The purpose of this study is to estimate an average treatment effect in the treated (ATT) of prescribing AFOs for inpatient people with subacute stroke on recovery of Functional Independence Measure (FIM) using propensity score analysis.
A total of 6932 people who were admitted to a stroke care unit and newly diagnosed with stroke between December 2016 and October 2022 were enrolled in this retrospective study. We screened 451 adults with lower-limb hemiplegia due to cerebral infarction or intracerebral hemorrhage who were transferred to two rehabilitation hospitals and then discharged, 71 of whom were prescribed AFOs. As the primary outcome, ATT was estimated as the mean difference in FIM-motor part scores (FIM-M) at discharge for those who were prescribed AFOs compared to as if when they had not been prescribed it. We collected 28 variables as baseline covariates which contained clinical factors (e.g. severity of stroke and paralysis, comorbidities, FIM), socio-demographic factors (e.g. age, sex, premorbid situations, insurance) and blood biomarkers on admission to acute care hospital via medical records. The FIM scores on admission to rehabilitation hospital were also contained as covariates. Propensity scores of prescribing AFOs were calculated using a generalized linear model, applying the aforementioned covariates. We used the optimal matching method to match those who were prescribed AFOs (AFO group) with those who were not (non-AFO group), then applied the double propensity score adjustment (DA) method to reduce residual confounding and avoid target bias. We also calculated the E-value to test the robustness of our results as a sensitivity analysis.
Among 71 matched pairs, 19/28 covariates balanced, and the median (IQR) for FIM-M scores at discharge were 69 (45, 86) and 76 (59, 82) in the AFOs and non-AFOs groups. After applying the DA method, the ATT was +7.19 points (95%CI: 2.87 to 11.5, p 0.01). The E-value was estimated to be 2.71 (95%CI: 1.75 to NA), suggesting that the results were moderately robust.
Our results indicated that prescribing AFOs for inpatient people with subacute hemiplegia significantly enhanced functional independence compared to when they were not prescribed. This is the first retrospective study to quantify the causal effect of prescribing AFOs on functional recovery during rehabilitative care. Further analysis is needed to identify the people, situations, and mechanisms to which AFOs contributed.
The AFOs should be prescribed not only for functional aspects such as gait and balance, but also for its contribution to recovery of functional independence at subacute phase.
Average treatment effect
Functional Independence Measure
