The association between dexmedetomidine as a single or adjuvant sedative versus other sedatives and the duration of mechanical ventilation and ICU stay in critically ill patients in a central South African ICU

Abstract

Background: Sedation is often required in the intensive care unit (ICU) but can be harmful if administered inappropriately or excessively. Dexmedetomidine offers a favourable, cooperative sedation profile, despite a higher relative cost. It also has analgesic and opioid-sparing properties. The multidisciplinary ICU at our central South African hospital adopted the use of dexmedetomidine through the course of 2016. The aim of the study was to determine whether this change in practice affected the ICU length of stay (LOS) and duration of mechanical ventilation at this unit. Methods: A retrospective cohort analysis of patients’ files, who were sedated with midazolam and propofol in 2015 and those sedated with dexmedetomidine in 2017, was conducted. The data gathered included the sedatives used, demographic and vital data, ICU LOS, duration of mechanical ventilation and treatment of side effects. Group 2015 and Group 2017 were also analysed for possible confounders where appropriate, and these confounders were excluded for a re-analysis. Descriptive statistics were used and results were analysed for range, median, interquartile range (IQR), percentage and frequency. For post-hoc analysis of the effect of confounders, the Spearman rank correlation coefficient was used to determine the association between duration and sedative exposure and either duration of ICU stay or mechanical ventilation. The null hypothesis was set at p < 0.05. Results: Group 2015 comprised 52 patients and Group 2017 60 patients. No difference between the groups was found regarding ICU LOS (median [IQR] 5 [2–14] vs 8.5 [5–12.5] days; p = 0.10) or mechanical ventilation (median [IQR] 91 [34–272] vs 129 [58–221] hours; p = 0.44). Those who were sedated with dexmedetomidine had better initial prognoses (median APACHE II score 13 vs 18, p = 0.01), were sedated for greater fractions of their total ICU stay (median 46% vs 25%, p < 0.01), and had a higher incidence of hypotension and bradycardia (36.7% vs 11.4%; p < 0.01); which did not relate to a higher mortality. Spearman’s rank correlation coefficients also showed a weak to moderate association with longer ICU stay and ventilation duration when the duration of sedation with midazolam or propofol was shorter in relation to ICU stay. Conclusion: We did not find a reduction in ICU LOS or mechanical ventilation with the advent of dexmedetomidine in our unit. The lack of regular documentation of sedation levels and scheduled sedation breaks might have contributed to these results. Dexmedetomidine has a role to play in the ICU setting, but it should only be used when clearly indicated. Vigilance for hypotension and bradycardia is required when using dexmedetomidine.