Early tracheostomy doesn't improve outcomes ... much (Meta-analysis, CHEST)
[poll id="3"] A 2005 meta-analysis of 5 studies (n=406) concluded that early tracheostomy reduced need for mechanical ventilation and ICU days. But then a 2006 randomized trial in trauma patients found no benefit to early trach, and an underpowered 2008 RCT also found no benefit. In a new meta-analysis and systematic review of 7 trials (n=1,044), Fei Wang et al take a fresh look including the new data, and conclude that early tracheostomy doesn't improve clinical outcomes in critically ill patients:
Short-term mortality, relative risk 0.86 (95% CI 0.65 - 1.13)
Long-term mortality, RR 0.84 (CI 0.68 - 1.04)
Incidence of VAP, RR 0.94 (CI 0.77 - 1.15)
They further concluded that early trach's did not "markedly" improve other important outcomes, noting that those getting early tracheostomy had:
4 more ventilator-free days (CI -9 days to +2 days)
7 fewer days of sedation (-15 days to +0.5 days)
7 fewer ICU days (-17 to +3 days)
1.45 fewer days in the hospital (-5 to +8 days)
Although these do look like markedly good trends, the wide confidence intervals preclude a definite conclusion of benefit in this analysis. (It's commonly accepted that trach's are more comfortable than endotracheal intubati0n, and reduce need for sedation; these trends validate that belief, to me.) A week's less time in an ICU would be great -- unless of course that week is your first week in the LTAC your ICU team shipped you to. In at least one of the included trials, such a transfer reduced the number of ICU days counted, contaminating this statistic. (Since weaning rates in LTACs are ~30-50% and 1 year mortality ~50-70% [ref], transfer to one should in no way be construed as a good outcome or successful weaning). Authors recommend -- you guessed it -- an adequately powered randomized controlled trial to answer the question "when to trach?" more definitively.
Clinical Takeaway: Early tracheotomy reduces need for sedation, and there's a suggestion (but no good proof) early trachs reduce time on the ventilator. However, in almost every randomized trial, a large proportion of early tracheotomies were unnecessary or futile (because many patients in the late-trach groups weaned from the ventilator or died before "needing" a tracheotomy).
If you consider tracheotomy a safe, benign procedure that enhances patient comfort and ICU throughput with minimal risk, trach early and often. If you consider tracheotomy rather brutal, believe its real-world risks are understated, or just feel that even low-risk procedures should be avoided where possible, wait to trach. And if you care what the lawyers are saying about it, read this.
Wang F et al. The timing of tracheotomy in critically ill patients undergoing mechanical ventilation: a systematic review and meta-analysis of randomized controlled trials. CHEST 2011; ePub September 22, 2011.