Web site designed and maintained by
Chris Cairns
© SICS EBM Group 2004
|
|
|
|
Percutaneous Tracheostomy
Principal Investigator: Dr Richard Price
Early tracheostomy: A review of the available evidence from a meta-analysis and its composite controlled studies.
Current State: Complete (July 2006)
Early tracheostomy: A review of the available evidence from a meta-analysis and its composite controlled studies.
Bottom Line: Early tracheostomy may be of benefit for ventilated patients, but this needs to be subject to rigorous randomised controlled trials.
Investigators: Drs RJ Price, D Swann and M Hughes on behalf of the SICS EBM group.
Correspondence: Richard Price, Spr, West of Scotland Anaesthetic Rotation: rjp@doctors.org.uk
Citation: EBM Reviews. Scottish Intensive Care Society EBM Group. Price RJ, Swann D, Hughes M. Early tracheostomy: A review of the available evidence from a meta-analysis and its composite controlled studies. June 2006.
Background: Tracheostomy is commonly performed in the ICU and has a number of perceived advantages. It is unclear, however, if the timing of tracheostomy has any effect on outcomes such as death; ventilation duration; incidence of pneumonia and ICU stay. There have been a large number of studies looking into this. Recently, the randomised controlled trials of Rumbak et al and Bouderka et al, and a meta-analysis by Griffiths et al have been published.
Objectives: to critically appraise the available evidence for early tracheostomy.
Search strategy and study selection: We reviewed the meta-analysis of Griffiths et al, and the individual studies that contributed to this.
Main Conclusions: Five composite studies looked at early tracheostomy. These were the only reasonable randomised controlled trials identified in the English language literature. Each study only looked at a selected group of patients rather than a heterogenous intensive care population. Each study had limitations. Overall, we concluded that three of these five studies had significant sources of bias, and we graded them as evidence level 1-. Two studies were methodologically sounder and we graded them as evidence level 1+. We considered that one study had such limitations that it should not have been included in the meta-analysis. One study did not address early vs late tracheostomy at all. Not all studies reported all outcomes; the total patient population was 406.
For each study, if we ask the question “is the only difference between the groups the intervention” then it is difficult to say “yes” with certainty for any of them.
The outcomes looked at are addressed more fully in the CAT of the meta-analysis and the individual studies. However, it may be possible to conclude:
Implications for future studies: Multi-centre RCTs are currently underway to address this issue. Taking lessons from these earlier trials, these studies could reduce sources of bias in several ways. At least, we should ask these questions when reviewing the results.
How were the patients selected to enter the trial? Were there any objective criteria? How many patients received a tracheostomy that were not in the trial?
Were the patients who kept their trans-laryngeal tube treated the same? Were they awake? Was there a comparable protocol for ventilation and weaning?
How was pneumonia diagnosed? Were the criteria accepted and appropriate? Was it independent of the investigators? Which strategies were used to prevent VAP?
Surely these studies need to have three arms: early tracheostomy, late tracheostomy and no tracheostomy? One potential problem with a multi-centre pragmatic study is that many of these factors are uncontrollable by the study design. Unlike a drug study there can be neither blinding nor placebo control. It will be unlikely that control and intervention groups are treated in comparable ways.
References and links to CATS:
Formatted by CC
©SICS EBM 2006
|
