Sedation practices and clinical outcomes in mechanically ventilated patients in a prospective multicenter cohort

Crit Care. 2019 Apr 17;23(1):130. doi: 10.1186/s13054-019-2394-9.

Abstract

Objectives: We sought to study the association between sedation status, medications (benzodiazepines, opioids, and antipsychotics), and clinical outcomes in a resource-limited setting.

Design: A longitudinal study of critically ill participants on mechanical ventilation.

Setting: Five intensive care units (ICUs) in four public hospitals in Lima, Peru.

Patients: One thousand six hundred fifty-seven critically ill participants were assessed daily for sedation status during 28 days and vital status by day 90.

Results: After excluding data of participants without a Richmond Agitation Sedation Scale score and without sedation, we followed 1338 (81%) participants longitudinally for 18,645 ICU days. Deep sedation was present in 98% of participants at some point of the study and in 12,942 ICU days. Deep sedation was associated with higher mortality (interquartile odds ratio (OR) = 5.42, 4.23-6.95; p < 0.001) and a significant decrease in ventilator (- 7.27; p < 0.001), ICU (- 4.38; p < 0.001), and hospital (- 7.00; p < 0.001) free days. Agitation was also associated with higher mortality (OR = 39.9, 6.53-243, p < 0.001). The most commonly used sedatives were opioids and benzodiazepines (9259 and 8453 patient days respectively), and the latter were associated with a 41% higher mortality in participants with a higher cumulative dose (75th vs 25th percentile, interquartile OR = 1.41, 1.12-1.77; p < 0.01). The overall cumulative dose of benzodiazepines and opioids was high, 774.5 mg and 16.8 g, respectively, by day 7 and by day 28; these doses approximately doubled. Haloperidol was only used in 3% of ICU days; however, the use of it was associated with a 70% lower mortality (interquartile OR = 0.3, 0.22-0.44, p < 0.001).

Conclusions: Deep sedation, agitation, and cumulative dose of benzodiazepines were all independently associated with higher 90-day mortality. Additionally, deep sedation was associated with less ventilator-, ICU-, and hospital-free days. In contrast, haloperidol was associated with lower mortality in our study.

Keywords: Clinical outcomes; Critical illness; Sedation.

Publication types

  • Multicenter Study
  • Research Support, N.I.H., Extramural

MeSH terms

  • APACHE
  • Adult
  • Aged
  • Aged, 80 and over
  • Analgesics / administration & dosage
  • Analgesics / adverse effects
  • Analgesics / therapeutic use
  • Antipsychotic Agents / administration & dosage
  • Antipsychotic Agents / adverse effects
  • Antipsychotic Agents / therapeutic use
  • Cohort Studies
  • Conscious Sedation / adverse effects
  • Conscious Sedation / methods
  • Conscious Sedation / standards*
  • Deep Sedation / adverse effects
  • Deep Sedation / methods
  • Deep Sedation / standards*
  • Female
  • Humans
  • Hypnotics and Sedatives / administration & dosage
  • Hypnotics and Sedatives / adverse effects
  • Hypnotics and Sedatives / therapeutic use
  • Intensive Care Units / organization & administration
  • Intensive Care Units / statistics & numerical data
  • Logistic Models
  • Longitudinal Studies
  • Male
  • Middle Aged
  • Neuromuscular Blocking Agents / administration & dosage
  • Neuromuscular Blocking Agents / adverse effects
  • Neuromuscular Blocking Agents / therapeutic use
  • Odds Ratio
  • Peru
  • Prospective Studies
  • Respiration, Artificial / methods
  • Treatment Outcome*

Substances

  • Analgesics
  • Antipsychotic Agents
  • Hypnotics and Sedatives
  • Neuromuscular Blocking Agents