This study rigorously evaluated the reliability and validity of Section GG, a component of the Centers for Medicare and Medicaid Services Inpatient Rehabilitation Facility Patient Assessment Instrument (CMS-IRF PAI) Version 1.4, focusing on its effectiveness in assessing self-care and mobility. The research compared Section GG with the established Functional Independence Measure (FIM) motor subscale (FIMm) to determine its accuracy and consistency in measuring functional independence in rehabilitation settings.
Conducted as a retrospective cohort study within a single inpatient rehabilitation facility, the research encompassed 1296 patients admitted for post-stroke rehabilitation between October 2016 and October 2019. This robust sample size allowed for a comprehensive analysis of Section GG’s performance. The study did not involve specific interventions but rather focused on comparing scores from Sections GG, B (Hearing, Speech, and Vision), and C (Cognitive Patterns) of the CMS-IRF PAI against the FIMm and cognitive subscale (FIMc). Spearman’s and Bland-Altman analyses were employed as the main outcome measures to rigorously compare these assessment tools.
The findings revealed a strong correlation between Section GG and FIMm scores, both at admission (ρ=0.919, P<.001) and discharge (ρ=0.929, P<.001). This high correlation indicates that Section GG is a reliable measure of motor function, performing comparably to the widely accepted FIMm. However, the study also highlighted a notable ceiling effect for Section GG at discharge (8.6%), suggesting that it may not effectively differentiate among patients with higher levels of functional independence as well as FIMm.
Bland-Altman analysis further indicated a bias towards higher scores with Section GG compared to FIMm, both at admission (Bias=2.3%, P<.001) and discharge (Bias=6.2%, P<.001). This bias suggests that Section GG tends to rate patient abilities slightly more favorably than FIMm. Furthermore, Section GG demonstrated a bias towards greater gains in function (Bias=3.9%, P<.001), particularly in areas of walking and stair climbing (bias=3.71%, P<.001). Interestingly, self-care items contributed less to the gains measured by Section GG compared to FIMm (bias=-7.5%, P<.001), indicating potential differences in how these tools capture improvements in specific functional areas.
In addition to Section GG, the study assessed the internal validity of a combined scale comprising Section B and C (B+C scale) for cognitive assessment. The B+C scale demonstrated good internal validity (Cronbach’s alpha=0.868) and a significant correlation with FIMc (ρ=0.745). However, similar to Section GG, the B+C scale also tended to rate patients at a higher cognitive level (bias=20.0%, P<.001) and exhibited a greater ceiling effect at admission (20.4%) compared to FIMc (0.6%). This suggests that while the B+C scale is a reliable measure of cognition, it may overestimate cognitive abilities, especially at higher levels of function.
In conclusion, this validation study demonstrates that both Section GG and the B+C scale of the CMS-IRF PAI are correlated with the established FIM subscales, indicating their potential utility in rehabilitation settings. However, the observed bias towards higher ratings and ceiling effects, particularly at higher functional levels, suggests limitations in their ability to discriminate among more independent patients. Further research is warranted to evaluate the clinical acceptability of Section GG total scores as a definitive outcome measure and predictor of long-term patient outcomes. These findings are crucial for clinicians and researchers in understanding the nuances of using Section GG and the B+C scale in patient assessment and outcome measurement within inpatient rehabilitation facilities.
