The Efficacy and Safety of in-ICU Leg Cycle Ergometry in Critically Ill Adults: A Systematic Review and Meta-Analysis.

BACKGROUND: Survivors of critical illness may experience physical function deficits following intensive care unit (ICU) discharge. In-ICU cycle ergometry may facilitate early mobilization and decrease functional impairment. OBJECTIVE: We conducted a systematic review and meta-analyses to understand the effect of in-ICU leg cycle ergometry on patient-important and clinically-relevant outcomes. DATA SOURCES: We searched eight electronic databases from inception until July 2019. DATA EXTRACTION: We included randomized (RCTs) and non-randomized studies of critically ill adults admitted to ICU for ≥24 hours, comparing cycling interventions to control arms that did not receive cycling. Main outcomes included physical function, mechanical ventilation (MV) duration, length of stay (LOS), quality-of-life (QoL), mortality, and safety. We conducted independent duplicate citation screening, data abstraction, and risk of bias assessments. We pooled RCTs using a random-effects model and calculated relative risk (RR), mean difference (MD), or standardized mean difference (SMD) with 95% confidence intervals [CI]. We assessed certainty of outcomes using GRADE. RESULTS: Of 6,531 citations, we included 12 RCTs and 2 non-randomized studies (n=926). Between cycling and control groups, there were no differences in physical function at hospital discharge (3 RCTs, n=225, SMD 0.07, -0.38 to 0.53, very low-certainty), MV duration (9 RCTs, n=676, MD 0.01 [-1.04 to 1.07] days, moderate-certainty), ICU LOS (10 RCTs, n=511, MD 0.23 [-1.44 to 1.89] days, moderate-certainty), hospital LOS (7 RCTs, n=393, MD -0.07 [-3.87 to 3.73] days, moderate-certainty), QoL at 6-months post-hospital discharge (2 RCTs, n=103, MD 9.13 [13.80 to 32.05] points higher, very low-certainty), or hospital mortality (7 RCTs, n=710, RR 1.09 [0.82 to 1.46], moderate-certainty). The adverse event rate in cycling sessions was 0.16% across studies (10 studies, 5/3117 sessions, very low-certainty). CONCLUSIONS: Cycling initiated in the ICU is probably safe, however we did not find any differences in physical function, MV duration, LOS, QoL or mortality compared to those not receiving cycling. Rigorously designed RCTs are needed to improve precision and further investigate the effect of cycling on patient-important outcomes.