Effect of Dexmedetomidine on Duration of Mechanical Ventilation in Septic Patients

A Systematic Review and Meta-Analysis

Peifen Chen; Jihong Jiang; Yunhe Zhang; Guobao Li; Zhihui Qiu; Mitchell M. Levy; Baoji Hu


BMC Pulm Med. 2020;20(42) 

In This Article


Sedation was essential for septic patients to tolerate mechanical ventilation.[4,18] According to the Canadian Agency for Drugs and Technologies in Health, DEX was associated with less duration of mechanical ventilation.[18] A most recent meta-analysis also revealed that DEX reduced the geometric mean respiratory support time by 22% in critically ill patients.[19] An RCT, conducted by two multi-center, revealed DEX shortened the median breathing support time in ICU patients.[6] However, the present meta-analysis showed that DEX was not associated with shortened duration of mechanical ventilation in adults. The subgroup analysis indicated that compared with propofol, DEX was not associated with shortened duration of mechanical ventilation either. It was confined to the limited trials on comparing DEX with midazolam, the subgroup analysis of that was not conducted.

A previous meta-analysis of critically ill patients, including medical, surgical, and trauma patients, revealed that compared with traditional sedative agents, DEX reduced the geometric mean duration of mechanical ventilation.[17] However, the study analyzed critical patients including septic patients but didn't do the subgroup analysis of septic group. In the present meta-analysis, we included four clinical trials involving 349 of septic patients. In other words, we focus on the effect of DEX on septic patients and we concluded that DEX resulted in improvement of ventilator-free days and 28-day mortality, but it did not reduce the duration of mechanical ventilation in septic patients. The results of the present meta-analysis were contrary to that conducted by Chen et al..[19] That may be attributed to differences in the participants (sepsis or septic shock patients vs. critically ill patients). Thus, in considering reductions in the ventilator duration in ICU patients, DEX could be better than other sedative agents, but it may not be the preferred agent in septic patients.

Ventilator-free day was defined as the number of days alive and free of mechanical ventilation in the first 28 days after enrollment.[11] The concept combines both mortality and duration of mechanical ventilation. It includes a binary variable of whether the patient is alive or not in the first 28 days and a continuous variable of the patient requiring mechanical ventilation.[20] A previous meta-analysis including two clinical trials involving 103 septic patients analyzed the number of mechanical ventilation free days during the 28-day period. The authors concluded that DEX had no significant effect on the duration of mechanical ventilation.[21] Nevertheless, the authors did not distinguish between mechanical ventilation free days and the duration of mechanical ventilation. There were not any solid data from the meta-analysis to show whether sedation with DEX affected the duration of mechanical ventilation. The present results suggested that DEX increased the number of ventilator-free days and reduced 28-day mortality, but it did not reduce the duration of mechanical ventilation. Because the number of ventilator-free days includes both mortality and the duration of mechanical ventilation, we may infer that the reduction in 28-day mortality contributed to the increase in the number of ventilator-free days.

There were also many limitations in the present meta-analysis. Firstly, there were only four trials included in the present meta-analysis that focused on septic patients. As the result, the outcome of the present study may not be used clinically and warrant further solid randomized control trials. Secondly, the study conducted by Kawazoe et al.[8] including patients on invasive and non-invasive mechanical ventilation which may aggrandize the bias of the overall outcome. Thirdly, we can't get the data of the duration of mechanical ventilation in three non-survivors in the included study conducted by Tasdogan et al.,[13] which may lead to a publication bias. Finally but not the least, because Tasdogan[13] and Kawazoe's[8] data were described as medians in the interquartile range, we estimated the means using medians because of the lack of individual patient data. Because estimating the sample mean and variance from the median, range, and size of the sample is a widely accepted practice in meta-analyses, we did not consider that this estimation would significantly affect the results of this meta-analysis.