Adverse Effects of Deep Sedation in Mechanically Ventilated Patients

Abstract & Commentary

By Linda L. Chlan, RN, PhD, Dean’s Distinguished Professor of Symptom Management Research, The Ohio State University, College of Nursing. Dr. Chlan reports that she receives grant/research support from the National Institutes of Health.

Synopsis: Deep sedation during the early period of mechanical ventilatory support delays extubation and increases mortality, yet is a modifiable risk factor that requires innovative intervention to reduce these adverse outcomes.

Source: Shehabi Y, et al. Sedation depth and long-term mortality in mechanically ventilated critically ill adults: A prospective longitudinal multicenter cohort study. Intensive Care Med 2013;39:910-918.

This multicenter, prospective, longitudinal, observational, non-interventional cohort study was conducted in 11 Malaysian ICUs. This replication study aimed to determine if the findings from the Australian New Zealand Sedation Practice in Intensive Care Evaluation (ANZ SPICE) trial on sedation practice and depth of sedation would be similar in the participating Malaysian ICUs. The relationship among current sedation practices — depth of sedation (assessed every 4 hours with the Richmond Agitation and Sedation Scale [RASS]) after early initiation of mechanical ventilation, time to extubation, delirium (assessed daily with the Confusion Assessment Method for the ICU [CAM-ICU]) — and 180-day mortality were evaluated. All of the participating ICUs had 24/7 physician coverage (intensivist or anesthesiologist), with 1:2 nurse: patient ratios in most of the ICUs. Patients (n = 259) were included who had mechanical ventilation initiated within the previous 24 hours, were receiving sedative medications, and were expected to receive ventilatory support for > 24 hours. Deep sedation was defined as RASS ≤ -3. Patients were followed for up to 28 days in the ICU. The early ventilation period was defined as the first 48 hours of ventilatory support. All aspects of this replication study were modeled after the ANZ SPICE trial, including the statistical analyses that were performed at the Australian research center.

Overall, 180-day mortality was 46.4% with 8.5% of patients lost to follow-up. Midazolam was the most commonly used sedative medication (93%), administered to patients on almost 40% of study days. Dexmedetomidine was used in 29.3% of patients, while propofol was used in 28% of patients. Morphine was used slightly more (77.6%) than fentanyl (60.2%). The RASS assessments conducted in the first 48 hours of ventilatory support indicated that 58% of patients were deeply sedated, with 39.3% in the light range and 2.8% in the 2-4 RASS range. A small number of ICU days saw the use of neuromuscular blockade (4.1%), physical restraints (21.1%), or self-extubations (0.4%). A full 61% of patients were evaluated after 48 hours as still deeply sedated. Clinical reasons for desiring deep sedation included controlled ventilation, muscle relaxation, and need to address severe agitation. Routine daily sedation interruption was implemented on only 2.3% of study days and was not a common practice in the Malaysian ICUs. Delirium was present overall in 44% of patients throughout the study period. In patients with > 8 days in the ICU, delirium was present in 50% of assessments, and increased to 68.9% in patients in the ICU ≥ 14 days.

Not surprisingly, patients who were initially deeply sedated had longer time to extubation with higher hospital and 180-day mortality. For every 4-hour assessment period spent at RASS -3 to -5, there was a 13% increase in risk of death at discharge, with an 8.5-hour delay in extubation, despite adjusting for sedative choice, diagnosis, and severity of illness. However, neither early deep sedation nor cumulative sedative medication doses were independent predictors of time to delirium. The Malaysian investigators concluded that their findings were similar to the ANZ SPICE trial findings and identified early, deep sedation as a potentially modifiable risk factor in delayed extubation and mortality. Shehabi et al suggest that early deep sedation may be a global problem associated with poor outcomes.


The findings of this replication study emphasize the deleterious effects of early, deep sedation on outcomes and mortality, regardless of the country of origin, something that appears to be a worldwide problem. Shehabi et al call for the delivery of interventions during the mechanical ventilatory support period to reduce the intensity and depth of sedation, particularly early in the ICU stay. However, this can be a very complicated, clinically vexing goal to attain.

Sedation management strategies need to be developed that allow mechanical ventilation synchrony and promote patient comfort, yet avoid deep patient sedation, unless medically indicated, at the same time. This may mean frequent ongoing evaluation of sedation goals, more than once daily during morning ICU rounds. The overall goal should be a comfortable, awake patient while achieving ventilator synchrony and adequate gas exchange. While this goal may not be achievable in all mechanically ventilated patients, it should be the target, nonetheless, given the mounting evidence that deep sedation does not provide any benefit to outcomes of this challenging and vulnerable group of ICU patients.