bmj.c7237

7/31/2019 bmj.c7237

1/14

RESEARCH

Use of weaning protocols for reducing duration ofmechanicalventilationincriticallyilladultpatients:Cochranesystematic review and meta-analysis

Bronagh Blackwood, lecturer in nursing,1 Fiona Alderdice, director,1 Karen Burns, clinician scientist,2 Chris

Cardwell, lecturer in medical statistics,3 Gavin Lavery, consultant in intensive care medicine,4 Peter

OHalloran, lecturer in nursing1

ABSTRACT

Objective To investigate the effects of weaning protocolson the total duration of mechanical ventilation, mortality,

adverse events, quality of life, weaning duration, and

length of stay in the intensive care unit and hospital.

Design Systematic review.

Data sources Cochrane Central Register of Controlled

Trials, Medline, Embase, CINAHL, LILACS, ISI Web of

Science, ISI Conference Proceedings, Cambridge

Scientific Abstracts, and reference lists of articles. We did

not apply language restrictions.

Review methods We included randomised and quasi-

randomised controlled trials of weaning from mechanical

ventilation with and without protocols in critically ill

adults.

Data selection Three authors independently assessed

trial quality and extracted data. A priori subgroup and

sensitivity analyses were performed. We contacted study

authors for additional information.

Results Eleven trials that included 1971 patients met the

inclusion criteria. Compared with usual care, the

geometric mean duration of mechanical ventilation in the

weaning protocol group was reduced by 25% (95%

confidence interval 9% to 39%, P=0.006; 10 trials); the

duration of weaning was reduced by 78% (31% to 93%,

P=0.009; six trials); and stay in the intensive care unit

length by 10% (2% to 19%, P=0.02; eight trials). There

was significant heterogeneity among studies for total

duration of mechanical ventilation (I2=76%, P

7/31/2019 bmj.c7237

2/14

through a ventilator circuit with the application of con-tinuous positive airway pressure; and combinations oftheseandneweroptions,suchasbi-levelpositiveairwaypressure. The evidence is equivocal as to which methodis superior, though it has been suggested that synchro-nised intermittent mechanical ventilation is the leasteffective method.12-14

Doctors have different experiences,skills, and wean-ing philosophies, and, in view of the potential for var-iation, there has been an increasing interest inproviding more consistent practice in intensive care

units by developing weaning protocols that providestructured guidance. Protocols are intended toimprove efficiency of practice by following an expertconsensus to reduce variation produced by the appli-cation of individual judgment and experience.15 Ingeneral, there are three components to a weaning pro-tocol. The first component is a list of objective criteria(often referred to as readiness to wean criteria) basedongeneralclinicalfactorstohelptodecideifapatientisready to breathe without the help of a ventilator, suchas that used by Ely and colleagues.16 The second com-ponent consists of structured guidelines for reducingventilatory support. This might be abrupt (for exam-

ple, spontaneous breathing trials on a T piece) or gra-dual with a stepwise reduction in mechanical support(for example, synchronised intermittent mechanicalventilation or pressure support ventilation) such asthat used by Brochard et al,12 Esteban et al,14 Kollef etal,17 and Marelich et al.18 The third component is a listof criteria for deciding if thepatient is ready forextuba-tion, such as that used by Hendrix et al.19 In manyintensive care units, protocols are presented as writtenguides or algorithms, and ventilator settings are manu-ally adjusted by healthcare professionals. Morerecently, progress in ventilator microprocessor tech-nology has enabled the development of computer

assisted management of ventilation and weaning.Computerised ventilatory management adapts the

ventilator output to the patients needs with closedloop systems. These systems measure and interpretrespiratory data in real time and provide continualadjustment of the level of assistance within targetedvalues. It is suggested that through enabling inter-action between the patient and the ventilator, theclosed loop systems can improve toleranceof mechan-ical ventilation and reduce the work of breathing.20

Multiple commercial computerised ventilation andweaning programs have been developed, includingadaptive support ventilation, proportional assist venti-lation, and pressure support ventilation (SmartCare).21

Several studies have explored the use of weaningprotocols in clinical practice and shown that they canbe safe and effective in reducing the time spent onmechanical ventilation.22 Other studies in variouspopulations, however, have not shown benefit.23-25

The discordant results of these studies might reflectthe fact that protocols vary in more ways than in com-position alone. While many protocols include criteriafor readiness to wean and guidelines for reducing ven-tilator support, the specific criteria and guidance vary.Furthermore, not all protocols include extubation cri-teria. Protocols are implemented in different environ-ments by healthcare professionals (including nurses,respiratory therapists, and doctors) and by automated(computerised) systems. Limited evidence suggeststhat nurses and allied health professionals mightadhere to protocols more than physicians.26 Conse-quently, recent studies have compared weaning proto-cols led by nurses or respiratory therapists withtraditional or medical directed weaning.16-18

We synthesised the best current evidence for the

effectiveness of weaning protocols compared with noprotocols in weaning critically ill adults from invasivemechanical ventilation. The protocol and the reviewcan be found in the Cochrane Database of SystematicReviews.27

METHODS

Criteria for inclusion of studies

Studies and participantsWe included randomised and quasi-randomised con-trolled trials. The study population included adultsreceiving invasive mechanical ventilation with a naso-tracheal or orotracheal tube. We excluded studies in

children, non-invasive ventilation as a weaning strat-egy, or patients with tracheotomies.

InterventionsWeaningperprotocolwasdefinedasamethodoflimit-ingthe duration of invasive ventilation that included atleast the first two of: a list of objective criteria based ongeneral clinical factors for deciding if a patient is readyto discontinue mechanical ventilation; structuredguidelines for reducing ventilatory support, such as atrial of spontaneous breathing or a stepwise reductionin support (for example, synchronised intermittentmechanical ventilation or pressure support ventila-

tion); and a list of criteria for deciding if the patient isready for extubation.

Records identified through database search(n=5987):CENTRAL (n=406)ISI Web of Science and Conference

Proceedings (n=1381)LILACS (n=776)CINAHL (n=107)Embase (n=2324)

Records identified through other sources (n=29):Cambridge Scientific abstracts (n=15)www.controlled-trials.com (n=14)Contact with first authors (n=0)Reference search of included studies (n=0)

Included (n=11)10 16-18 23 24 31-34 36

Records excluded dueto overlap or not

meeting inclusioncriteria (n=5973)

Excluded (n=5)43-46 49Excluded (n=4)41 42 47 48

Ongoing (n=1)

Full paperreview (n=14)

Records excludeddue to not

meeting inclusioncriteria (n=22)

Further informationsought (n=7)

Fig 1 | Identification of studies on weaning from mechanical ventilation

RESEARCH

page 2 of 14 BMJ | ONLINE FIRST | bmj.com

7/31/2019 bmj.c7237

3/14

We did not exclude studies that did not include for-mal extubation criteria as not all studies included thiscomponent. Furthermore, delay in extubation can becaused by organisational factors and not necessarily bydelays in weaning. Usual weaning practice wasdefinedas the usual practice in an intensive care unit (as stated

by the authors) where no written guidelines wereapplied.

Outcome measuresThe primary outcome measure was the duration ofmechanical ventilation. Secondary outcome measuresincluded mortality (intensive care unit and hospital);adverse events (re-intubation, tracheostomy, pro-tracted mechanical ventilation); weaning duration;length of stay in intensive care unit; length of stay inhospital; and cost.

Search methods for identification of studies

We used the standard search strategy of the CochraneAnaesthesia Review Group of the Cochrane Colla-boration. The search included the Cochrane CentralRegister of Controlled Trials (CENTRAL) (CochraneLibrary 2010, Issue1), Medline (1950 to January 2010),Embase (1988 to January 2010), CINAHL (1937 to

January 2010), ISI Web of Science and ConferenceProceedings (1970 to January 2010), and LILACS(1982 to January 2010). The search strategy and termsaredetailedintheonlinereviewathttp://onlinelibrary.wiley.com/o/cochrane/clsysrev/articles/CD006904/frame.html. In addition, we searched reference lists ofall identified study reports, contacted authors for

further information on ongoing trials, and searchedthe meta-register of controlled trials at www.controlled-trials.com. No language restrictions wereapplied.

Selection of studies, data extraction, and qualityassessmentTwoauthors(BBandPOH)independentlyscannedtitlesand abstracts identified by electronic searching, manualsearches, and contacts with experts. Three authors (BB,KB, POH) retrieved and evaluated the full text versionsof potentially relevant studies and independently extra-cteddatausingamodifiedpaperversionoftheCochrane

Anaesthesia Review Group

s data extraction form (ver-sion 3 January 2007). Data were extracted on study

design, setting and participants, inclusion and exclusioncriteria,andinterventions andoutcomes. In addition, weassessed risk of bias using the Cochrane Collaborationsdomain based evaluation tool for assessing the risk ofbias.28 We assessed adequacy of the generation of theallocation sequence; concealment of allocation; blindingprocedures; whether or not outcome data were ade-quately addressed; whether the study was free from sug-gestion of selective outcome reporting; and whether itwas free from other problems that could put it at risk ofbias. BB contacted authors of included studies if insuffi-cient information was available in the publications toobtain missing data.Disagreement was resolved throughconsultation with a fourth author (FA).

Data synthesis

Data were processed in accordance with the Cochranehandbook.29 Intervention effects were expressed withodds ratios and 95% confidence intervals for dichoto-mous data and mean differences and 95% confidenceintervals for continuous data. The degree of heteroge-neity was informally evaluated by visual inspection offorest plots, and more formally by measuring theimpact of heterogeneity with the I2 statistic (I2 >50%indicates significant heterogeneity), and tested withthe 2 statistic (P

7/31/2019 bmj.c7237

4/14

duration of mechanical ventilation. We could not do asubgroup analysis on type of protocol because onlytwo studies used the same protocol.1624

The data were entered into Review Manager soft-ware byBB, and POH checked data entry. All analyseswere conducted with Review Manager.40

RESULTS

Description of studies

Search results

The search of electronic databasesretrieved 6016 citations: 5987 references from thedatabase search and 29 relevant references from webbased sources. After reviewing the titles and abstracts,we identified and retrieved 14 database references infulltextforreviewandobtainedfurtherinformationonseven unpublished trials located on the controlledtrials website. Figure 1 summarises the study selectionprocess, and table 1 provides details of excludedstudies 41-49 and reasons for their exclusion.

Included studiesWe included 11 studies with 1971participants (table 2). The sample sizes ranged from15 to 357 participants. All studies took place in inten-

sive care units in hospitals. Trials were conducted inthe United States, 1016-182324 Brazil,34 Italy,3133

Germany,36 and Australia.32 Participants wererecruited from various intensive care units includingmedical,1016-1823 coronary,1634 surgical,1736 surgical/trauma,18 mixed (including medical, surgical/traumapatients),32 neurosurgical,2431 and cardiac surgical.33

Three trials were conducted in multiple units 16-18 andseven in single units. 102331-3436 One trial specified thepopulation (neurosurgical) rather than the unit.24 Fourstudies described the ventilatory modes used in usual

practice

in the control group, and these involved areduction in respiratory rate in synchronised inter-mittentmechanical ventilation and a reductionin pres-sure support in pressure support ventilation, 1934 areductionin positive end expiratory pressure and pres-sure support ventilation,32 and a reduction in pressuresupport ventilation.36 The seven remaining describedusual practice as weaning according to the physiciansdiscretion without describing what this constituted.

Protocols were delivered by registered nurses andrespiratory therapists,16-1823 respiratory therapists,24 orphysicians, registered nurses, and respiratorytherapists,31 or computer driven103236 or not

stated.3334

All studies used criteria on readiness towean for protocol entry, but the criteria varied greatly.

Table 2 | Summary of included studies of weaning in critically ill adults on mechanical ventilation

Study Methods No of patients Interventions Outcomes Country, setting

Ely, 199616 RCT 300 Protocol delivered b y RNs and RTs vphysicianjudgment

Total duration of MV, weaningduration, length of stay inICU, adverse events, ICU and hospital costs, length of stay

in hospital, mortality

US, one medical and onecoronary ICU, closed units

Kollef, 199717 RCT 357 Protocol delivered b y RNs and RTs vphysicianjudgment

Total duration of MV, reintubation, length of stay inhospital, hospital mortality, hospital cost, MV time before

weaning, protracted weaning >7 days

US, two medical and twosurgical ICUs

Krishnan, 200423 Quasi-RCT

299 Protocol delivered by RNs and RTs vphysicianjudgment

Total duration of MV, duration of SBT preceding MVdiscontinuation, length of stay in ICU, location after ICUdischarge, ICU and hospital mortality, reinstitution of MV

US, one medical ICU

Marelich, 200018 RCT 335 Protocol delivered b y RNs and RTs vphysician

judgment (medical ICU) and standardisedapproach (surgical ICU)

Total duration of MV, incidence of VAP, weaning duration,

ventilator discontinuation failure rate

US, one medical and one

surgical/trauma ICU

Namen, 200124 RCT 100 Protocol delivered by RTs vpractice(not stated) Total duration of MV, length of stay in ICU,time tosuccessful extubation, adverse events, ICU and hospital

costs

US, neurosurgical patientpopulation

Navalesi, 200831 RCT 318 Protocol vdaily evaluation and physicianjudgment

Rateof extubation,durationof MV,lengthof stayin ICUandhospital, ICU mortality,tracheostomy

Italy, one closed neuro ICU

Piotto, 200834 Quasi-RCT

36 Protocol delivered by RTvgradual reduction inRR and PS possible SBT according to RTphysician judgment

Reintubation rate, length of stay in CCU, time fromintubation to start of weaning,start of weaning toextubation, SBT to extubation, presence of respiratory

infection in patients requiring reintubation, mortality ofreintubated patients

Brazil, one CCU

Rose, 200832 RCT 1 02 C omputerised protocol (SmartCare) vweaningof PSand PEEPaccordingto usual local practice

Time to separation, total durationof MV, intubation to firstextubation and successful extubation,lengthof stayin ICUand hospital, ICU mortality, rate of successful extubation,rate of reintubation, rate of non-invasive ventilation afterextubation.

Australia, one mixed medical,surgical,trauma ICU

Simeone,2 00233 RCT 49 Protocol vphysician judgment Total durationof MV, lengthof stayin ICU,Noof

postoperative complications

Italy, one cardiac s urgical ICU

Stahl, 200936 RCT 60 C omputerised protocol (SmartCare) vweaningof PS according to physician judgment

Durationof ventilator weaning,totalduration of MV,lengthof stay in ICU,reintubation within 48 hours, physician andnursing workload, ICU and hospital mortality

Germany, one surgical ICU

Strickland,199310

RCT 15 C omputerised protocol (Superspor t model) vSIMV and PS weaning according to physicianjudgment

TimespentwithRR8 or>30,timespentwithTV

7/31/2019 bmj.c7237

5/14

7/31/2019 bmj.c7237

6/14

7/31/2019 bmj.c7237

7/14

Risk of bias in included studies

Most trials had low risk of bias across the six domains(fig 2). In eight trials, the allocation sequence was ade-quately generated and concealed. 1016-1831-3336 Two trialsused inadequateallocation generation and concealment:oneallocatedby using odd andeven hospital numbers,23

and one allocated sequentially on recruitment.34

Theremaining trial did not report the method used, and wewere unable to obtain this information.24 Given the nat-ure of the intervention, blinding of participants and staffto the intervention is not feasible, but in seven trials theoutcome assessors were independent from the indivi-duals administering the intervention: this was confirmedinpublications 1017andthrough personalcommunicationwithauthors.1618 31-3336 Blindingofoutcomeassessorswasunclear in one study,23 not done in one study,34 andcould not be confirmed in one study despite attempts toobtain this information.24 Most trials reported completeoutcome data: two trials insufficiently reported on

recruitment, attrition, and exclusion to permitjudgment.3334 Eight trials published the weaning

protocol,1016-18232431 34 and two described the automatedcomputer system3236 and reported all pre-specified out-comes:one trial published theweaning algorithmbut didnot pre-specify outcomes so there was insufficient infor-mation to permit a judgment. 33 Seven trials seemed freefrom other sources of bias as defined in the Cochrane

Collaboration

s domain based evaluation,1016-182331 32

two were stopped early for ineffectiveness,2436 onereported unsubstantiatedfindings,33 andonewasunpub-lished so there was insufficient information to permit a

judgment.34 Five studies conducted a priori calculationsof sample size,17243134 36 two studies mentioned powercalculations but were unclear,1823 and four studies didnot mention this.10163233

Effects of interventions

Total duration of mechanical ventilation (hours)Ten trials reported on the total duration of mechanicalventilation.16-1823 2431-3436 One trial did not report on

this outcome measure as the trial lasted only 48 hoursfor each patient.10 The pooled result for duration of

Mixed

Kollef 199717

Marelich 200018

Piotto 200834

Rose 200832

Subtotal (95% CI)

Test for heterogeneity: 2=10.20, df=3, P=0.02, I2=71%

Test for overall effect: z=1.42, P=0.16

Neurosurgical

Namen 200124

Navalesi 200831

Subtotal (95% CI)



Surgical

Simeone 200233

Stahl 200936

Subtotal (95% CI)



Medical

Ely 199616

Krishnan 200423

Subtotal (95% CI)



Total (95% CI)

Test for heterogeneity: 2

=37.71, df=9, P

7/31/2019 bmj.c7237

8/14

mechanical ventilation, with the random effects modelbecause of significant substantial heterogeneity(I=76%, P=0.0001), showed that weaning protocols

were associated with significantly reduced mean logtotal duration of mechanical ventilation (mean log0.29, 95% confidence interval 0.5 to 0.09 ;P=0.006), corresponding to a reduction of 25% (9% to39%) in the geometric mean (fig 3).

We performed a subgroup analysis to assess theimpact of type of intensive care unit on the total dura-tion of mechanical ventilation (fig 3). The subgroupswere small, with two to four studies in each, andincluded mixed units that incorporated medical, surgi-cal and trauma patients; neurosurgical units; and sur-gical units and medical units. The neurosurgicalintensive care unit subgroup was introduced post hoc

because we were unaware of these specific studieswhen writing the protocol and their weaning progressis different to other groups of patients because of neu-rological impairment. Pooled analysis of the four trialsinthemixedintensivecareunitgroup 17183234showedanon-significant reduction in the mean log in the wean-ing protocol group (mean log 0.23, 0.54 to 0.09;P=0.16), corresponding to a reduction of 21% (9% to42%) in thegeometricmean. Pooled analysis of thetwoneurosurgical studies2431 also showed a non-significantreduction in the mean log in the weaning protocolgroup (0.01, 0.2 to 0.18; P=0.93), corresponding toa reduction of 1% (20% to 18%) in the geometric

mean. The surgical intensive care units3236

showed asignificant reduction in the mean log in the weaning

protocol group (0.66, 1.25 to 0.06; P=0.03), corre-sponding to a reduction of 48% (6% to 71%) in the geo-metric mean; and the two medical intensive care

units1623 showed a non-significant reduction in themean log (0.35, 0.81 to 0.11; P=0.14), correspond-ing to a reduction of 30% (12% to 56%) in the geo-metric mean.

We also performeda subgroup analysis to assess theimpact of type of approach: professionalled or compu-ter driven (fig 4). The eight studies that used a profes-sional led approach 16-182324 313334 showed a significantreduction in the mean log, favouring the weaning pro-tocol group (mean log0.25, 0.43 to0.06;P=0.009),corresponding toa reduction of 22% (6% to35%)in thegeometric mean, and there was significant heterogene-ity(P=0.008,I2=63%).Thetwostudiesthatusedacom-puter driven approach 3236 showed a non-significantreduction in the mean log in the weaning protocolgroup (0.5, 1.42 to 0.42; P=0.28), corresponding toa reduction of 39% (52% to 76%) in the geometricmean. For this outcome, the average percentage differ-ence in geometric mean of 25% is consistent with esti-mates in all subgroups in both subgroup analyses (thatis, it is contained within the 95% confidence intervals).Therefore, the heterogeneity cannot be explained bytype of unit or type of approach.

Mortality

We found no significant differences between groups inhospital mortality (odds ratio 1.10, 0.86 to 1.41; sixtrials,

Professional led

Ely 199616

Kollef 199717

Krishnan 200423

Marelich 200018

Namen 200124

Navalesi 200831

Piotto 200834

Simeone 200233

Subtotal (95% CI)



Computer driven

Rose 200832

Stahl 200936

Subtotal (95% CI)

Test for heterogeneity: 2=15.71, df=1, P

7/31/2019 bmj.c7237

9/14

n=1368)16-18232436 or mortality in the intensive care unit(0.98, 0.48 to 2.02; four trials, n=508,)31323436 (fig 5).

Adverse eventsWe found no significant differences between groups(table 5) in the odds of reintubation (eight trials,n=1314),16172431-3436 self extubation (two trials, n=198),1624

andtracheostomy(six trials, n=1191).161824313234Theodds

of requirement for protracted weaning were significant forweaning lasting over 21 days in one trial (n=300)16 but notinanother (n=100)24andnot significantforweaning lastingover 14 days (n=102) 32 or seven days (n=357).17

Weaning duration (hours)In the random effects model for significant substantialheterogeneity (I=97%, P

7/31/2019 bmj.c7237

10/14

interpret clinically. To illustrate these findings in meandifferences that are clinically relevant we can use dataon ventilation times from a large epidemiologicalstudy (n=5183) of characteristics and outcomes in

patients receiving mechanical ventilation.

50

Themean duration of mechanical ventilation in this studywas 144 hours. The corresponding risk for a weaningprotocol is the mean that one would expect based onthe effect estimates in this review. With this method,reductions can be found in the total duration ofmechanical ventilation from an assumed risk of 144hours to a corresponding risk of 108 hours (88 to 131hours); duration of weaning from 96 hours to 21 hours(7to 66hours); and lengthof stayin intensive carefrom11 days to 10 days (9 to 11 days).

Though the data from the pooled summaries aloneseem to support benefit with weaning protocols, they

should be viewed with caution because of the signifi-cant heterogeneity among studies, particularly in rela-tion to total duration of mechanical ventilation(I2=76%) and duration of weaning (I2=97%). Weexplored heterogeneity through subgroup analyseson the impact of type of intensive care unit (mixed,neurosurgical, surgical, medical) and type of approach(professional led or computer driven). We foundinconsistency among results and little statistical evi-dence of difference in treatment effect, possiblybecauseof the small number of studies with subgroupsforanalysis. The use of protocols to guide weaning did

not adversely affect mortality in intensive care or hos-pital. We found no effect on adverse events includingreintubation, self extubation, tracheostomy, and pro-tracted weaning, though our meta-analysis was under-powered to investigate the impact of the interventionson these outcomes, which were infrequent. Further-

more, basic costing exercises in intensive care unitsand hospital in three US studies showed no significantdifference between the alternative weaning strategies.

Strengths and limitations of the review

In this systematic reviewand meta-analysis of weaningprotocols, most trials had sound methods and had alow risk of bias. Based on GRADE,51 however, thequality of evidence was low, mainly because of sub-stantial variability in the effect estimates. As a resultof this heterogeneity, our findings should be inter-preted with caution. The methods in trials were limitedby the inability to blind clinical staff to the method of

weaning; therefore it is possible that clinician

s deci-sions andactions could have been influenced, resultingin biased estimates of treatment effect. As it is not fea-sibletoblindstaffintheseweaningstudies,weassessedblinding of investigators collecting outcome data andfound risk of bias to be low in eight of 11 includedstudies. Six of the 11 studies originated in the US,which could limit the extent to which findings can begeneralised to other healthcare systems.

Implications for clinical practice

Ventilator weaning is a complex process, and it is noteasy to isolate the reasons for heterogeneity. The dis-

cordance in results among studies could be caused bycontextual factors (differences in populations ofpatientsandusualpracticewithinunits)orinterventionfactors (differences in determining readiness to wean,ventilator modes, and variables used in weaning pro-tocols).Thoughweattemptedtoexaminetheimpactofdifferent populations of patients on duration ofmechanical ventilation by exploring types of intensivecare units, we could not isolate populations in all stu-dies because some units were mixed and includedmedical, surgical, neurosurgical, and trauma patients.Clearly, the population of patients can affect the

Table 5 | Summary of adverse events associated with weaning from mechanical ventilation

with and without weaning protocol in critically ill adults on mechanical ventilation

Adverse eventNo of patients with events/total

No of e vents Odds ratio (95% CI), P value

Reintubation16 17 24 31-34 36 1314 0.76 (0.40 to 1.42), P=0.39

Self extubation16 300 0.40 (0.08 to 2.08), P=0.25

Self extubation24

100 0.50 (0.09 to 2.86), P=0.68Tracheostomy16 18 24 31 32 34 1191 0.74 (0.45 to 1.22), P=0.24

Protracted weaning (days):

>2116 300 0.42 (0.19 to 0.96), P=0.04

>2124 100 0.18 (0.02 to 1.63), P=0.21

>1432 102 0.68 (0.20 to 2.31), P=0.54

>717 357 0.63 (0.35 to 1.15), P=0.13

Ely 199616

Marelich 200018

Piotto 200834

Rose 200832

Stahl 200936

Strickland 199310

Total (95% CI)

Test for heterogeneity: 2=169.00, df=5, P

7/31/2019 bmj.c7237

11/14

duration of weaning. For example, weaning a surgical

patient in intensive care after elective major surgerymight be morestraightforward than weaning a medicalpatient in intensive care with respiratory failure afteracute exacerbation of chronic pulmonary disease. Inaddition, because of the wide variety of protocolsused in included studies, we could not examine theimpact of specific weaning protocols on specific popu-lations of patients. What remains unknown and war-rants further investigation is whether or not specificprotocols are more beneficial than others in particularpopulations of patients.

Another important contextual factor, and one thatcauses controversy in studies of non-pharmacological

interventions in intensive care units, is the use of theusual care group as a control in randomised trials.52

Usual care in intensive care units can encompass awide variety of practicesfor example, usual caremight be standardised around high level evidenceand thus represent best practice or it might be highlyvariable and include unfavourable practices.52 Conse-quently,ifthecultureofaunitissuchthatusualcareisastandardisedhighlevelapproachtoweaning,albeitnotformally laid out in guidelines, then it might not differgreatly from that delivered by a weaning protocol.Thus, in a trial of effectiveness, the gap between usualcare and weaning with a protocol might be too small to

show a significant difference between groups.Forexample,Marelichetalconductedastudyinonemedical and one surgical/trauma intensive care unitand reported variable practice between units: the med-ical unit had no standardised approach to weaningwhereas the surgical unit had a standardised approachto ventilator management, though extubation wasbased on the judgment of individual physicians.18

Thus, while combined data from both units showed areduction in the duration of mechanical ventilation,when we analysed data separately for each unit thereduction was significant only in the medical intensivecare unit where no standard approach to weaning

existed. Similarly, the study by Rose et al attributedtheir lack of effect to usual practice in their intensive

care unit, which comprised unlimited assessment of

weaning by experienced autonomous critical carenurses, a 1:1 nurse to patient ratio supported by 24hour medical staff, and twice daily rounds by anintensivist.32 The association between staffing in inten-sive care units and clinical outcomes has been studiedpreviously. High intensity of medical staffing withmandatory intensivist consultation (such as thatfound in intensive care units in the United Kingdom)has been associated with reduced mortality andreducedlengthofstayinhospitalandtheintensivecareunit53; higher doctor to patient ratios are significantlyassociated with higher rates of success of weaning andhome discharge in patients receiving prolonged

mechanical ventilation

54

; and an optimum number ofqualified intensive care nurses led to a reduction in theduration of weaning for patients with exacerbation ofchronic obstructivepulmonary disease.55 These exam-ples suggest that in units where the organisational cul-ture and context supports optimum trained staffnumbers, there might not be additional benefit fromthe use of weaning protocols compared with standar-disedhighlevel approaches to weaning. Notwithstand-ing, full descriptions of usual care in the control groupswere not provided in the included studies, and there-fore we cannot be certain that this is the case.

In relation to intervention factors, there were many

differences in methods among studies that could havecontributed to heterogeneity. The number and type ofcriteria used to determine readiness to wean withinprotocols varied considerably (ranging from five to17) and the broadness or restrictiveness of criteriaused could have contributed to differences in results.In relation to the protocols themselves, only two usedan identical weaning protocol.1824 Despite this, thesetrials reported conflicting results in both the durationof mechanical ventilation and weaning, possiblybecause of differences in the populations of patientsstudied or usual practice withinthe intensivecare units.

We focused solely on the impact of weaning proto-

cols, but it is worth noting that sedation practices influ-ence the duration of ventilation and must be

-0.75 -0.50 -0.25 0 0.25 0.50 0.75

Favoursweaning protocol

Favoursusual care

Ely 199616

Krishnan 200423

Namen 200124

Navalesi 200831

Piotto 200834

Rose 200832

Simeone 200233

Stahl 200936

Total (95% CI)



-0.06 (-0.57 to 0.45)

-0.24 (-0.46 to -0.02)

0.07 (-0.18 to 0.32)

-0.11 (-0.28 to 0.06)-0.09 (-0.54 to 0.36)

-0.09 (-0.37 to 0.19)

-0.37 (-0.70 to -0.04)

0.10 (-0.31 to 0.51)

-0.11 (-0.21 to -0.02)

4

19

16

315

12

9

6

100

Study

Weaning protocol Usual care

Mean difference(95% CI)


Weight(%)

3.7 (2.4)

4.7 (1.0)

5.9 (0.4)

4.9 (0.8)6.1 (0.7)

5.1 (0.7)

3.2 (0.5)

6.3 (0.8)

Mean (SD)log hours

149

154

49

16518

51

24

26

636

Total

3.8 (2.1)

5.0 (0.9)

5.8 (0.8)

5.0 (0.8)6.2 (0.7)

5.2 (0.8)

3.6 (0.7)

6.2 (0.7)

Mean (SD)log hours

151

145

51

15318

51

25

26

620

Total

Fig 7 | Length of stay in intensive care unit with and without weaning protocol. Mean difference calculated with fixed effects model

RESEARCH

BMJ | ONLINE FIRST | bmj.com page 11 of 14

7/31/2019 bmj.c7237

12/14

considered in trials of weaning protocols. The processofweaningcanbeaffectedbyuseofsedativeagents.Tothis end, the manner in which sedative agents are usedhas been shown to be just as important as the pharma-cological properties of the agents themselves. Shorter

durations of ventilation and length of stay in the inten-sive care unit and hospital have been associated withintermittent boluses of sedation and analgesia insteadof continuous infusions,56 daily interruption of infu-sions (sedation breaks) and subsequent assessment forweaning,57 and a protocol combining daily sedationbreaks with trials of spontaneous breathing.58 A recentpilot trial of a protocol for no sedation found thatpatients with no sedation had significantly more dayswithout ventilation than those receiving daily sedationbreaks (mean difference 4 days, 0.3 to 8; P=0.019).59

Sedation practice in intensive care units typically fol-lows an agreed protocol or guideline60 involving

assessment of the patients response to administrationof sedation/analgesia with a sedation scoring system

and an algorithm that uses sedation scores to modifydrug delivery. The studies included in our review pro-vided little or no information regarding their sedationpractices. To interpret changes in weaning success ortime to successful weaning, future studies shouldinclude detailed information on sedation practicesincluding the agents used, use of a sedation protocol(or lack of one) and scoring system, and whether ornot daily interruptions in sedation were permitted.

New developments in weaning

Weaning and sedation protocols have contributed tothe management of weaning in important ways overthe past 15 years. New developments in this specialtyare targeted at discontinuing invasive ventilatory sup-port in a timely manner by using automated systemsornon-invasive ventilation as a weaning strategy, andawakening and mobilising critically ill patients assoonas possible.Applyingprotocolsto real life clinicalpractice can be difficult because their effectivenessdepends on many factors, including their acceptabilityto clinicians, the workload of the intensive care unit,the requirement for frequent assessments, and moni-toring to ensure compliance. Thus, automated compu-

terised systems are increasingly being used in anattempt to improve the adaptation of mechanical

support to the needs of individual patients duringweaning and to reduce the time spent on ventilation,costs, and staff workload.61 Computers can continu-ously monitor changes in ventilation, interpret realtime physiological changes, and adapt ventilation in

response to these changes. As shown in this review,however, compared with usual care their efficacy inreducing the duration of mechanical ventilation hasyet to be established. Nevertheless, the use of weaningprotocols is increasing to the point where it is usualpractice in many units, and we are now beginning tosee more studies that compare automated weaningwith weaning with a protocol.414345 47-49

Non-invasive ventilation (where the endotrachealtube is removed, but the patient continues to receiveventilatory support delivered by face or nasal mask) isgaining popularity as a weaning strategy. To date thenumber of trials are small (around five), but a recent

systematic review has shown the clinical benefit ofthis strategy in reducing the total duration of mechan-ical ventilation support in cases of difficult weaning,particularly in patients with chronic obstructive pul-monary disease.62

In patients in intensive care, mechanical ventilationand immobilisation contribute to complications suchas delirium and weakness, which can affect the dura-tion of mechanical ventilation and length of stay inintensive care and hospital.63-65 A recent randomisedcontrolled trial of early mobilisation instigated duringperiods of daily sedation breaks showed significantlyshorter durations of delirium (median 2 v 4 days,

P=

0.02) and more ventilator-free days (median 23.5v21.1 days, P=0.05).66

Implications for research

The studies we included varied in the details they pre-sented regarding weaning protocols, the degree towhich they described usual practice within their inten-sive care units, and the settings in which they were con-ducted. In many studies, neither usual weaning practicenor organisational context (for example, staffing ratiosandfrequencyof medicalrounds) weredescribed in suf-ficient detail, thus it is difficult to ascertain the extent towhich weaning practice differed between the experi-

mental and control groups in the individual studies. Itis important that future trials fully report the details of

-0.75 -0.50 -0.25 0 0.25 0.50 0.75

Favoursweaning protocol

Favoursusual care

Ely 199616

Kollef 199717

Namen 200124

Rose 200832

Total (95% CI)



-0.10 (-0.33 to 0.13)

0.00 (-0.14 to 0.14)

0.23 (-0.07 to 0.53)

-0.09 (- 0.36 to 0.18)-0.01 (-0.11 to 0.10)

20

53

12

15100

Study

Weaning protocol Usual care



Weight(%)

5.8 (0.8)

5.6 (0.7)

6.9 (0.6)

2.9 (0.7)

Mean (SD)log hours

149

179

49

51428

Total

5.9 (1.2)

5.6 (0.7)

6.6 (0.9)

3.0 (0.7)

Mean (SD)log hours

151

178

51

51431

Total

Fig 8 | Length of stay in hospital with and without weaning protocol. Mean difference calculated with fixed effects model

RESEARCH

page 12 of 14 BMJ | ONLINE FIRST | bmj.com

7/31/2019 bmj.c7237

13/14

weaning protocols, usual weaning practice (includingsedation practice), and the context into which weaningprotocols are introduced (such as staffing ratios andorganisation of care) as this would enable clinicians togain a more accurate picture of the potential impact ofsuch protocols in their own environment. Moreover, asweaning protocols are complex with multiple inter-related and interdependent components,67 welldesigned clinical trials should take into account othercontextual and intervention factors that could have aneffect. These factors need to be described in sufficientdetail to enable accurate replication and comparisonsamong studies. We would argue that mixed methodsresearch is necessary to fully evaluate the componentsof complex interventions such as weaning. Future stu-dies of theefficacyof weaning protocols should follow aframework that incorporates process evaluation (such

as that advocated by the Medical Research Council

68

)to understand how context influences outcomes and toprovide insights to aid implementation in other settings.

Conclusion

Use of a weaning protocol can result in decreased totalduration of mechanical ventilation, weaning duration,andlength of stay in intensive care unit. The reductionin the duration of mechanical ventilation and weaningmight be because of consistent application of objectivecriteriafordeterminingreadinesstoweanandaguidedapproach to reducing support. Similarly, reducedlength of stay in intensive care might be attributable

to the reduction in mechanical ventilation. Reducedmechanical ventilation, in turn, might lead to reducedrequirements for tracheostomy. In settings whereobjective criteria and guided approaches are alreadyincorporatedinto standard weaning practice, however,further beneficial effects of weaning protocols on theseoutcomes might not be realised.

We thank Harald Herkner, H S Jeffrey Man, Carmen Silvia Valente Barbas,

and Nathan L Pace for their help and editorial advice during the

preparation of the review. The results of a Cochrane review can be

interpreted differently, depending on peoples perspectives and

circumstances. Please consider the conclusions presented carefully. They

are the opinions of review authors and are not necessarily shared by the

Cochrane Collaboration. This paper is based on a Cochrane review first

published in the Cochrane Library 2010 Issue 5 (www.thecochranelibrary.com). Cochrane reviews are regularly updated as new

evidence emerges and in response to feedback, and the Cochrane Library

should be consulted for the most recent version of the review.

Contributors: BB, FA, and POH were involved in the design of the review.BB developed the search strategy. BB, PO H, KB, and FA conducted the

screening, and BB, POH, and KB undertook assessment of risk of bias and

data extraction. BB, PO H and CC were involved in data entry and data

analysis. BB wrote the first draft of the review, and all authors contributed

to the various drafts of the report. BB is guarantor.

Funding: The review was funded through a Cochrane Fellowship Awardfor BB from the Research and Development Office, Northern Ireland, and

the Health Research Board, Ireland.

Competing interests: All authors have completed the Unified CompetingInterest form at www.icmje.org/coi_disclosure.pdf (available on request

from the corresponding author) and declare: no support from any

organisation for the submitted work, no financial relationships with any

organisations that might have an interest in the submitted work in the

previous three years, no other relationships or activ ities that could appear

to have influenced the submitted work.

Ethical approval: Not required.Data sharing: No additional data available.

1 Dries DJ. Weaning from mechanical ventilation.J Trauma1997;43:372-84.

2 Mancebo J. Weaning from mechanical ventilation. Eur Resp J1996;9:1923-31.

3 CookDJ, WalterSD, CookRJ, Griffith LE, GuyattGH, Leasa D,et al.Incidence of and risk factors for ventilator associated pneumonia incritically ill patients.Ann Intern Med1998;129:433-40.

4 PapazianL, Bregeon F,Thirion X, GregoireR, Saux P,Denis JP, etal.Effectof ventilator-associated pneumonia on mortality andmorbidity. Am J Resp Crit Care Med1996;154:91-7.

5 Vincent JL, BihariDJ, Suter PM, BruiningHA, White J,Nicolas-Chanoin MH, et al. The prevalence of nosocomial infection inintensive care units in Europe (EPIC).JAMA 1995;274:639-44.

6 MeadeMO, Cook DJ.The aetiology, consequencesand prevention ofbarotrauma: a critical review of the literature. Clin Intensive Care1995;6:166-73.

7 Meade MO, CookDJ, Kernerman P,BernardG. How touse articlesaboutharm:the relationshipbetweenhigh tidal volumes, ventilatingpressures, and ventilator-induced lung injury. Crit Care Med1997;25:1915-22.

8 Slutsky AS, Tremblay LN.Multiplesystem organ failure. Ismechanicalventilation a contributing factor?Am J RespCrit Care Med1998;157:1721-5.

9 Sahn SA,Lakshminarayan S. Bedside criteriafor discontinuation ofmechanical ventilation. Chest1973;63:1002-5.

10 Strickland JH Jr, Hasson JH. A computer-controlled, ventilatorweaning system. A clinical trial. Chest1993;103:1220-6.

11 Stroetz RW,HubmayerRD. Tidal volumemaintenance duringweaning with pressure support.Am J Resp Crit Care Med1995;152:1034-40.

12 BrochardL, Rauss A, BenitoS, Conti G,ManceboJ, Rekik N,et al.Comparison of three methods of gradual withdrawal fromventilatorysupport during weaning from mechanical ventilation.Am J Resp CritCare Med1994;150:896-903.

13 Esen F,Denkel T,Telci L,Kesecioglu J, Ttnc AS, Akpir K,et al.Comparison of pressure support ventilation (PSV) and intermittentmandatory ventilation (IMV) during weaning in patients with acuterespiratory failure.Adv Exp Med Biol 1992;317:371-6.

14 Esteban A, FrutosF, Tobin MJ, AliaI, Solsona JF, ValverduI, etal. Acomparison of four methods of weaning patientsfrom mechanical

ventilation. Spanish Lung Failure Collaborative Group. N EnglJ Med1995;332:345-50.

15 Murtagh D, Baum N. Protocols putmore pennies in your practicepockets or protocols leadto practice improvement. 2010. www.neilbaum.com.

16 Ely EW, Baker AM, Dunagan DP, Burke HL, Smith AC, Kelly PT, etal.Effecton the durationof mechanical ventilation of identifyingpatients capable of breathing spontaneously. N EnglJ Med1996;335:1864-9.

17 KollefMH,Shapiro SD, SilverP, StJohnRE,Prentice D,Sauer S,et al.A randomized, controlled trial of protocol directed versus physician-directed weaning from mechanical ventilation. Crit Care Med1997;25:567-74.

18 Marelich GP,MurinS, Battistella F, Inciardi J, VierraT, Roby M.Protocol weaning of mechanical ventilation in medical and surgicalpatients by respiratory care practitioners and nurses: effect onweaning time and incidence of ventilator-associated pneumonia.Chest2000;118:459-67.

19 Hendrix H, KaiserME, Yusen RD,Merk J. A randomizedtrial ofautomated versus conventionalprotocol-driven weaning from

WHAT IS ALREADY KNOWN ON THIS TOPIC

Prolonged mechanical ventilationcan be associated with adverse outcomes for patients

Weaning frommechanical ventilation according to standardised protocols is purported to besafe and effective in reducing the duration of mechanical ventilation, though trials haveconflicting results

WHAT THIS STUDY ADDS

Compared with usual care, use of weaning protocols can reduce the duration of mechanicalventilation by 25%, weaning duration by 78%, and length of stay in intensive care unit by10%

As there was significant heterogeneity in included trials and most were conducted in the US,the findings might not be generalisable

RESEARCH

BMJ | ONLINE FIRST | bmj.com page 13 of 14

7/31/2019 bmj.c7237

14/14

mechanical ventilation followingcoronary artery bypass surgery.Euro J CardiothoracSurg 2006;29:957-63.

20 Burns KEA, Lellouche F, Lessard MR. Automating theweaningprocess with advanced closed-loop systems. Intensive Care Med2008;34:1757-65.

21 Rose L,PresneillJJ, Cade JF.Updatein computer-drivenweaningfrommechanical ventilation.Anaesth Intensive Care 2007;35:213-21.

22 CookD, Meade M, Guyatt G, Griffith L, Booker L. Criteria forweaningfrom mechanical ventilation. Evidence report/technology

assessment No 23. AHRQ Publication No 01-E005, 2000.23 KrishnanJA,Moore D,Robeson C,RandCS,FesslerHE.A prospective,

controlled trial of a protocol-based strategy to discontinuemechanical ventilation.Am J Resp Crit Care Med2004;169:673-8.

24 Namen AM, Ely EW, Tatter SB, CaseLD, Lucia MA, Smith A, etal.Predictors of successful extubation in neurosurgical patients.Am JResp Crit Care Med2001;163:658-64.

25 Randolph AG, Wypij D, Venkataraman ST, Hanson JH,Gedeit RG,Meert KL, et al. Effect of mechanical ventilator weaning protocols onrespiratory outcomes in infants and children.JAMA2002;288:2561-8.

26 Lawton R, Parker D. Procedures and theprofessional: thecaseof theBritish NHS. Soc Sci Med 1999;48:353-61.

27 BlackwoodB, AlderdiceF, Burns KEA, Cardwell CR,Lavery G,OHalloran P. Protocolized versus non-protocolized weaning forreducing the durationof mechanical ventilation in critically ill adultpatients (review). Cochrane Database Syst Rev. 2009;5:CD006914.

28 HigginsJPT,Altman DG.Assessingriskof bias inincludedstudies.In:HigginsJPT, Green S, eds.Cochrane handbookfor systematic reviewsof interventions. Version 5.0.2. Cochrane Collaboration, 2009.

29 Higgins JPT,Green S, eds.Cochrane handbookfor systematic reviewsof interventions. Version 5.0.2. Cochrane Collaboration, 2009.

30 Higgins JP,ThompsonSG. Quantifying heterogeneity in a meta-analysis. Stat Med2002;21:1539-58.

31 NavalesiP, FrigerioP, Moretti MP,SommarivaM, Vesconi S,Baiardi P, et al. Rate of reintubationin mechanically ventilatedneurosurgical and neurologic patients: evaluation of a systematicapproach to weaning and extubation. Crit Care Med2008;36:2986-92.

32 RoseL, Presneill JJ, JohnstonL, CadeJF. A randomised,controlledtrialof conventional versus automated weaning from mechanicalventilation using SmartCare TM/PS. Intensive Care Med2008;34:1788-95.

33 Simeone F, BiagioliB, Scolletta S, Marullo ACM, MarchettiL,Caciorgna M, et al. Optimization of mechanical ventilation supportfollowing cardiac surgery.J Cardiovasc Surg 2002;43:633-41.

34 PiottoRF, MaiaLN, Machado MN, OrricoSRP.Effects of the use of

mechanical ventilation protocol in the coronarycare unit:randomized study. Rev Bras CirCardiovasc2010 (online publication ahead of print) www.rbccv.org.br/sgp/visualizar_artigo.asp?ObjSubmissao=1&cod_versao=3719&cod_fluxo=2171.

35 Higgins JPT, White IR,Anzures-CabreraJ. Meta-analysisof skeweddata: combining results reported on log transformed or raw scales.Stat Med2008;27:6072-92.

36 Stahl C, Dahmen G, ZieglerA, MuhlE. Protocolised automated versusnon-protocolised physician-directed weaning from mechanicalventilation: a controlled clinical trial. IntensivmedPrax2009;46:441-6.

37 Hozo SP,Djulbegovic B, Hozo I. Estimating themean andvariancefrom themedian,range andthe size of a sample. BMC Med ResMethodol 2004;5:13.

38 Higgins JPT,Green S, eds.Cochrane handbookfor systematic reviewsof interventions. Version 5.0.2. Cochrane Collaboration, 2009.

39 BlandJM, Altman DG.Theuse oftransformationwhencomparingtwomeans. BMJ1996;312:1153.

40 ReviewManager(RevMan), Version 5 forWindows [computerprogram]. Nordic Cochrane Centre, Cochrane Collaboration, 2008.

41 Beale R. Comparison of an automatedweaningprogramme andastandardclinical weaning protocol for weaning critically ill patients.www.controlled-trials.com/ISRCTN82559457.

42 ButlerRJ, MartinC, Gwadry-SridharF. A cluster randomized trialtoimprove weaning and extubation from mechanical ventilation incommunity hospitals. NCT00157287. www.clinicaltrials.gov.

43 Donglemans DA,VeeloDP, PaulusF, de MolBA, Korevar JC,Kudoga A, et al. Weaning automation with adaptive supportventilation: a randomized controlled trial in cardiothoracic surgerypatients. Anesth Analg 2009;108:565-71.

44 East TD,Heermann LK, Bradshaw RL,Lugo A, Sailors RM,Ershler L,et al. Efficacy of computerized decision support for mechanicalventilation: results of a prospective multi-center randomized trial.Proc AMIA Symp 1999;251-5.

45 Lellouche F, Mancebo J, Jolliet P, RoeselerJ, Schortgen F, Dojat M,et al. A multicenter randomized trial of computer-driven protocolisedweaning from mechanical ventilation.Am J Resp Crit Care Med2006;174:894-900.

46 McKinley BA,Moore FA,Sailors RM,Cocanour CS,Marquez A,Wright RK, et al. Computerized decision support for mechanicalventilation of trauma induced ARDS: results of a randomized clinicaltrial.J Trauma 2001;50:415-24.

47 Papirov G. Computer driven management of weaning following

prolonged mechanical ventilation. 2007. www.clinicaltrials.gov.NCT00502489.48 Scholz J. Automatic controlof pressuresupport ventilationin surgical

intensive care units. http://data.linkedct.org/page/trials/NCT00445289.

49 Taniguchi C, Eid RC, Saghabi C, Souza R,Silva E,Knobel E,Paes AT,Barbas CS. Automatic versus manual pressure support reduction inthe weaning of post-operative patients: a randomised controlledtrial. Crit Care 2009;13:R6.

50 Esteban A, Anzueto A, FrutosF, AliaI, BrochardL, Stewart TE, etal.Mechanical Ventilation International Study Group. Characteristicsand outcomes in adult patients receiving mechanical ventilation. A28-day international study.JAMA 2002;287:345-55.

51 Guyatt GH,OxmanAD, Kunz R,VistGE, Falck-YtterY, Schnemann HJ,forthe GRADE Working Group. Ratingquality of evidence andstrength of recommendations: what is quality of evidence andwhyis it important to clinicians? BMJ2008;336:995-8.

52 Thompson BT,Schoenfeld D. Usual care as thecontrolgroup inclinical trialsof nonpharmacologicinterventions. Proc AmThorac Soc

2007;4:577-82.53 Pronovost PJ,AngusDC, Dorman T, Robinson KA,Dremsizov TT,

YoungTL. Physician staffing patternsand clinical outcomes incritically ill patients: a systematic review. JAMA 2002;288: 2151-62.

54 Polverino E, NavaS, Ferrer M,CerianaP, Clini E,Spada E, etal.Patients characterization, hospital course and clinical outcomes infive Italian respiratory intensive care units. Intensive Care Med2010;36;137-42.

55 Thorens JB, Kaelin RM,Jolliet P, Chevrolet JC. Influence of thequalityof nursing on the durationof weaning frommechanical ventilation inpatients with chronic obstructive pulmonary disease. Crit Care Med1995;23:1807-15.

56 Kollef MH, Levy NT,Ahrens TS,Schaiff R, PrenticeD, Sherman G. Theuse of continuous IV sedation is associated with prolongationofmechanical ventilation. Chest1998;14:541-8.

57 Kress JP,Pohlman AS, OConnor MF, Hall JB. Daily interruptionofsedative infusionsin critically ill patients undergoing mechanicalventilation. N EnglJ Med2000;342:1471-7.

58 GirardTD, Kress JP, Fuchs BD,Thomason JWW, Schweickert WD,PunBT, et al.Efficacyand safetyof a pairedsedation and ventilatorweaning protocol for mechanically ventilated patientsin intensivecare (awakening and breathing controlled trial): a randomisedcontrolled trial. Lancet2008;371:126-34.

59 Strom T, Martinussen T, Toft P. A protocolof no sedation forcriticallyill patients receiving mechanical ventilation: a randomised trial.Lancet2010;375:475-80.

60 ReschreiterH, MaidenM, Kapila A. Sedationpracticein theintensivecare unit: a UK national survey. Crit Care 2008;12: R152.

61 LelloucheF, Brochard L. Advanced closedloops duringmechanicalventilation (PAV, NAVA, ASV, SmartCare). Best Pract Res Clin

Anaesthesiol 2009;23:81-93.62 Burns KEA, AdhikariNKJ, Keenan SP,Meade MO.Use of noninvasive

ventilation to wean critically ill adults frominvasive ventilation: ameta-analysis update. BMJ2009;338:b1574.

63 Schweickert WD,Hall J. ICU-acquiredweakness. Chest2007;131:1541-9.

64 Ely EW, GautamS, MargolinR, Francis J, May L,Speroff T,et al. The

impactof delirium in theintensive care unit on hospital length ofstay. Intensive Care Med 2001;27:1892-900.65 KressJP, OConnorMF, Pohlman AS, Olson D, Lavoie A, Toledano A,

et al. Sedation of critically ill patients during mechanical ventilation.A comparison of propofoland midazolam.Am J Respir Crit Care Med1996;153:1012-8.

66 Schweickert WD, Pohlman MC,Pohlman AS, Nigos C, PawlikAJ,Esbrook CL,et al.Early physicaland occupationaltherapyinmechanically ventilated, critically ill patients: a randomisedcontrolled trial Lancet2009;373:1874-82.

67 Blackwood B. Methodological issues in evaluating complexhealthcare interventions.J Adv Nurs 2006;54:612-22.

68 Medical Research Council. Developing and evaluating complexinterventions: new guidance. 2008. www.mrc.ac.uk/complexinterventionsguidance.

Accepted: 08 November 2010

RESEARCH

bmj.c7237

Documents