Key Characteristics of Rehabilitation Quality Improvement Publications: Scoping Review From 2010 to 2016.

OBJECTIVE

To characterize the peer-reviewed quality improvement (QI) literature in rehabilitation.

DATA SOURCES

Five electronic databases were searched for English-language articles from 2010 to 2016. Keywords for QI and safety management were searched for in combination with keywords for rehabilitation content and journals. Secondary searches (eg, references-list scanning) were also performed.

STUDY SELECTION

Two reviewers independently selected articles using working definitions of rehabilitation and QI study types; of 1016 references, 112 full texts were assessed for eligibility.

DATA EXTRACTION

Reported study characteristics including study focus, study setting, use of inferential statistics, stated limitations, and use of improvement cycles and theoretical models were extracted by 1 reviewer, with a second reviewer consulted whenever inferences or interpretation were involved.

DATA SYNTHESIS

Fifty-nine empirical rehabilitation QI studies were found: 43 reporting on local QI activities, 7 reporting on QI effectiveness research, 8 reporting on QI facilitators or barriers, and 1 systematic review of a specific topic. The number of publications had significant yearly growth between 2010 and 2016 (P=.03). Among the 43 reports on local QI activities, 23.3% did not explicitly report any study limitations; 39.5% did not used inferential statistics to measure the QI impact; 95.3% did not cite/mention the appropriate reporting guidelines; only 18.6% reported multiple QI cycles; just over 50% reported using a model to guide the QI activity; and only 7% reported the use of a particular theoretical model. Study sites and focuses were diverse; however, nearly a third (30.2%) examined early mobilization in intensive care units.

CONCLUSIONS

The number of empirical, peer-reviewed rehabilitation QI publications is growing but remains a tiny fraction of rehabilitation research publications. Rehabilitation QI studies could be strengthened by greater use of extant models and theory to guide the QI work, consistent reporting of study limitations, and use of inferential statistics.