Reporting of and factors associated with treatment fidelity of ICU-based physical rehabilitation

Abstract

Background: Physical function impairments associated with intensive care unit (ICU) admission can persist 5 years after ICU discharge. Early trials of ICU-based physical rehabilitation demonstrated improved physical function outcomes at hospital discharge. Recently, ICU-based physical rehabilitation trials have shown discordant results. Variability may be related to differences in study populations, intervention and comparator treatments or outcomes assessed; it may also be due to differences in treatment fidelity, the extent to which the protocol was delivered as planned. Objectives: The primary objectives of the two complementary studies were to determine: (1) the proportion of treatment fidelity components reported in ICU-based physical rehabilitation randomized controlled trials (RCT) and (2) the patient characteristics associated with treatment fidelity in a pilot RCT of early ICU-based physical rehabilitation. Methods: We conducted a scoping review with standard methods according to the Joanna Briggs Institute. We included RCTs if they enrolled adults or children admitted to an ICU that underwent a planned physical rehabilitation intervention. We did not limit comparator group or outcomes assessed. We calculated treatment fidelity scores across studies and by treatment groups as the proportion of reported National Institutes of Health-Behavioural Change Consortium components. We conducted a secondary analysis of the CYCLE Pilot RCT which enrolled adults admitted to an ICU to cycle ergometry + routine physiotherapy (PT) or routine PT alone. We used multiple linear regression to assess the relationship between patient characteristics and ICU exposures and treatment fidelity, defined as the proportion of planned intervention days when the randomized intervention was delivered or did not occur due medical status. Results: Our scoping review included 94 studies which met the inclusion criteria. Median [1st-3rd quartiles] treatment fidelity scores were low (19% [14%-26%]), but worse in comparator groups (Intervention: 24% [19%-33%]; Comparator: 14% [5%-24%]; p<0.01). In our secondary analysis, we found randomized group allocation to cycling was associated with 9.7% (95% confidence interval [CI] -17.8 to -1.5) lower treatment fidelity. Conclusion: Treatment fidelity reporting was limited, which challenges interpretation of ICU-based physical rehabilitation trial results. Researchers can use this thesis to guide study protocol development, study conduct, and manuscript preparation to optimize treatment fidelity. Readers can use this work to guide interpretation of ICU-based physical rehabilitation RCTs with a consideration of treatment fidelity.