A randomized trial of daily awakening in critically ill patients managed with a sedation protocol: a pilot trial.

ABSTRACT

OBJECTIVE: Protocolized sedation (PS) and daily sedative interruption (DI) in critically ill patients have both been shown to shorten the durations of mechanical ventilation (MV) and intensive care unit (ICU) stay. Our objective was to determine the safety and feasibility of a randomized trial to determine whether adults managed with both PS + DI have a shorter duration of MV than patients managed with PS alone. DESIGN: Prospective randomized, concealed, unblinded, multicenter, pilot trial. SETTING: Three university-affiliated medical-surgical ICUs. PATIENTS Sixty-five adults anticipated to require MV >48 hrs and receiving sedative/analgesic infusions. INTERVENTIONS: Patients were randomized to PS alone, or PS + DI. PS was implemented by bedside nurses; sedatives/analgesics were titrated to achieve Sedation Agitation Score (SAS) 3-4. The PS + DI group also had infusions interrupted daily until the patients awoke. MEASUREMENTS AND MAIN RESULTS: Diagnosis, age [mean +/- SD] (53 +/- 18.3 vs. 62.1 +/- 16.7 yrs) and Acute Physiology and Chronic Health Evaluation II (27.7 +/- 8.4 vs. 26.6 +/- 8.4) were similar in the PS and PS + DI groups, respectively. The median duration of MV in the PS and PS + DI groups was 8.0 vs. 10.5 days, and ICU stay was 10.0 vs. 13.0 days, respectively. The SAS was within target range (3-4) in 59% of 9,611 measurements, and within an acceptable range (2-5) in 86% of measurements. Self-assessed nursing and respiratory therapist workload was low in the majority of the cohort. Adverse events were similar in both groups. Patient recruitment was slower than projected (1.5 patients/mo). CONCLUSION: This pilot trial comparing PS vs. PS + DI confirmed the safety and acceptability of the sedation protocol and DI, and guided important modifications to the protocol, thus enhancing the feasibility of a future multicenter trial. This trial was not designed to detect small but significant differences in clinically important outcomes.