensitivity and responsiveness of the instruments. RESULTS Median age was 61 years (interquartile range [IQR], 49-73 years)], 60% were men, and the median Acute Physiology and Chronic Health Evaluation II score was 17 (IQR, 13-21). Cronbach's α was acceptable for the AQoL (α = 0.81) but not for the SF-6D (α = 0.65). The AQoL and SF-6D showed evidence of validity but, despite moderate agreement between their utilities, their scores were not interchangeable. This was likely due to the SF-6D's truncated scoring range. The AQoL was predictive of hospital readmission but the responsiveness and sensitivity of the instruments varied by clinical condition. CONCLUSIONS The AQoL and SF-6D demonstrated acceptable measurement properties in the ICU population, but the findings raised questions about the reliability and predictive power of the SF-6D. Further research is required to determine the most appropriate instrument for use in measuring health utility in the ICU population.