A meta-analysis of prospective trials comparing percutaneous and surgical tracheostomy in critically ill patients.

STUDY OBJECTIVES: Tracheostomy is one of the most commonly performed procedures in the patient receiving long-term mechanical ventilation. While percutaneous dilational tracheostomy (PDT) is becoming increasingly utilized as an alternative to conventional surgical tracheostomy, most literature evaluating these two techniques is neither prospective nor controlled. We performed a meta-analysis of available prospective controlled studies comparing PDT and surgical tracheostomy in critically ill patients to more fully understand the relative benefits and risks of these two procedures in this population.

DESIGN: Meta-analysis using Mantel-Haenszel fixed effect model.

INTERVENTIONS: We performed searches of MEDLINE, Current Contents, Best Evidence, Cochrane, and HealthSTAR databases from 1985 to present to identify prospective controlled studies comparing PDT and surgical tracheostomy in critically ill patients. After establishing clinical and statistical homogeneity (Q: statistic), studies were analyzed by a Mantel-Haenszel fixed effect model. For each clinical end point examined, PDT and surgical tracheostomy were compared by calculating either absolute differences or odds ratios (ORs) with 95% confidence intervals (CIs) for continuous or discrete variables, respectively.

MEASUREMENTS AND RESULTS: We pooled data from five studies (236 patients) satisfying our search criteria to analyze eight clinical end points. Operative time was shorter for PDT than surgical tracheostomy: absolute difference with 95% CI, 9. 84 min (7.83 to 10.85 min). There was no difference comparing PDT and surgical tracheostomy with respect to overall operative complication rates: OR with 95% CI, 0.732 (0.05 to 9.37). However, relative to surgical tracheostomy, PDT was associated with less perioperative bleeding (OR with 95% CI, 0.14 [0.02 to 0.39]), a lower overall postoperative complication rate (OR with 95% CI, 0.14 [0.07 to 0.29]), as well as a lower postoperative incidence of bleeding (OR with 95% CI, 0.39 [0.17 to 0.88]), and stomal infection (OR with 95% CI, 0.02 [0.01 to 0.07]). No difference was identified in days intubated prior to tracheostomy (absolute difference with 95% CI, 0.16 days [- 0.9 to 1.22 days]), overall procedure-related complications (OR with 95% CI, 0.73 [0.06 to 9.37]), or death (OR with 95% CI, 0.63 [0.18 to 2.20]) comparing these two techniques.

CONCLUSIONS: Despite its popularity, there are currently only a limited number of small studies prospectively evaluating PDT and surgical tracheostomy. Our meta-analysis of these studies suggests potential advantages of PDT relative to surgical tracheostomy, including ease of performance, and lower incidence of peristomal bleeding and postoperative infection. If confirmed by additional, adequately powered prospective trials, these findings support PDT as the procedure of choice for the establishment of elective tracheostomy in the appropriately selected critically ill patient.