Presence and severity of intensive care unit-acquired paresis at time of awakening are associated with increased intensive care unit and hospital mortality

Objectives: To assess whether the presence and severity of intensive care unit-acquired paresis are associated with intensive care unit and in-hospital mortality. Design: Prospective, observational study. Setting: Two medical, one surgical, and one medico-surgical intensive care units in two university hospitals and one university-affiliated hospital. Patients: A total of 115 consecutive patients were enrolled after > 7 days of mechanical ventilation. Interventions: None. Measurements and Main Results: The Medical Research Council score (from 0-60) was used to evaluate upper and lower limb strength at time of awakening, identified as the ability to follow five commands. Intensive care unit-acquired paresis was defined as a Medical Research Council score <48. Patients were followed-up until hospital discharge. The primary end point was hospital mortality. At awakening, median Medical Research Council score was 41 (interquartile range, 21-52), and 75 (65%) patients had intensive care unit-acquired paresis. Hospital non-survivors had a significantly lower Medical Research Council score at awakening (21 [11-43]) vs. 41 [28-53]; p = .008) and a significantly higher rate of intensive care unit-acquired paresis (85.1% vs. 58.4%; p = .02) compared to survivors. After multivariate risk adjustment, intensive care unit-acquired paresis was independently associated with higher hospital and intensive care unit mortality (odds ratio for hospital mortality, 2.02; 95% confidence interval, 1.03-8.03; p = .048). Each Medical Research Council point decrease was associated with a significantly higher hospital mortality (odds ratio, 1.03; 95% confidence interval, 1.01-1.05; p = .033). Conclusions: Both the presence and severity of intensive care unit-acquired paresis at the time of awakening are associated with increased intensive care unit and hospital mortality; the mechanisms underlying this association need further study. (Crit Care Med 2009; 37:3047-3053)