Multidisciplinary Extubation Protocol in Cardiac Surgical Patients Reduces Ventilation Time and Length of Stay in the Intensive Care Unit

Multidisciplinary Extubation Protocol in Cardiac Surgical Patients Reduces Ventilation Time and Length of Stay in the Intensive Care Unit

Multidisciplinary Extubation Protocol in Cardiac Surgical Patients Reduces Ventilation Time and Length of Stay in the Intensive Care Unit Matthew E. C...

471KB Sizes 2 Downloads 68 Views

Multidisciplinary Extubation Protocol in Cardiac Surgical Patients Reduces Ventilation Time and Length of Stay in the Intensive Care Unit Matthew E. Cove, MBChB, Chen Ying, BS, Juvel M. Taculod, RRT-NPS, Siow Eng Oon, RN, Pauline Oh, RN, Ramanathan Kollengode, MBBS, Graeme MacLaren, MBBS,* and Chuen Seng Tan, PhD* Cardiothoracic Intensive Care Unit, Department of Cardiothoracic and Vascular Surgery, and Division of Respiratory Medicine and Critical Care, Department of Medicine, National University Hospital, and Saw Swee Hock School of Public Health, National University of Singapore, Singapore

Background. Protocolized care bundles may improve patient care by reducing medical errors, minimizing practice variability, and reducing mortality. We hypothesized that the introduction of a multidisciplinary extubation protocol would reduce duration of mechanical ventilation and intensive care unit length of stay in a tertiary cardiothoracic intensive care unit. Methods. A multidisciplinary extubation protocol was created. The protocol was applied to all elective postoperative cardiac surgery patients. Data were collected 3 months before and 3 months after protocol initiation. Patients were excluded if they experienced events that contraindicated application of the protocol. Results. Two hundred one patients undergoing elective open cardiac surgery were included: 99 patients before protocol implementation (preprotocol) and 102

patients after implementation (postprotocol). Median extubation time was reduced by 35% (620 minutes versus 405 minutes; p < 0.001), whereas adjusted extubation time remained significantly reduced by 144 minutes (p < 0.001). Intensive care unit length of stay was reduced from 2 days preprotocol to 1 day postprotocol (p < 0.001). Reintubation rate was the same in both groups (2.06% versus 1.96%; p [ 1.0). Conclusions. A simple multidisciplinary extubation protocol is safe and associated with a significant reduction in the duration of mechanical ventilation and intensive care unit length of stay after elective cardiac surgery.

I

and increases cost [14, 15]. In contrast, early extubation (4 to 6 hours) after cardiac surgery results in shorter ICU stays and lower overall costs [16]. Consequently, postoperative ventilation exceeding 24 hours is an important performance measure published by the National Quality Forum in the United States [17]. The opportunity to perform early extubation is frequently missed; it was recently reported that as few as 12% of cardiac surgery patients are extubated within 6 hours [18]. However, standardized protocols can more than double the number of patients achieving extubation within 6 to 8 hours [18, 19]. In our ICU, we recently implemented a multidisciplinary extubation protocol. It was designed to ensure suitable postoperative patients were safely prepared for extubation, regardless of how busy or distracted ICU physicians were. We hypothesized that a team-based

ntensive care units (ICUs) are challenging, fast-paced environments, in which health care teams are constantly responding to new information and urgent problems [1]. Such environments are prone to errors of commission and failures of omission [2–6] as staff are constantly distracted [7]. Physicians are particularly prone; conflicting demands, long hours, and a need to focus on the most unstable patients can distract them from simple decisions that substantially affect patient care [6]. For example, timely removal of invasive devices may often be overlooked, but appropriate early device removal reduces device-related infection rates by up to 50% [8, 9]. In the ICU, delayed extubation increases the risk of ventilator-acquired pneumonia [10], prolongs the use of sedative medication [11, 12], delays rehabilitation [13],

(Ann Thorac Surg 2016;-:-–-) Ó 2016 by The Society of Thoracic Surgeons

Accepted for publication Feb 16, 2016. *Graeme MacLaren and Chuen Seng Tan contributed equally to this work. Address correspondence to Dr Cove, Division of Respiratory Medicine and Critical Care, Department of Medicine, National University Hospital, NUHS Tower Block Level 10, 1E Kent Ridge Rd, Singapore 119228; email: [email protected].

Ó 2016 by The Society of Thoracic Surgeons Published by Elsevier

The Appendices can be viewed in the online version of this article [http://dx.doi.org/10.1016/j.athoracsur.2016. 02.071] on http://www.annalsthoracicsurgery.org.

0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2016.02.071

2

COVE ET AL SIMPLE PROTOCOL REDUCES EXTUBATION TIME

protocol would reduce time to extubation and that this would translate into reduced length of stay in the ICU.

Material and Methods Extubation Protocol In December 2013, the Cardiothoracic ICU at National University Hospital, Singapore, a 19-bed ICU in a tertiary academic referral center, introduced an extubation protocol (Appendix A). Before implementation, bedside physicians would initiate a spontaneous breathing trial and decide when to extubate. The extubation protocol aimed to empower nurses and respiratory therapists to autonomously initiate sedation and ventilator weaning and standardize spontaneous breathing trial practices (Fig 1). Initial ICU ventilator settings were not defined in the protocol because all postoperative patients are placed on volume control, synchronous intermittent mandatory ventilation with a set tidal volume of 8 mL/kg of ideal body weight, pressure support of 10 cm H2O, positive end-expiratory pressure of 5 cm H2O, and fraction of inspired oxygen of 0.8, which is then titrated for oxygen saturations of 96% or greater, using an existing care pathway. All patients are ventilated with a Puritan Bennett 840 ventilator (Covidian, Boulder, CO), and neuromuscular blockade is not routinely reversed in the operating room or ICU. During the study period, cardiothoracic anesthesia guidelines were not changed or modified. All elective cardiac surgery patients were eligible for protocolized extubation unless they had a clinical reason for exclusion (Fig 1). Protocolized extubation empowered the nurse to wean sedation and analgesia (intravenous propofol and morphine) by reducing the dose in half and then removing the infusion completely if the patient awoke comfortably. Tramadol and paracetamol were administered before initiating sedation weaning. Nursing staff would judge readiness for a spontaneous breathing trial within 4 to 6 hours of arrival in the ICU. The respiratory therapist would then initiate a spontaneous breathing trial and, with the presence of ICU physicians, determine readiness to extubate after 30 minutes (Fig 1). To measure the impact of the protocol on extubation practices, we conducted a retrospective before and after observation study for 3 months before and after protocol implementation. Following approval from Singapore’s ethical review board to collect data with a waiver of consent (NHG DSRB 2014/00596), we obtained data from all postoperative elective cardiac surgery patients older than 21 years who were admitted to our ICU. Patients were excluded if they were not eligible for the extubation protocol (Fig 1) or were readmitted to ICU during the same hospitalization. The data collected included age, sex, logistic European System for Cardiac Operative Risk Evaluation (EuroSCORE) [20], cardiopulmonary bypass time, airway assessment, ICU admission time, admission ventilator settings, time of extubation, inotropic agent use, admission blood gases, ICU discharge date, and ICU

Ann Thorac Surg 2016;-:-–-

complications, such as reintubation. Data for December were excluded because protocol implementation occurred during this month. The nurse-to-patient ratio was 1:1 and respiratory therapist-to-patient ratio was 1:15 throughout the observational period. Daytime physician coverage included 3 fellows and a dedicated consultantintensivist, whereas overnight, 2 fellows were present with an on-call consultant off-site.

Statistical Analysis Our primary outcome measure was time taken to extubate a patient after arrival to the ICU. Mean or median was calculated for continuous variables with standard deviation or interquartile range, reported when appropriate. For continuous outcomes, unpaired Student’s t test or Mann-Whitney U test were used to assess for differences in parametric and nonparametric data, respectively. For categorical variables, odds ratios and their 95% confidence intervals were reported, and differences were assessed using Fisher’s exact test. To account for potential covariates, we performed multiple linear regression on extubation time and duration of pressure support, and multiple Poisson regression on ICU length of stay using complete cases. Three multivariate models were created. Model 1 used an agnostic approach to determine the covariates included in the model, performing forward variable selection with the likelihood ratio test to add covariates one at a time. At each variable selection step, the most significant covariate that was not in the current model was added if its probability value was at least 0.05. No further covariates were added if the remaining covariates had probability values greater than 0.05. Model 2 used prior knowledge, including covariates known to be associated with the outcome; logistic EuroSCORE, age, type of surgical procedure, pH, partial pressure of carbon dioxide, and lactate. Model 3 included all covariates in models 1 and 2 and represents a sensitivity analysis to ensure the robustness of model 1 and 2 findings. To improve the normality assumption for multiple linear regression, extubation time and duration of pressure support were logarithmically transformed. For length of stay in ICU, which is a count variable with no zeros, a Poisson regression was performed on the length of stay in ICU excluding the first day of stay. Although the estimated effect is a ratio for the three outcomes (Appendix B), we reported the estimated change in minutes or days, to facilitate interpretation, by multiplying median quantity before intervention by the estimated percentage change. When applicable, modifiers on the intervention were assessed by including the interaction terms into the model. A probability value of less than 0.05 was considered significant. Analysis was performed using R software version 3.1.1 (R-Foundation, Vienna, Austria).

Results During the preprotocol period, 128 patients were admitted to the ICU after cardiac surgery, of which 99 met extubation protocol criteria. During the postprotocol

Ann Thorac Surg 2016;-:-–-

COVE ET AL SIMPLE PROTOCOL REDUCES EXTUBATION TIME

3

Fig 1. Extubation protocol flow diagram. (A ¼ adrenaline; CABG ¼ coronary artery bypass grafting; FIO2 ¼ fraction of inspired oxygen; NA ¼ noradrenaline; PaO2 ¼ arterial partial pressure of oxygen; PEEP ¼ positive end-expiratory pressure; RT ¼ respiratory therapist; SaO2 ¼ arterial saturation of oxygen; SBT ¼ spontaneous breathing trial.)

period, 168 were admitted, with 102 meeting protocol criteria. Descriptive statistics are presented in Table 1. There were significantly more Malay participants and fewer Indian participants in the postprotocol group. Median pH was slightly lower on arrival to the ICU in the preprotocol group, and median partial pressure of carbon dioxide levels were slightly higher (7.37 versus 7.40; p < 0.001; and 40 versus 39 mm Hg; p ¼ 0.01, respectively). Lactate was lower in the postprotocol group (2.20 versus 3.56 mmol/L; p < 0.001). Quality-related events occurred more often in the preprotocol group, but this difference was not significant. However, significantly more patients in the preprotocol group received an intraaortic balloon pump (9 versus 2; p ¼ 0.031), although the proportion of patients requiring interventions for hypotension did not significantly differ (32.32% versus 22.6%; p ¼ 0.154). Patients receiving an intraaortic balloon pump had a higher median lactate than those without an intraaortic balloon pump (2.6 versus 3.5 mmol/L), but this difference was not statistically significant. The remaining characteristics were similar in both groups. Unadjusted median extubation time was reduced by 35% after implementation of the protocol (620 versus 405 minutes; p < 0.001). Excluding patients receiving an intraaortic balloon pump did not greatly change this observation (Table 2). When adjusted for covariates from models 1 and 2 (ie, model 3), extubation time was reduced

by 144 minutes (95% confidence interval, 70 to 209 minutes) compared with the preprotocol group (Table 3). Similarly, unadjusted median time spent on pressure support was reduced by 24% in the postprotocol group (130 versus 98 minutes; p < 0.001), which remained significant after adjustment (Table 3, model 3). Median ICU length of stay was reduced from 2 days to 1 day (p < 0.001), and remained significant after adjusting for confounding factors in model 3 (Table 3). Modifiers including low pH and ICU admission after-hours were assessed in the model. Although the interaction effects were not significant (Appendix B, Table 1), protocolized extubation had more impact in patients arriving out of hours, and less impact when patients arrived with low pH values (Table 3). The reintubation rate was unaffected, with 2 patients requiring reintubation in both groups (odds ratio, 1.03; 95% confidence interval, 0.07 to 14.48; p ¼ 1). Both patients in the postprotocol group were reintubated more than 5 days after extubation; in the preprotocol group both reintubations occurred within 12 hours.

Comment In this study we show that the introduction of a simple extubation protocol into the cardiothoracic ICU is associated with a 35% reduction in time to extubation, (620 versus 405 minutes) and a reduction in length of

4

COVE ET AL SIMPLE PROTOCOL REDUCES EXTUBATION TIME

Ann Thorac Surg 2016;-:-–-

Table 1. Demographic and Descriptive Statistics Preprotocol (n ¼ 99)

Variable Median age, y (IQR) Median EuroSCOREa (IQR) Potential difficult airway (%) Quality-related eventsb (%) Sex (%) Female Racea (%) Chinese Malay Indian Others Number of participants (%) Intraaortic balloon pump Hypotension Reintubationa Acidosis Median pHd (IQR) Median PCO2,d mm Hg (IQR) Median lactate,a,d mmol/L (IQR) Arrival time to ICU (%) Between 9 AM and 4:59 PM Between 5 PM and 8:59 AM Procedurea (%) Valve surgery onlye Bypass surgery with 1–2 vessels (w/out valve surgery) Bypass surgery with 3 vessels (w/out valve surgery)

60 2.54 11 18

Postprotocol (n ¼ 102)

(55, 68) (1.51, 4.38) (11.11) (18.18)

60 2.28 10 11

Mean, Median Difference/OR (95%)

(54, 69) (1.42, 4.69) (9.8) (10.78)

0 0.16 1.15 0.59

(3, 3) (0.67, 0.35) (0.42, 3.18) (0.30, 1.19)

Two-Sided p Value 0.901 0.489 0.82 0.162

20 (20.2)

21 (20.59)

1.02 (0.49, 2.16)

62 13 16 7

63 26 5 8

. 1.96 (0.88, 4.56) 0.31 (0.08, 0.96) 1.12 (0.33, 3.88)

0.018c 0.097 0.032 1

0.2 (0.021, 1.01) 0.61 (0.31, 1.19) 0.95 (0.068, 13.349)

0.031 0.154 1

(63.27) (13.27) (16.33) (7.14)

9 (9.09) 32 (32.32) 2 (2.06)

(61.76) (25.49) (4.9) (7.84)

2 (1.96) 23 (22.55) 2 (1.96)

7.37 (7.33, 7.41) 40 (38, 44) 3.56 (1.90, 5.90)

7.40 (7.36, 7.43) 39 (36.1, 42) 2.20 (1.40, 3.90)

0.03 (0.013, 0.05) 2 (3, 0.4) 1 (1.8, 0.48)

1

<0.001 0.01 <0.001

56 (56.57) 43 (43.43)

66 (64.71) 36 (35.29)

. 0.71 (0.39, 1.30)

. 0.21

21 (21.21) 16 (16.16)

19 (18.81) 24 (23.76)

. 1.65 (0.66, 4.42)

0.435c 0.37

62 (62.63)

58 (57.43)

1.03 (0.48, 2.26)

1

a b There are some missing values. Reoperation, bleeding, tamponade, bronchospasm, renal replacement therapy, respiratory acidosis, c d e death. Fisher’s exact test was performed. Values taken on ICU arrival. Includes aortic, mitral, and tricuspid valve replacement or repair, or any combination.

ICU ¼ intensive care unit;

IQR ¼ interquartile range;

PCO2 ¼ partial pressure of carbon dioxide.

ICU stay of 1 day. Early extubation is desirable in patients undergoing elective cardiac surgery as it reduces elective cancellations, ICU length of stay, and cost by as much as 25% [21]. As early as the 1970s, physicians explored early ventilator weaning after cardiac surgery [22], but there was little interest during the 1980s

because of the popularity of opioid-based anesthesia, which necessitated prolonged postoperative ventilation [16]. However, by the 1990s frequent interruptions to surgical schedules, resulting from a lack of available ICU beds, prompted surgical teams to revisit early extubation.

Table 2. Summary Statistics for Main Outcome Measures, Including and Excluding Patients With Intraaortic Balloon Pumps Variable Median time to extubation, min (IQR) Median time on pressure support,a min (IQR) Median number of days in ICUb (reported range) IABP patients excluded Median time to extubation, min (IQR) Median time on pressure support,a min (IQR) Median number of days in ICUb (IQR) a

There is 1 missing value in the postprotocol group.

IABP ¼ intraaortic balloon pump;

b

Preprotocol

Postprotocol

p Values

620 (395, 925) 130 (95, 220) 2 (1, 8)

405 (323, 571) 105 (80, 157) 1 (1, 2)

<0.001 0.015 <0.001

605 (385, 915) 130 (93, 210) 2 (1, 8)

403 (321, 574) 105 (80, 155) 1 (1, 2)

<0.001 0.053 <0.001

There are 3 missing values in the preprotocol group and 2 in the postprotocol group.

ICU ¼ intensive care unit;

IQR ¼ interquartile range.

Ann Thorac Surg 2016;-:-–-

COVE ET AL SIMPLE PROTOCOL REDUCES EXTUBATION TIME

5

Table 3. Unadjusted and Adjusted Effect of the Protocol on Log Transformed Minutes to Extubation, Minutes on Pressure Support, and Length of Stay in the Intensive Care Unita Variable Minutes to extubationd Reduction in minutes (95% CI) Minutes on pressure supportc Reduction in minutes (95% CI)

Unadjusted

Model 1

Model 2b

Model 3c

186 (116 to 246) p < 0.001

157 (87 to 218) p < 0.001

152 (78 to 216) p < 0.001

144 (70 to 209) p < 0.001

28 (1 to 49) p ¼ 0.047

27 (0 to 48) p ¼ 0.051

31 (3 to 53) p ¼ 0.032

31 (3 to 52) p ¼ 0.034

Number of days in CTICUe Reduction in days (95% CI)

1.62 (1.28 to 1.8) 1.55 (1.13 to 1.77) 1.56 (1.15 to 1.77) 1.54 (1.11 to 1.76) p < 0.001 p < 0.001 p < 0.001 p < 0.001 Including interaction effect between protocol status and arrival time to CTICU, and between protocol status and pH status for minutes to extubation If arrival time is normal work hours and arrival . 78 (204 to 264) 139 (59 to 208) 34 (272 to 235) pH is normal (95% CI) p ¼ 0.529 p ¼ 0.002 p ¼ 0.794 If arrival time is out of normal hours and arrival . 129 (32 to 209) . 105 (0 to 191) pH is normal (95% CI) p ¼ 0.012 p ¼ 0.051 If arrival time is normal work hours and arrival . 214 (153 to 407) 259 (36 to 397) 207 (162 to 401) pH is low (95% CI) p ¼ 0.199 p ¼ 0.029 p ¼ 0.214 If arrival time is out of normal work hours and . 252 (12 to 397) . 256 (23 to 399) arrival pH is low (95% CI) p ¼ 0.043 p ¼ 0.037

a

Median time before intervention has been multiplied by the estimated percentage reduction, and its 95% confidence interval (CI), to provide reduction in b c minutes and days. Model 2 includes logistic EuroSCORE, age, type of surgical procedure, pH, PCO2, and lactate. Model 1 includes age and low d Model 1 includes low blood pressure blood pressure status, identified by the forward selection; model 3 includes variables from models 1 and 2. e status, arrival time, pH, by the forward selection method; model 3 includes variables from models 1 and 2. Day 0 was excluded and then a Poisson regression was performed. Model 1 includes IABP, and categorized lactate values, by the forward selection method; model 3 includes variables from models 1 and 2. CI ¼ confidence interval;

CTICU ¼ cardiothoracic intensive care unit;

In 1993, Westaby and colleagues [23] published their experience of extubating carefully selected patients in a postoperative recovery area. They demonstrated these patients could be safely extubated within 2 to 3 hours and discharged directly to a postoperative ward. Subsequently, the term “fast-track cardiac surgery” was quickly adopted, describing a short postoperative ICU stay of less than 1 day in which extubation occurs within 4 hours of surgery. Fast-track protocols require significant workflow changes. In particular, anesthesia needs to be adjusted, relying less on high-dose opioid techniques [24] and keeping intraoperative patient temperatures greater than 32 C [23]. Implementation of protocols that direct intraoperative management, as well as ICU management, can be challenging, especially where anesthesia and ICU teams function independently. Implementation of our protocol was relatively simple because we designed an ICUcentric protocol that only directs care after the patient is admitted to the ICU, which is distinct from existing fasttrack approaches because no intraoperative changes were necessary. Our protocol appears safe because there was no effect on reintubation rates, which aligns with previous studies of both fast-track protocols and ventilator weaning protocols in general ICUs [16, 18, 25–27]. It is arguable our reintubation rate is too low, but a large safety margin is necessary because our consultant intensivist coverage, as well as consultant surgical coverage, is off-site after hours.

PCO2 ¼ partial pressure of carbon dioxide.

We also observed a reduction in length of ICU stay, but it is not entirely clear why a reduction of median extubation time from 620 to 405 minutes would be associated with a reduction in length of stay by 1 day. Previous studies of ICU-centric interventions, mostly conducted in general ICUs, have a variable impact on length of stay, with several showing no change [28, 29] leading some authors to conclude weaning protocols are unnecessary [30] and may prolong weaning [31]. In contrast to general ICU settings, postoperative cardiac surgery patients are more homogeneous and usually not in respiratory failure when placed on mechanical ventilation. This may partly explain our observed effect on shortened length of stay, and reduced extubation times have been associated with shorter ICU length of stay in other studies focusing on cardiac surgery patients [27, 32, 33]. The observed reduction in length of stay may be a consequence of workflow patterns in our cardiothoracic ICU, where more than a third of admissions occurred after 5 PM (Table 1). A median extubation time of 620 minutes (>10 hours) suggests many patients were not extubated until after the consultant ward round the following morning. Reducing extubation time to 405 minutes (<7 hours) allows most patients to be extubated before consultant review, potentially accelerating step-down decisions. This is supported by the multivariate analysis, in which protocol implementation had a larger effect on patients admitted after hours (Table 3). We do not know whether our observed reduction in ICU length of stay was associated with a

6

COVE ET AL SIMPLE PROTOCOL REDUCES EXTUBATION TIME

reduction in hospital stay, but a recent Cochrane review concluded that fast-track protocols have no effect on hospital length of stay [27]. Reduction in ICU length of stay may reduce cost. Our ICU patients are charged a daily ICU fee of $825, although additional ICU costs are incurred (eg, ventilation). The median hospital cost for coronary artery bypass grafting is $23,760 [34], and patients are counseled for an expected bill of $24,750 before insurance and government subsidies. Therefore, reducing ICU stay by 1 day potentially saved at least 3% of the overall cost.

Limitations As a retrospective, before and after observational study, several limitations deserve consideration. First, both groups were not identical. In particular, the preprotocol group received more intraaortic balloon pumps and had a lower median pH and higher median lactate on arrival to the ICU. Although most clinicians would argue that the statistically significant pH difference (7.37 versus 7.40) is not clinically significant, the difference in lactate (3.56 versus 2.2 mmol/L) might be considered clinically relevant. The lactate difference might be related to increased intraaortic balloon pump use in the preprotocol group because these patients had a higher lactate. However, after adjusting for these differences (Table 3, Appendix B Table 2), there was still a significant reduction in extubation time. A second limitation is the relatively small size of the two groups. Larger groups could have been achieved with a longer observational period, but this would increase the chance of introducing bias from other practice changes. During this short study period there were no other changes to ICU or anesthesia management. Third, this study design cannot exclude the potential for the Hawthorne effect to influence our findings. Another potential limitation is unequal variance, as the interquartile ranges are larger in the preprotocol group (Table 2). Although this can be accounted for in the statistical analysis, we also argue that reduction of variance is an important practice improvement goal, and has long been recognized in industry as an important quality marker [35].

Conclusions In this study we have shown that introduction of a simple, team-based extubation protocol for postoperative cardiac surgery patients is associated with reduced extubation time and reduced overall ICU length of stay without any increase in extubation failure. The authors wish to acknowledge Melvin J. Dajac, RRT-NPS, Arman Del Rosario, RRT-NPS, Joseph Gammad, MD, and Smriti Mahaju, MD, who helped with data collection, and Richard Tierney, MD, who clarified hospital cardiac anesthesia guidelines. This work was supported in part by a grant from the National Medical Research Council in Singapore (M.E.C.), NMRC/ TA/0015/2013, and Centre for Health Services and Policy Research SBRO14/NS01G from the National University Health Systems Pte Ltd (C.S.T.).

Ann Thorac Surg 2016;-:-–-

References 1. Vincent JL. A critical look at critical care. Lancet 2010;376: 1273. 2. Abramson NS, Wald KS, Grenvik AN, Robinson D, Snyder JV. Adverse occurrences in intensive care units. JAMA 1980;244:1582–4. 3. Weingart SN, Wilson RM, Gibberd RW, Harrison B. Epidemiology of medical error. BMJ 2000;320:774–7. 4. McMullin J, Cook D, Griffith L, et al. Minimizing errors of omission: behavioural reenforcement of heparin to avert venous emboli: the BEHAVE study. Crit Care Med 2006;34: 694–9. 5. Moreno RP, Rhodes A, Donchin Y, European Society of Intensive Care. Patient safety in intensive care medicine: the Declaration of Vienna. Intensive Care Med 2009;35:1667–72. 6. Pronovost PJ, Rinke ML, Emery K, Dennison C, Blackledge C, Berenholtz SM. Interventions to reduce mortality among patients treated in intensive care units. J Crit Care 2004;19:158–64. 7. See KC, Phua J, Mukhopadhyay A, Lim TK. Characteristics of distractions in the intensive care unit: how serious are they and who are at risk? Singapore Med J 2014;55:358–62. 8. Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med 2006;355:2725–32. 9. Meddings J, Rogers MA, Macy M, Saint S. Systematic review and meta-analysis: reminder systems to reduce catheterassociated urinary tract infections and urinary catheter use in hospitalized patients. Clin Infect Dis 2010;51:550–60. 10. Bonten MJ, Kollef MH, Hall JB. Risk factors for ventilatorassociated pneumonia: from epidemiology to patient management. Clin Infect Dis 2004;38:1141–9. 11. Hansen-Flaschen JH, Brazinsky S, Basile C, Lanken PN. Use of sedating drugs and neuromuscular blocking agents in patients requiring mechanical ventilation for respiratory failure. A national survey. JAMA 1991;266:2870–5. 12. Watling SM, Dasta JF, Seidl EC. Sedatives, analgesics, and paralytics in the ICU. Ann Pharmacother 1997;31:148–53. 13. Choi J, Tasota FJ, Hoffman LA. Mobility interventions to improve outcomes in patients undergoing prolonged mechanical ventilation: a review of the literature. Biol Res Nurs 2008;10:21–33. 14. Warren DK, Shukla SJ, Olsen MA, et al. Outcome and attributable cost of ventilator-associated pneumonia among intensive care unit patients in a suburban medical center. Crit Care Med 2003;31:1312–7. 15. Shorr AF, Kollef MH. Ventilator-associated pneumonia: insights from recent clinical trials. Chest 2005;128(5 Suppl 2): 583S–91S. 16. Cheng DC. Pro: early extubation after cardiac surgery decreases intensive care unit stay and cost. J Cardiothorac Vasc Anesth 1995;9:460–4. 17. National Voluntary Consensus Standards for Cardiac Surgery. National Quality Forum. Available at https://www. qualityforum.org/Publications/2005/01/National_Voluntary_ Consensus_Standards_for_Cardiac_Surgery.aspx. Accessed December 17, 2015. 18. Fitch ZW, Debesa O, Ohkuma R, et al. A protocol-driven approach to early extubation after heart surgery. J Thorac Cardiovasc Surg 2014;147:1344–50. 19. Briner C. EB42 Implementation of an early extubation protocol for cardiovascular surgery patients. Crit Care Nurse 2014;34:e4. 20. Roques F, Michel P, Goldstone AR, Nashef SA. The logistic EuroSCORE. Eur Heart J 2003;24:881–2. 21. Cheng DC, Karski J, Peniston C, et al. Early tracheal extubation after coronary artery bypass graft surgery reduces costs and improves resource use. A prospective, randomized, controlled trial. Anesthesiology 1996;85:1300–10. 22. Prakash O, Jonson B, Meij S, et al. Criteria for early extubation after intracardiac surgery in adults. Anesth Analg 1977;56:703–8.

Ann Thorac Surg 2016;-:-–-

23. Westaby S, Pillai R, Parry A, et al. Does modern cardiac surgery require conventional intensive care. Eur J Cardiothorac Surg 1993;7:313–8. 24. Mora CT, Dudek C, Torjman MC, White PF. The effects of anesthetic technique on the hemodynamic response and recovery profile in coronary revascularization patients. Anesth Analg 1995;81:900–10. 25. Robertson TE, Sona C, Schallom L, et al. Improved extubation rates and earlier liberation from mechanical ventilation with implementation of a daily spontaneous-breathing trial protocol. J Am Coll Surg 2008;206:489–95. 26. Chan PK, Fischer S, Stewart TE, et al. Practising evidencebased medicine: the design and implementation of a multidisciplinary team-driven extubation protocol. Crit Care 2001;5:349–54. 27. Zhu F, Lee A, Chee YE. Fast-track cardiac care for adult cardiac surgical patients. Cochrane Database Syst Rev 2012;10:CD003587. 28. Lellouche F, Mancebo J, Jolliet P, et al. A multicenter randomized trial of computer-driven protocolized weaning from mechanical ventilation. Am J Respir Crit Care Med 2006;174: 894–900. 29. Dries DJ, McGonigal MD, Malian MS, Bor BJ, Sullivan C. Protocol-driven ventilator weaning reduces use of

COVE ET AL SIMPLE PROTOCOL REDUCES EXTUBATION TIME

30.

31. 32.

33.

34.

35.

7

mechanical ventilation, rate of early reintubation, and ventilator-associated pneumonia. J Trauma 2004;56:943–51. Krishnan JA, Moore D, Robeson C, Rand CS, Fessler HE. A prospective, controlled trial of a protocol-based strategy to discontinue mechanical ventilation. Am J Respir Crit Care Med 2004;169:673–8. Tanios MA, Nevins ML, Hendra KP, et al. A randomized, controlled trial of the role of weaning predictors in clinical decision making. Crit Care Med 2006;34:2530–5. Camp SL, Stamou SC, Stiegel RM, et al. Can timing of tracheal extubation predict improved outcomes after cardiac surgery. HSR Proc Intensive Care Cardiovasc Anesth 2009;1: 39–47. Camp SL, Stamou SC, Stiegel RM, et al. Quality improvement program increases early tracheal extubation rate and decreases pulmonary complications and resource utilization after cardiac surgery. J Card Surg 2009;24:414–23. Ministry of Health Singapore. Hospital bill sizes. Available at https://www.moh.gov.sg/content/moh_web/home/costs_and_ financing/HospitalBillSize/heart_surgery_coronaryarterybypass graftwithoutangiography.html. Accessed December 17, 2015. FDA. Guidance for industry: process validation, general principles and practices. Silver Spring, MD: US Food and Drug Administration; 2011.