PURPOSE This study aimed to elucidate the impact of protocolized family support intervention on length of stay (LOS) in the intensive care unit (ICU) through a systematic review and meta-analysis.

METHODS Medline, EMBASE, the Cochrane Central Register of Controlled Trials, and other web-based databases were referenced since inception until November 26, 2018. We included randomized-controlled trials wherein protocolized family support interventions were conducted for enhanced communication and shared medical decision-making. LOS (in days) and mortality were evaluated using a random-effects model, and adjusted LOS was estimated using a mixed-effects model.

RESULTS We included seven randomized-controlled trials with 3477 patients. Protocolized family support interventions were found to significantly reduce the ICU LOS {mean difference = - 0.89 [95% confidence interval (CI) = - 1.50 to - 0.27]} and hospital LOS [mean difference = - 3.78 (95% CI = - 5.26 to - 2.29)]; the results of the mixed-effect model showed that they significantly reduced ICU LOS after adjusting for the therapeutic goal [mean difference = - 1.30 (95% CI = - 2.35 to - 0.26)], methods of measurement [mean difference = - 0.89 (95% CI = - 1.55 to - 0.22)], and timing of intervention [mean difference = - 1.05 (95% CI = - 2.05 to - 0.05)]. Similar results were found after adjusting for patients' disease severity [mean difference = - 1.21 (95% CI = - 2.03 to - 0.39)] and the trim-and-fill method [mean difference = - 0.86 (95% CI = - 1.44 to - 0.28)]. There was no difference in mortality rate in ICU and hospital between the protocolized intervention and control groups.

CONCLUSIONS Protocolized family support intervention for enhanced communication and shared decision-making with the family reduced ICU LOS in critically ill patients without impacting mortality.