Web site designed and maintained by Chris Cairns  © SICS EBM Group 2004                                  

Early tracheostomy 1

Citation: Rumbak MJ, Newton M, Truncale T, Schwartz SW, Adams JW, Hazard PB. A prospective, randomised study comparing early percutaneous dilational tracheotomy to prolonged translaryngeal intubation (delayed tracheotomy) in critically ill medical patients. Critical Care Medicine 2004; 32: 1689 – 1694.
 

Lead author's name and fax: Rumbak MJ. dmrumcake@aol.com

Three-part Clinical Question: Does early tracheostomy (2 days) in medical intensive care patients projected to require prolonged ventilation improve mortality; incidence of pneumonia and number of ventilator days when compared to delayed tracheostomy (14 days)?

Search Terms: This is one of the five studies that contributed to the meta-analysis of Griffiths et al.

The Study: Non-blinded concealed randomised controlled trial with intention-to-treat.
 

The Study Patients: Intubated medical intensive care patients with acute respiratory failure who were projected to need mechanical ventilation for more than 14 days. Patients were in three intensive care units in the United States. Patients with various contraindications for percutaneous tracheostomy were excluded. Projection of who would require a tracheostomy appears to be subjective.

Control group (N = 60; 60 analysed): Control patients underwent standard intensive care with the expectation of proceeding to percutaneous tracheostomy at 10-14 days. Ventilation, weaning and sedation were protocol driven.

 Experimental group (N = 60; 60 analysed): These patients proceeded to percutaneous tracheostomy after 48 hours of mechanical ventilation. This was under bronchoscopic control with experienced practitioners. Sedation was stopped 24 hours after the tracheostomy was sited.

The Evidence: (Control is the late tracheostomy group).

Comments and EBM questions:

1) Do the methods allow accurate testing of the hypothesis?

Yes. However the study compares early versus late tracheostomy, not prolonged trans-laryngeal intubation as suggested in the title. Of the five studies that contributed to the meta-analysis, this is the most meticulous in design and presentation. The projection of who needed a tracheostomy appears to have been subjective. Of those randomised for late tracheostomy, 50 of 60 proceeded to it in the end. The diagnosis of pneumonia rested on bronchoscopically obtained cultures. The diagnosis of pneumonia does not appear to have been independent of the investigators and may be subject to detection bias. The ICU stay needs to be interpreted with care since there was a separate weaning floor, and again, transfer to this unit may be prone to bias in this unblinded study. Baseline patient characteristics were presented in detail, and the two arms of the study appear well matched.

 

2)  Do the statistical tests correctly test the results to allow differentiation of statistically significant results?

Yes. The study was powered for a 50% reduction in the incidence of pneumonia with a baseline incidence of 50%. Dichotomous variables were assessed with the chi-squared test. Time based outcomes were compared with linear regression. Intention to treat is implied. There was real randomisation. Risk of death was adjusted with logistic regression to account for any baseline differences.

3)   Are conclusions valid in light of the results?

Yes, although as commented by the authors, this represents only a specific patient group: medical patients with an admission APACHE II score of >25 and where prolonged ventilation was anticipated. The authors did not state if any medical patients not projected to need prolonged ventilation subsequently proceeded to tracheostomy. The reason for an APACHE II score of >25 was not made clear. The authors also carried out bronchoscopic evaluation of the trachea in 100% of survivors at 10 weeks and did not find a difference between the groups. For this, however, the power was insufficient, and as they discussed, this may be prone to type II error. There was less orolabial injury in the early tracheostomy group.
 

4)   Did results get omitted, and why? No.

5)   Did they suggest areas of further research?

No. It may be worth considering if the outcome difference was due to a sedation effect, and this also applies to other studies on this topic.

6)   Did they make any recommendations based on the results and were they appropriate?

Yes. That the risk / benefit is in favour of early tracheostomy, and this appears to be appropriate in this patient group.

7)   Is the study relevant to my clinical practice?

Largely. Tracheostomy and prolonged ventilation are pertinent issues on the intensive care unit. These were sick medical patients only. There are similarities and differences between US and UK practices. The depth of sedation in control patients is unclear except “sedated but arousable” during orotracheal intubation. A daily sedation holiday allowed a T-piece trial “if the patients had a good level of consciousness”. There was a 10% self-extubation rate, which seems high. The major difference was the transfer of patients to a weaning unit after tracheostomy. Deaths in the control (late) group began to occur more frequently after tracheostomy – this is shown in the Kaplan-Meir curves. We can only speculate on the reasons for this. Deaths were mostly attributed to VAP, respiratory failure and septic shock.

8)   What level of evidence does this study represent?1⁺: RCT with low risk of bias.

9)   What grade of recommendation can I make on this result alone? B

10) What grade of recommendation can I make when this study is considered along with other available evidence?

This is the only prospective randomised study of its type in this patient group. There are mixed results from other patient groups and these have been subject to meta-analysis.

11) Should I change my practice because of these results?

Yes. It may be reasonable to proceed to early tracheostomy in this group of sick medical patients. The study demonstrates that early tracheostomy, in experienced hands is safe, and leads to a marked improvement in outcome. It also challenges the current practice of performing tracheotomy at 10-14 days. The accompanying editorial points out that the current 10-14 days practice is not evidence based.

12) Should I audit my current practice because of these results?

Yes. The timing of tracheotomy for various patient groups and also, the early use of tracheotomy tubes with interchangeable inner cannulae – as used in the study.

Appraised by: Richard Price; SpR Stirling Royal Infirmary. 21 October 2004. Edited 2^nd May 2006.
Email: rjp@doctors.org.uk

Kill or Update By: May 2011.

Printer friendly view 

©SICS EBM 2006