early versus late tracheostomy for critically ill patients · timing of tracheostomy for critically...

Early versus late tracheostomy for critically ill patients

(Review)

Gomes Silva BN, Andriolo RB, Saconato H, Atallah ÁN, Valente O

This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library2012, Issue 3

http://www.thecochranelibrary.com

Early versus late tracheostomy for critically ill patients (Review)

Copyright © 2012 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

http://www.thecochranelibrary.com

T A B L E O F C O N T E N T S

1HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2SUMMARY OF FINDINGS FOR THE MAIN COMPARISON . . . . . . . . . . . . . . . . . . .5BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Figure 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Figure 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

13DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Analysis 1.1. Comparison 1 Early versus late tracheostomy, Outcome 1 Mortality at maximal follow-up time available. 32Analysis 1.2. Comparison 1 Early versus late tracheostomy, Outcome 2 Pneumonia. . . . . . . . . . . . . 33

33ADDITIONAL TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43DIFFERENCES BETWEEN PROTOCOL AND REVIEW . . . . . . . . . . . . . . . . . . . . .

iEarly versus late tracheostomy for critically ill patients (Review)


[Intervention Review]


Brenda Nazaré Gomes Silva1, Régis B Andriolo2, Humberto Saconato3 , Álvaro N Atallah4, Orsine Valente1

1Brazilian Cochrane Centre, Universidade Federal de São Paulo, São Paulo, Brazil. 2Department of Public Health, Universidade doEstado do Pará, Belém, Brazil. 3Department of Medicine, Santa Casa de Campo Mourão, Campo Mourão, Brazil. 4Brazilian CochraneCentre, Universidade Federal de São Paulo / Escola Paulista de Medicina, São Paulo, Brazil

Contact address: Brenda Nazaré Gomes Silva, Brazilian Cochrane Centre, Universidade Federal de São Paulo, Rua Pedro de Toledo,598, Vl. Clementino, São Paulo, São Paulo, 04039-001, Brazil. [email protected].

Editorial group: Cochrane Anaesthesia Group.Publication status and date: New, published in Issue 3, 2012.Review content assessed as up-to-date: 14 December 2010.

Citation: Gomes Silva BN, Andriolo RB, Saconato H, Atallah ÁN, Valente O. Early versus late tracheostomy for critically ill patients.Cochrane Database of Systematic Reviews 2012, Issue 3. Art. No.: CD007271. DOI: 10.1002/14651858.CD007271.pub2.


A B S T R A C T

Background

Long-term mechanical ventilation is the most common situation where tracheostomy is indicated for patients in intensive care units(ICU). ’Early’ and ’late’ tracheostomies are two categories of the timing of tracheostomy. The evidence on the advantages attributed toearly over late tracheostomy is somewhat conflicting but includes shorter hospital stays and lower mortality rates.

Objectives

To evaluate the effectiveness and safety of early (≤ 10 days after intubation) versus late tracheostomy (> 10 days after intubation) incritically ill adult patients predicted to be on prolonged mechanical ventilation and with different clinical conditions.

Search methods

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2010, Issue 12); MEDLINE (viaPubMed) (1966 to December 2010); EMBASE (via Ovid) (from 1974 to December 2010); LILACS (1986 to December 2010); PEDro(Physiotherapy Evidence Database) at www.pedro.fhs.usyd.edu.au (1999 to December 2010) and CINAHL (1982 to December 2010).

Selection criteria

We included all randomized or quasi-randomized controlled trials which compared early tracheostomy (two to10 days after intubation)against late tracheostomy (> 10 days after intubation) for critically ill adult patients expected to be on prolonged mechanical ventilation.There was no language restriction.

Data collection and analysis

Two authors extracted data and conducted a quality assessment. Meta-analyses using the random-effects model were conducted formortality and pneumonia.

Main results

We included four studies, with a high risk of bias, in which a total of 673 patients were randomized to either early or late tracheostomy.We could not pool data in a meta-analysis because of clinical, methodological and statistical heterogeneity between the included studies.There is no strong evidence for real differences between early and late tracheostomy in the primary outcome of mortality. In one studya statistically significant result favouring early tracheostomy was observed in the outcome measuring time spent on ventilatory support(mean difference (MD) -9.80 days, 95% CI -11.48 to -8.12, P < 0.001).

1Early versus late tracheostomy for critically ill patients (Review)


mailto:[email protected]

Authors’ conclusions

Updated evidence is of low quality, and potential differences between early and late tracheostomy need to be better investigated bymeans of randomized controlled trials. At present there is no specific information about any subgroup or individual characteristicspotentially associated with better outcomes with either early or late tracheostomy.

P L A I N L A N G U A G E S U M M A R Y

Timing of tracheostomy for critically ill patients who are predicted to be on long-term artificial respiration

Tracheostomy is a surgical procedure where an external artificial opening is made in the trachea (windpipe). Long-term mechanicalventilation (where a machine is used to mechanically assist breathing) is the most common situation where tracheostomy is indicated forpatients in intensive care units (ICU). ’Early’ and ’late’ tracheostomies may be undertaken, categorised by the timing of tracheostomy.We included four studies in this systematic review, with a total of 673 patients randomized to either an early or late tracheostomy. Therewas no significant difference between early and late tracheostomies for patient deaths (mortality). The more significant effects of earlytracheostomy were the reduction in time spent on ventilatory support (by around nine days) and in the intensive unit care (by around11 days), but even so the results were limited because they were observed in individual studies that had a high risk of bias (systematicerror) and a relatively low number of patients. There was no significant difference between the comparison groups with regards toundesired events such as tracheal stenosis (narrowing of the trachea or windpipe) in hospital and sepsis (blood infection) because ofrespiratory infection. Possible differences between early and late tracheostomy still need to be adequately investigated by means offurther randomized controlled trials. Moreover, there is no information about the best indication for either early or late tracheostomyin patients with specific characteristics.



S U M M A R Y O F F I N D I N G S F O R T H E M A I N C O M P A R I S O N [Explanation]

Early tracheostomy compared to late tracheostomy for critically ill patients

Patient or population: patients with critically ill patients

Settings:

Intervention: Early tracheostomy

Comparison: late tracheostomy

Outcomes Illustrative comparative risks* (95% CI) Relative effect

(95% CI)

No of Participants

(studies)

Quality of the evidence

(GRADE)

Comments

Assumed risk/mean dif-

ference

Corresponding risk

Late tracheostomy Early tracheostomy

Mortality at maximal fol-

low-up time available

Follow-up: 1-12 months

193/301

(64.1%)

148/298

(49.7%)

Not estimable 599

(3 studies)

⊕⊕©©

low1,2,3,4,5,6,7

Time spent on mechani-

cal ventilation8

days

Follow-up: 30-70 days

The mean time spent on

mechanical ventilation in

the control groups was

17.4 days8

The mean time spent on

mechanical ventilation in

the intervention groups

was

9.8 lower

(8.12 to 11.48 lower)8

120

(1 study)

⊕⊕⊕©

moderate1,3,5,6,7,9,10

Pneumonia

Follow-up: 30 days

59/270

(21.9%)

33/269

(12.3%)

Not estimable 539

(2 studies)

⊕⊕©©

low1,2,3,5,6,7,11

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the

assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval;

3E

arly

versu

sla

tetra

ch

eo

stom

yfo

rcritic

ally

illp

atie

nts

(Revie

w)

Co

pyrig

ht

©2012

Th

eC

och

ran

eC

olla

bo

ratio

n.P

ub

lished

by

Joh

nW

iley

&S

on

s,L

td.

http://www.thecochranelibrary.com/view/0/SummaryFindings.html

GRADE Working Group grades of evidence

High quality: Further research is very unlikely to change our confidence in the estimate of effect.

Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.

Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.

Very low quality: We are very uncertain about the estimate.

1 See ’Risk of bias’ table found in Characteristics of included studies section2 Heterogeneity between studies was considered substantial, but all of individual estimate effects from primary studies had the same

direction which favoured the early tracheostomy group3 All studies had compared early versus late tracheostomy for critically ill patients4 There was appreciable but substantially heterogeneous estimate effects favouring early as compared to late tracheostomy5 There was no reason to suspect of publication bias6 Several clinical and methodological characteristics are suspected to change the estimate effects, including the type and severity of

baseline diseases, age range, methods available to perform the tracheostomy, etc. However, the effects of such factors could not yet be

precisely quantified. More studies are necessary.7 The question of this systematic review about the effectiveness of early versus late tracheostomy for critically ill patients was based

on the potential association between earlier tracheostomy and lower probabilitIes of undesired but relevant outcomes such as death,

unsuccessful weaning from mechanical ventilation and pneumonia8 No explanation was provided9 Additional randomized controlled trials are necessary to either confirm or not the evidences available at the moment this review was

finished10 The estimate effect was clinically relevant as showed by the MD of -9.80 days that favoured the early group and statistically significant

as showed by the 95% CI of -11.48 to -8.12 (P <0.00001)11 The estimate effects were appreciable and had the same direction favouring the early tracheostomy group. However, the effects from

individual studies were substantially heterogeneous between each other (inconsistency test >50%, P <0.10)

4E

arly

versu

sla

tetra

ch

eo

stom

yfo

rcritic

ally

illp

atie

nts

(Revie

w)

Co

pyrig

ht

©2012

Th

eC

och

ran

eC

olla

bo

ratio

n.P

ub

lished

by

Joh

nW

iley

&S

on

s,L

td.

B A C K G R O U N D

Description of the condition

Long-term mechanical ventilation is the most common situationwhere tracheostomy is indicated for patients in intensive care units(ICU) (Heffner 2001). Although the definition of prolonged ven-tilation can include periods as short as 24 hours (Criner 1994;Griffiths 2005), only patients who are foreseen to be on artifi-cial ventilation for around 10 days or more (Armstrong 1998;Plummer 1989) are generally subjected to elective tracheostomy.In this circumstance, tracheostomy is offered as a strategy to re-duce respiratory injury and other undesired consequences of pro-longed translaryngeal intubation. These include ventilator-associ-ated pneumonia (Ranes 2006), sinusitis (Holzapfel 1993) and tra-cheal stenosis (Cavaliere 2007). Some predictive systems have beenused to predict the duration of mechanical ventilation in variouspatient settings (Agle 2006; Gajic 2007; Légaré 2001; Sellers1997)but many of these systems are not appropriately validated. Severalother factors have also been shown to provide indications for tra-cheostomy: neuromuscular disease or trauma, age, injury sever-ity score, damage control laparotomy, and others (Frutos-Vivar2005; Goettler 2006). Other researchers propose that the decisionto perform tracheostomy should be based on objective measuresfrom spontaneous breathing trials or weaning from mechanicalventilation trials (Freeman 2008). Thus, the development of pre-dictive methods tailored for each clinical condition would be amajor advance in patient care.

Description of the intervention

Tracheostomy is a surgical procedure where an external, artifi-cial opening is made in the trachea (Stedman 1995). A numberof techniques are used to perform tracheostomy. The main onesare the classical standard surgical procedure (in a surgical room)and the percutaneous method performed at the patient’s bedside(Friedman 2006; Gullo 2007; Pappas 2011; Schultz 2007). Thesurgical and percutaneous procedures are usually performed bydifferent surgical specialists such as general, thoracic, ear, nose andthroat (ENT) or maxillofacial surgeons, while the percutaneousprocedures are usually but not exclusively performed by surgeonsand intensivists (Pappas 2011; Plummer 1989). There is also adiversity of materials (equipment and designs) used in performingtracheostomy (Björling 2007; Crimlisk 2006; Hess 2005). Thesecan potentially be associated with complications such as trachealulceration, distortion of soft tracheal tissue and airway obstruction(Tibballs 2006).Plummer 1989 used the translaryngeal route for patients expectedto be on mechanical ventilation for up to 10 days and tracheostomyfor those on artificial ventilation for longer than 21 days; however,tracheostomy is usually performed between the 10th and 14th dayof intubation (Armstrong 1998). Nowadays there are conflicting

views about the best time for tracheostomy (Heffner 2003). Rel-evant studies vary in design and the clinical condition (Ahmed2007; Barquist 2006). To circumvent this the literature offers thetwo categories of ’early’ and ’late’ for the timing of tracheostomy.Unfortunately these categories are not precisely defined and au-thors use different times that they characterize as ’early’ and ’late’,which results in some overlap between the categories (Aissaoui2007; Barquist 2006; Dunham 2006; Lesnik 1992). There is alsoconflicting evidence about the advantages of early over late tra-cheostomies. For example, some comparative studies have shownshorter hospital stays, lower mortality rates and other benefits withthe use of early tracheostomy as compared to late tracheostomy(Arabi 2004; Rodriguez 1990). Conversely, Clec’h 2007 observedno difference in mortality in ICU between patients with early andlate tracheostomies.

How the intervention might work

Potential benefits from tracheostomy include: lower airway resis-tance, easier and safer tracheal suction, greater patient comfort,better communication, oral feeding, faster weaning from the venti-lator and lower rates of ventilator-associated pneumonia (Heffner2001; Plummer 1989). On the other hand, some of the disad-vantages of tracheostomy include: dislodgement or obstruction,wound infection, scarring, a false passage, haemorrhage and sub-glottic and tracheal stenosis (Bartels 1998; Dollner 2002; Higgins2007; Norwood 2000).

Why it is important to do this review

The present review intended to systematically map the evidenceon timing of tracheostomy (early versus late) in mechanically ven-tilated, critically ill patients.

O B J E C T I V E S

To evaluate the effectiveness and safety of early (≤ 10 days af-ter tracheal intubation) versus late tracheostomy (> 10 days aftertracheal intubation) in critically ill adult patients predicted to beon prolonged mechanical ventilation and with different clinicalconditions.

M E T H O D S

Criteria for considering studies for this review



Types of studies

We included all randomized or quasi-randomized controlled trials(RCTs) published in any language. We included studies publishedin abstract form if sufficient information regarding their methodsand results were provided. We approached the principal authorsfor additional information wherever necessary.

Types of participants

Inclusion criteria

• Critically ill patients (for whom death is possible orimminent)

• Patients expected to be on prolonged mechanical ventilation• Adults (≥ 18 years)

We defined prolonged mechanical ventilation as ventilation forfrom 24 hours to 21 consecutive days, six or more hours per day(Divo 2010; Shirzad 2010).

Exclusion criteria

• Anatomical anomalies of the neck which would impair thetracheostomy procedure

• Patients already tracheostomized• Coagulation disturbances (e.g., thrombocytopenia)• Soft tissue infection of neck

Types of interventions

We considered the following comparison arms.1. Early tracheostomy, if no serious attempt was made to wean

the patients from the ventilator (tracheostomy based only onclinical or laboratory results, performed from two days to 10days after intubation).

2. Late tracheostomy, where weaning had not been successful,performed later than 10 days after intubation.

Types of outcome measures

We considered all outcomes measures found in the primary studies.For each outcome we accepted the definition used by the studyauthors. We discussed, where necessary, limitations such as theuse of non-validated instruments for evaluation or a divergence ofdefinitions.

Primary outcomes

1. Mortality (time to mortality or frequency of deaths at anytime point: in hospital, in ICU, or after discharge)

2. Duration of artificial ventilation

Secondary outcomes

1. Length of stay in ICU (or frequency of tracheostomy at anytime point)

2. Ventilator-associated pneumonia at any time point3. Laryngotracheal lesions at any time point (in epiglottis,

vocal cord, larynx; subglottic ulceration and inflammation;stenosis)For details about definitions see Appendix 1 (Glossary of terms).

Search methods for identification of studies

Electronic searches

We searched the following electronic databases: Cochrane Cen-tral Register of Controlled Trials (CENTRAL) (The CochraneLibrary 2010, Issue 12); MEDLINE (via the PUBMED inter-face) (1966 to December 2010); EMBASE (via Ovid interface)(from 1974 to December 2010); LILACS (1986 to Decem-ber 2010); PEDro (Physiotherapy Evidence Database) at http://www.pedro.fhs.usyd.edu.au) (1999 to December 2010) and CIN-HAL (1982 to December 2010).The search strategy for MEDLINE included terms for clinical con-dition and intervention as well as their synonyms (see Appendix2). This strategy was modified as required for other databases (seeAppendix 3 (CENTRAL); Appendix 4 (EMBASE); Appendix 5(LILACS); Appendix 6 (Current Controlled trials) Appendix 7(PEDro); and Appendix 8 (CINAHL). We used a highly sensitivesearch filter for randomized controlled trials in databases where itwas necessary (MEDLINE, EMBASE and LILACS) to optimizethe search process (Higgins 2011b).We did not impose any language restriction.

Searching other resources

We handsearched the references of relevant articles including nar-rative reviews and non-randomized controlled studies in the areaof mechanical ventilation.We searched for ongoing randomized controlled trials in theCurrent Controlled Trials database at http://www.controlled-trials.com/.

Data collection and analysis

Selection of studies

Two authors (HS and BGS) independently analysed the titles andabstracts of publications obtained by the search strategy. We (RAand BGS) acquired all studies that met our inclusion criteria asfull text.



http://www.pedro.fhs.usyd.edu.au



http://www.controlled-trials.com/




Data extraction and management

We (RA and BGS) extracted data using a specially designed dataextraction sheet (Appendix 9) containing information about meth-ods (study design), participants, interventions (for example sur-gical procedures, materials) and results. We resolved all disagree-ments by consensus. We contacted the authors of the primarystudies for further information about methodology and partici-pants, when necessary. Two authors (RA and BGS) abstracted thedata and entered all data into Review Manager (RevMan 5.1). Athird author (HS) rechecked all entries.

Assessment of risk of bias in included studies

Two authors (RA and BGS) assessed all included studies formethodological quality based on the criteria described in theCochrane Handbook for Systematic Reviews of Interventions (Higgins2011a).

1. Was the random allocation sequence adequately generated?2. Was allocation adequately concealed?3. Was knowledge of the allocated interventions adequately

prevented for data collectors, or they were independent from theresearchers who planned the study (blinding)?

4. Were incomplete outcome data adequately addressed?5. Are reports of the study free of suggestion of selective

reporting?6. Was the study apparently free of other bias?

We classified each of the items as low risk of bias, high risk of biasor unclear risk of bias.Because of the nature of the interventions of interest for this sys-tematic review, we considered item 3 (blinding) only at the datacollection level.

Measures of treatment effect

For comparable studies, dichotomous data were expressed as rela-tive risk (RR) with 95% confidence intervals (CI) using the ran-dom-effects model (Deeks 2001a). We calculated number neededto treat (NNT) where risk differences were statistically significant(Christensen 2006). For continuous data we calculated the meandifference using the random-effects model. We planned to cal-culate the standardized mean difference when the trials assessedthe same outcome but used different instruments or scales (Deeks2001b).

Unit of analysis issues

We based the unit of analysis on the individual patient (unit to berandomized for interventions to be compared) (Higgins 2011a).We did not expect to find cross-over study designs because of thecharacteristics of the intervention.

Dealing with missing data

Irrespective of the type of data, we planned to report dropout ratesin the table Characteristics of included studies and perform in-tention-to-treat (ITT) analysis only for dichotomous data (Deeks2005).

Assessment of heterogeneity

We presented data using a random-effects model (DerSimonian1986). We quantified inconsistency among the pooled estimatesusing the Chi2 statistic, for heterogeneity we used the I2 statistic(where I2 = [(Q - df )/Q] x 100%; Q is the Chi2 statistic and dfits degrees of freedom) . This illustrates the percentage of the vari-ability in effect estimates resulting from heterogeneity rather thanfrom sampling error (Higgins 2002; Higgins 2003). We decidednot to combine the studies in a meta-analysis when they presentedsubstantial clinical and methodological heterogeneity in conjunc-tion with statistical heterogeneity as indicated by the I2 statistic,according to the following thresholds:

• 0% to 40%: may not be important;• 30% to 60%: may represent moderate heterogeneity;• 50% to 90%: may represent substantial heterogeneity;• 75% to 100%: considerable heterogeneity.

Assessment of reporting biases

We planned to assess publication bias or a systematic differencebetween the smaller and larger studies (small study effects) bypreparing a funnel plot (trial effect versus trial size) if a sufficientnumber of studies were available (Copas 2000).

Data synthesis

We synthesized qualitative information relative to methods, risk ofbias, description of participants and outcomes measures and pre-sented them in the table ’Characteristics of included studies’. Forquantitative data, we planned to use the random-effects model inthe meta-analysis because of substantial clinical and methodologi-cal heterogeneity between studies, which by themselves could gen-erate substantial statistical heterogeneity. When data from primarystudies were not parametric (for example effects were reported asmedians, quartiles, etc) or were without sufficient statistical infor-mation (for example standard deviations, number of patients, etc)we planned to insert them into an ’Appendix’. Additionally, eachclinically relevant estimate effect was presented in a Summary offindings for the main comparison (Schünemann 2009).

Subgroup analysis and investigation of heterogeneity

We planned to stratify our analysis by the following independentvariables that are expected to be associated with heterogeneity.

• By clinical condition (e.g., trauma, pre-existingneurological and lung diseases).



• Different timing of ’early’ and ’late’ tracheostomies.• Type of tracheostomy, such as percutaneous or surgical

tracheostomy.

We planned to only conduct these analyses if data were availableeither in the report or by contacting the main authors of the studies.In spite of the number of defined subgroup analyses, the eventualstatistical heterogeneity observed across subgroups would not beassumed to be a true causal relationship between dependent andindependent variables but only as generating a hypothesis to betested in future trials.

Sensitivity analysis

If there were an adequate number of studies, we planned to per-form a sensitivity analysis to explore the causes of heterogeneityand the robustness of the results. We planned to consider the fol-lowing factors in the sensitivity analysis: quality of allocation con-cealment (adequate or unclear or inadequate); blinding (adequateor unclear or inadequate or not performed); analysis using bothrandom-effects or fixed-effect models; intention-to-treat analysisand available case analysis (only for dichotomous data), and stud-ies with different times of early and late tracheostomies from ourinclusion criteria.We did not plan to include the results obtained from subgroup andsensitivity analyses as conclusions. We intended that they would

be used for hypothesis generation to be tested in future adequatelydesigned studies.

R E S U L T S

Description of studies

See: Characteristics of included studies; Characteristics of excludedstudies; Characteristics of ongoing studies.

Results of the search

The search strategy retrieved 1433 citations across all electronicdatabases. We excluded duplicate references and thus retrieved1219 unique citations. Of these citations, we excluded a further1134, on the basis of title and abstract, because they were notspecifically related to the ’timing of tracheostomy’. From the re-maining 85 studies, we excluded a further 74 because of their studydesign. Thus, 11 studies had the potential to be included in thereview (Figure 1). Of those 11 studies, we contacted the main au-thors of one study for further information about the comparisongroups (Blot 2008). This study was later excluded for the reasonsoutlined in the Characteristics of excluded studies.



Figure 1. Flow diagram showing the studies retrieved from databases to inclusion and exclusion.



At the title and abstract stage of selection, the Kappa coefficients(Kc) to evaluate the concordances between the two observers (RAand BNG) were calculated in the databases where there was atleast one discordance (Latour 1997). The concordance levels wereconsidered excellent for three databases: Kc = 0.91 (CENTRAL),Kc = 0.85 (EMBASE), Kc = 0.94 (MEDLINE); and good forCINAHL (Kc = 0.63). For the other databases there was no dis-cordance between the observers.

Included studies

We included four studies available in English language journals(Barquist 2006; Dunham 1984; Rumbak 2004; Terragni 2010),with a total of 673 patients randomized to either early or latetracheostomy. The studies were diverse with respect to their in-clusion criteria, methods of tracheostomy and outcomes measures(see Characteristics of included studies).

Excluded studies

We excluded six studies because they compared early tracheostomywith prolonged endotracheal intubation (Blot 2008; Bouderka2004; El-Naggar 1976; Saffle 2002; Stauffer 1981; Sugerman1997). In one quasi-randomized study, the late tracheostomy wasperformed eight days after admission (< 10 days) breaching theselection criteria (> 10 days after intubation) for this review (Rodriguez 1990). For further details, see the Characteristics ofexcluded studies.

Risk of bias in included studies

We paid special attention to the description of randomization andallocation concealment, as the absence of adequate methodologicalaspects are associated with biased estimated effects (Schulz 1995).A synthesis of the assessment of all of the items of methodologicalquality described below are presented in Figure 2 and Figure 3.

Figure 2. Risk of bias graph: review authors’ judgements about each risk of bias item presented as

percentages across all included studies.



Figure 3. Risk of bias summary: review authors’ judgements about each risk of bias item for each included

study.

Allocation

Randomization

Two studies (Barquist 2006; Terragni 2010) reported computer-generated randomization, which we considered to possess a lowrisk of bias. Neither study found significant differences betweenthe comparison groups in terms of baseline characteristics.Dunham 1984 referred to randomization based upon the last digitof the patient’s hospital number, a method which we deemed in-dicative of resulting in a high risk of bias (quasi-randomized study).

Rumbak 2004 did not explicitly report the method of randomiza-tion; thus the study was considered to reflect a moderate risk ofbias.

Allocation concealment

Two studies (Barquist 2006; Rumbak 2004) utilized envelopes toconceal the allocation of participants. Terragni 2010 clearly re-ported a centralized process of randomization. These three stud-ies were therefore considered to have a low risk of bias. However,Dunham 1984, a quasi-randomized study, was considered to pos-



sess a high risk of bias associated with the allocation concealment.

Blinding

In three studies (Barquist 2006; Dunham 1984; Rumbak 2004),no information was given as to whether the data collectors wereindependent from the researchers who designed the study, or wereblinded to the allocations. These studies were thus considered tohave a moderate risk of bias associated with potential knowledgeabout the allocated interventions (blinding). In contrast, Terragni2010 clearly explained that a blinded clinician was responsible forlooking at the clinical charts remotely and for evaluating non-objective components of the Clinical Pulmonary Infection Score.Consequently, this study was deemed as possessing a low risk ofbias associated with the blinding.

Incomplete outcome data

Two studies (Barquist 2006; Rumbak 2004) were considered tohave a low risk of bias associated with incomplete outcome data,due to intention-to-treat (ITT) analysis and clear patient flow.Dunham 1984 was considered to possess a high risk of bias as,after randomization, only those patients who were intubated forat least seven days were included in the study. The authors did notindicate the percentages or number of patients not considered foranalysis after randomizationTerragni 2010 deserved special consideration because the authorsreported ITT, but the numbers of patients from the randomizationto the analysis of each outcome were not clearly reported for eachone of the comparison groups. Therefore this study was consideredto have a high risk of bias.

Selective reporting

Three studies were considered to have a low risk of bias based onthe relevant outcomes considered for evaluation, and the absenceof suspected selective outcome reporting (Barquist 2006; Rumbak2004; Terragni 2010). Dunham 1984 was deemed to be a studywith high risk of systematic error due to the absence of clinicallyrelevant outcomes (such as mortality rates).

Other potential sources of bias

Dunham 1984 evaluated 50% of the patients at four to six monthsafter extubation. The remaining patients were interviewed 12months after extubation but the exact number of patients per com-parison group was not specified. No indication was given of theabsence of substantial differences between the comparison groupsat baseline (comparable groups).Three studies were lacking clear information about the time offollow-up for all of the outcomes (Barquist 2006; Rumbak 2004;Terragni 2010). Thus, none of these studies were considered to befree of other biases.

Effects of interventions

See: Summary of findings for the main comparison Earlytracheostomy compared to late tracheostomy for critically illpatients

Primary outcomes

Mortality

Mortality at day 28 of the follow-up was lower in the group sub-jected to early tracheostomy (26.3%, 55/209) as compared to thelate tracheostomy group (31.4%, 66/210), with a statistically non-significant risk ratio (RR) of 0.84 (95% CI 0.62 to 1.13, P = 0.25)in one study (Terragni 2010), see Table 1 (line 1). An ITT analysisof combined mortality (at 28 days and at one year) showed 50%(127/209) for the early tracheostomy group and 70% (151/210)for the late tracheostomy group (RR 0.85, 95% CI 0.74 to 0.97,P = 0.02) (Terragni 2010), see Table 1 (line 2). Rumbak 2004 alsoreported a statistically significant result favouring the early tra-cheostomy (31.6%, 19/60) as compared to the late tracheostomy(61.6%, 37/60), with a risk ratio of 0.51 [0.34, 0.78, P = 0.002](Table 1, line 3). The direction of effect also favoured the earlytracheostomy in Barquist 2006 but without statistical significance(RR 0.43 [0.09, 2.03], P=0.29) (Table 1, line 4).After trying to combine the mortality rates (irrespective of follow-up time) in a meta-analysis, we found substantial heterogeneitybetween studies (Barquist 2006; Rumbak 2004; Terragni 2010) (I2 = 66%, P < 0.10). Such a statistical heterogeneity is explainedby the substantial clinical and methodological diversities betweenthe studies. Thus, data on mortality were presented in a forestplot with isolated study estimated effects and no meta-analysis(Analysis 1.1). The percentage of deaths in the early tracheostomygroup was 49.7% versus 64.1% in the late tracheostomy group.

Duration of artificial ventilation

Terragni 2010 measured the median values of ventilator-free daysat 28th day, which is considered a good outcome. The resultsshowed statistically significant (P = 0.02) estimated effects favour-ing early tracheostomy (median 11 days, interquartile range 0 to21 days) as compared to late tracheostomy (median 6 days, in-terquartile range 0 to 17 days) (Appendix 10). Such non-paramet-ric data could not be reported as means and standard deviations(SD). Thus, they were not appropriate for insertion in a forestplot. Rumbak 2004 reported a lower mean of days on mechanicalventilation in the early tracheostomy group (7.4 days, SD = 4)when compared with the late group (17.4 days, SD = 5.3), with astatistically significant mean difference (MD) of -9.80 (95% CI -11.48 to -8.12, P < 0.001) (Table 1, line 5). On the other hand,Dunham 1984 reported the risk of artificial ventilation for morethan 21 days, with a higher proportion of events observed within



the early tracheostomy group (52.9%, 18/34) as compared to thelate tracheostomy group (45%, 18/40), but the estimated effectwas not statistically significant (RR 0.85, 95% CI 0.53 to 1.36, P= 0.49) (Table 1, line 6).Terragni 2010 observed that 67.6% (142/210) of patients fromthe late tracheostomy group and 77% (161/209) from the earlygroup experienced a successful weaning. Thus, it was possible toestimate a statistically significant higher proportion of patientswith successful weaning in the early tracheostomy group as com-pared to the late tracheostomy group (RR 0.88, 95% CI 0.78 to0.99, P = 0.03, one study). The number of patients who needed tobe treated (NNT) in order to achieve just one successful weaningwas approximately 11 (95% CI 5.55 to 100) (Table 1, line 7).

Secondary outcomes

Length of stay in ICU

Patients subjected to early tracheostomy seemed to have a higherprobability of ICU discharge at 28 days after the randomization(48.3%, 101/209) as compared to the patients in the late tra-cheostomy group ( 39%, 82/210), but the risk ratio was not sta-tistically significant (RR 0.81, 95% CI 0.65 to 1.01, P = 0.06, onestudy) (Terragni 2010), see Table 1 (line 8). Moreover, Terragni2010 evaluated ICU-free days at 28 days after randomization, re-sulting in an absence of difference in the median values betweenthe comparison groups (median values of zero for both groups)(Appendix 10).In terms of the length of hospital stay, Terragni 2010 observedno clinically relevant difference between groups, as shown by themedian values of 31 days (interquartile range 17 to 39) in the earlytracheostomy group and 32 days (interquartile range 18 to 59) inthe late tracheostomy group (Appendix 10).Clinically and statistically significant estimates favouring the earlytracheostomy group were observed for the time spent in the ICU(MD -11.40 days, 95% CI -12.42 to -10.38, P<0.00001) in onestudy (Table 1, line 9).

Pneumonia

In the same way as for mortality, we tried to combine the pneu-monia rates in a meta-analysis. We again found substantial het-erogeneity between studies (Rumbak 2004; Terragni 2010) (I2 =0.73, P = 0.15). Such statistical heterogeneity was also justifiedbecause of the substantial clinical and methodological differencesbetween studies. By consensus, we decided to present the data onpneumonia in a forest plot with isolated estimated effects fromthe studies, excluding a meta-analysis (Analysis 1.2). The percent-age of events of pneumonia in the early tracheostomy group was12.3% versus 21.9% in the late tracheostomy group. The effect es-timate was even weaker (RR 0.69, 95% CI 0.45 to 1.05, P = 0.08)in Terragni 2010, which did not include patients with pneumonia

at the study entry, than that observed in Rumbak 2004 (RR 0.20,95% CI 0.06 to 0.66, P = 0.008).

Postoperative adverse events

The different studies included in this systematic review foundno clinically or statistically relevant difference between earlyand late tracheostomies in the following postoperative adverseevents: stoma inflammation; stoma infection; minor bleeding;major bleeding; pneumothorax; subcutaneous emphysema; tra-cheoesophageal fistula and cannula displacement or need for re-placement (Terragni 2010); significant laryngotracheal pathology;respiratory sepsis; major complications; complications (Dunham1984); tracheal stenosis, irrespective of the severity (in hospi-tal); tracheal stenosis > 50 (10 weeks postintubation) (Rumbak2004); and self extubation (Rumbak 2004). However, the follow-ing events occurred significantly more in the early tracheostomygroup: tracheal stenosis with a severity score from 0 to 20 in hos-pital and 10 weeks after intubation; tracheal stenosis, irrespectiveof the severity, 10 weeks after intubation (Rumbak 2004) (Table1, line 10).

Other potentially relevant outcomes not planned in the

protocol of this systematic review

No significant difference was found between comparison groupsrelated to the need for a long-term care facility; other outcomeslooked at were: days spent on sedation and on high dose vasopres-sor infusion, intraoperative adverse events and events followed bydeath (Appendix 11).


Because of the relative paucity of adequate studies, a sensitivityanalysis was not done.

D I S C U S S I O N

Summary of main results

Available evidence from randomized controlled trials showed nosignificant change in mortality at any time of follow-up in theearly tracheostomy group as compared to the late tracheostomygroup. Three trials (Barquist 2006; Rumbak 2004; Terragni 2010)showed lower proportions of mortality at any time of follow-up inthe early tracheostomy group as compared to late tracheostomy,but this is not strong evidence.Two observations from two different studies regarding the timespent on mechanical ventilation were also of importance. One



study indicated that early tracheostomy reduces the time on me-chanical ventilation by more than nine days (9.8 days) (Rumbak2004). The other study found more days ’free from the ventilator’,measured at day 28 of follow-up (Terragni 2010), in the early tra-cheostomy group. This study excluded patients with chronic ob-structive pulmonary disease, which is more likely to require long-term ventilation. Additionally, the same study (Terragni 2010)demonstrates that early tracheostomy is significantly associatedwith a higher rate of successful weaning.With respect to the secondary outcomes, the available evidencefrom only one study involving more than 400 patients (Terragni2010) demonstrates that early tracheostomy is significantly associ-ated with a higher rate of discharge from the ICU at day 28. Earlytracheostomy again proved to be significantly superior to late tra-cheostomy in terms of the mean time spent in the ICU, but nodifferences were observed for ICU-free days at 28 days, the needfor a long-term care facility or length of hospital stay (Terragni2010).There is no evidence to suggest that any one treatment is associatedwith lower probability of pneumonia, possibly because of the largeheterogeneity between the studies (Rumbak 2004; Terragni 2010);Terragni 2010 in fact excluded patients with chronic obstructivepulmonary disease and pneumonia at study entry.

Overall completeness and applicability ofevidence

There is not sufficient evidence to recommend either early or latetracheostomy in clinical practice since the findings of this system-atic review did not demonstrate significant differences between thecomparison groups measuring risk of mortality and pneumonia(Barquist 2006; Rumbak 2004; Terragni 2010). Although indi-vidual studies (Rumbak 2004; Terragni 2010) favoured early tra-cheostomy compared to late tracheostomy in the time spent onmechanical ventilation, time in ICU and successful weaning, thesestudies were judged to have high risk of bias. The same concernapplies to Rumbak 2004, which suggested a higher risk of trachealstenosis in the early tracheostomy group.

Quality of the evidence

Evidence in this review cannot be considered robust due to the lownumber of studies (Barquist 2006; Dunham 1984; Rumbak 2004;Terragni 2010) and the relatively low sample power for each of therelevant outcomes (599 patients at most for mortality). Althoughthe estimated effects of studies with relevant outcomes (mortal-ity and pneumonia) had the same direction, their intensities dif-fered significantly, as demonstrated by heterogeneity tests. Suchstatistical heterogeneity probably resulted from the clinical andmethodological diversity among the studies (for example, the date

of realization, inclusion criteria, geographic region, and methodsof tracheostomy).According to the Summary of findings for the main comparisonthe quality of the evidence was considered low for the outcomes ofdeath and pneumonia, and moderate for successful weaning frommechanical ventilation.

Potential biases in the review process

A high sensitivity search strategy was used in this systematic reviewso as to avoid missing any randomized controlled trial that com-pared early versus late tracheotomy in critically ill patients. Weprevented any language bias by not imposing language restrictionsupon the search.Other studies have been conducted but were not yet published(Dumire 2008; Huttner 2010; Kluge 2009; Ranieri 2009; Young2008) and their results may improve the evidence in this area.Thus, such ongoing studies will probably be included in futureversions of this review once their results are made available.

Agreements and disagreements with otherstudies or reviews

The findings of our systematic review and the reviews by Dunham2006 and Griffiths 2005 do not support either early or late tra-cheostomy for mortality. Scales 2008, in a large observationalstudy involving more than 10,000 patients, showed that earlytracheostomy was associated with significant advantages in termsof mortality over late tracheostomy for critically ill patients. Pre-vious systematic reviews, as well as other observational studiesand non-randomized controlled trials with lower methodologicalrigour, also showed a decreased amount of time spent on ventila-tory support (Arabi 2004; Arabi 2009; Blot 1995; Dunham 2006;Gandía-Martínez 2010; Griffiths 2005, Lesnik 1992; Zagli 2010);time in the ICU (Arabi 2004; Arabi 2009, Gandía-Martínez2010; Griffiths 2005, Lesnik 1992; Zagli 2010), at the hospital(Arabi 2004; Arabi 2009; Blot 1995); and lower probabilities ofpneumonia (Lesnik 1992; Gandía-Martínez 2010) and extuba-tion (El-Naggar 1976) with early tracheostomy as compared tolate tracheostomy. All these results have been observed in the faceof large clinical, regional, methodological and chronological di-versity among the studies.

A U T H O R S ’ C O N C L U S I O N S

Implications for practice

The evidence in this Cochrane review is insufficient to recommendeither early or late tracheostomy for critically ill patients. We have



insufficient information about any subgroup or individual charac-teristic(s) potentially associated with the best indications for eitherearly or late tracheostomy.

Implications for research

Additional high quality randomized controlled trials are neces-sary to evaluate possible differences between early and late tra-cheostomy for critically ill patients. Some trials have already be-gun, and we are awaiting their results for an updated version ofthis systematic review. Researchers would contribute significantlyto improving the evidence by considering the following outcomemeasures: mortality rates up to 12 months of follow-up, time spenton mechanical ventilation and in the ICU, length of hospital stay,

successful weaning, pneumonia and costs. All researchers inter-ested in this research area should work together to unify the rawdata collected in their studies in order to make more informationavailable for inferences about more precise indications.

A C K N O W L E D G E M E N T S

We thank Dr Mathew Zacharias (content editor), Cathal Walsh(statistical editor), Dr John Griffiths (peer reviewer), Dr JohnHeffner (peer reviewer), Dr Leila Bender (peer reviewer), DrBradley Freeman (peer reviewer) and Janet Wale (consumer) fortheir help and editorial advice during the preparation of this sys-tematic review.

R E F E R E N C E S

References to studies included in this review

Barquist 2006 {published data only}

Barquist ES, Amortegui J, Hallal A, Giannotti G, WhinneyR, Alzamel H, et al.Tracheostomy in ventilator dependenttrauma patients: a prospective, randomized intention-to-treat study. The Journal of Trauma 2006;60(1):91–7.[PUBMED: 16456441]

Dunham 1984 {published data only}

Dunham CM, LaMonica C. Prolonged tracheal intubationin the trauma patient. The Journal of Trauma 1984;24(2):120–4. [PUBMED: 6694236]

Rumbak 2004 {published data only}

Rumbak MJ, Newton M, Truncale T, Schwartz SW,Adams JW, Hazard PB. A prospective, randomized, studycomparing early percutaneous dilational tracheotomy toprolonged translaryngeal intubation (delayed tracheotomy)in critically ill medical patients. Critical Care Medicine2004;32(8):1689–94. [PUBMED: 15286545]

Terragni 2010 {published data only}

Terragni PP, Antonelli M, Fumagalli R, Faggiano C,Berardino M, Pallavicini FB, et al.Early vs late tracheotomyfor prevention of pneumonia in mechanically ventilatedadult ICU patients: a randomized controlled trial. JAMA

2010;303(15):1483–9. [PUBMED: 20407057]

References to studies excluded from this review

Blot 2008 {published data only}

Blot F. A study of early tracheostomy in patients undergoingprolonged mechanical ventilation [Etude de lintérêt dela trachéotomie precoce chez les malades sous ventilationmécanique prolongée]. Revue des Maladies Respiratoires2003;20(3 Pt 1):411–20. [PUBMED: 12910115]∗ Blot F, Similowski T, Trouillet JL, Chardon P, KorachJM, Costa MA, et al.Early tracheotomy versus prolongedendotracheal intubation in unselected severely ill ICU

patients. Intensive Care Medicine 2008;34(10):1779–87.[PUBMED: 18592210]

Bouderka 2004 {published data only}∗ Bouderka MA, Fakhir B, Bouaggad A, Hmamouchi B,Hamoudi D, Harti A. Early tracheostomy versus prolongedendotracheal intubation in severe head injury. The Journal

of Trauma 2004;57(2):251–4. [PUBMED: 15345969]

El-Naggar 1976 {published data only}

El-Naggar M, Sadagopan S, Levine H, Kantor H, CollinsVJ. Factors influencing choice between tracheostomy andprolonged translaryngeal intubation in acute respiratoryfailure: a prospective study. Anesthesia and Analgesia 1976;55(2):195–201. [PUBMED: 943979]

Rodriguez 1990 {published data only}

Rodriguez JL, Steinberg SM, Luchetti FA, Gibbons KJ,Taheri PA, Flint LM. Early tracheostomy for primary airwaymanagement in the surgical critical care setting. Surgery1990;108(4):655–9. [PUBMED: 2218876]

Saffle 2002 {published data only}

Saffle JR, Morris SE, Edelman L. Early tracheostomy doesnot improve outcome in burn patients. Journal of BurnCare and Rehabilitation 2002;23(6):431–8. [PUBMED:12432320]

Stauffer 1981 {published data only}

Stauffer JL, Olson DE, Petty TL. Complications andconsequences of endotracheal intubation and tracheotomy.A prospective study of 150 critically ill adult patients.The American Journal of Medicine 1981;70(1):65–76.[PUBMED: 7457492]

Sugerman 1997 {published data only}

Sugerman HJ, Wolfe L, Pasquale MD, Rogers FB, O’MalleyKF, Knudson M, DiNardo L, Gordon M, SchafferS. Multicenter, randomized, prospective trial of earlytracheostomy. The Journal of Trauma 1997;43(5):741–7.[PUBMED: 9390483]



References to ongoing studies

Dumire 2008 {published data only}

Dumire RD, Stephen LM. A Prospective, Randomized Trialof Early Versus Late Tracheostomy in Trauma Patients WithSevere Brain Injury. Current Controlled Trials 2008. [:NCT00292097]

Huttner 2010 {published data only}

Huttner HB, KÃhrmann M, Staykov D. WEANING-Study: Weaning by Early Versus Late Tracheostomy insupratentorIal intracerebral Bleedings. Current ControlledTrials 2010. [: NCT01176214]

Kluge 2009 {published data only}

Kluge S. Timing of Dilation Tracheostomy in MechanicallyVentilated Chronic Obstructive Pulmonary Disease(COPD) Patients. Current Controlled Trials 2009. [:NCT01021202]

Ranieri 2009 {published data only}

Ranieri VM. Efficacy of Early Tracheostomy to ReduceIncidence of Ventilator Acquired Pneumonia (VAP).Current Controlled Trials. [: NCT00262431]

Young 2008 {published data only}

Young D. Tracheostomy Management in Critical Care.Current Controlled Trials 2008. [: ISRCTN28588190]

Additional references

Agle 2006Agle SC, Kao LS, Moore FA, Gonzalez EA, VercruysseGA, Todd SR. Early predictors of prolonged mechanicalventilation in major torso trauma patients who requireresuscitation. American Journal of Surgery 2006;192(6):822–7. [PUBMED: 17161101]

Ahmed 2007Ahmed N, Kuo YH. Early versus late tracheostomy inpatients with severe traumatic head injury. Surgical Infections

(Larchmt) 2007;8(3):343–7. [PUBMED: 17635057]

Aissaoui 2007Aissaoui Y, Azendour H, Balkhi H, Haimeur C, KamiliDrissi N, Atmani M. Timing of tracheostomy andoutcome of patients requiring mechanical ventilation[Délai de la trachéotomie et devenir des patients sousventilation mécanique]. Annales Françaises d’Anesthèsie et de

Rèanimation 2007;26(6):496–501. [PUBMED: 17521853]

Arabi 2004Arabi Y, Haddad S, Shirawi N, Al Shimemeri A. Earlytracheostomy in intensive care trauma patients improvesresource utilization: a cohort study and literature review.Critical Care 2004;8(5):R347–52. [PUBMED: 15469579]

Arabi 2009Arabi YM, Alhashemi JA, Tamim HM, Esteban A, HaddadSH, Dawood A, et al.The impact of time to tracheostomyon mechanical ventilation duration, length of stay, andmortality in intensive care unit patients. Journal of CriticalCare 2009;24(3):435–40. [: 2010377724; PUBMED:19327302]

Armstrong 1998Armstrong PA, McCarthy MC, Peoples JB. Reduced use ofresources by early tracheostomy in ventilator-dependentpatients with blunt trauma. Surgery 1998;124(4):763–6.[PUBMED: 9780999]

Bartels 1998Bartels HE, Stein HJ, Siewert JR. Tracheobronchial lesionsfollowing oesophagectomy: prevalence, predisposing factorsand outcome. The British Journal of Surgery 1998;85(3):403–6. [PUBMED: 9529504]

Björling 2007Björling G, Axelsson S, Johansson UB, Lysdahl M,Markström A, Schedin U, et al.Clinical use and materialwear of polymeric tracheostomy tubes. Laryngoscope 2007;117(9):1552–9. [PUBMED: 17632426]

Blot 1995Blot F, Guiguet M, Antoun S, Leclercq B, Nitenberg G,Escudier B. Early tracheotomy in neutropenic, mechanicallyventilated patients: rationale and results of a pilot study.Support Care Cancer 1995;3(5):291–6. [PUBMED:8520874]

Cavaliere 2007Cavaliere S, Bezzi M, Toninelli C, Foccoli P. Managementof post-intubation tracheal stenoses using the endoscopicapproach. Monaldi Archives for Chest Disease 2007;67(2):73–80. [PUBMED: 17695689]

Christensen 2006Christensen PM, Kristiansen IS. Number-needed-to-treat (NNT) needs treatment with care. Basic & Clinical

Pharmacology & Toxicology 2006;99(1):12–6. [PUBMED:16867164]

Clec’h 2007Clec’h C, Alberti C, Vincent F, Garrouste-Orgeas M, deLassence A, Toledano D, et al.Tracheostomy does notimprove the outcome of patients requiring prolongedmechanical ventilation: A propensity analysis. Critical Care

Medicine 2007;35(1):132–8. [PUBMED: 17133180]

Copas 2000Copas J, Shi JQ. Meta-analysis, funnel plots and sensitivityanalysis. Biostatistics 2000;1(3):247–62. [PUBMED:12933507]

Crimlisk 2006Crimlisk JT, O’Donnell C, Grillone GA. Standardizingadult tracheostomy tube styles: what is the clinical andcost-effective impact?. Dimensions of Critical Care Nursing:DCCN 2006;25(1):35–43. [PUBMED: 16501371]

Criner 1994Criner GJ, Tzouanakis A, Kreimer DT. Overview ofimproving tolerance of long-term mechanical ventilation.Critical Care Clinics 1994;10(4):845–66. [PUBMED:8000930]

Deeks 2001aDeeks JJ, Altman DG. Effect measures for meta-analysis oftrials with binary outcomes. In: Egger M, Davey SmithG, Altman DG editor(s). Systematic reviews in health care:



Meta-analysis in context. 2th. BMJ Publication Group,2001.

Deeks 2001bDeeks JJ, Altman DG, Bradburn MJ. Statistacal methodsfor examining heterogeneity and combining results fromseveral studies in meta-analysis. In: Egger M, Davey SmithG, Altman DG editor(s). Systematic reviews in health care:

Meta-analysis in context. 2nd Edition. BMJ PublicationGroup, 2001.

Deeks 2005Deeks JJ, Higgins, JPT, Altman DG (editors). Analysingand presenting results. Cochrane Handbook for SystematicReviews of Interventions 4.2.5 [updated May 2005];Section 8. In: Higgins JPT, Green S editor(s). The Cochrane

Library. Chichester, UK: John Wiley & Sons, Ltd, 2005.

DerSimonian 1986DerSimonian R, Laird N. Meta-analysis in clinical trials.Controlled Clinical Trials 1986;7:177. [PUBMED:3802833]

Divo 2010Divo MJ, Murray S, Cortopassi F, Celli BR. Prolongedmechanical ventilation in Massachusetts: the 2006prevalence survey. Respiratory Care 2010;55(12):1693–8.[PUBMED: 21122178]

Dollner 2002Dollner R, Verch M, Schweiger P, Graf B, Wallner F. Long-term outcome after Griggs tracheostomy. The Journal ofOtolaryngology 2002;31(6):386–9. [PUBMED: 12593553]

Dunham 2006Dunham CM, Ransom KJ. Assessment of earlytracheostomy in trauma patients: a systematic review andmeta-analysis. American Surgeon 2006;72(3):276–81.[PUBMED: 16553133]

Freeman 2008Freeman BD, Kennedy C, Robertson TE, CoopersmithCM, Schallom M, Sona C, et al.Tracheostomy protocol:experience with development and potential utility. Critical

Care Medicine 2008;36(6):1742–8. [: 2009952685;PUBMED: 18496369]

Friedman 2006Friedman Y. Percutaneous versus surgical tracheostomy:The continuing saga. Critical Care Medicine 2006;34(8):2250–1. [PUBMED: 16883198]

Frutos-Vivar 2005Frutos-Vivar F, Esteban A, Apezteguia C, Anzueto A,Nightingale P, Gonzalez M, et al.Outcome of mechanicallyventilated patients who require a tracheostomy. CriticalCare Medicine 2005;33(2):290–8. [PUBMED: 15699830]

Gajic 2007Gajic O, Afessa B, Thompson BT, Frutos-Vivar F,Malinchoc M, Rubenfeld GD, Esteban A, Anzueto A,Hubmayr RD, Second International Study of MechanicalVentilation and ARDS-net Investigators. Prediction ofdeath and prolonged mechanical ventilation in acute

lung injury. Critical Care 2007;11(3):R53. [PUBMED:17493273]

Gandía-Martínez 2010Gandía-Martínez F, Martínez-Gil I, Andaluz-Ojeda D,Bobillo de Lamo F, Parra-Morais L, Díez-GutiérrezF. Analysis of early tracheostomy and its impact ondevelopment of pneumonia, use of resources andmortality in neurocritically ill patients [Análisis de latraqueotomía precoz y su impacto sobre la incidencia deneumonía, consumo de recursos y mortalidad en pacientesneurocríticos]. Neurocirugia (Astur) 2010;21(3):211–21.[PUBMED: 20571724]

Goettler 2006Goettler CE, Fugo JR, Bard MR, Newell MA, SagravesSG, Toschlog EA, et al.Predicting the need for earlytracheostomy: a multifactorial analysis of 992 intubatedtrauma patients. The Journal of Trauma 2006;60(5):991–6.[PUBMED: 16688060]

Griffiths 2005Griffiths J, Barber VS, Morgan L, Young JD. Systematicreview and meta-analysis of studies of the timing oftracheostomy in adult patients undergoing artificialventilation. BMJ 2005;330(7502):1243–8. [PUBMED:15901643]

Gullo 2007Gullo A, Sorbello M, Frova G. Percutaneous versus surgicaltracheostomy: an unfinished symphony. Critical Care

Medicine 2007;35(2):682–3. [PUBMED: 17251746]

Heffner 2001Heffner JE, Hess D. Tracheostomy management in thechronically ventilated patient. Clinics in Chest Medicine

2001;22(1):55–69. [PUBMED: 11315459]

Heffner 2003Heffner JE. Tracheotomy application and timing. Clinics

in Chest Medicine 2003;24(3):389–98. [PUBMED:14535212]

Hess 2005Hess DR. Tracheostomy tubes and related appliances.Respiratory Care 2005;50(4):497–510. [PUBMED:15807912]

Higgins 2002Higgins JPT, Thompson SG. Quantifying heterogeneity ina meta-analysis. Statistics in Medicine 2002;21:1539–58.[PUBMED: 12111919]

Higgins 2003Higgins JP, Thompson SG, Deeks JJ, Altman DG.Measuring inconsistency in meta-analysis. BMJ 2003;327:557–60. [PUBMED: 12958120]

Higgins 2007Higgins KM, Punthakee X. Meta-analysis comparison ofopen versus percutaneous tracheostomy. Laryngoscope 2007;117(3):447–54. [PUBMED: 17334304]

Higgins 2011aHiggins JPT, Green S (editors)Cochrane Handbook forSystematic Reviews of Interventions Version 5.1.0 [updated



March 2011] .. The Cochrane Collaboration, 2011.Available from www.cochrane-handbook.org..

Higgins 2011bHiggins JPT, Green S, editors. Highly sensitive searchstrategies for identifying reports of randomized controlledtrials in MEDLINE. Cochrane Handbook for SystematicReviews of Interventions Version 5.1.0 [updated March2011 ; Appendix 5b. The Cochrane Collaboration, 2011.Available from www.cochrane-handbook.org. Chichester,UK: John Wiley & Sons, Ltd.

Higgins 2011cHiggins JPT, Green S, editors. Assessment of studyquality. Cochrane Handbook for Systematic Reviews ofInterventions Version 5.1.0 [updated March 2011] ; Section6. The Cochrane Collaboration, 2011. Available fromwww.cochrane-handbook.org. Chichester, UK.

Holzapfel 1993Holzapfel L, Chevret S, Madinier G, Ohen F, DemingeonG, Coupry A, et al.Influence of long-term oro- ornasotracheal intubation on nosocomial maxillary sinusitisand pneumonia: results of a prospective, randomized,clinical trial. CritIcal Care Medicine 1993;21(8):1132–8.[PUBMED: 8339576]

Latour 1997Latour J, Abraira V, Cabello JB, López Sánchez J.Investigation methods in clinical cardiology. IV. Clinicalmeasurements in cardiology: validity and errors ofmeasurements [Las mediciones clínicas en cardiología:validez y errores de medición]. Revista Española de

Cardiología 1997;50(2):117–28. [PUBMED: 9091999]

Lesnik 1992Lesnik I, Rappaport W, Fulginiti J, Witzke D. The roleof early tracheostomy in blunt, multiple organ trauma.The American Surgeon 1992;58(6):346–9. [PUBMED:1596033 ]

Légaré 2001Légaré JF, Hirsch GM, Buth KJ, MacDougall C, SullivanJA. Preoperative prediction of prolonged mechanicalventilation following coronary artery bypass grafting.European Association for Cardio-thoracic Surgery 2001;20(5):930–6. [PUBMED: 11675177]

Norwood 2000Norwood S, Vallina VL, Short K, Saigusa M, FernandezLG, McLarty JW. Incidence of tracheal stenosis and otherlate complications after percutaneous tracheostomy. Annals

of Surgery 2000;332(2):233–41. [PUBMED: 10903603]

Pappas 2011Pappas S, Maragoudakis P, Vlastarakos P, AssimakopoulosD, Mandrali T, Kandiloros D, Nikolopoulos TP. Surgicalversus percutaneous tracheostomy: an evidence-basedapproach. European Archives of Oto-rhino-laryngology 2011;268(3):323–30. [PUBMED: 20957486]

Plummer 1989Plummer AL, Gracey DR. Consensus conference onartificial airways in patients receiving mechanical ventilation.Chest 1989;96(1):178–80. [PUBMED: 2500308]

Ramsay 2000Ramsay MA. Measuring level of sedation in the intensivecare unit. JAMA 2000;284(4):441–2. [PUBMED:10904502]

Ranes 2006Ranes JL, Gordon SM, Chen P, Fatica C, Hammel J,Gonzales JP, et al.Predictors of long-term mortality inpatients with ventilator-associated pneumonia. TheAmerican Journal of Medicine 2006;119(10):819.e13–9.[PUBMED: 17000224]

RevMan 5.1Copenhagen: The Nordic Cochrane Centre, The CochraneCollaboration. Review Manager (RevMan). Version 5.1.Copenhagen: The Nordic Cochrane Centre, The CochraneCollaboration, 2011.

Scales 2008Scales DC, Thiruchelvam D, Kiss A, Redelmeier DA. Theeffect of tracheostomy timing during critical illness on long-term survival. Critical Care Medicine 2008;36(9):2547–57.[PUBMED: 18679113]

Schultz 2007Schultz MJ, Veelo DP, Dongelmans DA. Percutaneoustracheostomies are preferable to surgical tracheostomies.Critical Care Medicine 2007;35(2):676–7.

Schulz 1995Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empiricalevidence of bias. Dimensions of methodological qualityassociated with estimates of treatment effects in controlledtrials. JAMA 1995;273(5):408–12. [PUBMED: 7823387]

Schünemann 2009Schünemann H, Bro ek J, Oxman A, editors. GRADEhandbook for grading quality of evidence and strength ofrecommendation. Version 3.2 [updated March 2009]. TheGRADE Working Group, 2009. Available from http://www.cc–ims.net/gradepro, 2009.

Sellers 1997Sellers BJ, Davis BL, Larkin PW, Morris SE, Saffle JR.Early prediction of prolonged ventilator dependence inthermally injured patients. The Journal of Trauma 1997;43(6):899–903. [PUBMED: 9420102]

Shirzad 2010Shirzad M, Karimi A, Ahmadi SH, Marzban M, TazikM, Aramin H. Predictors and early outcome of prolongedmechanical ventilation in contemporary heart valve surgery.Monaldi Archives for Chest Disease 2010;74(1):22–7.[PUBMED: 20925175]

Stedman 1995Stedman. Stedman’s Medical Dictionary. 26th Edition.Williams and Wilkins, 1995.



Tibballs 2006Tibballs J, Robertson C, Wall R. Tracheal ulceration andobstruction associated with flexible Bivona tracheostomytubes. Anaesthesia and Intensive Care 2006;34(4):495–7.[PUBMED: 16913349]

Zagli 2010Zagli G, Linden M, Spina R, Bonizzoli M, Cianchi G,Anichini V, et al.Early tracheostomy in intensive care unit:a retrospective study of 506 cases of video-guided CiagliaBlue Rhino tracheostomies. The Journal of Trauma 2010;68(2):367–72. [PUBMED: 20154550]

∗ Indicates the major publication for the study



C H A R A C T E R I S T I C S O F S T U D I E S

Characteristics of included studies [ordered by study ID]

Barquist 2006

Methods Study design: parallel randomized controlled trial with intention-to-treat analysis andsample size based in the following information: a total of 140 patients would be neededif the SD was nine ventilator days, the difference between means was three days and thepower was set at 90%Local/Setting: Division of Trauma and Surgical Critical Care, DeWitt Daughtry FamilyDepartment of Surgery, University of Miami School of Medicine, Miami, Florida

Participants • N=60 (early tracheostomy, n = 29 / late tracheostomy, n = 31)• Age range:18-87 / mean age: 51.8 (whole sample)• Gender: 46 male / 14 female• Ventilator dependent patients• Traumatic injury as the proximate cause of their ventilator dependence• Intubated at least 3 days when they were seven days after admission to the Trauma

ICU

Interventions 1. Early tracheostomy: before day 82. Late tracheostomy: after day 28

All tracheostomies (early and late) were performed by the open surgical technique

Outcomes • Mortality rates (time of data collection was not explicitly referred by the authors)• Mean ICU free days at 20 days:• Mean Ventilation free at day 30 (with extubation performed after spontaneous

breathing trial [CPAP at 5 cm water pressure with 5 cm water pressure support for 30minutes] with predefined criteria for passing [pO2 greater than 55 mm Hg, respiratoryrate less than 35 breaths/minute and no respiratory acidosis)]

• Mean ICU free at day 30• Ventilator-associated pneumonia at any time point (CDC CRITERIA: Centers

for Disease Control: elevated WBC, fevers, CXR infiltrate, and Broncho-AlveolarLavage [BAL] culture with greater than 10,000 colony-forming units per millilitre[CFU/mL])

• Single superficial surgical site infection• Major complications related to the tracheostomy (including stomal infection,

stomal haemorrhage, major vascular injury, pneumothorax, subglottic stenosis, andtracheo-oesophageal fistulae)

Notes Four patients in the ’late’ group had a surgical tracheostomy placed on days 17, 18, 19,and 21 to facilitate transfer to long-term careMethod to predict prolonged artificial ventilation: Not explicitly reported

Risk of bias

Bias Authors’ judgement Support for judgement



Barquist 2006 (Continued)

Random sequence generation (selectionbias)

Low risk Computer-generated tableThe authors referred no significant differ-ence between comparison groups on base-line characteristics

Allocation concealment (selection bias) Low risk After consent was obtained, an envelopewith the assigned group inside was opened

Blinding (performance bias and detectionbias)All outcomes

Unclear risk There was no information on whetherdata collectors were independent from theresearchers who designed the study orblinded to the allocations

Incomplete outcome data (attrition bias)All outcomes

Low risk The authors carried out the intention-to-treat analysis according to the appropriateddefinition

Selective reporting (reporting bias) Low risk No suspected selective outcome reporting,since relevant outcomes were evaluated

Other bias High risk Although times of follow-up were explicitlyannounced for only three outcomes (meanICU free days at 20 days, mean ventilationfree at day 30, mean ICU free at day 30),the authors were not explicit to report thefollow-up times for the other outcome data

Dunham 1984

Methods Study design: Parallel quasi-randomized controlled trial without intention-to-treat anal-ysisLocal/Setting: The Shock Trauma Center of the Maryland Institute for Emergency Med-ical Services Systems (MIEMSS)

Participants • N=74 (Early tracheostomy, n = 34 / Late tracheotomy, n = 40)• Age range: 17 to 75 years• Intubation for at least 7 days• Severe head injury• Respiratory insufficiency and/or a tenuous or incompetent airway secondary to

maxillofacial injury

Interventions 1. The early group underwent transtracheal intubation at three to four days afterinitiation of translaryngeal intubation

2. Patients assigned to the late group had transtracheal intubation performed 14 daysafter the initiation of translaryngeal intubation, if continued intubation was required.Trachestomy (early and late) method: the incision was standardized as a vertical softtissue incision and a vertical incision through the second and third tracheal rings andthe upper half of the fourth ring without removing any tracheal tissue



Dunham 1984 (Continued)

Outcomes • Significant laryngotracheal pathology (irrespective of type) that required surgeryand/or prolonged tracheal intubation beyond that required for the patient’s generalcondition

• Respiratory sepsis (tracheitis, pneumonia, lung abscess, and peristomal infection)• Major complications (not explicitly defined by the authors)• Complications (self-extubation, patient tolerance, respiratory hygiene, and

aspiration)• Proportion of patients intubated for until 21 days

Notes Method to predict prolonged artificial ventilation: patients were randomized into an earlyor late tracheostomy group If at the end of 48 to 72 hours of translaryngeal intubationthe attending surgeon felt that a patient needed at least 48 hours of additional trachealintubationGender: Not informed

Risk of bias



High risk Randomization was based on the last digitof the patient’s hospital number; evennumbers comprised the early tracheostomygroup and odd numbers comprised the latetracheostomy groupThere were substantial differences of fre-quency of baseline characteristics betweencomparison groups (rigid head injury, non-head injury and non-rigid, head injury)

Allocation concealment (selection bias) High risk The method of sequence generation usedin this study ideally permits anyone to fore-see the group by which each one of the par-ticipants would be allocated




High risk After randomization, only patients whowere intubated for at least seven days wereincluded. The authors did not inform per-centages or number of patients not consid-ered for analysis after randomization

Selective reporting (reporting bias) High risk Authors did not reported mortality



Dunham 1984 (Continued)

Other bias High risk Fifty per cent of the patients were evalu-ated at four to six months after extubation;the remaining patients were interviewed 12months after extubation and there was noindication about the number of patients byeach one of the comparison groupsStatistical differences at baseline not in-formed. No explicit information about thetimes of follow-up

Rumbak 2004

Methods Study design: Parallel randomized controlled trial without intention-to-treat analysisLocal/Setting: Medical Intensive Units at the Baptist Memorial Hospital, Universityof Tennessee, Memphis, TN, and Tampa General and the James A. Haley VeteransAdministration Hospital, University of South Florida, Tampa, FL

Participants • N: 120 (Early tracheostomy, n = 60 / Late tracheostomy, n = 60)• Mean age: 63 years• Gender: 65 male / 55 female• Projected to need ventilation support for >14 days• initial Acute Physiology and Chronic Health Evaluation (APACHE) II score > 25

Interventions 1. Tracheotomy within 48 hours after intubation2. Late tracheotomy at days 14 to16

Patients from both groups were subjected to percutaneous Dilational Tracheotomy Pro-cedure

Outcomes • Mortality• Mean intensive care stay:• Days mechanically ventilated:• Days sedated• Days on high-dose pressors• Pneumonia• Ventilator-associated pneumonia and death• Gastrointestinal bleed and death• Acute myocardial infarction and death• Pulmonary embolus and death• Intractable septic shock and death• Withdrawal of life support and death• Respiratory failure and death• Tracheal stenosis 0-20 (in-hospital)• Tracheal stenosis 21-50 (in-hospital)• Tracheal stenosis >50 (in-hospital)• Tracheal stenosis 0-20 (10-wk postintubation)• Tracheal stenosis 21-50 (10-wk postintubation)• Tracheal stenosis >50 (10-wk postintubation)• Self-extubation



Rumbak 2004 (Continued)

Notes Method to predict prolonged artificial ventilation: Not explicitly reported

Risk of bias



Unclear risk Not explicitly reported.The groups were similar in mean age, pro-portion of women and African Americans,APACHE II score, and underlying diseases

Allocation concealment (selection bias) Low risk Independent group randomization wasplaced in sequentially numbered envelopesto be open once consent was signed




Low risk Although there was no intention-to-treatanalysis, the patients flow within the studywas clear

Selective reporting (reporting bias) Low risk Clinically relevant outcomes were reportedby the authors.

Other bias High risk The authors (except SWS) performed thepercutaneous dilational tracheotomyAlthough, the airways were assessed fororal, laryngeal, and tracheal damage at 10weeks postintubation, there is no explicitinformation about time of follow-up for theother outcomes

Terragni 2010

Methods Study design: Parallel randomized controlled trial with intention-to-treat analysisLocal/Setting: Italian Intensive Care Units

Participants • N= 419 (Early tracheostomy, n = 209 / Late tracheostomy, n = 210)• Mean age: 61.5 years old• Gender: 138 male / 142 female• Mechanically ventilated for acute respiratory failure for 24 hours;• Simplified Acute Physiology Score II between 35 and 65;• Sequential organ failure assessment (SOFA) score > 5;• Without a pulmonary infection (estimated by a Clinical Pulmonary Infection

Score [CPIS] of <6), chronic obstructive pulmonary disease, anatomical deformity of



Terragni 2010 (Continued)

the neck (including thyromegaly) and cervical tumors; a history of oesophageal,tracheal, or pulmonary cancer; previous tracheotomy; soft tissue infection of the neck;hematological malignancy; or pregnant

• PaO2 ≤ 60 mm Hg• Fraction of inspired oxygen (FIO2) ≥0.5• Positive end-expiratory pressure (PEEP) ≥8 cm H2O• Acute clinical condition requiring ventilatory support and still unresolved• SOFA score ≥5

Interventions 1. Early tracheostomy: After six to eight days of laryngeal intubation2. Late tracheostomy: after 13 to15 days of laryngeal intubation

Patients from both groups were subjected to percutaneous tracheostomy

Outcomes • Mortality (28 days)• Mortality (at one year)• Need for a long-term care facility• Ventilator-free days (at day 28)• ICU-free days (at day 28)• ICU discharge• Successful weaning• Hospital length of stay• Ventilator associated pneumonia• Intraoperative adverse events (minor bleeding, significant bleeding, tube

dislocation, hypoxaemia, arrhythmia, cardiac arrest)• Postoperative adverse events (stoma inflammation, stoma infection, minor

bleeding, major bleeding, pneumothorax, subcutaneous emphysema,tracheoesophageal fistula, cannula displacement or need for replacement)

Notes Method to predict prolonged artificial ventilation: mechanically ventilated for acuterespiratory failure for 24 hours; Simplified Acute Physiology Score II between 35 and65; sequential organ failure assessment (SOFA) score ≥ 5

Risk of bias



Low risk Computer-generated randomizationschedule.Baseline characteristics at admission or be-fore randomization did not differ betweenthe two groups

Allocation concealment (selection bias) Low risk Randomization conducted centrally us-ing a computer generated randomizationschedule


Low risk According to the authors, ”a clinicianblinded to patient allocation looked at theclinical charts remotely and evaluated the



Terragni 2010 (Continued)

nonobjective components of the CPIS (qual-ity of secretions, chest x-ray, evidence of acuterespiratory distress syndrome)”.


High risk Intention-to-treat analysis (ITT) was an-nounced by the authors, but the exactmethod was not clearly reported. The num-bers of patients from the randomization tothe analysis of each outcomes were not co-herent. The authors of this systematic re-view could not perform the most acceptedITT analysis performed by imputing theundesired outcome for all patients whodropped-out and/or for those who with-drawn from the planned protocolThe addition of the number of patientswho died at 28 days and one year (reportedin the article) plus the number of with-drawals from protocol exceeds the numberof randomized patients. Similarly, the addi-tion of all explicitly reported deaths at the28 days and at one year of follow-up plusthe difference between the total of random-ized participants minus the number of par-ticipants “who left the hospital alive” is in-compatible with the total number of ran-domized participantsThe difference between the total of ran-domized patients minus the number of par-ticipants “who left the hospital alive” is in-coherent with the number of death at 28days

Selective reporting (reporting bias) Low risk None suspected. Clinically relevant out-comes were analysed.

Other bias High risk There was no explicit information aboutthe time of follow-up for some postop-erative adverse events: stoma inflamma-tion, stoma infection, minor bleeding, ma-jor bleeding, pneumothorax, subcutaneousemphysema, tracheoesophageal fistula)



Characteristics of excluded studies [ordered by study ID]

Study Reason for exclusion

Blot 2008 Early tracheotomy versus prolonged endotracheal intubation. No data available for patients subjected to the latetracheostomy

Bouderka 2004 Early tracheostomy versus prolonged endotracheal intubation.

El-Naggar 1976 Early tracheostomy versus prolonged endotracheal intubation.

Rodriguez 1990 Late tracheostomy >8 days after admission (before 10 days).

Saffle 2002 Early tracheostomy versus continued endotracheal intubation with no data available specifically for patients sub-jected to the late tracheostomy

Stauffer 1981 Early tracheostomy versus continued endotracheal intubation.

Sugerman 1997 Early tracheostomy versus continued endotracheal intubation with no data available for patients subjected to thelate tracheostomy

Characteristics of ongoing studies [ordered by study ID]

Dumire 2008

Trial name or title A prospective, randomized trial of early versus conventional conversion from endotracheal intubation topercutaneous tTracheostomy for ventilatory support of trauma patients with severe brain injury

Methods Open label parallel randomized controlled trial

Participants - 18 years of age or older

- TBI defined as penetrating or blunt brain injury including1. subarachnoid haemorrhage2. subdural haemorrhage3. epidural haemorrhage4. brain contusion5. diffuse axonal injury

- mechanically ventilated by endotracheal intubation

- projected to need ventilation support for more than 14 days according to: GCS measured in field less thanor equal to eight and a GCS on day three which remains less than or equal to eight

- informed consent obtained from patient or legal representative

Interventions 1. Early tracheostomy (less than or equal to 72 hours)2. Late tracheostomy (10 to 14 days).



Dumire 2008 (Continued)

Outcomes • Total number of mechanical ventilation days until discharged• Total number of hospital days until discharged• Incidence of ventilator-associated pneumonia until discharged• Incidence of accidental extubation until discharged• Incidence of death until discharged

Starting date February 2006

Contact information Pennsylvania, United StatesMemorial Medical CenterJohnstownPennsylvania15905

Notes

Huttner 2010

Trial name or title WEANING-Study: “Weaning by early versus late tracheostomy in supratentorIal intracerebral bleedings”


Participants - Patients requiring intubation/mechanical ventilation

- Supratentorial intracerebral haemorrhage (including:)

- primary spontaneous ICH (lobar/deep)

- ICH related to anticoagulant therapy

- with or without intraventricular haemorrhage

- with or without occlusive and/or communicating hydrocephalus

- Haematoma volume > 0 ml and < 60 ml

- Age 18 - 85 years

- Informed consent (legal representative)

Interventions “Early” tracheostomy within 72 hours after hospital admission“Late” tracheostomy (= control group; undergoing conventional tracheostomy between day 12 to14 if extu-bation fails). Both groups receive plastic tracheostomy

Outcomes • Cumulative time requiring mechanical ventilation and Overall duration of neurocritical care 30 days• Incidence of respirator-associated pneumonia 30 days• Cumulative consumption of sedative drugs 30 days• Incidence of episodes with increased intracranial pressure 30 days



Huttner 2010 (Continued)

• In-hospital mortality 30 days• Three-months functional outcome (mRS) 90 days No functional outcome after three months using

the modified Rankin Scale

Starting date July 2010

Contact information Hagen B. Huttner, MDtel: [email protected]

Notes

Kluge 2009

Trial name or title Early versus late percutaneous dilation tracheostomy in mechanically ventilated patients with chronic ob-structive pulmonary disease


Participants - >18 years old- Diagnosis of COPD (GOLD stage III or IV)- Suspected long-time invasive mechanical ventilation due to ARF (> 10 days)- Informed consent of the patient or legal guardian

Interventions Early tracheostomy: tracheostomy at the next possible opportunity but not later than 72 h after initiation ofinvasive ventilationPatients of the control group will be invasively ventilated at least until Day 10

Outcomes • Cumulative duration of mechanical ventilation (in days) Day 1 to 28• All-cause mortality Day 28, 90 and end of ICU stay• Length of stay on ICU / hospital end of ICU / hospital stay• Infections (ventilator-associated pneumonia, spectrum of pathogens in BALF, infectious

complications) Day 1 to28• Cumulative use of sedatives Day 1 to 28• Quality of Life discharge from ICU, day 28 and day 90

Starting date October 2009

Contact information Stefan Kluge, MDtel: +4940 7410 ext.: [email protected]

Notes



http://mailto:[email protected]



Ranieri 2009

Trial name or title Efficacy of early tracheostomy to reduce incidence of ventilator acquired pneumonia (VAP)


Participants - Oro/nasotracheal intubation for less than three days- Simplified Acute Physiology Score (SAPS II) between 35 and 65 upon admission toIntensive Care Unit (ICU)

Interventions Early tracheostomy on day three to fiveLate tracheostomy on days 10-12

Outcomes • Increase of “ventilator associated pneumonia-free days”. Follow-up terminates on day 28 from the dateof oro/nasotracheal intubation

• Increase of “ventilator-free days”. Follow-up terminates on day 28 from the date of oro/nasotrachealintubation

• Reduction of mortality one year

Starting date June 2004

Contact information ItalyUniversity of Turin, Department of Anesthesia and Intensive Care MedicineTurin10126

Notes

Young 2008

Trial name or title Tracheostomy Management in Critical Care

Methods Randomized controlled trial

Participants Eligible patients:a. Are intubatedb. High chance of requiring a further seven days or more of ventilatory support during their ICU stayc. Have been in the intensive care unit for less than four daysd. The recruiting consultant is uncertain about whether an ’early’ or ’late’ tracheostomy is more appropriatefor this patient

Interventions Group 1: ’Early tracheostomy’Tracheostomy on day one to four post ICU admissionGroup 2: ’Late tracheostomy’No tracheostomy before day 10 post ICU admission

Outcomes • Mortality 30 days after randomisation.• Mortality rate at discharge from hospital• ICU length of stay• Hospital length of stay (acute hospitals)• Mortality rate at (first) discharge from ICU



Young 2008 (Continued)

• Number of days receiving any sedative medication• Number of antibiotic-free days

Starting date 01/10/2004

Contact information c/o Lesley MorganKadoorie Centre for Critical Care Research and EducationLevel 3John Radcliffe HospitalOxfordUnited KingdomOX3 9DU

Notes



D A T A A N D A N A L Y S E S

Comparison 1. Early versus late tracheostomy

Outcome or subgroup titleNo. ofstudies

No. ofparticipants Statistical method Effect size

1 Mortality at maximal follow-uptime available

3 Risk Ratio (M-H, Random, 95% CI) Totals not selected

2 Pneumonia 2 Risk Ratio (M-H, Random, 95% CI) Totals not selected

Analysis 1.1. Comparison 1 Early versus late tracheostomy, Outcome 1 Mortality at maximal follow-up

time available.

Review: Early versus late tracheostomy for critically ill patients

Comparison: 1 Early versus late tracheostomy

Outcome: 1 Mortality at maximal follow-up time available

Study or subgroup Early tracheostomy Late tracheostomy Risk Ratio Risk Ratio

n/N n/N

M-H,Random,95%

CI

M-H,Random,95%

CI

Barquist 2006 2/29 5/31 0.43 [ 0.09, 2.03 ]

Rumbak 2004 19/60 37/60 0.51 [ 0.34, 0.78 ]

Terragni 2010 127/209 151/210 0.85 [ 0.74, 0.97 ]

0.1 0.2 0.5 1 2 5 10

Favours early trach Favours late trach



Analysis 1.2. Comparison 1 Early versus late tracheostomy, Outcome 2 Pneumonia.

Review: Early versus late tracheostomy for critically ill patients

Comparison: 1 Early versus late tracheostomy

Outcome: 2 Pneumonia

Study or subgroup Early tracheostomy Late tracheostomy Risk Ratio Risk Ratio

n/N n/N

M-H,Random,95%

CI

M-H,Random,95%

CI

Rumbak 2004 3/60 15/60 0.20 [ 0.06, 0.66 ]

Terragni 2010 30/209 44/210 0.69 [ 0.45, 1.05 ]

0.01 0.1 1 10 100

Favours early trach Favours late trach

A D D I T I O N A L T A B L E S

Table 1. Continuous and dichotomous outcomes not possible to be combined in a meta-analysis

Outcome n Estimate effect (MD or RR,95% CI, P, NNT, CI 95% forNNT)

Favoured group Study

1. Mortality at 28 days 419 RR 0.84 [0.62, 1.13, P = 0.25] Early tracheostomy Terragni 2010

2. ITT analysis of com-bined mortality (at 28days and at one year)

419 RR 0.85 [0.74, 0.97, P = 0.02] Early tracheostomy Terragni 2010

3. Mortality (in hospital-30 days)

60 RR 0.51 [0.34, 0.78, P = 0.002] Early tracheostomy Rumbak 2004

4. Mortality (in hospital) 120 RR 0.43 [0.09, 2.03, P = 0.29] Early tracheostomy Barquist 2006

5.Time spent on mechani-cal ventilation (days)

120 MD -9.80 [-11.48, -8.12, P<0.001]

Early tracheostomy Rumbak 2004

6. Intubation for morethan 21 days

74 RR 0.85 [0.53, 1.36, P = 0.49] Late tracheostomy Dunham 1984

7. Successful weaning 419 RR 0.88 [0.78, 0.99, P = 0.03,NNT=11]

Early tracheostomy Terragni 2010

8. ICU discharge at day28

419 RR 0.81 [0.65, 1.01, P = 0.06,NNT =11]

Early tracheostomy Terragni 2010



Table 1. Continuous and dichotomous outcomes not possible to be combined in a meta-analysis (Continued)

9. Time spent on ICU(days)

120 -11.40 [-12.42, -10.38, P<0.001]


10. Postoperative adverse events

10.1 Stoma inflamma-tion

264 RR 1.00 [0.57, 1.78, P = 0.99] Late tracheostomy Terragni 2010

10.2 Stoma infection 264 RR 1.06 [0.41, 2.75, P = 0.91] Late tracheostomy Terragni 2010

10.3 Minor bleeding 264 RR 1.09 [0.39, 3.07, P = 0.86] Late tracheostomy Terragni 2010

10.4 Major bleeding 264 RR 0.82 [0.17, 3.99, P = 0.81] Early tracheostomy Terragni 2010

10.5 Pneumothorax 264 RR 2.47 [0.10, 59.98, P = 0.58] Late tracheostomy Terragni 2010

10.6 Subcutaneous em-physema


10.7 Tracheoesophagealfistula


10.8 Cannula displace-ment or need for replace-ment

264 RR 4.11 [0.20, 84.78, P=0.36] Late tracheostomy Terragni 2010

10.9 Significant laryngo-tracheal pathology

74 RR 1.41 [0.47, 4.22, P=0.54] Late tracheostomy Dunham 1984

10.10 Respiratory sepsis 74 RR 1.18 [0.77, 1.79, P=0.45] Late tracheostomy Dunham 1984

10.11 Major complica-tions

74 RR 1.41 [0.47, 4.22, P=0.54] Late tracheostomy Dunham 1984

10.12 Complications 74 RR 1.41 [0.47, 4.22, P=0.54] Late tracheostomy Dunham 1984

10.13 Tracheal stenosis0-20 (in hospital)

120 RR 1.27 [1.04, 1.55, P = 0.02,NNT=10]

Late tracheostomy Rumbak 2004

10.14 Tracheal stenosis21-50 (in hospital)


10.15 Tracheal stenosis>50 (in hospital)


10.16 Tracheal stenosisirrespective of the sever-ity (in hospital)

120 RR 1.03 [0.98, 1.09, P = 0.24] Late tracheostomy Rumbak 2004



Table 1. Continuous and dichotomous outcomes not possible to be combined in a meta-analysis (Continued)

10.17 Tracheal stenosis0-20 (10-wk postintuba-tion)

120 RR 2.00 [1.14, 3.51, P = 0.02,NNT = 4.54]


10.18 Tracheal stenosis21-50 (10-wk postintu-bation)


10.19 Tracheal stenosis>50 (10-wk postintuba-tion)


10.20 Tracheal stenosisirrespective of the sever-ity (10-wk postintuba-tion)

120 RR 1.78 [1.24, 2.57, P = 0.002,NNT = 3.33]


10.21 Self-extubation 120 RR 0.08 [0.00, 1.34, P = 0.08] Early tracheostomy Rumbak 2004

A P P E N D I C E S

Appendix 1. Glossary of terms

Term Definition

COPD (Chronic obstructive pulmonary disease) A disease of chronic diffuse irreversible airflow obstruction. Subcategories ofCOPD include chronic bronchitis and pulmonary emphysema

Critically ill adults Adults with a disease or in a state in which death is possible or imminent

Early tracheostomy Although not precisely defined, it usually refers to a tracheostomy performedfrom two days to 10 days after intubation

Late tracheostomy Although not precisely defined, it usually refers to a tracheostomy performedafter 10 days of intubation

Percutaneous tracheostomy Usually a tracheostomy based on: 1) needle-guide wire airway access followed byserial dilations with sequentially larger dilators; 2) guide wire dilating forceps; 3)mini tracheostomy only for emergency airway access or for aspiration of retainedbronchopulmonary secretions



(Continued)

Pneumomediastinum or mediastinal emphysema Presence of air in the mediastinal tissues due to leakage of air from the tracheo-bronchial tree, usually as a result of trauma

Pneumothorax An accumulation of air or gas in the pleural space, which may occur spontaneouslyor as a result of trauma or a pathological process, or be deliberately introduced

Prolonged mechanical ventilation At least 21 consecutive days for six or more hours per day of any method of artificalbreathing that employs mechanical or non-mechanical means to force the air intoand out of the lungs. Artificial respiration or ventilation is used in individualswho have stopped breathing or have respiratory insufficiency to increase theirintake of oxygen (O2) and excretion of carbon dioxide (CO2).

Ramsay score (Ramsay 2000) Numeric scale of responses to verbal, tactile or nociceptive stimuli

Self-extubations Unplanned removal of an endotracheal airway tube by a patient

Severe hypoxia Referred to as low oxygen levels or anoxia and is a relatively common cause ofinjury to the central nervous system. Prolonged brain anoxia may lead to braindeath or a persistent vegetative state. Histologically, this condition is characterizedby neuronal loss which is most prominent in the hippocampus; globus pallidus;cerebellum; and inferior olives

Surgical tracheostomy Tracheostomy performed by surgeons in the operating theatre using an opentechnique

Tracheal aspiration Aspiration or suctioning of oropharyngeal secretions past tracheal cuffs into thelungs in mechanically ventilated patients (usually in the intensive care unit (ICU))

Ventilator-associated pneumonia Serious inflammation of the lungs in patients who required the use of a pulmonaryventilator. it is usually caused by cross-bacterial infections in hospitals (nosocomialinfections)

Appendix 2. Search strategy - MEDLINE (via the PubMed interface)

#1 (“Tracheostomy”[mesh] OR Tracheostomies OR Tracheotomy OR Tracheotomies) AND (((early or precocious or premature) and(late or tardy)) OR (Artificial Respiration) OR (Artificial Respirations) OR (“Respirations, Artificial”[mesh]) OR (Ventilation, Mechan-ical) OR (Mechanical Ventilation) OR (Mechanical Ventilations) OR (Ventilations, Mechanical) OR (High-Frequency-Ventilation)OR (High-Frequency Ventilations) OR (Ventilations, High-Frequency) OR (Ventilation, High Frequency) OR (Ventilation, High-Frequency) OR (High Frequency Ventilation) OR (High Frequency Ventilations) OR (Ventilations, High Frequency) OR (High-Fre-quency Oscillation Ventilation) OR (High-Frequency Oscillation Ventilations) OR (Oscillation Ventilation, High-Frequency) OR (Os-cillation Ventilations, High-Frequency) OR (Ventilation, High-Frequency Oscillation) OR (Ventilations, High-Frequency Oscillation)OR (High Frequency Oscillation Ventilation) OR (High-Frequency Positive Pressure Ventilation) OR (High Frequency Positive Pres-sure Ventilation) OR (High-Frequency Jet Ventilation) OR (High-Frequency Jet Ventilations) OR (Jet Ventilation, High-Frequency)OR (Jet Ventilations, High-Frequency) OR (Ventilation, High-Frequency Jet) OR (Ventilations, High-Frequency Jet) OR (Ventilation,High Frequency Jet) OR (High Frequency Jet Ventilation) OR (Ventilator-Weaning) OR (Ventilator Weaning) OR (Weaning, Ventila-tor) OR (Respirator Weaning) OR (Weaning, Respirator) OR (Mechanical Ventilator Weaning) OR (Ventilator Weaning, Mechanical)OR (Weaning, Mechanical Ventilator) OR (Ventilators, Mechanical) OR (Mechanical Ventilator) OR (Mechanical Ventilators) OR



(Ventilator, Mechanical) OR (Ventilator, Pulmonary) OR (Pulmonary Ventilators) OR (Pulmonary Ventilator) OR (Respirators) OR(Respirator) OR (Ventilators, Pulmonary) OR (Ventilators) OR (Ventilator) OR (Respiration, Artificial) OR (Artificial Respiration)OR (Artificial Respirations) OR (Respirations, Artificial) OR (Ventilation, Mechanical) OR (Mechanical Ventilation) OR (MechanicalVentilations) OR (Ventilations, Mechanical) OR (Chest Tubes) OR (Chest Tubes) OR (Chest Tube) OR (Tube, Chest) OR (Tubes,Chest) OR (Laryngeal Masks) OR (Laryngeal Mask) OR (Mask, Laryngeal) OR (Masks, Laryngeal) OR (Laryngeal Mask Airway) OR(Airway, Laryngeal Mask) OR (Airways, Laryngeal Mask) OR (Laryngeal Mask Airways) OR (Ventilators, Mechanical) OR (Mechan-ical Ventilator) OR (Mechanical Ventilators) OR (Ventilator, Mechanical) OR (Ventilator, Pulmonary) OR (Pulmonary Ventilators)OR (Pulmonary Ventilator) OR (Respirators) OR (Respirator) OR (Ventilators, Pulmonary) OR (Ventilators) OR (Ventilator))#2 ((randomized controlled trial [pt]) OR (controlled clinical trial [pt]) OR (randomized [tiab]) OR (placebo [tiab]) OR (drug therapy[sh]) OR (randomly [tiab]) OR (trial [tiab]) OR (groups [tiab])) AND (humans [mh])#3 #1 and #2

Appendix 3. Search strategy - The Cochrane Library

(Tracheostomy OR Tracheostomies OR Tracheotomy OR Tracheotomies) AND (((early or precocious) and (late)) OR (Artificial Res-piration) OR (Artificial Respirations) OR (Ventilation, Mechanical) OR (Mechanical Ventilation) OR (Mechanical Ventilations) OR(High-Frequency-Ventilation) OR (High-Frequency Ventilations) OR (High Frequency Ventilation) OR (High Frequency Ventila-tions) OR (High-Frequency Oscillation Ventilation) OR (High-Frequency Oscillation Ventilations) OR (High Frequency OscillationVentilation) OR (High-Frequency Positive Pressure Ventilation) OR (High Frequency Positive Pressure Ventilation) OR (High-Fre-quency Jet Ventilation) OR (High-Frequency Jet Ventilations) OR (High Frequency Jet Ventilation) OR (Ventilator-Weaning) OR(Ventilator Weaning) OR (Respirator Weaning) OR (Mechanical Ventilator Weaning) OR (Mechanical Ventilator) OR (MechanicalVentilators) OR (Pulmonary Ventilators) OR (Pulmonary Ventilator) OR (Respirators) OR (Respirator) OR (Ventilators) OR (Venti-lator) OR (Artificial Respiration) OR (Artificial Respirations) OR (Mechanical Ventilation) OR (Mechanical Ventilations) OR (ChestTubes) OR (Chest Tubes) OR (Chest Tube) OR (Laryngeal Masks) OR (Laryngeal Mask) OR (Laryngeal Mask Airway) OR (LaryngealMask Airways) OR (Mechanical Ventilator) OR (Mechanical Ventilators) OR (Pulmonary Ventilators) OR (Pulmonary Ventilator)OR (Respirators) OR (Respirator) OR (Ventilators) OR (Ventilator))

Appendix 4. Search strategy - EMBASE

#1 (early OR precocious AND late OR ’high frequency ventilation’/exp OR ’high frequency ventilation’ OR (’high frequency’ ANDventilations) OR (ventilations, AND ’high frequency’) OR (ventilation, AND high AND frequency) OR (ventilation, AND ’highfrequency’) OR (high AND frequency AND (’ventilation’/exp OR ventilation)) OR (high AND frequency AND ventilations) OR(ventilations, AND high AND frequency) OR (’high frequency’ AND (’oscillation’/exp ORoscillation) AND (’ventilation’/exp ORventilation)) OR (’high frequency’ AND (’oscillation’/exp OR oscillation) AND ventilations) OR (’oscillation’/exp OR oscillation ANDventilation, AND ’high frequency’) OR (’oscillation’/exp OR oscillation AND ventilations, AND ’high frequency’) OR (ventilation,AND ’high frequency’ AND (’oscillation’/exp OR oscillation)) OR (ventilations, AND ’high frequency’ AND (’oscillation’/exp ORoscillation)) OR (high AND frequency AND (’oscillation’/exp OR oscillation) AND (’ventilation’/exp OR ventilation)) OR (’highfrequency’ AND positive AND (’pressure’/exp OR pressure) AND (’ventilation’/exp OR ventilation)) OR (high AND frequencyANDpositive AND (’pressure’/exp OR pressure) AND (’ventilation’/exp OR ventilation)) OR (’high frequency’ AND jet AND(’ventilation’/exp ORventilation)) OR (’high frequency’ AND jet AND ventilations) OR (jet AND ventilation, AND ’high frequency’)OR (jet AND ventilations, AND’high frequency’) OR (ventilation, AND ’high frequency’ AND jet) OR (ventilations, AND ’highfrequency’ AND jet) OR (ventilation, AND highAND frequency AND jet) OR (high AND frequency AND jet AND (’ventilation’/exp OR ventilation)) OR ’ventilator weaning’/exp OR ’ventilator weaning’ OR (’ventilator’/exp OR ventilator AND (’weaning’/exp OR weaning)) OR (weaning, AND (’ventilator’/exp OR ventilator)) OR (’respirator’/exp OR respirator AND (’weaning’/expOR weaning)) OR (weaning, AND (’respirator’/exp OR respirator)) OR (mechanical AND (’ventilator’/exp OR ventilator) AND(’weaning’/exp OR weaning)) OR (’ventilator’/exp OR ventilator AND weaning, AND mechanical) OR (weaning, AND mechanicalAND (’ventilator’/exp OR ventilator)) OR (respiration, AND artificial) OR (artificial AND (’respiration’/exp OR respiration)) OR(artificial AND respirations) OR (respirations, AND artificial) OR (ventilation, AND mechanical) OR (mechanical AND (’ventilation’/exp OR ventilation)) OR (mechanical AND ventilations) OR (ventilations, AND mechanical) OR (’chest’/exp OR chest ANDtubes)OR (’chest’/exp OR chest AND (’tube’/exp OR tube)) OR (tube, AND (’chest’/exp OR chest)) OR (tubes, AND (’chest’/exp OR chest))OR (laryngeal AND (’masks’/exp OR masks)) OR (laryngeal AND (’mask’/exp OR mask)) OR (mask, AND laryngeal) OR (masks,ANDlaryngeal) OR (laryngeal AND (’mask’/exp OR mask) AND (’airway’/exp OR airway)) OR (airway, AND laryngeal AND (’mask’/



exp OR mask)) OR (airways, AND laryngeal AND (’mask’/exp OR mask)) OR (laryngeal AND (’mask’/exp OR mask) AND airways)OR (ventilators, ANDmechanical) OR (mechanical AND (’ventilator’/exp OR ventilator)) OR (mechanical AND ventilators) OR(ventilator, AND mechanical) OR (ventilator, AND pulmonary) OR (pulmonary AND ventilators) OR (pulmonary AND (’ventilator’/exp OR ventilator)) OR respirators OR’respirator’/exp OR respirator OR (ventilators, AND pulmonary) OR ventilators OR ’ventilator’/exp OR ventilator) AND (’tracheostomy’/exp OR tracheostomy OR tracheostomies OR ’tracheotomy’/exp OR tracheotomy ORtracheotomies)#2 (random$ OR factorial$ OR crossover$ OR (cross AND over$) OR ’cross over$’ OR placebo$ OR (doubl$ AND adj AND blind$)OR (singl$AND adj AND blind$) OR assign$ OR allocat$ OR volunteer$ OR ’crossover procedure’/exp OR ’crossover procedure’OR (’double blind’ ANDprocedure) OR (randomized AND controlled AND trial) OR (’single blind’ AND procedure))#3 #1 and #2

Appendix 5. Search strategy - LILACS

(Tracheostomy OR Traqueostomia OR Tracheostomies OR Traqueostomias OR Tracheotomy OR Tracheotomies) AND (((early orprecocious) and (late)) OR (Artificial Respiration) OR (Artificial Respirations) OR (Ventilation, Mechanical) OR (Mechanical Ven-tilation) OR (Mechanical Ventilations) OR (High-Frequency-Ventilation) OR (High-Frequency Ventilations) OR (High FrequencyVentilation) OR (High Frequency Ventilations) OR (High-Frequency Oscillation Ventilation) OR (High-Frequency Oscillation Ven-tilations) OR (High Frequency Oscillation Ventilation) OR (High-Frequency Positive Pressure Ventilation) OR (High FrequencyPositive Pressure Ventilation) OR (High-Frequency Jet Ventilation) OR (High-Frequency Jet Ventilations) OR (High Frequency JetVentilation) OR (Ventilator-Weaning) OR (Ventilator Weaning) OR (Respirator Weaning) OR (Mechanical Ventilator Weaning) OR(Mechanical Ventilator) OR (Mechanical Ventilators) OR (Pulmonary Ventilators) OR (Pulmonary Ventilator) OR (Respirators) OR(Respirator) OR (Ventilators) OR (Ventilator) OR (Artificial Respiration) OR (Artificial Respirations) OR (Mechanical Ventilation)OR (Mechanical Ventilations) OR (Chest Tubes) OR (Chest Tubes) OR (Chest Tube) OR (Laryngeal Masks) OR (Laryngeal Mask) OR(Laryngeal Mask Airway) OR (Laryngeal Mask Airways) OR (Mechanical Ventilator) OR (Mechanical Ventilators) OR (PulmonaryVentilators) OR (Pulmonary Ventilator) OR (Respirators) OR (Respirator) OR (Ventilators) OR (Ventilator))

Appendix 6. Search strategy - Current Controlled Trials

(tracheostomy or tracheostomies) and (timing or ((early or precocious) and (late or later)))

Appendix 7. Search strategy - PEDro

tracheostomy or tracheostomies

Appendix 8. Search strategy - CINAHL

(Tracheostomy OR Tracheostomies) and (early and late)

Appendix 9. Extraction Sheet


Study ID: Date of Study (year): Review ID:

Reviewer:

Author (last name):Local of study:



I - ACTION

Contact author for:

II - PARTICIPANTS

Participantsa. N:b. Age (mean):c. Diagnosis (e.g., burning, lung disease, etc):d. Method to characterize patients as ’critically ill’e. Method to predict prolonged artificial ventilation:f. Gender:g. Setting:h. Statistical differences at baseline:

III - INTERVENTIONS

Early tracheostomy:Timing of tracheotomies [days (e.g., from day 0 of mechanical ventilation)]:

Type of tracheostomy, (e.g., percutaneous or surgical tracheostomy)Late tracheostomy:Timing of tracheotomies [days (e.g., from day 0 of mechanical ventilation)]:

Type of tracheostomy, (e.g., percutaneous or surgical tracheostomy)

IV - OUTCOMES

(final or change from baseline values)

Primary outcomes1. mortality (time to mortality or frequency at any time point: in hospital, in ICU, after discharge);

2. duration of artificial ventilation.

Secondary outcomes1. length stay in the ICU (or frequency at any time point);

2. ventilator-associated pneumonia at any time point;

3. laryngotracheal lesions at any time point (in epiglottis, vocal cord, larynx, subglottic ulceration and inflammation);

4. eating/vocal/speech problems.



V - METHODOLOGICAL QUALITY OF STUDY

Please, mark the appropriated itemWas the random allocation sequence adequately generated?Low risk:High risk:UnclearWas allocation adequately concealed?Low riskHigh riskUnclearBlinding: was knowledge of the allocated interventions adequately prevented during the study?Low riskHigh riskUnclearWere incomplete outcome data adequately addressed?Low riskHigh riskUnclearAre reports of the study free of suggestion of selective reporting?Low riskHigh riskUnclearOther bias?Low risk:High riskUnclear

VI - Observation (including non-published data)

Appendix 10. Non-parameric estimate effects not possible to be combined in a meta-analysis

Study ID Comparison groups Median Interquartile range P-value

Outcome: Ventilator free-days (at day 28)

Terragni 2010 Early tracheostomy 11 0-21 0.02

Late tracheostomy 6 0-17

Outcome: ICU free-days (at day 28)

Terragni 2010 Early tracheostomy 0 0-13

Late tracheostomy 0 0-8 0.02



(Continued)

Outcome: Hospital length of stay (days)

Terragni 2010 Early tracheostomy 31 17-39 not available

Late tracheostomy 32 18-59

Appendix 11. Other potentially relevant outcomes not planned in the protocol of this systematicreview

Outcome n Estimate effect (MD or RR,95% CI, P, NNT, CI 95% forNNT)

Favoured group Study

1. Need for a long-termcare facility

292 RR 1.09 [0.81, 1.46, P=0.59] Late tracheostomy Terragni 2010

2. Time spent on seda-tion (days)

120 MD -10.90 [-11.64, -10.16,P<0.00001]


3. Time spent on high-dose pressors (days)

120 MD 0.50 [-1.02, 2.02, P=0.52] Late tracheostomy Rumbak 2004

4. Any event followed by death

4.1 Ventilator-associatedpneumonia

120 RR 0.22 [0.05, 0.99, P=0.05,NNT=4.54]


4.2 Gastrointestinalbleed

120 RR 0.33 [0.04, 3.11, P=0.34] Early tracheostomy Rumbak 2004

4.3 Acute myocardial in-farction


4.4 Pulmonary embolus 120 RR 1.00 [0.06, 15.62, P=1.00] Early tracheostomy Rumbak 2004

4.5 Intractable septicshock


4.6 Withdrawal of lifesupport

120 RR 2.00 [0.19, 21.47, P=0.57] Late tracheostomy Rumbak 2004

4.7 Respiratory failure 120 RR 0.64 [0.26, 1.53, P=0.31] Early tracheostomy Rumbak 2004

5. Intraoperative adverse events



(Continued)

5.1 Minor bleeding 264 RR 0.55 [0.09, 3.22, P=0.50] Early tracheostomy Terragni 2010

5.2 Significant bleeding 264 no event in both groups - Terragni 2010

5.3 Tube dislocation 264 RR 0.55 [0.09, 3.22, P=0.50] Early tracheostomy Terragni 2010

5.4 Hypoxaemia 264 RR 1.15 [0.37, 3.53, P=0.81] Late tracheostomy Terragni 2010

5.5 Arrhyhtmia 264 no event in both groups - Terragni 2010

H I S T O R Y

Protocol first published: Issue 3, 2008

Review first published: Issue 3, 2012

Date Event Description

19 May 2010 Amended Contact details updated.

C O N T R I B U T I O N S O F A U T H O R S

Conceiving the review: Brenda NG Silva (BGS) and Álvaro N Atallah (ANA)

Co-ordinating the review: BGS and Humberto Saconato (HS)

Undertaking manual searches: BGS

Screening search results: BGS

Organizing retrieval of papers: BGS

Screening retrieved papers against inclusion criteria: BGS, RA and HS

Appraising quality of papers: BGS, Regis B Andriolo (RA) and HS

Abstracting data from papers: BGS, RA and Orsine Valente (OV)

Writing to authors of papers for additional information: BGS and RA

Providing additional data about papers: BGS and RA

Obtaining and screening data on unpublished studies: BGS and RA

Data management for the review: BGS and RA

Entering data into Review Manager (RevMan 5.1): BGS and RA

RevMan statistical data: BS and RA

Other statistical analysis not using RevMan: RA



Double entry of data: (data entered by person one - RA)

Interpretation of data: BGS, OV and RA

Statistical inferences: BGS, HS and RA

Writing the review: BGS, OV and RA

Guarantor for the review (one author): BGS

Persons responsible for reading and checking review before submission: OV and ANA

D E C L A R A T I O N S O F I N T E R E S T

The lead author (Brenda NG Silva) has been working as a respiratory therapist since 2002.

All other authors: none known.

D I F F E R E N C E S B E T W E E N P R O T O C O L A N D R E V I E W

Types of outcome measures

The following sentence was inserted in the final version ’Types of outcome measures’.

Evaluation of the internal validity of included studies

At the time the protocol was prepared, the items of internal validity were compatible with previous versions of the Cochrane Handbook(Higgins 2011c), as listed below. However, the first full version of this review was prepared according to the updated Cochrane Handbook(Higgins 2011a).

Selection bias

Was allocation concealment adequate and were data similar at baseline?A: adequate allocation concealment and similar descriptive data between arms at baseline;B: not described;C: not adequate.

Detection bias

Was there a blinded assessment of outcomes?Met: assessors unaware of the assigned treatment when collecting outcome measures;Unclear: blinding of assessor not reported and cannot be verified by contacting investigators;Not met: assessors aware of the assigned treatment when collecting outcome measures.



Attrition bias

Were there any withdrawals described and were they acceptable?Met: no substantial loss of participants after randomization or difference between comparison groups not statistically significant;Unclear: losses not reported by the authors;Not met: substantial loss of participants after randomization or statistically significant difference in losses between comparison groups.

Performance bias

We will not use blinding of providers and patients as a criterion to assess internal validity of included trials because of the nature of theintervention.


The inclusion of a study with different times of early and late tracheostomies than our inclusion criteria was considered in a sensitivityanalysis.



early versus late tracheostomy for critically ill patients · timing of tracheostomy for critically...

Documents