Effect of Timing of Tracheotomy on Clinical Outcomes:an Update Meta-analysis Including 11 Trials

Liang Shan*,Rui Zhang,and Lian-di Li

Neurological Intensive Care Unit,the Affiliated Hospital of Medical College,Qingdao University,Qingdao,Shandong 266003,China

THE reasons of artificial airway for most of critically ill patients were mainly due to prolonged mechanical ventilation (MV) or airway protection,such as neurological patients.Comparing with endotracheal intubation,tracheotomy enhanced patient comfort and had shorter intensive care unit (ICU) or hospital stay,which may have supportive data.The other possible advantages of procedure were lowering airway resistance,improving pulmonary toilet and fewer lung infection,benefit for oral feeding,which might also be associated with less sedative administration.1,2

Percutaneous dilatation tracheotomy (PDT) with or without bronchoscopic control has been used widely and increasingly because it is easy and convenient implementation since 1985.3-5Comparing the method of tracheotomy itself,in spite of surgical or percutaneous techniques,the timing of tracheotomy was more important in terms of the outcomes.But the appropriate time of diverting artificial airway from translaryngeal endotracheal intubation to tracheotomy remained controversial,and also there was no consensus in the literature about what should be considered as an early tracheotomy.

In 1989,The National Association of Medical Directors of Respiratory Care has published a recommendation based on only expert opinion that tracheotomy should be performed in patients who still required artificial ventilation 21 days after admission.6But the clinicians often made the selection for tracheotomy timing according to their individual favors in fact.In the two retrospective observational studies,median time from onset of MV to tracheotomy was 12 days.7,8In the past decade well designed observational studies or randomized control trials (RCTs) have been completed gradually,so it was necessary to pull all studies available together and find the exact answer again.Without a large Meta-analysis,the question of whether early tracheotomy should be considered after MV could not be well addressed.Thus,we conducted a Meta-analysis of results of studies available comparing different time of tracheotomy on clinical impact and exploring the potential benefit of early tracheotomy on adult patients admitted to ICU who required prolonged MV.

MATERIALS AND METHODS

Search source and strategy

A systematic Medline and PubMed database search was performed (up to February 27,2013).Furthermore,other databases were also checked including the China National Knowledge Internet,EMBASE,J-STAGE,the Cochrane Library (Cochrane Central Register of Controlled Trials),International Pharmaceutical Abstracts,ISI Web of Science,Global Health,and the grey literature (SIGLE) databases.

The following key literatures were used to search studies of interest∶[tracheotomy]AND [artificial respiration OR mechanical ventilation]with no restriction or subheadings.The reference lists of all retrieved literatures were checked for other potentially relevant citations.

Selection criteria

Studies were included in the present Meta-analysis if they met all the following criteria∶(1) critically ill patients enrolled underwent MV;(2) they compared early tracheotomy with late tracheotomy or prolonged endotracheal intubation;(3)they were written in English.Both RCTs and observational studies were included to increase sample size,population coverage,and statistical power.Identification of early tracheotomy was according to individual definition.

Quality assessment and data extraction

We assessed the quality of the studies included with a well-established scale developed by Jadadet al.9The range of possible scores were 0-5.The two authors selected trials according to the inclusion criteria independently and blindly.Then they assessed trial quality independently and extracted information from published reports by using a standardized protocol and reporting form∶study design,first author’ first name,year of publication,number of enrolled patients,timing of tracheotomy,events of end-points.If there were several mortality endpoints,we selected the longer follow-up results.For example,in the ICU mortality,hospital mortality,90-day mortality,or one-year mortality,we chose oneyear mortality as the final endpoint event.Diagnosis criteria of hospital pneumonia depended on respective definition.Disagreement was resolved by discussion and arbitration by the third author (Lian-di Li) if necessary.

Statistical analysis

RevMan 5.0.25 software,developed by the Cochrane Collaboration (http∶//www.cc-ims.net/revman),was used for the Meta-analysis.The heterogeneity between selected articles was tested with use ofI2andchi-squaretests.We applied the fixed-effects model when there was no statistically significant difference between the results and the random-effects was applied when there was a significant difference.The publication bias was evaluated by using funnel plots and the fail-safe number (Nfs).Any calculated Nfsvalue smaller than the number of observed studies indicated publication bias that might influence the Meta-analysis results.We calculated the Nfs0.05according to the formula Nfs0.05=(∑Z/1.64)2–k,wherekis the number of reports of studies included in the Meta-analysis.P＜0.05 was considered as significant.

RESULTS

Through the MEDLINE electronic database search,a total of nine RCTs were found and three of them were excluded because the patients enrolled did not receive MV.10-18Figure 1 summarizes the selection of reports of eligible clinical trials.Seven retrospective observational studies were found and five of them were included.19-23The other two retrospective studies conducted by Combes Alain and Arabi Yaseen M were excluded because they compared tracheotomy and non-tracheotomy and the average time of procedure was 12 days.7,8A total of 13 705 subjects were enrolled in the studies,of which 1034 were from six RCTs11-13,15,17,18and 12 671 from five observational studies.19-23The quality assessment,detailed characteristics of subjects,and therapy information of the eligible studies are given in Table 1.

No heterogeneity with regard to mortality outcome was found in ten studies (P＞0.39,andI2=1%),so a fixed effects model was used for analysis for this variable.However,heterogeneity was found in terms of hospital acquired pneumonia,length of ICU stay,and duration of MV,so a random effects model was then used for their analysis.

We could compare data for four clinical end-points.The risk of mortality,length of ICU stay,and duration of MV differed between early tracheotomy group and late tracheotomy group (Figs.2,3,4).

Figure 1.Flowchart of literature selection process.

Table 1.Chief characteristics of studies included in the Meta-analysis

Primary endpoints

Mortality was the primary endpoint event and reviewed for ten trials.11-13,15,17,18,20-23The risk of mortality in early tracheotomy group was significantly lower than that in late tracheotomy [33.3%vs.36.3%,respectively;relative risk (RR)∶0.92,95% confidence interval (CI)∶0.88,0.97;P=0.004;Fig.2,Table 2].To exclude impact of potential confounding factors related to mortality,we analyzed each cluster of exposed and not exposed patients to allow an accurate estimation(Table 3).Study design and time of tracheotomy influenced the outcome of mortality.Observational study,performing the procedure and PDT had more possibility to induce the positive results.

Secondary endpoints

Secondary endpoints were length of ICU stay,duration of MV and hospital pneumonia.The length of ICU stay reviewed seven trials11-13,18,19,22,23in the early tracheotomy group was shorter than that in the late group (mean difference∶-6.55 days;95%CI∶-8.19,-4.90;P＜0.00001;Fig.3,Table 2).The duration of MV reviewed six trials12,13,18,19,22,23in the early tracheotomy group was also shorter comparing the late tracheotomy group (mean difference∶-6.53 days;95%CI∶-11.43,-1.63;P=0.009;Fig.4,Table 2).There was no significant difference on hospital acquired pneumonia between early tracheotomy and late tracheotomy in nine trials (21.9%vs.21.0%;RR∶0.85,95%CI∶0.68,1.06;P=0.15;Fig.5,Table 2).11-13,15,17,18,21-23

Sensitivity/subgroup analyses

In the sensitivity analyses,removal of the largest study21produced no substantial alterations in pooledRRs in terms of mortality (data not shown).

The between-study heterogeneity was explained in part by the variability in study design,type of technique and time of tracheotomy,which had not been well addressed by sensitivity analyses.So we performed subgroup analyses respectively with regard to the end-point of mortality(Table 3).

Publication bias

The funnel plot of the publication bias of the ten trials in terms of mortality events (Fig.6).The funnel plot showed no publication bias for studies included in the Meta-analysis in terms of mortality.We calculated the Nfs0.05for each end-point event and found the Nfs0.05values for length of ICU stay (Nfs0.05=9619),duration of MV (Nfs0.05=12 114),hospital pneumonia (Nfs0.05=22) and mortality (Nfs0.05=46)to be greater than the number of studies included in the Meta-analysis.

Figure 2.Meta-analysis of 6 RCT and 4 observational studies estimating the impact of early tracheotomy on mortality event comparing late tracheotomy.

Figure 3.Meta-analysis of 4 RCT and 3 observational studies estimating the impact of early tracheotomy on length of ICU stay comparing late tracheotomy.

Table 2.Meta-analysis of risk of clinical endpoint events

Table 3.Subgroup analyses with regard to the risk of mortality

Figure 5.Meta-analysis of 6 RCT and 3 observational studies estimating the impact of early tracheotomy on hospital acquired pneumonia event comparing late tracheotomy.

Figure 6.Funnel plot of publication bias in terms of mortality.

DISCUSSION

Our Meta-analysis pooled all RCTs and observational studies together for analysis and revealed that earlytracheotomy not only reduced duration of MV and length of ICU stay,but also decreased the incidence of overall mortality which considerably differed with the previous Meta-analysis results.In 2005 Griffiths and his colleagues found in a Meta-analysis that an earlier tracheotomy had some advantages such as reduced duration of artificial ventilation and length of stay in adult patients undergoing artificial ventilation,but the hospital mortality and pneumonia did not differ between early and late-tracheotomy.24Wanget al25also found negative results about outcomes of early tracheotomy.While in our analysis,there was significant difference with regard to mortality and this was the most interesting discovery.Our study enrolled larger sample of subjects including 13 705 patients which enlarged the accuracy of results.The six of all ten trials including in our analysis published after 2005,so in a way the trials enrolled were advanced and well designed,and also represented the new status.RCTs were generally lower sample and well designed,but that was the idealization which hardly easily being performed in fact.Observational studies commonly had larger subject samples and reflected real clinical status.The decision of procedure often depends on clinical team’ discussion and judgment.In this analysis,observational studies had more positive results.Subgroup analysis performed subsequently in terms of mortality showed that early tracheotomy had a trend of decreased mortality in six RCTs trials,and in observational studies including four trials showed an absolute advantage of reducing mortality.

Another favorable outcome well being recognized was lower duration of MV and length of ICU stay,which also mean reduced ICU resources and cost-saving.Several physiologic factors explained why early tracheotomy may facilitate weaning,such as reduced dead space,less airway resistance,decreased work of breathing,better secretion removal by suctioning,less likelihood of tube obstruction,and better glottic function.Clinicians’ behavior also affected discontinuation of respiratory support when encountered different patients of prolonged endotracheal intubation or tracheotomy,and the more rapid weaning were often seen in the latter.2,26

We found that incidence of hospital pneumonia was not affected by tracheotomy itself or time of that,which was in accordance with the previous three Meta-analyses reports.24,25,27Performing a tracheotomy itself may potentially lead to an increased incidence of pneumonia because it destroyed the structure and protection role of natural airway.In ICU circumstance,frequent invasive handling or contaminating air by pathogenic microorganism may counteract the benefit of tracheotomy such as pulmonary toilet or convenient nursing care.Other possible complications regarding tracheotomy included incision bleeding,laryngeal or tracheal injury,increased nursing care,et al.However,these outcomes mentioned above were not assessed because of unsatisfactory data collection.Several studies paying attention to these problems found that there were significant more oral-labial and laryngeal damages in the prolonged translaryngeal intubation patients.11,13

Generally performing a tracheotomy was considered as early tracheotomy if the procedure occurred ＜ 7 days after artificial ventilation,but the exact time remained debated.Subgroup analyses for timing of tracheotomy showed us that more earlier procedure (＜4 days) did not show additional benefits and 5-7 days after MV may be an appropriate time for patients in whom weaning and extubation was not likely before day 14 (Table 3).1

The main limitation of the present study was heterogeneity.Although both RCTs and observational study were enrolled together to enlarge the sample of study,which also resulted significant heterogeneity.Apart from this,another heterogeneity was that we studied targeted population included trauma,burn,medical,and surgery patients,which may mask some specific benefits and limit generalizability of results.Finally,there was different definition about early tracheotomy which varied greatly from immediate to day 8 after intubation.Considering items mentioned above,interpretation of results should be cautioned.

Our Meta-analysis results may be significant and helpful for ICU physicians when considering this procedure.Further investigation focusing on subgroup-specific population,for instance,neurological critically ill patients,should be advocated.

1.De Leyn P,Bedert L,Delcroix M,et al.Tracheotomy∶Clinical review and guidelines.Eur J Cardiothorac Surg 2007;32∶412-21.

2.Durbin CG Jr.Tracheostomy∶Why,when,and how.Respir Care 2010;55∶1056-68.

3.Fikkers BG,Fransen GA,der Hoeven JG v,et al.Tracheostomy for long-term ventilated patients∶A postal survey of ICU practice in the Netherlands.Intensive Care Med 2003;29∶1390-3.

4.Cox CE,Carson SS,Holmes GM,et al.Increase in tracheostomy for prolonged mechanical ventilation in North Carolina,1993-2002.Crit Care Med 2004;32∶2219-26.

5.Anon JM,Escuela MP,Gomez V,et al.Use of percutaneous tracheostomy in intensive care units in Spain.Results of a national survey.Intensive Care Med 2004;30∶1212-5.

6.Plummer AL,Gracey DR.Consensus conference on artificial airways in patients receiving mechanical ventilation.Chest 1989;96∶178-80.

7.Combes A,Luyt CE,Nieszkowska A,et al.Is tracheostomy associated with better outcomes for patients requiring long-term mechanical ventilation.Crit Care Med 2007;35∶802-7.

8.Arabi YM,Alhashemi JA,Tamim HM,et al.The impact of time to tracheostomy on mechanical ventilation duration,length of stay,and mortality in intensive care unit patients.J Crit Care 2009;24∶435-40.

9.Jadad AR,Moore RA,Carroll D,et al.Assessing the quality of reports of randomized clinical trials∶Is blinding necessary.Control Clin Trials 1996;17∶1-12.

10.Dunham CM,LaMonica C.Prolonged tracheal intubation in the trauma patient.J Trauma 1984;24∶120-4.

11.Sugerman HJ,Wolfe L,Pasquale MD,et al.Multicenter,randomized,prospective trial of early tracheostomy.J Trauma 1997;43∶741-7.

12.Saffle JR,Morris SE,Edelman L.Early tracheostomy does not improve outcome in burn patients.J Burn Care Rehabil 2002;23∶431-8.

13.Rumbak MJ,Newton M,Truncale T,et al.A prospective,randomized,study comparing early percutaneous dilational tracheotomy to prolonged translaryngeal intubation(delayed tracheotomy) in critically ill medical patients.Crit Care Med 2004;32∶1689-94.

14.Bouderka MA,Fakhir B,Bouaggad A,et al.Early tracheostomyversusprolonged endotracheal intubation in severe head injury.J Trauma 2004;57∶251-4.

15.Blot F,Similowski T,Trouillet JL,et al.Early tracheotomyversusprolonged endotracheal intubation in unselected severely ill ICU patients.Intensive Care Med 2008;34∶1779-87.

16.Schneider GT,Christensen N,Doerr TD.Early tracheotomy in elderly patients results in less ventilator-associated pneumonia.Otolaryngol Head Neck Surg 2009;140∶250-5.

17.Terragni PP,Antonelli M,Fumagalli R,et al.Earlyvs.late tracheotomy for prevention of pneumonia in mechanically ventilated adult ICU patients∶a randomized controlled trial.JAMA 2010;303∶1483-9.

18.Trouillet JL,Luyt CE,Guiguet M,et al.Early percutaneous tracheotomyversusprolonged intubation of mechanically ventilated patients after cardiac surgery∶A randomized trial.Ann Intern Med 2011;154∶373-83.

19.Moller MG,Slaikeu JD,Bonelli P,et al.Early tracheostomyversuslate tracheostomy in the surgical intensive care unit.Am J Surg 2005;189∶293-6.

20.Flaatten H,Gjerde S,Heimdal JH,et al.The effect of tracheostomy on outcome in intensive care unit patients.Acta Anaesthesiol Scand 2006;50∶92-8.

21.Scales DC,Thiruchelvam D,Kiss A,et al.The effect of tracheostomy timing during critical illness on long-term survival.Crit Care Med 2008;36∶2547-57.

22.Zagli G,Linden M,Spina R,et al.Early tracheostomy in intensive care unit∶A retrospective study of 506 cases of video-guided Ciaglia Blue Rhino tracheostomies.J Trauma 2010;68∶367-72.

23.Tong CC,Kleinberger AJ,Paolino J,et al.Tracheotomy timing and outcomes in the critically ill.Otolaryngol Head Neck Surg 2012;147∶44-51.

24.Griffiths J,Barber VS,Morgan L,et al.Systematic review and Meta-analysis of studies of the timing of tracheostomy in adult patients undergoing artificial ventilation.BMJ 2005;330∶1243.

25.Wang F,Wu Y,Bo L,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;140∶1456-65.

26.Pierson DJ.Tracheostomy and weaning.Respir Care 2005;50∶526-33.

27.Dunham CM,Ransom KJ.Assessment of early tracheostomy in trauma patients∶A systematic review and meta-analysis.Am Surg 2006;72∶276-81.