Psychometric Properties of the Occupation-Based Practice Assessment.

Only two therapists were observed, so the reliability estimates come from a very small sample of raters. Every interaction was an initial evaluation; no intervention sessions were rated, so the tool's reliability during treatment is unknown. Both therapists worked in the ED and ICU under a co-treatment model with a physical therapist, which may not reflect how most OTs practice in outpatient, inpatient rehab, or school settings. The engaged-participation subscale fell below the moderate-agreement threshold in both the observer-observer and self-observer comparisons, meaning that particular score is not yet dependable. No gold-standard measure of occupation-based practice exists, so concurrent validity could not be tested. The study was conducted at a single hospital in one region of the US.

The authors declare no financial support and no conflicts of interest.

Do not read the good overall reliability (ICC = .843) as meaning every part of the OBPA is equally dependable. The engaged-participation subscale, which captures how the therapist facilitates the client's active involvement, showed agreement below the moderate threshold (ICC = .487 between observers; ICC = .462 to .578 in self-observer comparisons), and the authors themselves note this item drove the discrepancy. A score on that subscale should not be used for individual performance feedback or program evaluation until its reliability is re-established in a larger, more varied sample.