Comparator Groups in ICU-Based Studies of Physical Rehabilitation: A Scoping Review of 125 Studies.

To characterize comparator groups (CGs) in ICU-based studies of physical rehabilitation (PR), including the type, content, and reporting.

Data sources: We followed a five-stage scoping review methodology, searching five databases from inception to June 30, 2022. Study selection and data extraction were completed independently, in duplicate.

Study selection: We screened studies by title and abstract, then full-text. We included prospective studies with greater than or equal to two arms enrolling mechanically ventilated adults (≥ 18 yr), with any planned PR intervention initiated in the ICU.

Data extraction: We conducted a quantitative content analysis of authors' description of CG type and content. We categorized similar CG types (e.g., usual care), classified content into unique activities (e.g., positioning), and summarized these data using counts (proportions). We assessed reporting using Consensus on Exercise Reporting Template (CERT; proportion of reported items/total applicable).

Data synthesis: One hundred twenty-five studies were included, representing 127 CGs. PR was planned in 112 CGs (88.2%; 110 studies), representing four types: usual care (n = 81, 63.8%), alternative treatment than usual care (e.g., different from intervention; n = 18, 14.2%), alternative treatment plus usual care (n = 7, 5.5%), and sham (n = 6, 4.7%). Of 112 CGs with planned PR, 90 CGs (88 studies) reported 60 unique activities, most commonly passive range of motion (n = 47, 52.2%). The remaining 22 CGs (19.6%; 22 studies) reported vague descriptions. PR was not planned in 12 CGs (9.5%; 12 studies), and three CGs (2.4%; three studies) reported no details. Studies reported a median (Q1-Q3) of 46.6% (25.0-73.3%) CERT items. Overall, 20.0% of studies reported no detail to understand planned CG activities.

Conclusions: The most common type of CG was usual care. We identified heterogeneity in planned activities and CERT reporting deficiencies. Our results could help guide the selection, design, and reporting of CGs in future ICU-based PR studies.