THE STATE OF THE SCIENCE OF HRQOL ASSESSMENT IN CANCER- FINDINGS FROM THE CANCER OUTCOMES MEASUREMENT WORKING GROUP
Author(s)
Lipscomb J1, Gotay C2, Snyder C1, 1National Cancer Institute, Bethesda, MD, USA; 2Cancer Research Center of Hawaii, Honolulu, HI, USA
OBJECTIVE: To assess the state of the science of HRQOL assessment in cancer. METHODS: The US National Cancer Institute formed the Cancer Outcomes Measurement Working Group (COMWG) made up of 35 leading experts in the field to review the state of the science of assessing three outcomes (HRQOL, patient satisfaction, economic burden) in four cancers (breast, prostate, lung, colorectal) across the continuum of care (screening/prevention, treatment, survivorship, end of life). In this presentation, the findings of the COMWG are evaluated using the Medical Outcomes Trust Instrument Review Criteria. RESULTS: The results are as follows: 1) There is currently no consensus definition or conceptual model for HRQOL; developing a stronger theoretical base would improve HRQOL measurement and interpretation; 2) While there are now well-validated HRQOL instruments, validity can be further enhanced by an integrated application of the tools from modern psychometrics - especially item response theory (IRT) modeling - survey research, and cognitive psychology; 3) a number of instruments demonstrate adequate internal reliability according to classical test theory, but reliability could be evaluated more precisely at different points along the HRQOL continuum using IRT approaches; 4) interpretation of HRQOL findings has benefited from recent research on defining and identifying clinically important differences in instrument scores; both anchor-based and distribution-based approaches suggest changes of about 7% of a scale's breadth are important to patients and clinicians; 5) responsiveness of a measure is demonstrated when it is shown to detect statistically significant differences in HRQOL also large enough to be clinically important; 6) while current measurement techniques have generally proven feasible and acceptable, IRT-based computer adaptive testing may significantly decrease respondent burden without sacrificing precision; and 7) for cross-cultural assessment of HRQOL, several instruments currently meet minimum criteria. CONCLUSIONS: The current state of the science of HRQOL assessment in cancer is strong. Further development of a theoretical base and use of modern analytic techniques, including IRT, will accelerate progress.
Conference/Value in Health Info
2004-05, ISPOR 2004, Arlington, VA, USA
Value in Health, Vol. 7, No. 3 (May/June 2004)
Code
CN1
Topic
Patient-Centered Research
Topic Subcategory
Patient-reported Outcomes & Quality of Life Outcomes
Disease
Oncology