Scientific Poster Session I - Original Research

Abstract

The 2018 PADIS guideline made moderate recommendation about non-benzodiazepine (BZD) sedatives over BZDs and suggested dexmedetomidine for mechanically ventilated patients with delirium and agitation. However, BZDs are still widely used in ICUs and the efficacy of dexmedetomidine to reduce delirium in real-world conditions is still unclear.

Research Question or Hypothesis:

Our study aimed to analyze prescription pattern of sedatives and how the clinical concerns affected guideline adherence. Also, we aimed to determine the correlation between non-BZD utilization and delirium under real-world circumstances.

Study Design:

This was a single-center, retrospective cohort study conducted in a mixed medical-surgical ICU.

Methods:

Patients were divided into three groups based on their first regular sedative prescription upon ICU admission: midazolam, propofol, or midazolam+propofol. Primary outcome of interest was antipsychotics new exposure after regular sedation with propofol or midazolam, in patients with no antipsychotic record within 6 months before admitted to ICU.

Results:

Among the 2,240 included patients, 857 were admitted to MICU, and 1,383 to SICU. The preferred sedative in MICU was midazolam (59.5%) while propofol in SICU (96.5%). Patients who received midazolam in SICU had a higher proportion of SBP <90 or HR <60 (p<.0001), and in MICU had history of conductive cardiovascular diseases (p=0.0132), and previous use of cardiovascular agents (p=0.0057) and sedative-hypnotics (p=0.05). Multivariate analysis revealed that regular midazolam was significantly associated with higher risk of antipsychotic utilization (aOR 1.265, 95% CI 1.1-1.454, p=0.0009), while adding dexmedetomidine did not reduce the risk after sedation with midazolam (aOR 1.45, 95% CI 1.062-1.98, p=0.0193).

Conclusion:

This study demonstrated differences in the preferred sedatives between MICU and SICU, with deviations from guideline due to patient comorbidities, past medication use, and hemodynamic instability. Furthermore, midazolam was associated with higher risk of antipsychotic utilization, and dexmedetomidine may not be effective in reducing the needs of antipsychotics after BZD sedation.

Presenting Author

Authors