Health outcomes methodology symposium: summary and recommendations

Interest in the philosophy and techniques of the assessment of health outcomes has burgeoned, prompting research funding agencies and others to examine traditional and emerging methods for outcome measurement.

This report summarizes the presentations and discussions at and research recommendations stemming from an invitational symposium on health outcomes methodology convened in September 1999.

The summary is based on the preliminary drafts of all formal reports and discussions, transcripts of all presentations and plenary discussions, and notes from breakout groups.

Existing health outcomes measures drawn from classic test theory and emerging approaches based on item response theory offer exciting opportunities for appreciably expanded applications in biomedical and health services research, clinical practice and decision making, and policy development. The major research agenda reflects the significance of this field of endeavor, its widening acceptance both at home and abroad, and its increasing applicability to many different patient and user communities. Of particular moment are the following: (1) refining and expanding of measurement techniques that rely on item response theory and making these approaches more understandable to potential users; (2) improving measurement tools to make them more culturally appropriate for diverse populations and more conceptually and psychometrically equivalent across such groups; (3) addressing longstanding issues in preference- and utility-based approaches, particularly in the elicitation of preference responses and scoring instruments; and (4) enhancing the ways in which data from outcomes measurement tools are calibrated against commonly understood clinical and lay metrics, are interpreted, and are made usable for different decision-makers.