Early Versus Late Tracheostomy On Patients Undergoing Prolinged Mechanical Ventilation: A Meta Analysis

Objective：In recent years, some research have focused on the clinicalsignificance between early and late tracheostomy in critically ill patientswho are expected to require prolonged mechanical ventilation. Butcontroversy still exists on mortality, the duration of mechanical ventilationand so on. Our study aims to assess the different clinical significancebetween early and late tracheostomy in critically ill patients who areexpected to require prolonged mechanical ventilation, and provide a basisfor the treatment of critically ill patients to some extent.Methods：An electronic search (last search: June1,2013) was conducted usinguniform requirements and the clinical data satisfied the inclusion criteriawere extracted. The modified Jadad score was used to evaluated the qualityof literature. Statistical analysis was performed by Revman5.2.Therelevant graphs were draw and the merge results were described by forest flots.Results：Ten RCTs (including2,121adult patients, of whom LT wereconducted on1,059) were extracted. Six RCTs are high quality studies withJadad scores of above four points, four RCTs are low quality studies. In ourmeta analysis, the main merged outcome turned out that ET groups didn’talter the short term mortality when it was compared with the LT groups（RR=0.85,95%CI=0.62to1.03; P=0.07>0.05）、long term mortality（RR=0.96,95%CI=0.85to1.08; P=0.47>0.05）、VAP（RR=0.88,95%CI=0.71to1.10; P=0.27>0.05）、complications（RR=1.03,95%CI=0.86to1.24; P=0.73>0.05）and duration of MV（WMD=2.04,95%CI=5.85to1.77; P=0.29>0.05）. When the four low quality studies are excluded, Asimilar findings was observed regarding short term mortality（RR=0.88,95%CI=0.75to1.02; P=0.09>0.05）as the same as VAP（RR=0.86,95%CI=0.60to1.21; P=0.38>0.05）、complications or side effects（RR=1.08,95%CI=0.88to1.32; P=0.46>0.05）；but ET can short the duration of MV（RR=1.64,95%CI=2.86to0.41; P=0.0009<0.05）. ET, however,significantly increased the successful weaning rate (RR=1.18;95%CI=1.05to1.31; P=0.004) and ICU discharge rate (RR=1.27;95%CI=1.06to1.53; P=0.009<0.05). The Nfs0.05of long term mortality, rate ofsuccessful weaning, rate of ICU discharge was respectively96.58,38.25,25.52. The Nfs0.05of MV duration was72.44. Conclusion：（1）ET could not obviously reduce the mortality，the rate of VAP andcomplications or side effects.（2）ET can shorten the duration of MV.（3）ET significantly increased the rate of successful weaning and the rate ofICU discharge.