Construct Validity, Criterion Validity, and Internal Consistency of GOT-Cog®, a Novel Occupational Therapy Cognitive Screen

Abstract

Abstract Importance No single cognitive screen adequately captures all cognitive domains that are important for inpatient occupational therapy treatment planning. Objective To assess the construct validity of the Gaylord Occupational Therapy Cognitive (GOT-Cog©) screen, a novel cognitive screen developed to better inform inpatient occupational therapy treatment planning. Design: Randomized cross-over controlled study design using the St. Louis University Mental Status (SLUMS) exam as a comparator. Setting Long-term acute care hospital. Participants Participants were recruited from inpatients admitted to Gaylord Hospital who were prescribed occupational therapy services. Intervention: During participants’ initial occupational therapy evaluation, either the SLUMS exam or the GOT-Cog screen was randomly delivered; the screen not delivered on admission was delivered 22–26 hours later by the same or different clinician. Outcomes and Measures Primary outcome measures were GOT-Cog and SLUMS total scores, individual item scores, and domain scores. Results 129 Participants were approached; 98 participants yielded sufficient data for analysis. Total GOT-Cog and SLUMS scores were shown to positively correlate (p < 0.0001). All shared domains between GOT-Cog and SLUMS significantly correlated (p ≤ 0.0155); similarly all unique domains showed significant correlations to both GOT-Cog and SLUMS total score (p ≤ 0.0194). No ordering effects were observed (p ≥ 0.8081), and despite containing 11 more items, GOT-Cog only took 6 minutes longer to complete (10 versus 16 minutes, p < 0.0001). GOT-Cog and SLUMS were found to share a similar internal consistency. Conclusions and Relevance: Together, this work indicates GOT-Cog has overall excellent construct and criterion validity. Going forward, we will test the rater reliability and responsiveness of GOT-Cog.