Is patient-reported function reliable for monitoring postacute outcomes?

A major challenge in the development of a comprehensive measurement system to evaluate effectiveness across a broad range of postacute care settings is the stability and consistency of outcomes measures across respondents and settings. The objective of this study was to investigate the test-retest and subject-proxy reliability of activity scores for use in a new postacute care outcome instrument using an interview format across different care settings.

Twenty-five subjects were randomly selected from a larger study of 485 individuals and were interviewed on two occasions within 1 to 4 days to assess self-reported test-retest reliability of summary scores of the Activity Measure-Post-Acute Care item pool. Proxy reliability was tested by interviewing the primary physical or occupational therapist or family member using an identical questionnaire in addition to the subject in 45 patients.

Test-retest and subject-proxy reliability was acceptable for the three domains of the activity construct: physical and movement, personal and instrumental, and applied cognition with intraclass correlation coefficients of the summary scores of each of the three domains ranging between 0.91 and 0.97 for test-retest and 0.68 and 0.90 for subject-proxy.

Reliability is adequate to justify use of these activity scales across respondents and settings.