Implementation of an Evidence-based Extubation … of an evidence-based extubation readiness bundle...

1

Implementation of an evidence-based extubation readiness bundle in 499 brain-

injured patients - a before-after evaluation of a quality improvement project

Antoine Roquilly, M.D. (1), Raphaël Cinotti, M.D. (2), Samir Jaber, M.D.,Ph.D. (3), Mickael Vourc’h (2), Florence Pengam (1), Pierre Joachim Mahe, M.D. (1), Karim Lakhal, M.D. (2), Dominique Demeure Dit Latte, M.D. (1), Nelly Rondeau, M.D. (2), Olivier Loutrel, M.D. (1), Jérôme Paulus, M.D. (1), Bertrand Rozec, M.D., Ph.D. (2), Yvonnick Blanloeil, M.D., Ph.D. (2), Marie-Anne Vibet (4,5), Véronique Sebille, Ph.D. (5,6), Fanny Feuillet (5,6), Karim Asehnoune, M.D., Ph.D. (1) Affiliations 1. Intensive Care Unit, Anesthesia and Critical Care Department, Hôtel Dieu - HME, University

Hospital of Nantes, France 2. Intensive Care Unit, Anesthesia and Critical Care Department, Laennec, University Hospital of

Nantes, France 3. Intensive Care Unit, Anesthesia and Critical Care Department Saint Eloi University Hospital of

Montpellier, France 4. INSERM U 657, Paris, France 5. Plateforme de Biométrie, Cellule de promotion de la recherche clinique, University Hospital of

Nantes, France 6. EA 4275 SPHERE "Biostatistics, Pharmacoepidemiology & Human Science Research", UFR

des Sciences Pharmaceutiques, Nantes University, France Corresponding Author

Karim Asehnoune, M.D.,Ph.D., CHU de Nantes, Service d’Anesthésie Réanimation, 1 place Alexis Ricordeau 44093 Nantes Cedex 1. e-mail address [email protected] Tel (+33) 240 08 3005. Fax (+33) 240 08 4682

The work was performed at the Nantes University Hospital 1 place Alexis Ricordeau 44093 Nantes Cedex 1. Authors’ Contributions: AR, RC, SJ, KA contributed to the study design, data analyses, interpretation of results, writing and revision of the manuscript; MAV, FF, VS contributed to the statistical analysis and revised the manuscript; MV, FP, PJM, KL, DDDL, NR, OL, JP, BR, YB contributed to data collection, manuscript writing and revision. All authors have approved the final manuscript for publication. Funding sources: institutional funds Short Running head: Mechanical ventilation in neurological patients

Descriptor: 10.12 ICU Management/Outcome

Word count text only: 2927

of 44 AJRCCM Articles in Press. Published on 08-August-2013 as 10.1164/rccm.201301-0116OC

Copyright © 2013 by the American Thoracic Society

2

AT A GLANCE COMMENTARY (88 words): Scientific knowledge on the subject: Complications associated with prolonged mechanical

ventilation frequently develop in brain-injured patients. Strategies to decrease ventilator time have

been poorly documented to date in this specific population.

What this study adds to the field: Implementation of an evidence-based extubation readiness bundle

associating protective ventilation, early enteral nutrition, local protocol for the probabilistic treatment

of hospital-acquired pneumonia (HAP) and systematic approach to extubation, was associated with a

reduction in the duration of mechanical ventilation and in the rates of HAP as well as of unplanned

extubation.



3

Abstract

Rationale: Mechanical ventilation is associated with morbidity in brain-injured patients. This study

aims to assess the effectiveness of an extubation readiness bundle to decrease ventilator time in brain-

injured patients.

Methods: Before/after design in two intensive care units (ICUs) in one university hospital. Brain-

injured patients ventilated > 24 hours were evaluated during two phases (a 3-year control phase

followed by a 22-month intervention phase). Bundle components were: protective ventilation, early

enteral nutrition, standardization of antibiotherapy for hospital-acquired pneumonia and systematic

approach to extubation. The primary endpoint was the duration of mechanical ventilation.

Results: 299 and 200 patients respectively were analyzed in the control and the intervention phases of

this before/after study. The intervention phase was associated with lower tidal volume (P<0.01), higher

PEEP (P<0.01), and higher enteral intake in the first 7 days (P=0.01). The duration of mechanical

ventilation was 14.9±11.7 days in the control phase and 12.6±10.3 days in the intervention phase

(P=0.02). The hazard ratio (HR) for extubation was 1.28 (95% confidence interval (95%CI) 1.04-1.57;

P=0.02) in the intervention phase. Adjusted HR was 1.40 (95%CI, 1.12-1.76, P<0.01) in multivariate

analysis and 1.34 (95%CI, 1.03-1.74; P=0.02) in propensity score-adjusted analysis. ICU-free days at

day-90 increased from 50±33 in the control phase versus 57±29 in the intervention phase (P<0.01).

Mortality at day 90 was 28.4% in the control phase and 23.5% in the intervention phase (P=0.22).

Conclusion: The implementation of an evidence-based extubation readiness bundle was associated

with a reduction in the duration of ventilation in brain-injured patients.

Key words: Mechanical Ventilation, Brain injury, Weaning, Ventilator-induced lung injury,

extubation.



4

Introduction

Weaning from mechanical ventilation is a major issue in the intensive care unit (ICU) (1). Brain-

injury is a major cause of respiratory failure and a frequent cause of prolonged mechanical ventilation

(2). Although primary brain insults are the main determinants of outcomes in critically ill neurologic

patients, mechanical ventilation morbidity has been associated with poor neurological recovery and

death in this specific population (3-5). Reduction of mechanical ventilation duration could improve

health-related quality of life and it reduces the medical costs associated with critically ill patients (6,

7).

The standardization of medical care has been shown to improve the outcomes of critically ill

patients. We hypothesized that the systematic application of a limited number of specific interventions

could reduce ventilator time in brain-injured patients. In this setting, four main targets appeared

attractive in this population. First, protective ventilation, defined as low tidal volume and positive end-

expiratory pressure (PEEP), is attractive since brain injury potentiates ventilator-induced lung injuries

(8, 9) and since PEEP does not alter brain perfusion (10, 11). Second, early enteral feeding decreases

the rate of infection as well as the duration of mechanical ventilation in surgical and trauma patients

(12, 13). Third, optimal protocols of probabilistic antibiotherapy to treat hospital-acquired pneumonia

(HAP) are mandated since inadequate antibiotherapy is associated with alteration of outcomes in ICU

patients (14). Consideration of local epidemiology is necessary when implementing international

guidelines for antibiotherapy of HAP (15-17). Fourth, extubation of patients without complete

awakeness could be safe (19, 20), reducing the risk of unplanned extubation (18, 19) and decreasing

the duration of ICU stay (20).

We hypothesized that the implementation of an evidence-base care bundle including protective

ventilation, early enteral nutrition, standardization of antibiotherapy and systemic approach to

extubation can accelerate extubation readiness in brain-injured patients (21). We therefore conducted a

before/after study to investigate the implementation of an evidence-based bundle to decrease ventilator



5

time in brain-injured patients. Some of the results of this study have been previously reported in the

form of an unpublished abstract (22).



6

Materials and methods Ethics statement

The protocol was approved by the local ethics committee of Nantes hospital (Groupe Nantais

d’Ethique dans le Domaine de la Santé, Nantes, France). Consent was waived because the project was

a prospective improvement process within a collaborative institutional quality improvement program

applied to all brain-injured patients hospitalized in our ICUs.

Patients and setting

The present study was performed in two ICUs of a university hospital. Data from all adult brain-

injured patients (traumatic brain-injured, subarachnoid hemorrhage, stroke or other) who required

mechanical ventilation for more than 24 hours were collected and analyzed. Brain injury was

considered when a Glasgow Coma Scale 12 was associated with at least one anomaly related to an

acute process on head tomographic tomodensitometry (extradural hematoma, subarachnoid

hemorrhage, brain contusion, hematoma, brain edema, skull fracture, stroke, abscess). Exclusion

criteria were early decision to withdraw care (taken in the first 24 hours in ICU), death in the first 24

hours or inclusion in a randomized controlled trial.

General Care for brain-injured patients

As previously described (23, 24), all patients were sedated with a continuous intravenous infusion of

fentanyl (2-5 g.kg-1.hr-1) and midazolam (initiated at 0.1 mg.kg-1.hr-1 and increased up to 0.2-0.5

mg.kg-1.hr-1 in case of intracranial hypertension). Intracranial pressure was monitored with an

intraparenchymal probe placed in the most affected side (Codman, Johnson and Johnson Company,

Raynham, Mass., USA.) in patients who were considered at increased risk of intracranial hypertension

(25). When control of intracranial pressure was poor despite osmotherapy or ventricular drainage,

sodium thiopental was used with a loading dose (2-3 mg.kg-1) followed by continuous administration

(2-3 mg.kg-1.hr-1) adapted to intracranial pressure evolution and to serum level monitoring (blood

level of thiopental between 20 and 30 g.ml-1). No daily interruption of sedation was performed during

the period of intracranial hypertension. Sedation was discontinued as soon as the risk of intracranial



7

hypertension was considered as low on head tomographic tomodensitometry (no cortical sulcal

effacement, no midline shift and no compressed peri-mesencephalic cisterns).

Study protocol

Study design

A before/after study design was used. The before period (control phase) consisted of all consecutive

patients with severe brain-injury who were admitted to the participating ICUs three years before the

educational program began (January 2007-December 2009). The intervention was introduced over a 12

month period (January-December 2010), during which no patient data were collected. The intervention

phase consisted of all consecutive severe brain-injured patients admitted to the participating ICUs

during a 22 month period (January 2011-October 2012).

Control phase

During a 3 year control phase, ventilatory setting, enteral nutrition, probabilistic antibiotherapy

for HAP and decision for extubation were performed at the discretion of the attending physician. We

performed a retrospective analysis of prospectively recorded data regarding: demographic

characteristics, ventilatory setting, nutritional support, pathogens involved in HAP, duration of

mechanical ventilation, hospital stay and mortality. The ideal body weight was calculated from the

patient’s height (ideal body weight of male patients was calculated as equal to 50+ 0.91(centimeters of

height - 152.4); that of female patients was calculated as equal to 45.5+0.91(centimeters of height-

152.4)).

Interphase

During a 12 month period, we developed a 4 point bundle to decrease ventilator time based on a

review of the literature and on analysis of the pathogens involved in HAP in our ICUs. Four targets

were defined: protective ventilation, early enteral nutrition, optimization of the probabilistic

antibiotherapy for HAP and a systematic approach to extubation (Table 1). During this phase, all

physicians, residents, physiotherapists and nurses received formal training for the processes and

procedures related to the 4 point bundle. Based on the prospective follow-up of the pathogens involved



8

in HAP during the control phase in our ICUs, two criteria have been defined for prescription of

probabilistic antibiotherapy against multiple-drug-resistant bacteria (previous 48-hour antibiotherapy

and/or hospitalization lasting more than 9 days). We also considered that cephalosporinase is the main

mechanism of resistance for Enterobacteriacae in our ICUs, and we therefore recommend the use of

cefepime, which is not affected by these enzymes. Moreover, methicillin-resistant Staphylococcus

aureus (S. aureus) was infrequent in our ICUs and this bacterium was not considered in probabilistic

antibiotherapy. This protocol is reassessed annually. During the control phase, sedation was already

standardized according to international guidelines (26, 27) and the management of sedation was

therefore not included in the bundle interventions.

Intervention phase

During a 22-month period, the recommended procedure was to conform to the extubation

readiness bundle (Table 1) and patients were followed up until day 90.

Outcomes

The primary outcome was the duration of mechanical ventilation. Secondary outcomes were the

compliance with bundle elements, the number of ventilator-free days at days 28 and 90, the rates of

extubation failure, of unplanned extubation, of tracheotomy, of HAP; ICU length of stay, the number

of ICU-free days at day 90 and the mortality at day 90.

Compliance with antibiotherapy for HAP was not applicable in the control phase. Compliance

with the systematic approach to extubation (undelayed extubation) was considered in the two phases

when tube removal was performed within 48 hours after all of the requirements for extubation were

met (inspiratory support < 10 cm H2O, Glasgow Coma Scale 10 and cough) (20).

The number of ventilator-free days was defined as the number of days from day 1 to day 28 (or

day 90 as stated) on which a patient breathed spontaneously and was alive.

Pneumonia was considered when two signs (among body temperature > 38°C; leukocytosis > 12

000/mL or leukopenia < 4000/mL and purulent pulmonary secretions) were associated with the

appearance of a new infiltrate or changes in an existing infiltrate on chest x-ray (28). Diagnosis was



9

confirmed by a respiratory tract sample using a quantitative culture with a predefined positive

threshold of 104 colony-forming units per milliliter (CFU/mL) for a bronchoalveolar lavage or non-

bronchoscopic sample, 106 CFU/mL for an endotracheal aspirate. Hospital-acquired pneumonia was

defined as pneumonia that occurred 48 hours after admission that had not been incubating at the time

of admission.

Statistical analysis

In a previous study by Mascia et al. (8), brain-injured patients who developed VILI had a 25%

decrease in the number of ventilator-free days. Our study was designed to detect a 25% relative

reduction in the duration of mechanical ventilation, which was estimated at 13 ± 10 days in the control

phase (29) with a power of 90% and a type I error of 5%. Considering this power analysis, 398 patients

are needed. The control phase has been a priori set to a 3-year duration and 299 patients were included

in this phase. Finally, the preintervention to postintervention ratio was set as 3:2, and a total of 500

patients were included in the two phases.

The primary analysis consisted in comparing the duration of mechanical ventilation between the

two phases by means of a Student’s t-test. A stratified log-rank test was then applied. To consider

death as a competing event, individuals presenting this event were censored at this time-point

(mechanical ventilation equal to the duration of survival).

A sensitivity analysis with a Fine and Gray model was a posteriori applied in order to consider

the potential bias associated with the treatment of death. Extubation was defined as the principal event

and death before extubation was defined as a competing risk (30). Proportional risk assumption was

not verified for the treatment phase, so a time-dependent variable was included in the model. A cut-off

was observed in the graphics and an optimal cut-off was obtained by minimizing the AIC criterion.

In order to consider potential bias associated with the before/after study design, we decided a

posteriori to perform two complementary sensitivity analyses with an adjustment for demographic

characteristics: 1/ A multivariate analysis was performed using a Cox regression model including

variables differing between the two phases in bivariate analyses (with a p value <0.10). Variables used



10

for adjustment were : cardiac failure, simplified acute physiology score-II, Glasgow Coma Scale,

initial pathology, extra-ventricular drainage for hydrocephalus and decompressive craniectomy; 2/ A

propensity score method was applied to balance covariates in the two phases and to reduce bias.

Propensity score is defined as the conditional probability of being in the intervention phase given the

covariates. It was estimated using logistic regression, including variables recorded on ICU admission.

A propensity score method was applied to balance covariates in the two phases and to reduce bias.

Propensity score was defined as the conditional probability of being in the intervention phase given the

covariates. It was estimated using logistic regression, including variables recorded on ICU admission.

At each step of the algorithm an intervention subject i was randomly selected and the difference

between his propensity score and the propensity score of all control subjects was calculated (ic,

i=1,…,m where m was the number of intervention subjects; c=1,...,n where n was the number of

control subjects). In order to select the lowest distance between propensity scores for all intervention

subjects (Di1, Di2,…, Din for subject i, i=1,…,m), a distance caliper was used whose size is a quarter

of a standard deviation of the logit of the propensity score. All control subjects with a distance ic

lower than the caliper size were retained for each intervention subject (control subjects for which ic

is higher than the caliper size are not retained). Finally, Mahalanobis distances were calculated

between each previously retained control subject and the intervention subject. The closest control and

intervention subject, regarding Mahalanobis distance, was then removed from the pool, and the process

was repeated for all intervention subjects (31). This caliper matching method was used to match 153

(76.5%) patients in the intervention phase and 153 patients in the control phase. A stratified log-rank

test and a Cox regression model were performed on this matched sample.

In order to assess treatment effect heterogeneity across initial pathology, a subgroup analysis was

performed using a Cox regression model (traumatic brain injury, subarachnoid hemorrhage, stroke and

other conditions). Interaction between phase and initial pathology subgroups was tested. To estimate

time trends effect, a time series analysis on aggregated data was performed. Segmented linear

regression was used. In order to determine the effect of adherence to best practices, the factors



11

associated with the duration of mechanical ventilation were investigated through a multivariate Cox

regression model.

The normality of the variables was tested with a Kolmogorov-Smirnoff test. Continuous

parametric data were expressed as the mean (SD), and non-parametric data were expressed as the

median and interquartile range (IQR). Categorical data were expressed as numbers and percentage, as

well as the absolute difference (95% confidence interval, 95%CI). For demographic characteristics and

secondary endpoint 2 test, Student’s t-test or Wilcoxon rank-sum test were used as appropriate. All

statistical tests were two-sided. A two-sided p value less than 0.05 was considered statistically

significant. An independent statistician performed the analyses with SAS statistical software (SAS 9.3

Institute, Cary, NC) and R statistical software (R 3.0).



12

Results

Population

Demographic characteristics at baseline are reported in Table 2. Of 7,590 patients admitted in the

participating ICUs, 299 brain-injured patients (49.3% traumatic brain-injured, 30.9% subarachnoid

hemorrhage, 18.8% stroke, 1% cancer or meningitis) were included in the control phase, and 200 in the

intervention phase (39% traumatic brain-injured, 36.5% subarachnoid hemorrhage, 16% stroke, 8.5%

cancer or meningitis) P<0.01) (see flow chart, Figure 1). In the intervention phase, extra-ventricular

drainage (P<0.01) and decompressive craniectomy (P=0.03) were performed more frequently than in

the control phase.

Compliance with bundle elements

Intervention adherence is reported in Table 3. The rates of adherence to tidal volume and PEEP

recommendations were 33.6% and 37.8% in the control phase respectively, and 42.6% and 62.6% in

the intervention phase (P=0.09 and P<0.01). When considering tidal volume and PEEP as continuous

values, the mean tidal volume was significantly lower and the mean PEEP was significantly higher in

the intervention phase than in the control phase (P=0.02 and P<0.01 respectively, Figures 2A-B).

These differences lasted during the first 5 days of mechanical ventilation (data not recorded afterward).

The rates of enteral nutrition on day-1 and of input target reached on day-3 were significantly higher in

the intervention phase than in the control phase (P<0.01 for both comparisons). Probabilistic

antibiotherapy was conformed to the protocol in 79 (83.2%) episodes of HAP. The percentage of

patients extubated within the first 48 hours after all requirements were met (ventilatory weaning,

Glasgow Coma Scale 10 and cough) increased from 57.5% patients in the control phase to 68.9% in

the intervention phase (P=0.03). The percentage of patients in whom care complied with the whole set

of best practices increased from 6.0% in the control phase to 21.1% in the intervention phase (P<0.01).

Primary outcome: Duration of mechanical ventilation

The mean duration of mechanical ventilation was 14.9 ± 11.7 days in the control phase and 12.6 ±

10.3 days in the intervention phase (P=0.02). No secular time trends unrelated to the intervention



13

under study was highlighted (see supplemental analysis). When considering death as a competitive

event (stratified log-rank test), the median duration of mechanical ventilation was 17 (9-25) days in the

control phase and 12 (7-19) days in the protocol-guided phase (P=0.02), with an hazard ratio (HR) for

weaning from the mechanical ventilation in the intervention phase of 1.28 (95% confidence interval

(95%CI), 1.04-1.57; P=0.02; Figure 3.A). In sub-group analysis, the benefit was consistent under all

conditions, and the test for interaction between phase and initial pathology subgroups was not

significant (P for interaction = 0.55; Figure 3.B). Using the Fine and Gray model with a time-

dependent effect, a cut-off was defined at 10 days. Before 10 days of mechanical ventilation, the HR

was 1.04 (95%CI, 0.89-1.21, P=0.64). After 10 days of mechanical ventilation, the HR was 1.28

(95%CI, 1.01-1.62, P=0.04). In order to account for potential confounding factors associated with a

before/after design, three sensitivity analyses were performed to assess the robustness of this result.

Using a cox multivariate proportional hazards model adjusted on variables unbalanced between the 2

phases (P<0.10, Table 1), the adjusted HR for being alive and free of mechanical ventilation was 1.40

(95%CI, 1.12-1.76, P<0.01) in the intervention phase. After propensity matching based on variables

potentially associated with the intervention phase (see Table 2), the HR was 1.34 (95%CI, 1.03-1.74,

P=0.02).

We then estimated the contribution of best practices to the probability of extubation (Table 4).

Only tidal volume 8 ml.kg-1 (HR 1.38, 95%CI 1.08-1.76; P<0.01) and PEEP > 3 cmH2O (HR 1.30,

95%CI 1.00-1.69; P=0.05) were associated with a reduction in the duration of mechanical ventilation.

Secondary outcomes

Secondary outcomes are provided in Table 5. The percentage of patients with HAP decreased

after bundle implementation. At day-90, both ventilator-free days and ICU-free days increased in the

intervention phase compared with the control phase. No significant difference in the mortality rate at

day-90 was apparent between the two phases.

Extubation-related complications



14

In the control phase, 27 (9%) reintubations were required compared with 22 (13.8%) in the

intervention phase (P=0.11, Table 5). The rate of unplanned extubation decreased from 9.4%, to 4.5%

(P<0.01). The rate of tracheostomy was 10% in the control phase and 8.7% in the intervention phase

(P=0.62).



15

Discussion

The present study shows for the first time that implementation of an evidence-based extubation

readiness bundle in brain-injured patients can reduce the duration of ventilator support. This bundle

can be mainly effective in patients with a duration of mechanical ventilation exceeding 10 days ,

corresponding to patients with a difficult or prolonged weaning (32). The rates of HAP and of

unplanned extubation were also lower in the intervention phase as compared with the control phase.

The reduction in mechanical ventilation support was well tolerated regarding mortality.

When caring for critically ill patients, standardized protocols have been shown to improve

outcomes in several conditions such as sedation (33), sepsis management (34) or surgical procedures

(35). In acute stroke units, the implementation of an evidence-based protocol to prevent extra-cerebral

complications has been shown to improve patient outcomes after discharge (36). We reported the

effectiveness of a bundle developed to reduce the duration of mechanical ventilation in brain-injured

patients.

Protective ventilation relies on the association of a low tidal volume with a positive end-

expiratory pressure. Lung protective ventilation was initially shown to reduce the duration of

mechanical ventilation in patients with acute respiratory distress syndrome (37). Few data are available

in brain-injured patients because they are most likely excluded from randomized clinical trials

evaluating such strategies. Thus, no consensus has been reached regarding protective ventilation in

neurologic critically ill patients (38). As a result, protective ventilation was poorly applied in

neurological patients in a recent large observational study (2) and the application of high tidal volume

is still recommended as a means to induce moderate hypocapnia (39). In the experimental setting,

brain injuries were shown to potentiate ventilator-induced lung injury (9) and a tidal volume higher

than 9 ml.kg-1 was associated with the subsequent development of acute lung injury and with brain

hypoxemia in patients with subarachnoid hemorrhage (8, 40). In a recent meta-analysis pooling studies

with relatively short durations of mechanical ventilation (mean duration: 6.9 hours) in patients without

ARDS, protective ventilation was associated with a decreased of ARDS/ALI rate and of mortality risk



16

(41). In this meta-analysis, protective ventilation was associated with a low increase in capnia (from 38

to 41 mmHg) which could be damaging in brain-injured patients. However, the lung compliance of

these patients is frequently not altered and we assume that the prevention of hypercapnia could easily

be achieved in brain-injured patients by increasing the respiratory rate instead of increasing the tidal

volume. In the current study, low tidal volume and PEEP were the only interventions independently

associated with the probability of successful extubation. Since tidal volumes (> 10 ml/kg) were

associated with ARDS and death in stroke patients (42), interventions aiming to reduce the use of such

tidal volumes could be particularly relevant in brain-injured patients. Interestingly, the favorable

effects of protective ventilation on the duration of mechanical ventilation were consistent for all of the

neurological conditions studied, but the effects could have been more pronounced in stroke patients.

Critically ill patients frequently experience episodes of enteral feeding interruption, and

nutritional goals are frequently not reached, thereby requiring early nutritional support. Early enteral

nutrition remains a controversial issue mainly because of potential problems regarding tolerance and

risk of HAP. Nutrition guidelines recommend achieving only 65% of calorie goals in the first week

(43). Early enteral nutrition provided within the first 48 hours after ICU admission has not been shown

prospectively to alter mortality (44). However, we have previously reported that early enteral nutrition

was a protective factor regarding the risk of HAP in traumatic brain-injured patients (29). For these

reasons, we chose early initiation of enteral nutrition and to reach the targeted input in the first 5 days.

Consistent with two meta-analyses on early enteral nutrition in ICU (13, 45), the current management

bundle was associated with a decrease in HAP that may have contributed to the reduction in the

duration of ventilatory support.

Trauma and central nervous system diseases are recognized as major independent risk

factors for pneumonia in critically ill patients (46). In several randomized studies, HAP reached

up to 40-50% of brain-injured patients (47-49). This high percentage of infection can partially

be explained by the prolonged duration of mechanical ventilation in this population (1).

Moreover, a central nervous system injury-induced immune deficiency alters host response to



17

pathogens and increases the rate of pneumonia in a neuro-ICU (50). Given the frequency and the

morbidity of HAP, optimization of antibiotherapy is a major objective when caring for brain-

injured patients. Indeed, delay in the administration of appropriate antibiotic therapy for HAP is

associated with worse outcomes in the ICU (14). Optimization of empirical antibiotic therapy

decreases the duration of mechanical ventilation in septic critically ill patients (51) and may

therefore reduce ventilator time for brain-injured patients. However, implementation of

international guidelines for antibiotherapy of HAP is controversial (16, 17), probably because it

is essential to consider local epidemiology (15) and patient conditions when choosing

probabilistic antibiotherapy (52). In the current study, adaptation of international guidelines to

local epidemiology was associated with a non-significant trend toward a decrease in

probabilistic antibiotherapy failure.

The decision to extubate is particularly challenging in comatose patients (53). On one hand,

extubation failure increases hospitalization duration and mortality in general ICU patients (54)

and on the other hand, delayed extubation is associated with unplanned extubation and with

altered outcomes of brain-injured patients (20). Thus, the prediction of successful extubation is

particularly important since up to 10% of neurological patients require re-intubation (55).

Interestingly, Coplin et al. suggested that prolonged intubation should be avoided when the only

concern is an impaired neurological status (20). Several studies have reported a low risk for

extubation failure in patients with a Glasgow Coma Scale between 8 and 10 provided that

patients are able to cough (19, 56). In a large interventional study, the rate of re-intubation

decreased when extubation was mandated as soon as the Glasgow Coma Scale was above 8 and

associated with cough (18). Using more liberal criteria, we currently report a decrease in the rate

of unplanned extubation. Since unplanned extubations are associated with duration of

hospitalization and death (57), this result may partially explain the beneficial effects of our

bundle. However, the decrease in unplanned extubation was approximately equal to the increase

in extubation failure, suggesting that the Glasgow Coma Scale could be a poor criterion for the



18

prediction of extubation success. Finally, the overall rate of extubation-related complications

was not altered in the intervention phase compared with the control phase. This result argues for

the development of scores for early detection of brain-injured patients at risk of extubation

failure. Such a score should allow the early extubation of patients with a low risk of extubation

failure, without exposing some high-risk patients to post-extubation respiratory failure.

Several limitations should be noted here. First, this two-phased intervention study did not

make it possible to demonstrate causality. This design was chosen because it was not possible to

randomly assign the bundle in the same ICU without significant cross-contamination. Moreover,

before/after studies provide a good level of evidence (58), with an acceptable estimation of

intervention effect compared with randomized control trials (59, 60). However, a before/after

design may be confounded by secular trends and we cannot exclude that some care

improvements other than the bundles elements (such as management of sedation) participated in

the reduction in morbidity. We therefore performed a time series analysis that did not

demonstrate a secular time trends and did not conclude to a difference in the mean duration of

mechanical ventilation between the two periods. This result did not contradict our conclusion

based on individual analysis. Indeed, interrupted time-series methods lose power because data

are aggregated at the monthly level and moreover, the autoregressive error model cannot control

all of the variability in the number of patients per month. Second, even if the rate of

antibiotherapy failure was low in the intervention phase, the adaption of antibiotherapy to the

local epidemiology observed in the preceding years limits the generalizability of this specific

element. However, ATS/IDSA guidelines state that the initial empiric therapy for HAP is more

likely to be appropriate if a protocol for antibiotic selection is adapted to local patterns of

antibiotic resistance, with each ICU collecting this information and updating it on a regular basis

(28). Third, no intervention targeted an improvement in the management of sedation, though it is

a major determinant of the duration of mechanical ventilation. Moreover, the reported doses of

sedative drugs exceeded those frequently reported in studies evaluating scale-based sedation (61,



19

62). However, in most trials evaluating sedation, hypnotics are titrated to achieve light sedation

(RASS scores between -2 and +1), when a heavy sedation is recommended for the prevention of

intracranial hypertension in brain-injured patients (27). Finally, similar doses of sedative drugs

were described in a recent meta-analysis reporting the sedation of brain-injured patients (63).

Fourth, overall bundle compliance ranged from 42.6% to 83.2% according the process-of-care

variables, which is similar to other before/after studies which reported ranges from 5% to 87%

(64). Some low rates of compliance have demonstrated that barriers for implementing bundles

remain important. However, these interventions were selected because they implied

multidisciplinary teamwork (nurses, clinicians, physiotherapists) which has been shown to

synergistically improve patient outcomes and health-care processes (65). Moreover, a group of

evidence-based treatments, instituted rapidly and together over a specific timeframe, is

frequently associated with better outcomes than when they are executed individually.

In conclusion, the implementation of a four-point extubation readiness bundle in brain-injured

patients, including a strategy for protective ventilation, early enteral nutrition, adaptation of

antibiotherapy to local epidemiology of HAP and a systematic approach to extubation was associated

with a reduction in the duration of mechanical ventilation and in the rate of associated complications.

The non-significant decrease in hospital mortality pleads for a large multi-center trial investigating the

effects of such a strategy.



20

Acknowledgement Delphine Flattres Duchaussoy, Magali Aziza and nurses of participating ICUs for technical support. Conflict of interest

Authors have no related conflict of interest to disclose.

Previous presentation

Some of the results of this study have been previously reported in the form of an unpublished abstract (Société Française d’Anesthésie Réanimation annual meeting, Paris, France, Septembre 19-22, 2012).



21

References 1. Esteban A, Anzueto A, Frutos F, Alía I, Brochard L, Stewart TE, Benito S, Epstein SK,

Apezteguía C, Nightingale P, Arroliga AC, Tobin MJ, Mechanical Ventilation International

Study Group. Characteristics and outcomes in adult patients receiving mechanical ventilation: a

28-day international study. JAMA 2002;287:345–355.

2. Pelosi P, Ferguson ND, Frutos-Vivar F, Anzueto A, Putensen C, Raymondos K, Apezteguía C,

Desmery P, Hurtado J, Abroug F, Elizalde J, Tomicic V, Cakar N, González M, Arabi Y,

Moreno R, Esteban A, for the Ventila Study Group. Management and outcome of mechanically

ventilated neurologic patients. Crit Care Med 2011;39:1482–1492.

3. Wartenberg KE, Schmidt JM, Claassen J, Temes RE, Frontera JA, Ostapkovich N, Parra A,

Connolly ES, Mayer SA. Impact of medical complications on outcome after subarachnoid

hemorrhage. Crit Care Med 2006;34:617–23– quiz 624.

4. Kahn JM, Caldwell EC, Deem S, Newell DW, Heckbert SR, Rubenfeld GD. Acute lung injury in

patients with subarachnoid hemorrhage: Incidence, risk factors, and outcome. Crit Care Med

2006;34:196–202.

5. Zygun DA, Kortbeek JB, Fick GH, Laupland KB, Doig CJ. Non-neurologic organ dysfunction in

severe traumatic brain injury. Crit Care Med 2005;33:654–660.

6. Combes A, Costa M-A, Trouillet J-L, Baudot J, Mokhtari M, Gibert C, Chastre J. Morbidity,

mortality, and quality-of-life outcomes of patients requiring 14 days of mechanical

ventilation. Crit Care Med 2003;31:1373–1381.

7. Dasta JF, McLaughlin TP, Mody SH, Piech CT. Daily cost of an intensive care unit day: The

contribution of mechanical ventilation. Crit Care Med 2005;33:1266–1271.

8. Mascia L, Zavala E, Bosma K, Pasero D, Decaroli D, Andrews P, Isnardi D, Davi A, Arguis MJ,

Berardino M, Ducati A. High tidal volume is associated with the development of acute lung

injury after severe brain injury: An international observational study. Crit Care Med

2007;35:1815–1820.



22

9. López-Aguilar J, Villagrá A, Bernabé F, Murias G, Piacentini E, Real J, Fernández-Segoviano P,

Romero PV, Hotchkiss JR, Blanch L. Massive brain injury enhances lung damage in an

isolated lung model of ventilator-induced lung injury. Crit Care Med 2005;33:1077–1083.

10. Mascia L, Grasso S, Fiore T, Bruno F, Berardino M, Ducati A. Cerebro-pulmonary interactions

during the application of low levels of positive end-expiratory pressure. Intensive Care Med

2005;31:373–379.

11. Huynh T, Messer M, Sing RF, Miles W, Jacobs DG, Thomason MH. Positive end-expiratory

pressure alters intracranial and cerebral perfusion pressure in severe traumatic brain injury. J

Trauma 2002;53:488–92– discussion 492–3.

12. Taylor SJ, Fettes SB, Jewkes C, Nelson RJ. Prospective, randomized, controlled trial to determine

the effect of early enhanced enteral nutrition on clinical outcome in mechanically ventilated

patients suffering head injury. Crit Care Med 1999;27:2525–2531.

13. Marik PE, Zaloga GP. Early enteral nutrition in acutely ill patients: a systematic review. Crit

Care Med 2001;29:2264–2270.

14. Iregui M, Ward S, Sherman G, Fraser VJ, Kollef MH. Clinical importance of delays in the

initiation of appropriate antibiotic treatment for ventilator-associated pneumonia. Chest

2002;122:262–268.

15. Rello J, Sa-Borges M, Correa H, Leal SR, Baraibar J. Variations in etiology of ventilator-

associated pneumonia across four treatment sites: implications for antimicrobial prescribing

practices. Am J Respir Crit Care Med 1999;160:608–613.

16. Lambert M-L, Suetens C, Savey A, Palomar M, Hiesmayr M, Morales I, Agodi A, Frank U,

Mertens K, Schumacher M, Wolkewitz M. Clinical outcomes of health-care-associated

infections and antimicrobial resistance in patients admitted to European intensive-care units: a

cohort study. Lancet infect dis 2011;11:30–38.

17. Kett DH, Cano E, Quartin AA, Mangino JE, Zervos MJ, Peyrani P, Cely CM, Ford KD, Scerpella

EG, Ramirez JA, Improving Medicine through Pathway Assessment of Critical Therapy of



23

Hospital-Acquired Pneumonia (IMPACT-HAP) Investigators. Implementation of guidelines for

management of possible multidrug-resistant pneumonia in intensive care: an observational,

multicentre cohort study. Lancet infect dis 2011;11:181–189.

18. Navalesi P, Frigerio P, Moretti MP, Sommariva M, Vesconi S, Baiardi P, Levati A. Rate of

reintubation in mechanically ventilated neurosurgical and neurologic patients: evaluation of a

systematic approach to weaning and extubation. Crit Care Med 2008;36:2986–2992.

19. Namen AM, Ely EW, Tatter SB, Case LD, Lucia MA, Smith A, Landry S, Wilson JA, Glazier

SS, Branch CL, Kelly DL, Bowton DL, Haponik EF. Predictors of successful extubation in

neurosurgical patients. Am J Respir Crit Care Med 2001;163:658–664.

20. Coplin WM, Pierson DJ, Cooley KD, Newell DW, Rubenfeld GD. Implications of extubation

delay in brain-injured patients meeting standard weaning criteria. Am J Respir Crit Care Med

2000;161:1530–1536.

21. Thille AW, Richard J-CM, Brochard L. The Decision to Extubate in the Intensive Care Unit. Am

J Respir Crit Care Med 2013;187:1294–1302.

22. A. Roquilly, R. Cinotti, M. Vourch, F. Pengam, PJ Mahe, O Loutrel, B Rozec, Y Blanloeil, K

Asehnoune. Prevention multi-modale des complications respiratoires chez les patients

cérébrolésés [unpublished abstract]. Société Française d’Anesthésie Réanimation annual

meeting, Paris, France, septembre 19-22, 2012;R320

23. Rondeau N, Cinotti R, Rozec B, Roquilly A, Floch H, Groleau N, Michel P, Asehnoune K,

Blanloeil Y. Dobutamine-Induced High Cardiac Index did not Prevent Vasospasm in

Subarachnoid Hemorrhage Patients: A Randomized Controlled Pilot Study. Neurocrit Care

2012;17:183–190.

24. Roquilly A, Mahe PJ, Latte DDD, Loutrel O, Champin P, Di Falco C, Courbe A, Buffenoir K,

Hamel O, Lejus C, Sebille V, Asehnoune K. Continuous controlled-infusion of hypertonic

saline solution in traumatic brain-injured patients: a 9-year retrospective study. Crit care

2011;15:R260.



24

25. Andrews PJD, Citerio G, Longhi L, Polderman K, Sahuquillo J, Vajkoczy P. NICEM consensus

on neurological monitoring in acute neurological disease. Intensive Care Med 2008;34:1362–

1370.

26. Bederson JB, Connolly ES, Batjer HH, Dacey RG, Dion JE, Diringer MN, Duldner JE, Harbaugh

RE, Patel AB, Rosenwasser RH, American Heart Association. Guidelines for the management

of aneurysmal subarachnoid hemorrhage: a statement for healthcare professionals from a

special writing group of the Stroke Council, American Heart Association. Stroke 2009;40:125-

139.

27. Brain Trauma Foundation, American Association of Neurological Surgeons, Congress of

Neurological Surgeons. Guidelines for the management of severe traumatic brain injury. J

Neurotrauma 2007;24 Suppl 1:S1–106.

28. American Thoracic Society, Infectious Diseases Society of America. Guidelines for the

management of adults with hospital-acquired, ventilator-associated, and healthcare-associated

pneumonia. Am J Respir Crit Care Med 2005;171:388-416.

29. Lepelletier D, Roquilly A, Demeure dit latte D, Mahe PJ, Loutrel O, Champin P, Corvec S, Naux

E, Pinaud M, Lejus C, Asehnoune K. Retrospective analysis of the risk factors and pathogens

associated with early-onset ventilator-associated pneumonia in surgical-ICU head-trauma

patients. J Neurosurg Anesthesiol 2010;22:32–37.

30. Lau B, Cole SR, Gange SJ. Competing risk regression models for epidemiologic data. Am J

Epidemiol 2009;170:244–256.

31. D'Agostino RB. Propensity score methods for bias reduction in the comparison of a treatment to a

non-randomized control group. Stat Med 1998;17:2265–2281.

32. Peñuelas O, Frutos-Vivar F, Fernández C, Anzueto A, Epstein SK, Apezteguía C, González M,

Nin N, Raymondos K, Tomicic V, Desmery P, Arabi Y, Pelosi P, Kuiper M, Jibaja M, Matamis

D, Ferguson ND, Esteban A, Ventila Group. Characteristics and outcomes of ventilated



25

patients according to time to liberation from mechanical ventilation. Am J Respir Crit Care

Med 2011;184:430–437.

33. Girard TD, Kress JP, Fuchs BD, Thomason JWW, Schweickert WD, Pun BT, Taichman DB,

Dunn JG, Pohlman AS, Kinniry PA, Jackson JC, Canonico AE, Light RW, Shintani AK,

Thompson JL, Gordon SM, Hall JB, Dittus RS, Bernard GR, Ely EW. Efficacy and safety of a

paired sedation and ventilator weaning protocol for mechanically ventilated patients in

intensive care (Awakening and Breathing Controlled trial): a randomised controlled trial.

Lancet 2008;371:126–134.

34. Ferrer R, Artigas A, Levy MM, Blanco J, González-Díaz G, Garnacho-Montero J, Ibáñez J,

Palencia E, Quintana M, la Torre-Prados de MV, Edusepsis Study Group. Improvement in

process of care and outcome after a multicenter severe sepsis educational program in Spain.

JAMA 2008;299:2294–2303.

35. Haynes AB, Weiser TG, Berry WR, Lipsitz SR, Breizat A-HS, Dellinger EP, Herbosa T, Joseph

S, Kibatala PL, Lapitan MCM, Merry AF, Moorthy K, Reznick RK, Taylor B, Gawande AA. A

Surgical Safety Checklist to Reduce Morbidity and Mortality in a Global Population. N Engl J

Med 2009;360:491–499.

36. Middleton S, McElduff P, Ward J, Grimshaw JM, Dale S, D'Este C, Drury P, Griffiths R, Cheung

NW, Quinn C, Evans M, Cadilhac D, Levi C, QASC Trialists Group. Implementation of

evidence-based treatment protocols to manage fever, hyperglycaemia, and swallowing

dysfunction in acute stroke (QASC): a cluster randomised controlled trial. Lancet

2011;378:1699–1706.

37. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung

injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome

Network. N Engl J Med 2000;342:1301–1308.

38. Lowe GJ, Ferguson ND. Lung-protective ventilation in neurosurgical patients. Curr Opin Crit

Care 2006;12:3–7.



26

39. Brain Trauma Foundation. The use of hyperventilation in the acute management of severe

traumatic brain injury. J Neurotrauma 1996;13:715-723.

40. Oddo M, Nduom E, Frangos S, MacKenzie L, Chen I, Maloney-Wilensky E, Kofke WA, Levine

JM, LeRoux PD. Acute lung injury is an independent risk factor for brain hypoxia after severe

traumatic brain injury. Neurosurgery 2010;67:338–344.

41. Serpa Neto A, Cardoso SO, Manetta JA, Pereira VGM, Espósito DC, Pasqualucci M de OP,

Damasceno MCT, Schultz MJ. Association between use of lung-protective ventilation with

lower tidal volumes and clinical outcomes among patients without acute respiratory distress

syndrome: a meta-analysis. JAMA 2012;308:1651–1659.

42. Elmer J, Hou P, Wilcox SR, Chang Y, Schreiber H, Okechukwu I, Pontes-Neto O, Bajwa E, Hess

DR, Avery L, Duran-Mendicuti MA, Camargo CA, Greenberg SM, Rosand J, Pallin DJ,

Goldstein JN. Acute Respiratory Distress Syndrome After Spontaneous Intracerebral

Hemorrhage. Crit Care Med 2013;doi:10.1097/CCM.0b013e31828a3f4d.

43. Martindale RG, Mcclave SA, Vanek VW, Mccarthy M, Roberts P, Taylor B, Ochoa JB,

Napolitano L, Cresci G. Guidelines for the provision and assessment of nutrition support

therapy in the adult critically ill patient: Society of Critical Care Medicine and American

Society for Parenteral and Enteral Nutrition: Executive Summary. Crit Care Med

2009;37:1757–1761.

44. The National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS)

Clinical Trials Network, Rice TW, Wheeler AP, Thompson BT, Steingrub J, Hite RD, Moss M,

Morris A, Dong N, Rock P. Initial Trophic vs Full Enteral Feeding in Patients With Acute

Lung Injury: The EDEN Randomized Trial. JAMA 2012;307:795–803.

45. Doig GS, Heighes PT, Simpson F, Sweetman EA, Davies AR. Early enteral nutrition, provided

within 24 h of injury or intensive care unit admission, significantly reduces mortality in

critically ill patients: a meta-analysis of randomised controlled trials. Intensive Care Med

2009;35:2018–2027.



27

46. Cook DJ, Walter SD, Cook RJ, Griffith LE, Guyatt GH, Leasa D, Jaeschke RZ, Brun-Buisson C.

Incidence of and risk factors for ventilator-associated pneumonia in critically ill patients. Ann

Intern Med 1998;129:433–440.

47. Acosta-Escribano J, Fernández-Vivas M, Grau Carmona T, Caturla-Such J, Garcia-Martinez M,

Menendez-Mainer A, Solera-Suarez M, Sanchez-Payá J. Gastric versus transpyloric feeding in

severe traumatic brain injury: a prospective, randomized trial. Intensive Care Med

2010;36(9):1532-9

48. Seguin P, Tanguy ML, Laviolle B, Tirel O, Mall Dant Y. Effect of oropharyngeal

decontamination by povidone-iodine on ventilator-associated pneumonia in patients with head

trauma. Crit Care Med 2006;34:1514–1519.

49. Lingnau W, Berger J, Javorsky F, Lejeune P, Mutz N, Benzer H. Selective intestinal

decontamination in multiple trauma patients: prospective, controlled trial. J Trauma

1997;42:687–694.

50. Meisel C, Schwab JM, Prass K, Meisel A, Dirnagl U. Central nervous system injury-induced

immune deficiency syndrome. Nat Rev Neurosci 2005;6:775–786.

51. Kollef MH, Sherman G, Ward S, Fraser VJ. Inadequate antimicrobial treatment of infections: a

risk factor for hospital mortality among critically ill patients. Chest 1999;115:462–474.

52. Agbaht K, Lisboa T, Pobo A, Rodriguez A, Sandiumenge A, Diaz E, Rello J. Management of

ventilator-associated pneumonia in a multidisciplinary intensive care unit: does trauma make a

difference? Intensive Care Med 2007;33:1387–1395.

53. Epstein SK. Decision to extubate. Intensive Care Med 2002;28:535–546.

54. Epstein SK, Ciubotaru RL. Independent effects of etiology of failure and time to reintubation on

outcome for patients failing extubation. Am J Respir Crit Care Med 1998;158:489–493.

55. Karanjia N, Nordquist D, Stevens R, Nyquist P. A clinical description of extubation failure in

patients with primary brain injury. Neurocrit Care 2011;15:4–12.



28

56. Salam A, Tilluckdharry L, Amoateng-Adjepong Y, Manthous CA. Neurologic status, cough,

secretions and extubation outcomes. Intensive Care Med 2004;30:1334–1339.

57. Epstein SK, Nevins ML, Chung J. Effect of unplanned extubation on outcome of mechanical

ventilation. Am J Respir Crit Care Med 2000;161:1912–1916.

58. Barton S. Which clinical studies provide the best evidence? The best RCT still trumps the best

observational study. BMJ 2000;321:255–256.

59. Concato J, Shah N, Horwitz RI. Randomized, controlled trials, observational studies, and the

hierarchy of research designs. N Engl J Med 2000;342:1887–1892.

60. Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N

Engl J Med 2000;342:1878–1886.

61. Riker RR, Shehabi Y, Bokesch PM, Ceraso D, Wisemandle W, Koura F, Whitten P, Margolis

BD, Byrne DW, Ely EW, Rocha MG, SEDCOM (Safety and Efficacy of Dexmedetomidine

Compared With Midazolam) Study Group. Dexmedetomidine vs midazolam for sedation of

critically ill patients: a randomized trial. JAMA 2009;301:489–499.

62. Mehta S, Burry L, Cook D, Fergusson D, Steinberg M, Granton J, Herridge M, Ferguson N,

Devlin J, Tanios M, Dodek P, Fowler R, Burns K, Jacka M, Olafson K, Skrobik Y, Hébert P,

Sabri E, Meade M, SLEAP Investigators, Canadian Critical Care Trials Group. Daily sedation

interruption in mechanically ventilated critically ill patients cared for with a sedation protocol:

a randomized controlled trial. JAMA 2012;308:1985–1992.

63. Roberts DJ, Hall RI, Kramer AH, Robertson HL, Gallagher CN, Zygun DA. Sedation for

critically ill adults with severe traumatic brain injury: a systematic review of randomized

controlled trials. Crit Care Med 2011;39:2743–2751.

64. Jaber S, Jung B, Corne P, Sebbane M, Muller L, Chanques G, Verzilli D, Jonquet O, Eledjam J-J,

Lefrant J-Y. An intervention to decrease complications related to endotracheal intubation in the

intensive care unit: a prospective, multiple-center study. Intensive Care Med 2009;36:248–255.



29

65. Jaeschke R, Jankowski M, Brozek J, Antonelli M. How to develop guidelines for clinical practice.

Minerva Anestesiol 2009;75:504–508.



30

Table 1 – Evidence based weaning bundle Intervention

Lung protective ventilation 1. Tidal volume 2. PEEP 3. Respiratory rate

1. 6-8 ml.kg-1 of ideal body weight 2. > 3 cmH2O 3. Set to achieve normocapnia or optional moderate hypocapnia

Nutrition support 1. Day of initiation 2. Input target

1. Day-1 2. 25 kCal.kg-1.day-1 before Day-3

Probabilistic antibiotherapy for hospital acquired pneumonia*

1. Low risk of MDR bacteria 2. High risk of MDR bacteria

1. Amoxicillin - clavulanic acid (2 g x 3/day iv.) 2. Cefepime (2 g x 3/day iv.) + tobramycin (5-6 mg/kg/day iv. once a day)

Systematic approach to extubation 1. Ventilatory weaning

2. Tube removal

1. Early spontaneous breathing: tolerance of inspiratory support < 10 cm H2O or spontaneous breathing and Fi0240% for 30 min. 2. Glasgow Coma Scale 10 and cough (spontaneous or caused)

* Adapted to local epidemiology of pathogen involved in hospital-acquired pneumonia: low risk of

MDR was defined as pneumonia before day-10 and previous antibiotherapy shorter than 48 hours; High risk of MDR bacteria was considered until day-10 or after previous antibiotherapy longer than 48 hours.

MDR: Multiple Drug Resistant, PEEP= positive end expiratory pressure. The management of the sedation was standardized and no formal modification was performed between the two periods of the study.



31

Table 2 - Population characteristics

Whole cohort Propensity-matched population

Control phase

Intervention phase

P-value Control phase

Intervention phase

P-value

Number of patients 299 200 153 153 Age, years, mean (SD) 50 (19) 52 (17) 0.15 52 (17) 52 (18) 0.92

Male, N (%) 188 (62.9) 124 (62.0) 0.84 91 (59.5) 96 (62.8) 0.56

Weight, kg, mean (SD) 72.2 (15.2) 73.8 (17.4) 0.30 71 (15) 73.10 (17) 0.27

Height, cm, mean (SD) 168 (9) 168 (10) 0.54 167 (7) 168 (10) 0.24

Medical history, N (%) Cardiac failure Respiratory failure Diabetes mellitus Smoking

100 (33.4) 36 (12.0) 23 (7.7)

74 (24.8)

30 (15.0) 25 (12.5) 24 (12.0) 47 (23.5)

<0.01 0.88 0.11 0.73

35 (22.9) 20 (13.1)

9 (5.9) 38 (24.8)

25 (16.3) 14 (9.2)

16 (10.5) 37 (24.2)

0.15 0.28 0.14 0.89

Severity Scores, mean (SD) SAPS II Glasgow coma scale

38 (13)

8 (4)

41 (13) 9 (4)

0.06 0.05

40.1 (13.9)

7 (3)

39.7 (11.9)

7 (4)

0.78

0.07

Initial pathology, N (%) Traumatic brain injury Sub-arachnoid hemorrhage Stroke Other (cancer, meningitis)

147 (49.3) 92 (30.9) 56 (18.8)

3 (1.0)

78 (39.0) 73 (36.5) 32 (16.0) 17 (8.5)

<0.01

67 (43.8) 61 (39.9) 23 (15.0)

2 (1.3)

72 (47.1) 56 (36.6) 23 (15.0)

2 (1.3)

0.71

Neurological status on ICU admission, N (%) Cerebellum injury Extra-ventricular drainage for hydrocephalus Decompressive craniectomy

33 (11.0) 72 (24.2)

14 (4.7)

18 (9.0)

70 (35.0)

19 (9.5)

0.46 0.01

0.04

17 (11.1) 53 (34.6)

10 (6.5)

13 (8.5)

50 (32.7)

10 (6.5)

0.44 0.71

1

ICU: intensive care unit, SAPS II: simplified acute physiology score.



32

Table 3 - Measures of compliance

Control phase Intervention phase P value

Number of patients 299 200 Lung protective ventilation on day 1

1. Tidal volume, N (%) 0.02

<6 ml.kg-1 of ideal body weight 7 (2.5) 7 (3.9) 6-8 ml.kg-1 of ideal body weight 87 (30.9) 70 (38.7) 8-10 ml.kg-1 of ideal body weight 136 (48.2) 88 (48.6) > 10 ml.kg-1 of ideal body weight 52 (18.4) 16 (8.8)

Tidal volume, mean (SD) 8.9 (1.9) 8.1 (1.4) <0.01

2. PEEP, N (%) <0.01

0 cmH2O 90 (30.6) 11 (5.6) 1-3 cmH2O 93 (31.6) 56 (28.3) > 3 cmH2O 111 (37.8) 131 (66.2)

PEEP, mean (SD) 2.7 (2.1) 4.2 (1.6) <0.01

Nutrition support

1. Initiation <0.01

Day 1, N (%) 80 (33.2) 88 (56.1) Day 2, N (%) 161 (66.8) 65 (43.9) Day of initiation, mean (SD) 2.6 (1.8) 1.5 (1.2) <0.01

2. Time to achieve an input of 25 kCal.kg-1.day-1 <0.01

Day 3 49 (25.1) 88 (65.7) Days 4-5 64 (32.8) 17 (12.7) Day 6 82 (42.1) 29 (21.6) Mean time (SD) 6.1 (3.9) 3.7 (3.2) <0.01

3. Total enteral intake in the first 7 days (kCal), mean (SD) 4300 (2650) 8300 (12300) 0.01

Glasgow coma scale on extubation day /

Mean (SD) 11 (2) 11 (3) 0.39

Glasgow 10, N (%) 38 (17.9) 58 (39.2) <0.01

Glasgow 11-12, N (%) 129 (60.9) 48 (32.4)

Glasgow 13-15, N (%) 45 (21.2) 42 (28.4)

Compliance with bundle elements, N (%)

1) Ventilatory setting Tidal volume 8 ml.kg-1 on day 1 PEEP > 3 cm H2O in day 1

94 (33.6) 111 (37.8)

77 (42.6)

131 (66.2)

0.09

< 0.01

2) Nutrition support Enteral nutrition initiated on Day-1 Target of enteral support reached Day-3

80 (33.2) 49 (25.1)

88 (56.1) 88 (65.7)

< 0.01 < 0.01

3) Antibiotic therapy for HAP NA 79 (83.2) NA

4) Systematic approach to extubation (undelayed extubation) (&)

122 (57.5) 102 (68.9) 0.03

Overall bundle compliance 18 (6.0) 42 (21.1) <0.01

Assessed in 95 HAP Assessed in 212 patients in the control phase and 148 patients in the intervention phase (not applicable for death prior to extubation or tracheotomy); (&) The compliance with the systematic approach to extubation (undelayed extubation) was considered when tube removal was performed within 48 hours after all requirements are met (ventilatory weaning, Glasgow coma scale 10 and cough); (&&) The compliance was considered when bundle interventions were not applicable: in patients without HAP (for the antibiotherapy protocol); in dead or tracheotomized patients (for the systematic approach to extubation). HAP: hospital acquired pneumonia, PEEP: positive end-expiratory pressure, NA: not applicable



33

Table 4 – Association of best practice with probability of weaning from mechanical ventilation

Risk factors Hazard Ratio [95 % CI] P value

Tidal volume 8 ml.kg-1 1.38 [ 1.08; 1.76] <0.01

PEEP > 3 cm H2O 1.30 [1.00; 1.69] 0.05

Early enteral nutrition 1.18 [0.92; 1.51] 0.19

Undelayed extubation (&) 1.07 [0.82; 1.40] 0.63

Compliance with all bundle elements 1.49 [0.97 ; 2.27] 0.07

PEEP: positive end-expiratory pressure. The association of the bundle elements with the duration of mechanical was assessed with an univariate analysis. (&) Undelayed extubation was considered when tube removal was performed within 48 hours after all requirements were met (inspiratory support < 10 cm H2O, Glasgow Coma Scale 10 and cough)



34

Table 5 - Outcomes

Control phase Intervention phase p-value

Number of patients 299 200

Duration of mechanical ventilation, days, mean (SD)* Duration of mechanical ventilation, days, median (IQR)** Ventilator-free days at day 28, mean (SD) Ventilator-free days at day 90, mean (SD)

14.9 (11.7) 17 (9-25) 3.7 (5.3) 54 (33)

12.6 (10.3) 12 (8-19) 5.5 (6.5) 64 (29)

0.02 0.02 0.01 0.01

Extubation management, N (%) Overall extubation-related complication (&) Extubation failure Unplanned extubation

Tracheotomy

55 (18.4) 27 (9.0) 28 (9.4)

29 (10.0)

36 (18)

27 (13.5) 9 (4.5)

17 (8.7)

0.91 0.11

<0.01 0.62

Hospital acquired pneumonia, N (%) Probabilistic antibiotherapy failure, N (%)

172 (57.5) 14 (27.5)

95 (47.5) 8 (8.4)

0.03 0.15

Acute respiratory distress syndrome, N (%) 19 (6) 7 (3.5) 0.16

ICU length of stay, days, median (IQR)* ICU-free days at day 90, mean (SD)

20 (11-32) 50 (33)

18 (12-29) 57 (29)

0.48 0.01

Death, N (%) in ICU at day 90

75 (25.1) 85 (28.4)

45 (22.5) 47 (23.5)

0.51 0.22

ICU: intensive care unit * gross analysis (Student’s t-test), ** calculated from stratified log-rank test with consideration of death, (&) defined as extubation failure and/or unplanned extubation.



35

Figure 1 - Flow chart MV: mechanical ventilation, RCT: randomized clinical trial

Figure 2 - Evolution of tidal volume and of positive end-expiratory pressure during the first 5

days of mechanical ventilation

Tidal volume and positive end-expiratory pressure (PEEP) were recorded at 8:00 am from days 1 to 5.

IBW: ideal body weight. Data represent mean ± SD. ** p<0.01 versus control phase

Figure 3 - Kaplan-Meier curves for mechanical ventilation

(A) The hazard ratio for weaning from the mechanical ventilation was 1.28 (95%CI, 1.04-1.57;

p=0.02) using a stratified log-rank analysis, 1.40 (95%CI, 1.12-1.76; p<0.01) using a multivariate cox

model and 1.34 (95%CI, 1.03-1.74, p=0.02) after matching with a propensity analysis. Multivariate

cox model was adjusted on cardiac failure, simplified acute physiology score II, Glasgow Coma Scale,

initial pathology, extra-ventricular drainage for hydrocephalus and decompressive craniectomy.

(B) Hazard ratio for weaning from mechanical ventilation in traumatic brain injured patients,

subarachnoid hemorrhage patients, stroke patients and other conditions.



Implementation of an Evidence-based Extubation … of an evidence-based extubation readiness bundle...

Documents

Transcript of Implementation of an Evidence-based Extubation … of an evidence-based extubation readiness bundle...