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


The present review study was performed according to Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA), which is the preferred system for reporting items for conducting systematic reviews and meta-analyses.[9]

Eligibility Criteria

The definition of sepsis was revised in February 2016.[10] Hence, the studies included in this review involved adults with sepsis and/or septic shock and at least two systemic inflammatory response syndrome (SIRS) criteria due to infection, which was defined by the investigators. All studies were prospective randomized control trials (RCTs), contained data on ventilation duration and/or ventilator-free duration. The exclusion criteria were as follows: pediatric; patients with SIRS by other causes, such as burn or trauma; and studies without a clear sepsis subgroup.

Identification of Studies

We searched the following databases: PubMed (1993 to 20 January 2019), Cochrane (2007 to 20 January 2019), and EMBASE (1990 to 20 January 2019). There was no language restriction. The search term "Clinical Trial" was used in searching the databases. The Endnote X8 citation manager was used to compile the references. Duplicates were filtered using the "Find Duplicates" feature, and then the data were searched manually. Two groups of search terms were combined in this study. The first group included "sepsis," "septic shock," "systemic inflammatory response," and "SIRS." The second group included "Alpha-2 agonists" and "dexmedetomidine" (Additional file 1). When they were identified using the above search strategies, the references list of RCTs and the relevant review articles were manually checked to include other potentially eligible trials.

Analysis of Outcomes

The primary outcome of this study was the duration of mechanical ventilation. The secondary outcomes were 28-day mortality and ventilator-free days, which was defined as the number of days alive and successfully weaning from mechanical ventilation in the first 28 days after enrollment in the trials.[11]

We also evaluated the methodological quality of this meta-analysis separately by using the "risk of bias table" tool in Manager (Revman) (Version 5.3. Copenhagen: The Nordic Cochrane Center, the Cochrane Collaboration, 2014).

Study Selection and Data Extraction

Two reviewers (i.e., Chen and Zhang) independently screened the titles and abstracts yielded by the search strategies and selected the potentially relevant trials. Then the full texts of relevant trials were assessed according to the eligibility criteria. Chen and Jiang extracted the data from the included studies independently. The details about the study designs and outcomes were entered in Microsoft Office Excel 2007 and then checked by the third author (Hu). Any discrepancy was resolved by either discussion or according to advice from other authors. The original authors were contacted if data were not present in the relevant articles.

Quality Assessment

The Grading of Recommendations Assessment, Development, and Evaluation (GRADE) methodology was used to assess the quality of the studies.[12] In brief, the quality of the evidence was analyzed and then categorized in one of four domains: "very low," "low," "moderate," or "high." All studies included in this meta-analysis were RCTs that provided high-quality evidence. In some cases, the quality of the evidence was decreased for several reasons, including reporting bias, imprecision, inconsistency, indirectness of evidence, and publication limitations.

Statistical Analysis

The mean values and standard deviation (SD) of the duration of mechanical ventilation and 28-day ventilator-free days were extracted for the outcome analysis. Because Tasdogan[13] and Kawazoe[8] expressed the data in the form of median and interquartile range, we emailed the first and corresponding authors but failed to obtain the raw data; therefore, we followed the recommendations of Wan et al..[14] and Luo et al.[15] to estimate the mean values and SD of Tasdogan and Kawazoe's data.

An inverse variance model with a 95% confidence interval was used to analyze the continuous outcome. The risk ratio (RR) and 95% confidence interval (CI) were used to analyze the dichotomous outcomes. A P value of less than 0.05 was considered significant. Significant heterogeneity was identified when the P value determine by the chi-square test was less than 0.10 and I 2 was greater than 50%. A fixed-effect model was employed to calculate the pooled effect when there was no statistically significant heterogeneity. Otherwise, a random-effects model was used. Publication bias was evaluated by a funnel plot. All statistical analyses were performed using the Review Manager software.