Various quality improvement strategies decrease injurious falls in the elderly

Référence

Tricco AC, Thomas SM, Veroniki AA, et al. Quality improvement strategies to prevent falls in older adults: a systematic review and network meta-analysis. Age Ageing 2019;48(3):337-346.

Sommaire

These authors performed a systematic review and network meta-analysis on 126 RCTs with 84,307(!!!) participants. The trials were of various quality improvement/system engineering strategies to decrease falls in the elderly. To find these studies, the authors searched several databases and registries, conference abstracts, and reference lists of included studies. After having been trained to a high level of proficiency, pairs of authors assessed potential studies for inclusion and methodologic quality. If disagreements occurred, a third author resolved them. Since quality improvement activities come in many shapes, sizes, and level of complexity, the authors employed a coder who independently applied a pre-established classification framework. Most of the participants were women and two thirds had no prior fall history. More than half the trials did not report medication use. Most of the studies (57%) took place in the home, 28% took place in clinics, 27% in the community, 15% in hospitals, 16% in chronic care facilities, and 8% in retirement homes (some studies took place in multiple settings). The RCTs ranged from 22 to 5310 participants; the cluster RCTs ranged from 80 to 10,558 participants. Overall, the studies were of modest quality. Using network meta-analysis, the authors determined that changing the structure or organization of care teams (addition of a team member, use of multidisciplinary teams, or the expansion or revision of roles) had a greater effect on reducing injurious falls than usual care (odds ratio [OR] 0.57 [95% CI 0.33 - 0.99). Compared with usual care, the number of patients who fell was much lower when there was a combination of case management, patient reminders, and staff education (OR 0.18; 0.07 - 0.47) and when there was a combination of case management and patient reminders (OR 0.36; 0.13 - 0.97). The authors were unable to perform network analyses on hospitalizations, emergency department visits, or outpatient visits.