- Email: [email protected]
An enhanced recovery pathway reduces duration of stay and complications after open pulmonary lobectomy
An enhanced recovery pathway reduces duration of stay and complications after open pulmonary lobectomy Amin Madani, MD,a,b Julio F. Fiore, Jr, PhD,b Yifan Wang, BSc,b Jimmy Bejjani, MD,b Lojan Sivakumaran, BSc,b Juan Mata, MD,b Debbie Watson, RN,a Franco Carli, MD, MPhil,c David S. Mulder, MD,a Christian Sirois, MD,a Lorenzo E. Ferri, MD,a,b and Liane S. Feldman, MD,a,b Montreal, Quebec, Canada
Background. Few studies have investigated the effectiveness of enhanced recovery pathways (ERP) for lung resection. This study estimates the impact of an ERP for lobectomy on duration of stay, complications, and readmissions. Methods. Patients undergoing open lobectomy were identified from an OR database between 2011 and 2013. Beginning September 2012, all patients were managed according to a 4-day multidisciplinary ERP with written daily patient education treatment plans, multimodal analgesia, early diet, structured mobilization and standardized drain management. Pre-pathway (PRE) and post-pathway (POST) patients were compared in terms of duration of stay, complications, and readmissions. Results. We identified 234 patients (PRE, 127; POST, 107). Groups were similar with respect to age, gender, American Society of Anesthesiologists score, and baseline pulmonary function. Compared with the PRE group, the POST group had decreased duration of stay (median, 6 [interquartile range (IQR), 5–7] vs 7 [6–10] days; P < .05), total complications (40 [37%] vs 64 [50%]; P < .05), urinary tract infections (3 [3%] vs 15 [12%]; P < .05), and chest tube duration (median, 4 [IQR, 3–6] vs 5 [4–7] days; P < .05), with no difference in readmissions (7 [7%] vs 6 [5%]; P < .05) or chest tube reinsertion (4 [4%] vs 6 [5%]; P < .05). Decreased duration of stay was driven by patients without complications (median, 5 [IQR, 4–6] vs 6 [5–7] days; P < .05). Conclusion. Implementation of a multimodal ERP for lobectomy was associated with decreased duration of stay and complications with no difference in readmissions. (Surgery 2015;158:899-910.) From the Department of Surgery,a Steinberg-Bernstein Centre for Minimally Invasive Surgery and Innovation, b and the Department of Anaesthesia,c McGill University, Montreal, Quebec, Canada
LUNG CANCER is the leading cause of cancer death worldwide.1 Despite advances in multimodal therapy, lung resection remains the cornerstone of curative intent management of localized disease, with an estimated 40,000 patients undergoing
Funded by an investigator-initiated research grant from Ethicon Canada. The Steinberg-Bernstein Centre for Minimally Invasive Surgery and Innovation (McGill University Health Centre, Montreal, Canada) is funded through an unrestricted educational grant from Covidien Canada. Accepted for publication April 5, 2015. Reprint requests: Liane S. Feldman, MD, FRCS, FACS, McGill University Health Centre, 1650 Cedar Avenue, Rm L9-309, Montreal, Quebec H3G 1A4, Canada. E-mail: liane.feldman@ mcgill.ca. 0039-6060/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2015.04.046
anatomic lung resection every year in the United States.2 Yet, postoperative complications remain high, ranging from 20 to 50%,3-6 leading to higher costs,7 prolonged recovery, and poorer long-term outcomes.8 Enhanced recovery pathways (ERPs) are multimodal, evidence-based protocols including step-by-step management plans throughout the perioperative period. Evidence suggests that ERPs improve postoperative recovery and decrease morbidity, hospital duration of stay, and cost of care for a variety of operative procedures915 ; however, few reports describe the impact of an ERP on outcomes of lung resection.16-20 To improve quality and recovery for patients undergoing lung resection, we implemented a multidisciplinary ERP, including patient educational material and an evidence- and consensus-based, standard perioperative management protocol. The purpose of this study was to estimate the extent to SURGERY 899
900 Madani et al
which an ERP for elective open lobectomies impacts hospital duration of stay, and 30-day complication and readmission rates in comparison with traditional care. In addition, we estimated the extent to which adherence to specific perioperative care elements of the ERP are associated with improved outcomes. MATERIALS AND METHODS This study is reported according to the STROBE criteria for observational studies (available at: www. strobe-statement.org). The study protocol was approved by the institutional review board and conforms to the Canadian Tri-Council Policy Statement of Ethical Conduct. Study design, setting, and participants. In this retrospective cohort study, patients undergoing lung resection for cancer at a single academic center between August 2011 and October 2013 were identified from an operating room database. Adult patients (>18 years old) were included if they underwent an open, scheduled lobectomy for either primary or secondary lung cancer. Exclusion criteria included age <18, benign disease, nonelective operations, pneumonectomy, bronchoplasty, nonanatomic lung resection, and extended resection of adjacent organs (eg, carina, chest wall). Although video-assisted thoracoscopic surgery (VATS) is the treatment of choice in many centers across North America, it was also excluded because only a minority of lobectomies at our institution are performed using this approach and inclusion would have confounded the results of this study. Beginning in September 2012, patients undergoing lung resection were managed according to a standardized pathway protocol; before this date, patients were managed according to surgeon preference. Participants were categorized according to the date of their operation into a prepathway group (PRE) or post-pathway group (POST) and compared in terms of demographics, duration of stay, short-term (30-day) complications, hospital readmissions, and adherence to elements of the pathway. Development and implementation of ERP for lung resection. In an effort to improve patient outcomes, decrease cost, and reduce variability, a multidisciplinary team was established at our institution in 2008 to develop and implement standardized ‘‘enhanced recovery’’ management protocols for high prevalence operative procedures. The team is led by a general surgeon and anesthesiologist and includes expertise in nursing (inpatient and outpatient), physiotherapy, pharmacy, pain service, nutrition, and medical informatics. A full-time care pathway coordinator is
Surgery October 2015
dedicated to the program. The goal of the enhanced recovery program is to facilitate widespread uptake of evidence-based interventions shown to decrease physiologic stress (eg, preoperative carbohydrate, exercise, and afferent neural blockade) and decrease the use of interventions that hamper recovery without improving outcomes (eg, prolonged preoperative fasting, prolonged insertion of urinary drains, and delayed feeding). Improving patient recovery and reducing morbidity decreases the need for hospitalization, reflected in shorter durations of hospital stay. The program has implemented pathways for a variety of procedures, including colorectal, esophageal, and hepatic surgeries. For the establishment of the lung resection pathway, the team also included 3 thoracic surgeons along with specific nursing expertise for thoracic patients. The objective of the ERP was to introduce a standard multimodal approach, including patient education and counselling, opioid-sparing pain control, early and structured mobilization, early feeding and optimization of nutritional status, standardized drain management, and target discharge with written patient goals for each postoperative day (POD). A medical librarian assisted with systematic literature searches to identify evidence-based interventions for the perioperative management of thoracic patients. Elements of the pathway were discussed extensively by various team members and stakeholders, and evaluated in terms of feasibility and adoptability amongst patients and caregivers. Once consensus was reached, the elements of the ERP were used to create physician orders (Table I), customized nursing records for each POD, and patient education material (available at: http:// muhcpatienteducation.ca). Implementation of this pathway did not require additional allied health care workers, such as nurses, physiotherapists, or other individuals involved in preoperative or postoperative care. Intervention. Before their operation, all patients underwent preoperative medical evaluation by an internist, anesthesiologist, and nurse in the preoperative clinic. After the ERP was implemented in September 2012, nurses also provided personalized patient education using an information booklet that reviewed the procedure using simplified terminology and illustrations. This included a review of daily goals for pain control, diet, drain and fluid management, structured daily physical activity, and target discharge date on POD 4. A physiotherapist provided instructions on the use of an incentive spirometer in preparation for the operation, along with general recommendations to increase physical
Madani et al 901
Surgery Volume 158, Number 4
Table I. Daily milestones of the enhanced recovery pathway for lung resection Traditional care (PRE) Preoperative Patient education Intraoperative Analgesia Extubation Postoperative Analgesia Urinary drain
Chest tube
Nutrition
Chest x-ray
Mobilization
Chest rehabilitation Target discharge
Enhanced recovery (POST)
Nonstandardized education given in surgeon’s office
Standardized preoperative education protocol Information booklet with daily goals
Thoracic epidural inserted Based on anesthesiologist preference
Preferred extubation in the operating room or in the postanesthesia care unit
Thoracic epidural stop test performed on the day the last chest tube is removed Nonstandardized management POD 1: drain removed if adequate urine output based on surgeon preference If no urine output after 8 hours of removal, a bladder scan is performed an a urinary retention protocol is followed POD 0: maintained at 20 cmH20 POD 0: maintained at 20 cmH20 suction suction POD 1: remove suction Weaning based on surgeon POD 2: remove chest tube#1 if <300 mL/24 h, preference nonchylous and no air leak POD 3: remove chest tube#2 if <300 mL/24 h, nonchylous and no air leak No nasogastric tube No nasogastric tube Diet advanced progressively based POD 0: clear fluid diet on surgeon preference POD 1: diet as tolerate After either chest tube suction removal or chest After either chest tube suction tube removal removal or chest tube removal No clamp test Clamp test based on surgeon preference Physical activity encouraged by POD 0: up in chair with assistance as tolerated health care provider POD 1: up in chair 3 times per day for all meals + 30–60 minutes each time, ambulate in hallway 2 times per day with assistance POD 2: out of bed for all meals and $8 hours during the day, walking in hallway 17.5–35 meters 3 times per day with assistance POD 3: increase ambulation to 75 meters 3–5 times per day Spirometry 10 times every hour while awake Chest physiotherapy every 4 hours None POD 3 if 1 chest tube POD 4 if 2 chest tubes
POD, Postoperative day.
activity and performance status. When relevant, consultations were made to the appropriate service for comorbidities. If the patient was a smoker, recommendations and referral to a smoking cessation program were made. All open lobectomies were performed by 3 thoracic surgeons with patients under general anesthesia, using a double-lumen endotracheal intubation for lung isolation and a thoracic epidural catheter for postoperative pain control. Before closure, 1 or 2 chest tubes were inserted based on surgeon preference. Patients were extubated in the operating room or on arrival to the
postanesthesia care unit before transfer to the thoracic surgery ward after observation for 2–4 hours and performance of a chest radiograph. Postoperative management also differed significantly between the 2 groups. In the PRE group, orders were highly variable and based on surgeon preference. There was no target discharge date and protocols for feeding, fluid, drain (urine and chest tube), and epidural management were according to surgeon discretion. Epidural catheters were removed typically after the last chest tube was removed. Although the team encouraged physical activity, there was no structured mobilization plan.
902 Madani et al
In the POST group, structured daily plans served to guide patient mobilization, pain control, feeding, weaning of chest tubes and urine catheter, and chest wall exercises with planned discharge on POD 4 if there were no complications (although patients could leave earlier if appropriate). ERP milestones are summarized in Table I. Outcomes and data collection. Data on demographics, preoperative evaluation, postoperative course, achievement of daily milestones, and hospital visits were obtained from both paper and electronic hospital charts. Postoperative complications were defined a priori according to the Seely et al modification of the Dindo–Clavien grading system.5 Pulmonary complications were defined as pneumonia (abnormal radiograph and sputum production with fever, leukocytosis, or both), pleural effusion or pneumothorax requiring drainage, atelectasis with sublobar collapse requiring bronchoscopy, respiratory failure requiring endotracheal intubation, pulmonary embolus, prolonged air leak (>7 days), chylothorax, hemoptysis, aspiration pneumonitis and pulmonary edema requiring diuresis. Primary outcome was hospital duration of stay and secondary outcomes were morbidity, readmission, emergency department visits within 30 days, and adherence to each element of the pathway. Statistical analysis. The PRE and POST groups were compared by univariate analysis using Pearson’s v2 (for categorical variables), Student’s t test (for normally distributed continuous variables), or Wilcoxon’s signed-rank test (for non-normally distributed continuous variables). Data were expressed as n (%), mean (SD), median (interquartile range [IQR]), as appropriate. Multivariate linear and logistic regression was used to estimate the adjusted association between the presence of an ERP and duration of stay, postoperative complications and hospital readmissions. Adherence to 6 individual ERP elements (standardized preoperative education, introduction of oral diet, urinary catheter removal, epidural analgesia catheter removal, chest tube removal, and intravenous fluid discontinuation) were examined for their association with duration of stay using a stepwise logistic regression with backward elimination of variables with a P value >.1. Multiple imputations by chained equation (10 imputed datasets) were used to replace missing data on adherence rates and adjustment variables. Sample size. For the purposes of power calculation, the difference in duration of stay before and after the pathway was used. In the year preceding the study, average duration of stay after lobectomy
Surgery October 2015
was 8.2 ± 6 days with only 19% of patients discharged on or before the target discharge date of POD 4. Assuming duration of stay decreased to 6.5 ± 6 days,17 a minimum sample size of 80 per group was calculated. RESULTS A total of 448 patients underwent elective lung resections between August 2011 and October 2013. Two hundred fourteen patients were excluded for VATS or nonanatomic resections (n = 185), benign disease (n = 7), extended resections (n = 10) or pneumonectomy (n = 12), with 234 meeting the study inclusion criteria. One hundred twentyseven patients were treated before the implementation of the pathway on September 2012 (PRE group) with the subsequent 107 patients receiving care after the pathway was implemented (POST group; Figure). Groups were similar in terms of demographics, baseline pulmonary function, or comorbidities (Table II). Missing data ranged from <1% for age to 55% for forced expiratory volume in 1 second (FEV1). Hospital duration of stay was significantly reduced after implementation of the ERP (6 [IQR, 5–7] vs 7 [6–10] days; P < .01; Table III). This decrease was mostly owing to decreased duration of stay in patients with no complications (5 [IQR, 4–6] vs 6 [5–7] days; P < .01). Compared with the PRE group, the POST group had a higher proportion of patients discharged by the POD 4 target (22% vs 5%; P < .01), with no difference in readmissions, emergency department visits after discharge, or proportion of patients with prolonged duration of stay (>14 days; Table III). The POST group also had a greater adherence to most individual elements of the ERP, with the exception of introduction of solid food (Table IV). Overall short-term (30-day) morbidity was 50% before and 37% after the implementation of the ERP (P = .03), including a significantly lower rate of urinary tract infections (UTI; 12% vs 3%, P < .01; Table V). There was no difference in pulmonary complications (31% vs 25%; P = .38). Despite earlier removal of chest tubes in the POST group compared with the PRE group (4 days [IQR, 3–6] vs 5 [4–7]; P < .01), there were no differences in incidence of pneumothorax (2 [2%] vs 0) or pleural effusion (2 [2%] vs 6 [5%]) requiring tube reinsertion. One patient in the POST group died on POD 3 from sepsis and subsequent respiratory failure and multiorgan failure. There were no signs of aspiration on chest x-132#ray owing to early feeding. Multivariate linear and logistic regression was used to estimate the impact of the ERP on
Surgery Volume 158, Number 4
Madani et al 903
Figure. Study design. ERP, Enhanced recovery pathway; POST, post-pathway group; PRE, pre-pathway group; VATS, video-assisted thoracoscopic surgery.
duration of stay, postoperative complications, and hospital readmissions (Table VI). After adjusting for age, gender, body mass index, and American Society of Anesthesiologists score, the ERP was associated with decreased duration of stay (average 18% shorter duration of stay; P < .01) and overall 30-day complications, with no association with readmissions (associations with duration of stay and readmissions were also adjusted for complications). Associations between adherence to individual ERP elements and duration of stay using stepwise linear regression identified urinary catheter removal on POD 1 (b, 0.23; 95% CI, 0.43 to 0.02) and removal of the last chest tube on or before POD 3 (b, 0.26; 95% CI, 0.44 to 0.08) as independent predictors of duration of stay (Table VII). DISCUSSION ERPs are coordinated, multidisciplinary, evidence-based protocols including detailed management plans for the preoperative, intraoperative, and postoperative phases of patient care. ERPs have been shown to decrease morbidity, hospital duration of stay, and cost of patient care. Nonetheless, few studies describe the impact of an ERP on patients undergoing lung resection.16-20 In this study, implementation of an integrated ERP for patients undergoing open lobectomy decreased duration of stay and short-term morbidity, with no
difference in either readmissions or emergency department visits after discharge. The patient cohorts before and after ERP implementation were clinically comparable with similar demographics, baseline comorbidities, and pulmonary reserve. Furthermore, perioperative care was directed by the same team of thoracic surgeons, nurses, nutritionists, and physiotherapists on the thoracic surgery ward. Compared with the traditional approach dictated by individual surgeon preference, patients treated after implementation of the ERP had significantly shorter duration of stay by 1 day, with a greater proportion of patients discharged by the target date (POD 4). The main driver of decreased duration of stay was earlier discharge of patients with no complications (Table III). This observation suggests that 1 important effect of the pathway is organizational, allowing for more efficient care of patients who have an uncomplicated postoperative course, whereas patients who experience complications have similar management and duration of hospital stay. However, the ERP was also associated with decreased overall morbidity, primarily driven by fewer urinary tract infections and surgical site infections, resulting in more uncomplicated patients eligible for the earlier discharge date. Nonetheless, only 31% of patients without complications were actually discharged by the POD 4 target. Despite earlier discharge, readmissions and emergency
904 Madani et al
Surgery October 2015
Table II. Patient demographics of patients in the pre-pathway (PRE) and post-pathway (POST) groups Characteristic
PRE (n = 127)
Age (y) Gender (male), n (%) BMI (kg/m2) OR time (min) American Society of Anesthesiologists score, n (%) I II III IV Alcohol history, n (%) Smoking history, n (%) Baseline pulmonary function FEV1 (% predicted) FVC (% predicted) DLCO (% predicted) Diabetes mellitus, n (%) Hypertension, n (%) Thyroid disorder, n (%) Cerebrovascular accident, n (%) Coronary artery disease, n (%) Dyslipidemia, n (%) Chronic obstructive pulmonary disease, n (%) Asthma, n (%) Anemia, n (%) Other cancer <5 years, n (%)
64 57 27 153 5 80 42 0 15 69 92 118 78 20 66 18 4 35 43 33 17 14 15
POST (n = 107)
(11) (45) (5) (56) (4) (63) (33) (0) (12%) (54) (22) (19) (20) (16) (52) (14) (3) (28) (34) (26) (13) (11) (12)
P value
67 65 27 149
(10) (61) (6) (50)
.08 .79 .65 .57
1 64 39 0 18 65
(1) (60) (36) (0) (17) (61)
.07 .65 .72 > 0.99 .37 .25
87 100 80 14 50 12 3 23 47 32 13 18 22
(24) (21) (28) (13) (47) (11) (3) (21) (44) (30) (12) (17) (20)
.27 .37 .70 .58 .43 .20 .76 .07 .09 .89 .80 .45 .13
Data are expressed as mean (standard deviation) or n (%). Missing data: age = 1, body mass index = 24, OR time = 16, American Society of Anesthesiologists score = 4, FEV1 = 130, FVC = 85, DLCO = 94. BMI, Body mass index; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; DLCO, diffusion capacity of the lung for carbon monoxide.
Table III. Duration of stay, readmissions and emergency visits of patients in the pre-pathway (PRE) and post-pathway (POST) groups Characteristic Duration of stay (d) Duration of stay by complication severity (d) None Minor (Clavien I–II) Major (Clavien III–IV) Discharge by target date* Overall No complications Any complications Prolonged duration of stay (>14 d) Readmission Emergency department visit
PRE (n = 127)
POST (n = 107)
P value
7 [6–10]
6 [5–7]
<.01
6 [5–7] 8 [6–12] 16 [10–20]
5 [4–6] 7 [6–10] 11 [7–17]
<.01 .58 .72
(22%) (31%; n = 67) (8%; n = 40) (7%) (6%) (5%)
<.01 <.01 <.01 .20 .20 .43
6 5 1 16 6 4
(5%) (8%; n = 63) (2%; n = 64) (13%) (5%) (3%)
24 21 3 8 7 5
*Target date = postoperative day 4. Data expressed as median [interquartile range] or n (%). One patient died in hospital and was excluded from the analysis. There were no missing data for duration of stay, complications, readmissions and emergency department visits.
visits after discharge remained stable. Therefore, the ERP likely improves flow and organization, especially for routine patients, without compromising care.
The introduction of the pathway was an independent predictor of postoperative complications. The ERP was developed specifically after synthesizing the available evidence with the help of a
Madani et al 905
Surgery Volume 158, Number 4
Table IV. Adherence to individual elements of the enhanced-recovery pathway protocol in the pre-pathway (PRE) and post-pathway (POST) groups Adherence rate
Pathway elements (target POD) Foley removal (POD 1) Solid diet (POD 1) Intravenous fluid discontinuation (POD 1) Ambulation (POD 1) Last chest tube removal (POD 3) Thoracic epidural removal (POD 3)
Median POD
PRE (n = 127), n (%)
POST (n = 107), n (%)
PRE (n = 127), median [interquartile range]
POST (n = 107), median [interquartile range]
62 33 20 39 65 75
88 36 52 63 75 83
2 3 3 2 5 5
1 3 2 1 4 4
(49)* (26) (16)* (31)* (51)* (59)*
(82)* (34) (49)* (59)* (70)* (78)*
[1–2]* [1–4] [2–5]* [2–3]* [4–6]* [4–6]*
[1–1]* [1–4] [1–3]* [1–2]* [3–5]* [3–5]*
*P < .05 between PRE and POST groups. Results shown as proportion (%) of patients meeting the milestone and median postoperative day (POD). Missing data: solid diet = 15, Foley removal = 2, thoracic epidural removal = 4, last chest tube removal = 5 and intravenous fluid discontinuation = 23, ambulation = 64.
Table V. Short-term (30-day) morbidity of patients in the pre-pathway (PRE) and post-pathway (POST) groups, classified by highest grade complication per patient Characteristic Any complications Any pulmonary complications* Minor complications (Clavien I–II) Atelectasis Pneumonia Surgical site infection Urinary tract infection Cardiac tachyarrhythmia Prolonged air leak (>7 d) Postoperative delirium Urinary retention Pulmonary edema Miscellaneousy Major complications (Clavien III-IV) Pleural effusion requiring drainage Pneumothorax requiring drainage Respiratory failure Miscellaneousz Mortality (Clavien V)
PRE (n = 127), n (%) 64 39 51 4 21 7 15 8 13 7 11 10 21 15 6 0 3 10 0
(50) (31) (40) (3) (17) (6) (12) (6) (10) (6) (9) (8) (17) (12) (5) (0) (2) (8) (0)
POST (n = 107), n (%) 40 27 34 2 15 1 3 6 7 7 9 12 10 9 2 2 4 7 1
(37) (25) (32) (2) (14) (1) (3) (6) (7) (7) (8) (11) (9) (8) (2) (2) (4) (7) (1)
P value .03 .38 .22 .69 .72 .07 <.01 > 0.99 .36 .79 > 0.99 .50 .12 .39 .30 .21 .71 .80 .46
*Includes pneumonia, respiratory failure, pleural effusion or pneumothorax requiring drainage, atelectasis, respiratory failure, pulmonary embolus, prolonged air leak, chylothorax, hemoptysis, aspiration pneumonitis and pulmonary edema. yThree or fewer people per group for chronic obstructive pulmonary disease exacerbation, deep vein thrombosis, pulmonary embolus, bradyarrhythmia, acute renal failure, chylothorax, syndrome of inappropriate antidiuretic hormone, ileus, Clostridium difficile colitis, pancreatitis, and aspiration pneumonitis. zThree or fewer people per group for sepsis, heart failure, myocardial infarction, hemoptysis, postoperative hemorrhage, deep wound dehiscence and gastrointestinal bleed.
medical librarian to implement interventions designed to attenuate the surgical stress response, such as opioid-sparing anesthesia, and avoidance of surgical traditions that may delay recovery, such as prolonged use of drains, delayed feeding, and delayed mobilization. Notably, there was a significant, 4-fold decrease in urinary tract infections after the pathway was implemented. This was not surprising, given that indwelling urinary catheters
in postoperative patients are associated with nosocomial infections21 and that urinary catheter removal on POD 1 can be accomplished safely with use of a bladder scan–based urinary retention protocol, even when a thoracic epidural remains in place.22 However, the existence of this evidence alone was not enough to change behavior; having standardized orders allowed for better adherence.
906 Madani et al
Surgery October 2015
Table VI. Impact of an enhanced recovery pathway implementation on duration of stay and complication and readmission rates Characteristic
Estimate*
95% CI
P value
Duration of stayy Complications Readmissions
b 0.18* OR 0.46 OR 1.59
0.29 to 0.06 0.26 to 0.81 0.49 to 5.18
<.01 <.01 .44
*Analysis adjusted for age, gender, body mass index, and American Society of Anesthesiologists score. Analyses of duration of stay and readmissions were also adjusted for complications. yDuration of stay was log transformed and should be interpreted as the percentage change in duration of stay between groups. One patient died in hospital and was excluded from the analysis. Imputations: age = 1, body mass index = 24, American Society of Anesthesiologists score = 4.
Table VII. Linear regression model estimating the impact of complications and adherence to individual elements of the enhanced recovery pathway on duration of stay* Estimate (b)*
Any complicationy Foley removal (POD 1)z Last chest tube removal (POD 3)z Solid diet (POD 1)z Thoracic epidural removal (POD 3)z
0.38 0.23
0.22 to 0.55 <.01 0.43 to 0.02 .03
0.26
0.44 to 0.08 <.01
0.19 0.20
95% CI
P value
Characteristic
0.39 to 0.01 0.41 to 0.01
.06 .06
Age, gender, body mass index, American Society of Anesthesiologists score, preoperative education, and intravenous fluid discontinuation were removed from the model (P > .1). One patient died in hospital and was excluded from the analysis. Imputations: age = 1, gender = 0, body mass index = 13, American Society of Anesthesiologists score = 4, complications = 0, standardized preoperative education = 0, solid diet = 8, Foley removal = 2, thoracic epidural removal = 4, last chest tube removal = 5 and intravenous fluid discontinuation = 5. POD, Postoperative day. *Duration of stay was log transformed and should be interpreted as the percentage change in length of stay between patients without and with complications (y), and patients not adhering and adhering to the enhanced recovery pathway elements (z).
Another critical element of the pathway was the removal of chest tubes using a threshold of <300 mL of drainage in 24 hours for removal as long as there was no evidence of ongoing air leak or chylothorax, which resulted in earlier tube removal by 1 day and decreased the duration of stay. The choice to use a threshold of 300 mL was relatively arbitrary in that there are few data to support 1 threshold over another. Also, there were differences in opinion between the 3 surgeons and this number was based on compromise. It is
not uncommon for thoracic patients to have their hospital discharge delayed because drainage of >200–250 mL in 24 hours is considered too high and there is an associated fear that early removal with high drainage volumes may subject patients to unnecessary morbidity by requiring a tube to be reinserted for symptomatic effusions. However, readmissions and interventions to manage recurrent effusions using thresholds as high as 500 mL in 24 hours are very uncommon.23-25 In fact, most patients who experience complications requiring tube reinsertion have chest tube outputs of <200 mL in 24 hours before the removal of the initial chest tubes.23 This suggests that drainage volume is unlikely to play a significant role and the practice of leaving chest tubes until very low drainage volumes have be achieved is not supported in the literature. Our findings support this notion: despite the earlier removal of chest tubes and resultant decrease in duration of stay, there was no change in pulmonary complications requiring an intervention or overall pulmonary complications. However, the study was underpowered to detect to differences of this magnitude. After the ERP was implemented, adherence was greater for most pathway milestones, with the exception of introduction of solid food (Table IV). Analysis of individual surgeons, however, revealed that this was owing to 1 surgeon’s preference for introduction of solid food on POD 3, regardless of the pathway protocol, because surgeons have the ability to ‘‘override’’ the pathway should they feel it is appropriate in a given context. Irrespective, this diet milestone was not a predictor of duration of stay. In fact, there is limited evidence suggesting the impact of each individual intervention of the pathway in isolation, or which element was most responsible for improving patient outcomes. In our study, we found that urinary catheter removal on POD 1 and removal of the last chest tube on or before POD 3 were independent predictors of decreased duration of stay (Table VII). This study estimates the impact of an ERP for lung resection that comprehensively addresses patient care from the preoperative to postoperative period with interventions for patient education, feeding, pain control, mobilization, and tube/ catheter removal. The ERP at our institution was multidisciplinary, achieved through active involvement and consensus by all stakeholders. After implementation, weekly meetings of the surgical recovery team allow the team to make ongoing improvements to the ERP, based on feedback from
Madani et al 907
Surgery Volume 158, Number 4
the managing team and new evidence from the literature. The results of this audit study identified current successes as well as challenges and will drive future refinements in the pathway protocol related to chest tube drainage, diet and mobilization. An important limitation of this study is the exclusion of VATS lobectomies as a potential confounder, despite its increasing popularity throughout many centers in North America. However, the major strength of an ERP seems to be the improvement in flow and organization (eg, early drain removal), which would benefit patients regardless of the operative approach. For instance, evidence from colorectal surgery suggests that implementation of an ERP improves outcomes for both open and minimally invasive surgery.14,22 Therefore, it is likely that an ERP would also benefit patients undergoing VATS lobectomy. Another limitation is the before–after study design and the retrospective nature, which could have introduced unknown confounders that may have biased our analysis and also precluded investigation of reasons for poor adherence with certain pathway elements. Although a prospective trial with a contemporary control arm could have minimized this effect, it would have been difficult to perform given that the treating teams would have to use different protocols for different patients on the same ward. Nonetheless, we limited selection bias by including consecutive patients throughout the duration of the study. Also, the results of this study are based on 2 cohorts that were clinically similar and multivariate analyses that adjusted for potential confounders. Last, target hospital discharge was only achieved in a minority of patients, even among those who had an uncomplicated hospital course and managed to meet all discharge criteria (median POD 5). This difference of 1 day is likely owing to the discrepancy between ‘‘time to readiness for discharge’’ and actual hospital duration of stay, the latter of which tends to be influence by various personal, societal, and organizational factors that are not related to surgical recovery, including hospital culture, patient preference, and availability of caregiver at home.26 Several studies have characterized this phenomenon and estimate a difference of 1 day,26,27 which seems to correspond with the difference between our target discharge POD 4 and the median POD 5 discharge for uncomplicated patients. Additional prospective studies are needed to elucidate the reasons behind this discrepancy. In conclusion, implementation of a multimodal ERP for open lobectomy is associated with decreased duration of stay and short-term
complications, with no difference in readmissions, emergency department visits after discharge, or complications related to early chest tube removal. Early removal of the chest tube and urinary catheter were independent predictors of decreased duration of hospital stay. The authors acknowledge the contribution of all members of the Surgical Recovery (SURE) team for the development and implementation of an enhancedrecovery pathway for lung resections. A.M. is supported by the Quebec Health Science Research Scholarship (FRQ-S) and the McGill Surgeon-Scientist Program. The Steinberg-Bernstein Centre for Minimally Invasive Surgery and Innovation is supported by an unrestricted educational grant from Covidien.
REFERENCES 1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011;61:69-90. 2. McKenna RJ Jr, Houck W, Fuller CB. Video-assisted thoracic surgery lobectomy: experience with 1,100 cases. Ann Thorac Surg 2006;81:421-5. 3. Chen FF, Zhang D, Wang YL, Xiong B. Video-assisted thoracoscopic surgery lobectomy versus open lobectomy in patients with clinical stage non-small cell lung cancer: a meta-analysis. Eur J Surg Oncol 2013; 39:957-63. 4. Phillips JD, Merkow RP, Sherman KL, DeCamp MM, Bentrem DJ, Bilimoria KY. Factors affecting selection of operative approach and subsequent short-term outcomes after anatomic resection for lung cancer. J Am Coll Surg 2012;215:206-15. 5. Seely AJ, Ivanovic J, Threader J, Al-Hussaini A, Al-Shehab D, Ramsay T, et al. Systematic classification of morbidity and mortality after thoracic surgery. Ann Thorac Surg 2010;90: 936-42. 6. Ivanovic J, Seely AJ, Anstee C, Villeneuve PJ, Gilbert S, Maziak DE, et al. Measuring surgical quality: comparison of postoperative adverse events with the American College of Surgeons NSQIP and the Thoracic Morbidity and Mortality classification system. J Am Coll Surg 2014;218:1024-31. 7. Lacin T, Swanson S. Current costs of video-assisted thoracic surgery (VATS) lobectomy. J Thorac Dis 2013;5(Suppl 3): S190-3. 8. Andalib A, Ramana-Kumar AV, Bartlett G, Franco EL, Ferri LE. Influence of postoperative infectious complications on long-term survival of lung cancer patients: a population-based cohort study. J Thorac Oncol 2013;8: 554-61. 9. Kehlet H, Wilmore DW. Evidence-based surgical care and the evolution of fast-track surgery. Ann Surg 2008;248: 189-98. 10. Li C, Ferri LE, Mulder DS, Ncuti A, Neville A, Lee L, et al. An enhanced recovery pathway decreases duration of stay after esophagectomy. Surgery 2012;152:606-14. 11. Lee L, Li C, Landry T, Latimer E, Carli F, Fried GM, et al. A systematic review of economic evaluations of enhanced recovery pathways for colorectal surgery. Ann Surg 2014; 259:670-6.
908 Madani et al
12. Walter CJ, Collin J, Dumville JC, Drew PJ, Monson JR. Enhanced recovery in colorectal resections: a systematic review and meta-analysis. Colorectal Dis 2009;11:344-53. 13. Gustafsson UO, Hausel J, Thorell A, Ljungqvist O, Soop M, Nygren J, et al. Adherence to the enhanced recovery after surgery protocol and outcomes after colorectal cancer surgery. Arch Surg 2011;146:571-7. 14. Larson DW, Lovely JK, Cima RR, Dozois EJ, Chua H, Wolff BG, et al. Outcomes after implementation of a multimodal standard care pathway for laparoscopic colorectal surgery. Br J Surg 2014;101:1023-30. 15. Nicholson A, Lowe MC, Parker J, Lewis SR, Alderson P, Smith AF. Systematic review and meta-analysis of enhanced recovery programmes in surgical patients. Br J Surg 2014; 101:172-88. 16. Muehling BM, Halter GL, Schelzig H, Meierhenrich R, Steffen P, Sunder-Plassmann L, et al. Reduction of postoperative pulmonary complications after lung surgery using a fast track clinical pathway. Eur J Cardiothorac Surg 2008; 34:174-80. 17. Numan RC, Klomp HM, Li W, Buitelaar DR, Burgers JA, Van Sandick JW, et al. A clinical audit in a multidisciplinary care path for thoracic surgery: an instrument for continuous quality improvement. Lung Cancer 2012;78:270-5. 18. Wright CD, Wain JC, Grillo HC, Moncure AC, Macaluso SM, Mathisen DJ. Pulmonary lobectomy patient care pathway: a model to control cost and maintain quality. Ann Thorac Surg 1997;64:299-302. 19. Salati M, Brunelli A, Xiume F, Refai M, Pompili C, Sabbatini A. Does fast-tracking increase the readmission rate after pulmonary resection? A case-matched study. Eur J Cardiothorac Surg 2012;41:1083-7. 20. Maruyama R, Miyake T, Kojo M, Aoki Y, Suemitsu R, Okamoto T, et al. Establishment of a clinical pathway as an effective tool to reduce hospitalization and charges after video-assisted thoracoscopic pulmonary resection. Jpn J Thorac Cardiovasc Surg 2006;54:387-90. 21. Wald HL, Ma A, Bratzler DW, Kramer AM. Indwelling urinary catheter use in the postoperative period: analysis of the national surgical infection prevention project data. Arch Surg 2008;143:551-7. 22. Carli F, Charlebois P, Baldini G, Cachero O, Stein B. An integrated multidisciplinary approach to implementation of a fast-track program for laparoscopic colorectal surgery. Can J Anaesth 2009;56:837-42. 23. Cerfolio RJ, Bryant AS. Results of a prospective algorithm to remove chest tubes after pulmonary resection with high output. J Thorac Cardiovasc Surg 2008;135: 269-73. 24. Bjerregaard LS, Jensen K, Petersen RH, Hansen HJ. Early chest tube removal after video-assisted thoracic surgery lobectomy with serous fluid production up to 500 ml/day. Eur J Cardiothorac Surg 2014;45:241-6. 25. Zhang Y, Li H, Hu B, Li T, Miao JB, You B, et al. A prospective randomized single-blind control study of volume threshold for chest tube removal following lobectomy. World J Surg 2014;38:60-7. 26. Fiore JF Jr, Faragher IG, Bialocerkowski A, Browning L, Denehy L. Time to readiness for discharge is a valid and reliable measure of short-term recovery after colorectal surgery. World J Surg 2013;37:2927-34. 27. Maessen JM, Dejong CH, Kessels AG, von Meyenfeldt MF. Enhanced Recovery After Surgery Group. Length of stay: an inappropriate readout of the success of enhanced recovery programs. World J Surg 2008;32:971-5.
Surgery October 2015
DISCUSSION Dr L. Michael Brunt (St. Louis, MO): The development of care pathways to enhanced recovery after surgery is gaining a lot of interest in the surgical community. The McGill group has made several recent contributions to this area. With the present study, they show modest improvements following implementation of an enhanced recovery pathway after lung resection compared with a control group from before the implementation. I have a few questions for you. Can you comment on the resources and costs of those resources necessary to implement this pathway, such as nursing, physiotherapy, preop counseling, et cetera? Did your institution help support the cost of the program in any way, recognizing the potential cost savings obtained by shorter length of stay, fewer complications, et cetera? Smoking cessation was a part of the preop counseling in the enhanced recovery path group. Did you observe any differences in smoking cessation rates between the 2 groups? Could that have been a factor in the wound infection rate (6% vs 1% post pathway implementation, which approached statistical significance)? Your hospital length of stay did not reach your target time point of four days, which was obtained by only 22% in the enhanced recovery group. The difference between the 2 groups was only 1 day. Why do you think you did not see a greater difference? Did you look at impact post discharge, such as return to full activity, employment, exercise, or other patient-centered outcomes that may have been impacted by this approach? Have you made any changes to your protocol since the study to improve and streamline the periop process? You mentioned VATS, and it is discussed in your paper, but I was wondering if you could comment on VATS and whether you think you will see similar improvements as you did with the open procedures. Dr Amin Madani (Montreal, Quebec, Canada): Our institution was involved in helping us fund this program. First of all, the booklets cost approximately $13,000 to design. This cost was funded through a patient education office at the McGill University Health Center. There was also opportunity costs, setting up the infrastructure and hiring a pathway coordinator, a nurse practitioner, whose full-time job is to run this pathway. All of these costs amount to approximately $120,000 annually.
Surgery Volume 158, Number 4
When you consider that in our institution we have approximately 1,200 patients who undergo the pathway annually, it comes out to approximately $100 per patient. We did not have to hire additional staff (allied health care workers, nurses and physiotherapists) for implementation of the pathway. It was the same staff that worked with patients before the pathway. We had very good buy-in from the allied health care workers. We got them involved very early on in setting up the pathway. They shared their concerns. Their feedback was very valuable and was implemented into the pathway. With respect to the trend of decreased surgical site infections, we can only speculate as to why that is the case. I think probably decreased length of stay played a role. Perhaps it was the fact that they had a decreased length-of-stay, less exposure to hospital pathogens and nosocomial infections. You mentioned cigarette smoking as a possible factor. This is a possibility, but unfortunately we did not look at smoking cessation rates before and after pathway implementation. Since smoking cessation was systematically implemented in the preoperative clinic for these patients, there is a really good chance that that could have played a role in decreasing rates of surgical site infections. Your next question makes an extremely important point: the target discharge was postop day 4, but only a quarter of the patients were actually discharged by postop day 4. The first explanation we could think of when we analyzed this data was that complications were responsible for increasing length-of-stay beyond the target of postop day 4. However, among patients who had no complications, two-thirds of them still left on postop day 5 or later, even though they managed to achieve their milestones by postop day 4. The reason behind this discrepancy, I think, is because our outcome was length of stay. There is a difference between actual discharge date and being medically fit for discharge. Looking at the literature, this discrepancy is usually about 1 day, which may explain why the median postoperative discharge day was postop day 5. Was it the case caregivers were not ready, it was late in the day, or it was a weekend? It is hard to tell from a retrospective study. Your question about post-discharge return to work is currently being evaluated in a subset of these patients who were evaluated prospectively. These patients are a part of another study in which we are looking at the cost-effectiveness analysis of the implementation of our pathway. These data have not been analyzed yet. Looking at previous
Madani et al 909
data in colorectal patients, we have seen that caregiver burden has improved and patients experience faster return to work. Given those results, we expect a similar trend. Dr Richard Anderson (Peoria, IL): Intraoperatively, did you do anything to decrease your air leaks in the postoperative period, such as any fibrin glue or Progel? My other comment is that I think the reason why you did not reach your mark of 4 days is your goals were not high enough. I put 1 chest tube in and remove it as soon as there is no air leak and <300 mL of drainage, regardless of what postop day it is. I think if you go back to what you are doing right now and just set your goals a little bit lower, I think you will probably achieve them. Dr Amin Madani: We do not systematically use any adjuncts such as fibrin glue to decrease air leak rates. Dr Gerald Larson (Louisville, KY): I would like to ask first, was this a surgeon-driven project or a hospital administration-driven project? Did the administrator come to you and say, can we shorten our hospital stay? If it was surgeon driven, I think that’s more laudatory. Second, removal of the Foley catheter and the chest tube had predictive value for length of stay. When the chest tube should be removed is a very personal decision for some surgeons. What were your new guidelines for chest tube removal? How much pushback did you have? How did you manage that issue? Dr Amin Madani: First, with respect to the impetus behind the pathway, I think it is probably both. There is a lot of pressure to decrease costs and length of stay at our institution, but I think the impetus behind it is largely driven by the surgeons and a multidisciplinary group of people working in a culture of trying to improve efficiency and quality of care. The surgical recovery group, so-called SuRe Group at our institution, is led by a surgeon champion, Dr Liane Feldman, and an anesthesiologist champion. It is a group of like-minded individuals coming together to improve efficiency. I think that is really a key message. You brought up a very controversial point with respect to the chest tube output. We chose 300 mL in 24 hours. As long as there was <300 mL, no chylous drainage, and no air leak, chest tubes were allowed to be removed. Why did we choose 300 mL? It was a very arbitrary number. In fact, it was a compromise amongst the 3 thoracic surgeons in the group. There is 1 surgeon who is usually
910 Madani et al
more conservative than the other 2, so we had to compromise. The existing literature on chest tube removal thresholds suggests that the practice of keeping a chest tube for prolonged periods until low drainage amounts have reached is not unsupported by any strong evidence. When I was a junior resident, it was ingrained in my head that you cannot take a chest tube out until there is <200 mL. We are seeing increasing studies showing you can go up to thresholds as high as 500 mL in 24 hours and the rate of tube reinsertion does not change. We have looked at our results and have actually increased our threshold to 450 mL in 24 hours. Dr Jai Prasad (Detroit, MI): I am talking from my experience in using enhanced recovery for colorectal surgery patients. The 2 things that we found very effective were duration of NPO before surgery and the use of intraoperative intravenous fluids. Were there any specific things that you changed as part of your enhanced recovery program in the preoperative or intraoperative period? I know you made a reference to smoking cessation and air leaks. Were there any other factors that were incorporated as part of the enhanced recovery protocol? Dr Amin Madani: Yes, the preoperative visits were very protocolized. We had the patient meet with a preoperative nurse. They go through an extensive amount of counseling. We teach them all their expected goals on every postoperative day. We really try to make an effort to engage the
Surgery October 2015
patient in their own care. They meet with the physiotherapist. They learn about how to do spirometry. We talked about a smoking cessation program. It is a very systematic approach. Intraoperatively, you talked about limiting fluids. In thoracic patients, although there is no specific number that we use, we usually try to limit fluids to avoid complications such as pulmonary edema. Dr James Madura (Phoenix, AZ): I think you could just as easily say that evidence-based medicine and patient expectations reduce duration of stay and complications, because I think that is what these enhanced recovery programs really are. Lung resections may be a little different than GI surgery, but what I have noticed is that there are populations of patients who undergo colon or small bowel resections in whom we can predict will not meet the enhanced recovery program goals. For example, an 80-year-old, diabetic lady who underwent a large cancer operation. Have you identified any of these factors as predictors of not meeting your goals in the enhanced recovery program? Dr Amin Madani: That’s a very good point. I agree that there are some patients who you cannot really fast track. Although we have these protocolized pathways, the surgeon still reserves the capacity and judgment to not adhere to the pathway if they feel it is appropriate. We did not look into specific patient factors that would predict adherence to the pathway milestones. Perhaps we can do that in the future.
Background. Few studies have investigated the effectiveness of enhanced recovery pathways (ERP) for lung resection. This study estimates the impact of an ERP for lobectomy on duration of stay, complications, and readmissions. Methods. Patients undergoing open lobectomy were identified from an OR database between 2011 and 2013. Beginning September 2012, all patients were managed according to a 4-day multidisciplinary ERP with written daily patient education treatment plans, multimodal analgesia, early diet, structured mobilization and standardized drain management. Pre-pathway (PRE) and post-pathway (POST) patients were compared in terms of duration of stay, complications, and readmissions. Results. We identified 234 patients (PRE, 127; POST, 107). Groups were similar with respect to age, gender, American Society of Anesthesiologists score, and baseline pulmonary function. Compared with the PRE group, the POST group had decreased duration of stay (median, 6 [interquartile range (IQR), 5–7] vs 7 [6–10] days; P < .05), total complications (40 [37%] vs 64 [50%]; P < .05), urinary tract infections (3 [3%] vs 15 [12%]; P < .05), and chest tube duration (median, 4 [IQR, 3–6] vs 5 [4–7] days; P < .05), with no difference in readmissions (7 [7%] vs 6 [5%]; P < .05) or chest tube reinsertion (4 [4%] vs 6 [5%]; P < .05). Decreased duration of stay was driven by patients without complications (median, 5 [IQR, 4–6] vs 6 [5–7] days; P < .05). Conclusion. Implementation of a multimodal ERP for lobectomy was associated with decreased duration of stay and complications with no difference in readmissions. (Surgery 2015;158:899-910.) From the Department of Surgery,a Steinberg-Bernstein Centre for Minimally Invasive Surgery and Innovation, b and the Department of Anaesthesia,c McGill University, Montreal, Quebec, Canada
LUNG CANCER is the leading cause of cancer death worldwide.1 Despite advances in multimodal therapy, lung resection remains the cornerstone of curative intent management of localized disease, with an estimated 40,000 patients undergoing
Funded by an investigator-initiated research grant from Ethicon Canada. The Steinberg-Bernstein Centre for Minimally Invasive Surgery and Innovation (McGill University Health Centre, Montreal, Canada) is funded through an unrestricted educational grant from Covidien Canada. Accepted for publication April 5, 2015. Reprint requests: Liane S. Feldman, MD, FRCS, FACS, McGill University Health Centre, 1650 Cedar Avenue, Rm L9-309, Montreal, Quebec H3G 1A4, Canada. E-mail: liane.feldman@ mcgill.ca. 0039-6060/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2015.04.046
anatomic lung resection every year in the United States.2 Yet, postoperative complications remain high, ranging from 20 to 50%,3-6 leading to higher costs,7 prolonged recovery, and poorer long-term outcomes.8 Enhanced recovery pathways (ERPs) are multimodal, evidence-based protocols including step-by-step management plans throughout the perioperative period. Evidence suggests that ERPs improve postoperative recovery and decrease morbidity, hospital duration of stay, and cost of care for a variety of operative procedures915 ; however, few reports describe the impact of an ERP on outcomes of lung resection.16-20 To improve quality and recovery for patients undergoing lung resection, we implemented a multidisciplinary ERP, including patient educational material and an evidence- and consensus-based, standard perioperative management protocol. The purpose of this study was to estimate the extent to SURGERY 899
900 Madani et al
which an ERP for elective open lobectomies impacts hospital duration of stay, and 30-day complication and readmission rates in comparison with traditional care. In addition, we estimated the extent to which adherence to specific perioperative care elements of the ERP are associated with improved outcomes. MATERIALS AND METHODS This study is reported according to the STROBE criteria for observational studies (available at: www. strobe-statement.org). The study protocol was approved by the institutional review board and conforms to the Canadian Tri-Council Policy Statement of Ethical Conduct. Study design, setting, and participants. In this retrospective cohort study, patients undergoing lung resection for cancer at a single academic center between August 2011 and October 2013 were identified from an operating room database. Adult patients (>18 years old) were included if they underwent an open, scheduled lobectomy for either primary or secondary lung cancer. Exclusion criteria included age <18, benign disease, nonelective operations, pneumonectomy, bronchoplasty, nonanatomic lung resection, and extended resection of adjacent organs (eg, carina, chest wall). Although video-assisted thoracoscopic surgery (VATS) is the treatment of choice in many centers across North America, it was also excluded because only a minority of lobectomies at our institution are performed using this approach and inclusion would have confounded the results of this study. Beginning in September 2012, patients undergoing lung resection were managed according to a standardized pathway protocol; before this date, patients were managed according to surgeon preference. Participants were categorized according to the date of their operation into a prepathway group (PRE) or post-pathway group (POST) and compared in terms of demographics, duration of stay, short-term (30-day) complications, hospital readmissions, and adherence to elements of the pathway. Development and implementation of ERP for lung resection. In an effort to improve patient outcomes, decrease cost, and reduce variability, a multidisciplinary team was established at our institution in 2008 to develop and implement standardized ‘‘enhanced recovery’’ management protocols for high prevalence operative procedures. The team is led by a general surgeon and anesthesiologist and includes expertise in nursing (inpatient and outpatient), physiotherapy, pharmacy, pain service, nutrition, and medical informatics. A full-time care pathway coordinator is
Surgery October 2015
dedicated to the program. The goal of the enhanced recovery program is to facilitate widespread uptake of evidence-based interventions shown to decrease physiologic stress (eg, preoperative carbohydrate, exercise, and afferent neural blockade) and decrease the use of interventions that hamper recovery without improving outcomes (eg, prolonged preoperative fasting, prolonged insertion of urinary drains, and delayed feeding). Improving patient recovery and reducing morbidity decreases the need for hospitalization, reflected in shorter durations of hospital stay. The program has implemented pathways for a variety of procedures, including colorectal, esophageal, and hepatic surgeries. For the establishment of the lung resection pathway, the team also included 3 thoracic surgeons along with specific nursing expertise for thoracic patients. The objective of the ERP was to introduce a standard multimodal approach, including patient education and counselling, opioid-sparing pain control, early and structured mobilization, early feeding and optimization of nutritional status, standardized drain management, and target discharge with written patient goals for each postoperative day (POD). A medical librarian assisted with systematic literature searches to identify evidence-based interventions for the perioperative management of thoracic patients. Elements of the pathway were discussed extensively by various team members and stakeholders, and evaluated in terms of feasibility and adoptability amongst patients and caregivers. Once consensus was reached, the elements of the ERP were used to create physician orders (Table I), customized nursing records for each POD, and patient education material (available at: http:// muhcpatienteducation.ca). Implementation of this pathway did not require additional allied health care workers, such as nurses, physiotherapists, or other individuals involved in preoperative or postoperative care. Intervention. Before their operation, all patients underwent preoperative medical evaluation by an internist, anesthesiologist, and nurse in the preoperative clinic. After the ERP was implemented in September 2012, nurses also provided personalized patient education using an information booklet that reviewed the procedure using simplified terminology and illustrations. This included a review of daily goals for pain control, diet, drain and fluid management, structured daily physical activity, and target discharge date on POD 4. A physiotherapist provided instructions on the use of an incentive spirometer in preparation for the operation, along with general recommendations to increase physical
Madani et al 901
Surgery Volume 158, Number 4
Table I. Daily milestones of the enhanced recovery pathway for lung resection Traditional care (PRE) Preoperative Patient education Intraoperative Analgesia Extubation Postoperative Analgesia Urinary drain
Chest tube
Nutrition
Chest x-ray
Mobilization
Chest rehabilitation Target discharge
Enhanced recovery (POST)
Nonstandardized education given in surgeon’s office
Standardized preoperative education protocol Information booklet with daily goals
Thoracic epidural inserted Based on anesthesiologist preference
Preferred extubation in the operating room or in the postanesthesia care unit
Thoracic epidural stop test performed on the day the last chest tube is removed Nonstandardized management POD 1: drain removed if adequate urine output based on surgeon preference If no urine output after 8 hours of removal, a bladder scan is performed an a urinary retention protocol is followed POD 0: maintained at 20 cmH20 POD 0: maintained at 20 cmH20 suction suction POD 1: remove suction Weaning based on surgeon POD 2: remove chest tube#1 if <300 mL/24 h, preference nonchylous and no air leak POD 3: remove chest tube#2 if <300 mL/24 h, nonchylous and no air leak No nasogastric tube No nasogastric tube Diet advanced progressively based POD 0: clear fluid diet on surgeon preference POD 1: diet as tolerate After either chest tube suction removal or chest After either chest tube suction tube removal removal or chest tube removal No clamp test Clamp test based on surgeon preference Physical activity encouraged by POD 0: up in chair with assistance as tolerated health care provider POD 1: up in chair 3 times per day for all meals + 30–60 minutes each time, ambulate in hallway 2 times per day with assistance POD 2: out of bed for all meals and $8 hours during the day, walking in hallway 17.5–35 meters 3 times per day with assistance POD 3: increase ambulation to 75 meters 3–5 times per day Spirometry 10 times every hour while awake Chest physiotherapy every 4 hours None POD 3 if 1 chest tube POD 4 if 2 chest tubes
POD, Postoperative day.
activity and performance status. When relevant, consultations were made to the appropriate service for comorbidities. If the patient was a smoker, recommendations and referral to a smoking cessation program were made. All open lobectomies were performed by 3 thoracic surgeons with patients under general anesthesia, using a double-lumen endotracheal intubation for lung isolation and a thoracic epidural catheter for postoperative pain control. Before closure, 1 or 2 chest tubes were inserted based on surgeon preference. Patients were extubated in the operating room or on arrival to the
postanesthesia care unit before transfer to the thoracic surgery ward after observation for 2–4 hours and performance of a chest radiograph. Postoperative management also differed significantly between the 2 groups. In the PRE group, orders were highly variable and based on surgeon preference. There was no target discharge date and protocols for feeding, fluid, drain (urine and chest tube), and epidural management were according to surgeon discretion. Epidural catheters were removed typically after the last chest tube was removed. Although the team encouraged physical activity, there was no structured mobilization plan.
902 Madani et al
In the POST group, structured daily plans served to guide patient mobilization, pain control, feeding, weaning of chest tubes and urine catheter, and chest wall exercises with planned discharge on POD 4 if there were no complications (although patients could leave earlier if appropriate). ERP milestones are summarized in Table I. Outcomes and data collection. Data on demographics, preoperative evaluation, postoperative course, achievement of daily milestones, and hospital visits were obtained from both paper and electronic hospital charts. Postoperative complications were defined a priori according to the Seely et al modification of the Dindo–Clavien grading system.5 Pulmonary complications were defined as pneumonia (abnormal radiograph and sputum production with fever, leukocytosis, or both), pleural effusion or pneumothorax requiring drainage, atelectasis with sublobar collapse requiring bronchoscopy, respiratory failure requiring endotracheal intubation, pulmonary embolus, prolonged air leak (>7 days), chylothorax, hemoptysis, aspiration pneumonitis and pulmonary edema requiring diuresis. Primary outcome was hospital duration of stay and secondary outcomes were morbidity, readmission, emergency department visits within 30 days, and adherence to each element of the pathway. Statistical analysis. The PRE and POST groups were compared by univariate analysis using Pearson’s v2 (for categorical variables), Student’s t test (for normally distributed continuous variables), or Wilcoxon’s signed-rank test (for non-normally distributed continuous variables). Data were expressed as n (%), mean (SD), median (interquartile range [IQR]), as appropriate. Multivariate linear and logistic regression was used to estimate the adjusted association between the presence of an ERP and duration of stay, postoperative complications and hospital readmissions. Adherence to 6 individual ERP elements (standardized preoperative education, introduction of oral diet, urinary catheter removal, epidural analgesia catheter removal, chest tube removal, and intravenous fluid discontinuation) were examined for their association with duration of stay using a stepwise logistic regression with backward elimination of variables with a P value >.1. Multiple imputations by chained equation (10 imputed datasets) were used to replace missing data on adherence rates and adjustment variables. Sample size. For the purposes of power calculation, the difference in duration of stay before and after the pathway was used. In the year preceding the study, average duration of stay after lobectomy
Surgery October 2015
was 8.2 ± 6 days with only 19% of patients discharged on or before the target discharge date of POD 4. Assuming duration of stay decreased to 6.5 ± 6 days,17 a minimum sample size of 80 per group was calculated. RESULTS A total of 448 patients underwent elective lung resections between August 2011 and October 2013. Two hundred fourteen patients were excluded for VATS or nonanatomic resections (n = 185), benign disease (n = 7), extended resections (n = 10) or pneumonectomy (n = 12), with 234 meeting the study inclusion criteria. One hundred twentyseven patients were treated before the implementation of the pathway on September 2012 (PRE group) with the subsequent 107 patients receiving care after the pathway was implemented (POST group; Figure). Groups were similar in terms of demographics, baseline pulmonary function, or comorbidities (Table II). Missing data ranged from <1% for age to 55% for forced expiratory volume in 1 second (FEV1). Hospital duration of stay was significantly reduced after implementation of the ERP (6 [IQR, 5–7] vs 7 [6–10] days; P < .01; Table III). This decrease was mostly owing to decreased duration of stay in patients with no complications (5 [IQR, 4–6] vs 6 [5–7] days; P < .01). Compared with the PRE group, the POST group had a higher proportion of patients discharged by the POD 4 target (22% vs 5%; P < .01), with no difference in readmissions, emergency department visits after discharge, or proportion of patients with prolonged duration of stay (>14 days; Table III). The POST group also had a greater adherence to most individual elements of the ERP, with the exception of introduction of solid food (Table IV). Overall short-term (30-day) morbidity was 50% before and 37% after the implementation of the ERP (P = .03), including a significantly lower rate of urinary tract infections (UTI; 12% vs 3%, P < .01; Table V). There was no difference in pulmonary complications (31% vs 25%; P = .38). Despite earlier removal of chest tubes in the POST group compared with the PRE group (4 days [IQR, 3–6] vs 5 [4–7]; P < .01), there were no differences in incidence of pneumothorax (2 [2%] vs 0) or pleural effusion (2 [2%] vs 6 [5%]) requiring tube reinsertion. One patient in the POST group died on POD 3 from sepsis and subsequent respiratory failure and multiorgan failure. There were no signs of aspiration on chest x-132#ray owing to early feeding. Multivariate linear and logistic regression was used to estimate the impact of the ERP on
Surgery Volume 158, Number 4
Madani et al 903
Figure. Study design. ERP, Enhanced recovery pathway; POST, post-pathway group; PRE, pre-pathway group; VATS, video-assisted thoracoscopic surgery.
duration of stay, postoperative complications, and hospital readmissions (Table VI). After adjusting for age, gender, body mass index, and American Society of Anesthesiologists score, the ERP was associated with decreased duration of stay (average 18% shorter duration of stay; P < .01) and overall 30-day complications, with no association with readmissions (associations with duration of stay and readmissions were also adjusted for complications). Associations between adherence to individual ERP elements and duration of stay using stepwise linear regression identified urinary catheter removal on POD 1 (b, 0.23; 95% CI, 0.43 to 0.02) and removal of the last chest tube on or before POD 3 (b, 0.26; 95% CI, 0.44 to 0.08) as independent predictors of duration of stay (Table VII). DISCUSSION ERPs are coordinated, multidisciplinary, evidence-based protocols including detailed management plans for the preoperative, intraoperative, and postoperative phases of patient care. ERPs have been shown to decrease morbidity, hospital duration of stay, and cost of patient care. Nonetheless, few studies describe the impact of an ERP on patients undergoing lung resection.16-20 In this study, implementation of an integrated ERP for patients undergoing open lobectomy decreased duration of stay and short-term morbidity, with no
difference in either readmissions or emergency department visits after discharge. The patient cohorts before and after ERP implementation were clinically comparable with similar demographics, baseline comorbidities, and pulmonary reserve. Furthermore, perioperative care was directed by the same team of thoracic surgeons, nurses, nutritionists, and physiotherapists on the thoracic surgery ward. Compared with the traditional approach dictated by individual surgeon preference, patients treated after implementation of the ERP had significantly shorter duration of stay by 1 day, with a greater proportion of patients discharged by the target date (POD 4). The main driver of decreased duration of stay was earlier discharge of patients with no complications (Table III). This observation suggests that 1 important effect of the pathway is organizational, allowing for more efficient care of patients who have an uncomplicated postoperative course, whereas patients who experience complications have similar management and duration of hospital stay. However, the ERP was also associated with decreased overall morbidity, primarily driven by fewer urinary tract infections and surgical site infections, resulting in more uncomplicated patients eligible for the earlier discharge date. Nonetheless, only 31% of patients without complications were actually discharged by the POD 4 target. Despite earlier discharge, readmissions and emergency
904 Madani et al
Surgery October 2015
Table II. Patient demographics of patients in the pre-pathway (PRE) and post-pathway (POST) groups Characteristic
PRE (n = 127)
Age (y) Gender (male), n (%) BMI (kg/m2) OR time (min) American Society of Anesthesiologists score, n (%) I II III IV Alcohol history, n (%) Smoking history, n (%) Baseline pulmonary function FEV1 (% predicted) FVC (% predicted) DLCO (% predicted) Diabetes mellitus, n (%) Hypertension, n (%) Thyroid disorder, n (%) Cerebrovascular accident, n (%) Coronary artery disease, n (%) Dyslipidemia, n (%) Chronic obstructive pulmonary disease, n (%) Asthma, n (%) Anemia, n (%) Other cancer <5 years, n (%)
64 57 27 153 5 80 42 0 15 69 92 118 78 20 66 18 4 35 43 33 17 14 15
POST (n = 107)
(11) (45) (5) (56) (4) (63) (33) (0) (12%) (54) (22) (19) (20) (16) (52) (14) (3) (28) (34) (26) (13) (11) (12)
P value
67 65 27 149
(10) (61) (6) (50)
.08 .79 .65 .57
1 64 39 0 18 65
(1) (60) (36) (0) (17) (61)
.07 .65 .72 > 0.99 .37 .25
87 100 80 14 50 12 3 23 47 32 13 18 22
(24) (21) (28) (13) (47) (11) (3) (21) (44) (30) (12) (17) (20)
.27 .37 .70 .58 .43 .20 .76 .07 .09 .89 .80 .45 .13
Data are expressed as mean (standard deviation) or n (%). Missing data: age = 1, body mass index = 24, OR time = 16, American Society of Anesthesiologists score = 4, FEV1 = 130, FVC = 85, DLCO = 94. BMI, Body mass index; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; DLCO, diffusion capacity of the lung for carbon monoxide.
Table III. Duration of stay, readmissions and emergency visits of patients in the pre-pathway (PRE) and post-pathway (POST) groups Characteristic Duration of stay (d) Duration of stay by complication severity (d) None Minor (Clavien I–II) Major (Clavien III–IV) Discharge by target date* Overall No complications Any complications Prolonged duration of stay (>14 d) Readmission Emergency department visit
PRE (n = 127)
POST (n = 107)
P value
7 [6–10]
6 [5–7]
<.01
6 [5–7] 8 [6–12] 16 [10–20]
5 [4–6] 7 [6–10] 11 [7–17]
<.01 .58 .72
(22%) (31%; n = 67) (8%; n = 40) (7%) (6%) (5%)
<.01 <.01 <.01 .20 .20 .43
6 5 1 16 6 4
(5%) (8%; n = 63) (2%; n = 64) (13%) (5%) (3%)
24 21 3 8 7 5
*Target date = postoperative day 4. Data expressed as median [interquartile range] or n (%). One patient died in hospital and was excluded from the analysis. There were no missing data for duration of stay, complications, readmissions and emergency department visits.
visits after discharge remained stable. Therefore, the ERP likely improves flow and organization, especially for routine patients, without compromising care.
The introduction of the pathway was an independent predictor of postoperative complications. The ERP was developed specifically after synthesizing the available evidence with the help of a
Madani et al 905
Surgery Volume 158, Number 4
Table IV. Adherence to individual elements of the enhanced-recovery pathway protocol in the pre-pathway (PRE) and post-pathway (POST) groups Adherence rate
Pathway elements (target POD) Foley removal (POD 1) Solid diet (POD 1) Intravenous fluid discontinuation (POD 1) Ambulation (POD 1) Last chest tube removal (POD 3) Thoracic epidural removal (POD 3)
Median POD
PRE (n = 127), n (%)
POST (n = 107), n (%)
PRE (n = 127), median [interquartile range]
POST (n = 107), median [interquartile range]
62 33 20 39 65 75
88 36 52 63 75 83
2 3 3 2 5 5
1 3 2 1 4 4
(49)* (26) (16)* (31)* (51)* (59)*
(82)* (34) (49)* (59)* (70)* (78)*
[1–2]* [1–4] [2–5]* [2–3]* [4–6]* [4–6]*
[1–1]* [1–4] [1–3]* [1–2]* [3–5]* [3–5]*
*P < .05 between PRE and POST groups. Results shown as proportion (%) of patients meeting the milestone and median postoperative day (POD). Missing data: solid diet = 15, Foley removal = 2, thoracic epidural removal = 4, last chest tube removal = 5 and intravenous fluid discontinuation = 23, ambulation = 64.
Table V. Short-term (30-day) morbidity of patients in the pre-pathway (PRE) and post-pathway (POST) groups, classified by highest grade complication per patient Characteristic Any complications Any pulmonary complications* Minor complications (Clavien I–II) Atelectasis Pneumonia Surgical site infection Urinary tract infection Cardiac tachyarrhythmia Prolonged air leak (>7 d) Postoperative delirium Urinary retention Pulmonary edema Miscellaneousy Major complications (Clavien III-IV) Pleural effusion requiring drainage Pneumothorax requiring drainage Respiratory failure Miscellaneousz Mortality (Clavien V)
PRE (n = 127), n (%) 64 39 51 4 21 7 15 8 13 7 11 10 21 15 6 0 3 10 0
(50) (31) (40) (3) (17) (6) (12) (6) (10) (6) (9) (8) (17) (12) (5) (0) (2) (8) (0)
POST (n = 107), n (%) 40 27 34 2 15 1 3 6 7 7 9 12 10 9 2 2 4 7 1
(37) (25) (32) (2) (14) (1) (3) (6) (7) (7) (8) (11) (9) (8) (2) (2) (4) (7) (1)
P value .03 .38 .22 .69 .72 .07 <.01 > 0.99 .36 .79 > 0.99 .50 .12 .39 .30 .21 .71 .80 .46
*Includes pneumonia, respiratory failure, pleural effusion or pneumothorax requiring drainage, atelectasis, respiratory failure, pulmonary embolus, prolonged air leak, chylothorax, hemoptysis, aspiration pneumonitis and pulmonary edema. yThree or fewer people per group for chronic obstructive pulmonary disease exacerbation, deep vein thrombosis, pulmonary embolus, bradyarrhythmia, acute renal failure, chylothorax, syndrome of inappropriate antidiuretic hormone, ileus, Clostridium difficile colitis, pancreatitis, and aspiration pneumonitis. zThree or fewer people per group for sepsis, heart failure, myocardial infarction, hemoptysis, postoperative hemorrhage, deep wound dehiscence and gastrointestinal bleed.
medical librarian to implement interventions designed to attenuate the surgical stress response, such as opioid-sparing anesthesia, and avoidance of surgical traditions that may delay recovery, such as prolonged use of drains, delayed feeding, and delayed mobilization. Notably, there was a significant, 4-fold decrease in urinary tract infections after the pathway was implemented. This was not surprising, given that indwelling urinary catheters
in postoperative patients are associated with nosocomial infections21 and that urinary catheter removal on POD 1 can be accomplished safely with use of a bladder scan–based urinary retention protocol, even when a thoracic epidural remains in place.22 However, the existence of this evidence alone was not enough to change behavior; having standardized orders allowed for better adherence.
906 Madani et al
Surgery October 2015
Table VI. Impact of an enhanced recovery pathway implementation on duration of stay and complication and readmission rates Characteristic
Estimate*
95% CI
P value
Duration of stayy Complications Readmissions
b 0.18* OR 0.46 OR 1.59
0.29 to 0.06 0.26 to 0.81 0.49 to 5.18
<.01 <.01 .44
*Analysis adjusted for age, gender, body mass index, and American Society of Anesthesiologists score. Analyses of duration of stay and readmissions were also adjusted for complications. yDuration of stay was log transformed and should be interpreted as the percentage change in duration of stay between groups. One patient died in hospital and was excluded from the analysis. Imputations: age = 1, body mass index = 24, American Society of Anesthesiologists score = 4.
Table VII. Linear regression model estimating the impact of complications and adherence to individual elements of the enhanced recovery pathway on duration of stay* Estimate (b)*
Any complicationy Foley removal (POD 1)z Last chest tube removal (POD 3)z Solid diet (POD 1)z Thoracic epidural removal (POD 3)z
0.38 0.23
0.22 to 0.55 <.01 0.43 to 0.02 .03
0.26
0.44 to 0.08 <.01
0.19 0.20
95% CI
P value
Characteristic
0.39 to 0.01 0.41 to 0.01
.06 .06
Age, gender, body mass index, American Society of Anesthesiologists score, preoperative education, and intravenous fluid discontinuation were removed from the model (P > .1). One patient died in hospital and was excluded from the analysis. Imputations: age = 1, gender = 0, body mass index = 13, American Society of Anesthesiologists score = 4, complications = 0, standardized preoperative education = 0, solid diet = 8, Foley removal = 2, thoracic epidural removal = 4, last chest tube removal = 5 and intravenous fluid discontinuation = 5. POD, Postoperative day. *Duration of stay was log transformed and should be interpreted as the percentage change in length of stay between patients without and with complications (y), and patients not adhering and adhering to the enhanced recovery pathway elements (z).
Another critical element of the pathway was the removal of chest tubes using a threshold of <300 mL of drainage in 24 hours for removal as long as there was no evidence of ongoing air leak or chylothorax, which resulted in earlier tube removal by 1 day and decreased the duration of stay. The choice to use a threshold of 300 mL was relatively arbitrary in that there are few data to support 1 threshold over another. Also, there were differences in opinion between the 3 surgeons and this number was based on compromise. It is
not uncommon for thoracic patients to have their hospital discharge delayed because drainage of >200–250 mL in 24 hours is considered too high and there is an associated fear that early removal with high drainage volumes may subject patients to unnecessary morbidity by requiring a tube to be reinserted for symptomatic effusions. However, readmissions and interventions to manage recurrent effusions using thresholds as high as 500 mL in 24 hours are very uncommon.23-25 In fact, most patients who experience complications requiring tube reinsertion have chest tube outputs of <200 mL in 24 hours before the removal of the initial chest tubes.23 This suggests that drainage volume is unlikely to play a significant role and the practice of leaving chest tubes until very low drainage volumes have be achieved is not supported in the literature. Our findings support this notion: despite the earlier removal of chest tubes and resultant decrease in duration of stay, there was no change in pulmonary complications requiring an intervention or overall pulmonary complications. However, the study was underpowered to detect to differences of this magnitude. After the ERP was implemented, adherence was greater for most pathway milestones, with the exception of introduction of solid food (Table IV). Analysis of individual surgeons, however, revealed that this was owing to 1 surgeon’s preference for introduction of solid food on POD 3, regardless of the pathway protocol, because surgeons have the ability to ‘‘override’’ the pathway should they feel it is appropriate in a given context. Irrespective, this diet milestone was not a predictor of duration of stay. In fact, there is limited evidence suggesting the impact of each individual intervention of the pathway in isolation, or which element was most responsible for improving patient outcomes. In our study, we found that urinary catheter removal on POD 1 and removal of the last chest tube on or before POD 3 were independent predictors of decreased duration of stay (Table VII). This study estimates the impact of an ERP for lung resection that comprehensively addresses patient care from the preoperative to postoperative period with interventions for patient education, feeding, pain control, mobilization, and tube/ catheter removal. The ERP at our institution was multidisciplinary, achieved through active involvement and consensus by all stakeholders. After implementation, weekly meetings of the surgical recovery team allow the team to make ongoing improvements to the ERP, based on feedback from
Madani et al 907
Surgery Volume 158, Number 4
the managing team and new evidence from the literature. The results of this audit study identified current successes as well as challenges and will drive future refinements in the pathway protocol related to chest tube drainage, diet and mobilization. An important limitation of this study is the exclusion of VATS lobectomies as a potential confounder, despite its increasing popularity throughout many centers in North America. However, the major strength of an ERP seems to be the improvement in flow and organization (eg, early drain removal), which would benefit patients regardless of the operative approach. For instance, evidence from colorectal surgery suggests that implementation of an ERP improves outcomes for both open and minimally invasive surgery.14,22 Therefore, it is likely that an ERP would also benefit patients undergoing VATS lobectomy. Another limitation is the before–after study design and the retrospective nature, which could have introduced unknown confounders that may have biased our analysis and also precluded investigation of reasons for poor adherence with certain pathway elements. Although a prospective trial with a contemporary control arm could have minimized this effect, it would have been difficult to perform given that the treating teams would have to use different protocols for different patients on the same ward. Nonetheless, we limited selection bias by including consecutive patients throughout the duration of the study. Also, the results of this study are based on 2 cohorts that were clinically similar and multivariate analyses that adjusted for potential confounders. Last, target hospital discharge was only achieved in a minority of patients, even among those who had an uncomplicated hospital course and managed to meet all discharge criteria (median POD 5). This difference of 1 day is likely owing to the discrepancy between ‘‘time to readiness for discharge’’ and actual hospital duration of stay, the latter of which tends to be influence by various personal, societal, and organizational factors that are not related to surgical recovery, including hospital culture, patient preference, and availability of caregiver at home.26 Several studies have characterized this phenomenon and estimate a difference of 1 day,26,27 which seems to correspond with the difference between our target discharge POD 4 and the median POD 5 discharge for uncomplicated patients. Additional prospective studies are needed to elucidate the reasons behind this discrepancy. In conclusion, implementation of a multimodal ERP for open lobectomy is associated with decreased duration of stay and short-term
complications, with no difference in readmissions, emergency department visits after discharge, or complications related to early chest tube removal. Early removal of the chest tube and urinary catheter were independent predictors of decreased duration of hospital stay. The authors acknowledge the contribution of all members of the Surgical Recovery (SURE) team for the development and implementation of an enhancedrecovery pathway for lung resections. A.M. is supported by the Quebec Health Science Research Scholarship (FRQ-S) and the McGill Surgeon-Scientist Program. The Steinberg-Bernstein Centre for Minimally Invasive Surgery and Innovation is supported by an unrestricted educational grant from Covidien.
REFERENCES 1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011;61:69-90. 2. McKenna RJ Jr, Houck W, Fuller CB. Video-assisted thoracic surgery lobectomy: experience with 1,100 cases. Ann Thorac Surg 2006;81:421-5. 3. Chen FF, Zhang D, Wang YL, Xiong B. Video-assisted thoracoscopic surgery lobectomy versus open lobectomy in patients with clinical stage non-small cell lung cancer: a meta-analysis. Eur J Surg Oncol 2013; 39:957-63. 4. Phillips JD, Merkow RP, Sherman KL, DeCamp MM, Bentrem DJ, Bilimoria KY. Factors affecting selection of operative approach and subsequent short-term outcomes after anatomic resection for lung cancer. J Am Coll Surg 2012;215:206-15. 5. Seely AJ, Ivanovic J, Threader J, Al-Hussaini A, Al-Shehab D, Ramsay T, et al. Systematic classification of morbidity and mortality after thoracic surgery. Ann Thorac Surg 2010;90: 936-42. 6. Ivanovic J, Seely AJ, Anstee C, Villeneuve PJ, Gilbert S, Maziak DE, et al. Measuring surgical quality: comparison of postoperative adverse events with the American College of Surgeons NSQIP and the Thoracic Morbidity and Mortality classification system. J Am Coll Surg 2014;218:1024-31. 7. Lacin T, Swanson S. Current costs of video-assisted thoracic surgery (VATS) lobectomy. J Thorac Dis 2013;5(Suppl 3): S190-3. 8. Andalib A, Ramana-Kumar AV, Bartlett G, Franco EL, Ferri LE. Influence of postoperative infectious complications on long-term survival of lung cancer patients: a population-based cohort study. J Thorac Oncol 2013;8: 554-61. 9. Kehlet H, Wilmore DW. Evidence-based surgical care and the evolution of fast-track surgery. Ann Surg 2008;248: 189-98. 10. Li C, Ferri LE, Mulder DS, Ncuti A, Neville A, Lee L, et al. An enhanced recovery pathway decreases duration of stay after esophagectomy. Surgery 2012;152:606-14. 11. Lee L, Li C, Landry T, Latimer E, Carli F, Fried GM, et al. A systematic review of economic evaluations of enhanced recovery pathways for colorectal surgery. Ann Surg 2014; 259:670-6.
908 Madani et al
12. Walter CJ, Collin J, Dumville JC, Drew PJ, Monson JR. Enhanced recovery in colorectal resections: a systematic review and meta-analysis. Colorectal Dis 2009;11:344-53. 13. Gustafsson UO, Hausel J, Thorell A, Ljungqvist O, Soop M, Nygren J, et al. Adherence to the enhanced recovery after surgery protocol and outcomes after colorectal cancer surgery. Arch Surg 2011;146:571-7. 14. Larson DW, Lovely JK, Cima RR, Dozois EJ, Chua H, Wolff BG, et al. Outcomes after implementation of a multimodal standard care pathway for laparoscopic colorectal surgery. Br J Surg 2014;101:1023-30. 15. Nicholson A, Lowe MC, Parker J, Lewis SR, Alderson P, Smith AF. Systematic review and meta-analysis of enhanced recovery programmes in surgical patients. Br J Surg 2014; 101:172-88. 16. Muehling BM, Halter GL, Schelzig H, Meierhenrich R, Steffen P, Sunder-Plassmann L, et al. Reduction of postoperative pulmonary complications after lung surgery using a fast track clinical pathway. Eur J Cardiothorac Surg 2008; 34:174-80. 17. Numan RC, Klomp HM, Li W, Buitelaar DR, Burgers JA, Van Sandick JW, et al. A clinical audit in a multidisciplinary care path for thoracic surgery: an instrument for continuous quality improvement. Lung Cancer 2012;78:270-5. 18. Wright CD, Wain JC, Grillo HC, Moncure AC, Macaluso SM, Mathisen DJ. Pulmonary lobectomy patient care pathway: a model to control cost and maintain quality. Ann Thorac Surg 1997;64:299-302. 19. Salati M, Brunelli A, Xiume F, Refai M, Pompili C, Sabbatini A. Does fast-tracking increase the readmission rate after pulmonary resection? A case-matched study. Eur J Cardiothorac Surg 2012;41:1083-7. 20. Maruyama R, Miyake T, Kojo M, Aoki Y, Suemitsu R, Okamoto T, et al. Establishment of a clinical pathway as an effective tool to reduce hospitalization and charges after video-assisted thoracoscopic pulmonary resection. Jpn J Thorac Cardiovasc Surg 2006;54:387-90. 21. Wald HL, Ma A, Bratzler DW, Kramer AM. Indwelling urinary catheter use in the postoperative period: analysis of the national surgical infection prevention project data. Arch Surg 2008;143:551-7. 22. Carli F, Charlebois P, Baldini G, Cachero O, Stein B. An integrated multidisciplinary approach to implementation of a fast-track program for laparoscopic colorectal surgery. Can J Anaesth 2009;56:837-42. 23. Cerfolio RJ, Bryant AS. Results of a prospective algorithm to remove chest tubes after pulmonary resection with high output. J Thorac Cardiovasc Surg 2008;135: 269-73. 24. Bjerregaard LS, Jensen K, Petersen RH, Hansen HJ. Early chest tube removal after video-assisted thoracic surgery lobectomy with serous fluid production up to 500 ml/day. Eur J Cardiothorac Surg 2014;45:241-6. 25. Zhang Y, Li H, Hu B, Li T, Miao JB, You B, et al. A prospective randomized single-blind control study of volume threshold for chest tube removal following lobectomy. World J Surg 2014;38:60-7. 26. Fiore JF Jr, Faragher IG, Bialocerkowski A, Browning L, Denehy L. Time to readiness for discharge is a valid and reliable measure of short-term recovery after colorectal surgery. World J Surg 2013;37:2927-34. 27. Maessen JM, Dejong CH, Kessels AG, von Meyenfeldt MF. Enhanced Recovery After Surgery Group. Length of stay: an inappropriate readout of the success of enhanced recovery programs. World J Surg 2008;32:971-5.
Surgery October 2015
DISCUSSION Dr L. Michael Brunt (St. Louis, MO): The development of care pathways to enhanced recovery after surgery is gaining a lot of interest in the surgical community. The McGill group has made several recent contributions to this area. With the present study, they show modest improvements following implementation of an enhanced recovery pathway after lung resection compared with a control group from before the implementation. I have a few questions for you. Can you comment on the resources and costs of those resources necessary to implement this pathway, such as nursing, physiotherapy, preop counseling, et cetera? Did your institution help support the cost of the program in any way, recognizing the potential cost savings obtained by shorter length of stay, fewer complications, et cetera? Smoking cessation was a part of the preop counseling in the enhanced recovery path group. Did you observe any differences in smoking cessation rates between the 2 groups? Could that have been a factor in the wound infection rate (6% vs 1% post pathway implementation, which approached statistical significance)? Your hospital length of stay did not reach your target time point of four days, which was obtained by only 22% in the enhanced recovery group. The difference between the 2 groups was only 1 day. Why do you think you did not see a greater difference? Did you look at impact post discharge, such as return to full activity, employment, exercise, or other patient-centered outcomes that may have been impacted by this approach? Have you made any changes to your protocol since the study to improve and streamline the periop process? You mentioned VATS, and it is discussed in your paper, but I was wondering if you could comment on VATS and whether you think you will see similar improvements as you did with the open procedures. Dr Amin Madani (Montreal, Quebec, Canada): Our institution was involved in helping us fund this program. First of all, the booklets cost approximately $13,000 to design. This cost was funded through a patient education office at the McGill University Health Center. There was also opportunity costs, setting up the infrastructure and hiring a pathway coordinator, a nurse practitioner, whose full-time job is to run this pathway. All of these costs amount to approximately $120,000 annually.
Surgery Volume 158, Number 4
When you consider that in our institution we have approximately 1,200 patients who undergo the pathway annually, it comes out to approximately $100 per patient. We did not have to hire additional staff (allied health care workers, nurses and physiotherapists) for implementation of the pathway. It was the same staff that worked with patients before the pathway. We had very good buy-in from the allied health care workers. We got them involved very early on in setting up the pathway. They shared their concerns. Their feedback was very valuable and was implemented into the pathway. With respect to the trend of decreased surgical site infections, we can only speculate as to why that is the case. I think probably decreased length of stay played a role. Perhaps it was the fact that they had a decreased length-of-stay, less exposure to hospital pathogens and nosocomial infections. You mentioned cigarette smoking as a possible factor. This is a possibility, but unfortunately we did not look at smoking cessation rates before and after pathway implementation. Since smoking cessation was systematically implemented in the preoperative clinic for these patients, there is a really good chance that that could have played a role in decreasing rates of surgical site infections. Your next question makes an extremely important point: the target discharge was postop day 4, but only a quarter of the patients were actually discharged by postop day 4. The first explanation we could think of when we analyzed this data was that complications were responsible for increasing length-of-stay beyond the target of postop day 4. However, among patients who had no complications, two-thirds of them still left on postop day 5 or later, even though they managed to achieve their milestones by postop day 4. The reason behind this discrepancy, I think, is because our outcome was length of stay. There is a difference between actual discharge date and being medically fit for discharge. Looking at the literature, this discrepancy is usually about 1 day, which may explain why the median postoperative discharge day was postop day 5. Was it the case caregivers were not ready, it was late in the day, or it was a weekend? It is hard to tell from a retrospective study. Your question about post-discharge return to work is currently being evaluated in a subset of these patients who were evaluated prospectively. These patients are a part of another study in which we are looking at the cost-effectiveness analysis of the implementation of our pathway. These data have not been analyzed yet. Looking at previous
Madani et al 909
data in colorectal patients, we have seen that caregiver burden has improved and patients experience faster return to work. Given those results, we expect a similar trend. Dr Richard Anderson (Peoria, IL): Intraoperatively, did you do anything to decrease your air leaks in the postoperative period, such as any fibrin glue or Progel? My other comment is that I think the reason why you did not reach your mark of 4 days is your goals were not high enough. I put 1 chest tube in and remove it as soon as there is no air leak and <300 mL of drainage, regardless of what postop day it is. I think if you go back to what you are doing right now and just set your goals a little bit lower, I think you will probably achieve them. Dr Amin Madani: We do not systematically use any adjuncts such as fibrin glue to decrease air leak rates. Dr Gerald Larson (Louisville, KY): I would like to ask first, was this a surgeon-driven project or a hospital administration-driven project? Did the administrator come to you and say, can we shorten our hospital stay? If it was surgeon driven, I think that’s more laudatory. Second, removal of the Foley catheter and the chest tube had predictive value for length of stay. When the chest tube should be removed is a very personal decision for some surgeons. What were your new guidelines for chest tube removal? How much pushback did you have? How did you manage that issue? Dr Amin Madani: First, with respect to the impetus behind the pathway, I think it is probably both. There is a lot of pressure to decrease costs and length of stay at our institution, but I think the impetus behind it is largely driven by the surgeons and a multidisciplinary group of people working in a culture of trying to improve efficiency and quality of care. The surgical recovery group, so-called SuRe Group at our institution, is led by a surgeon champion, Dr Liane Feldman, and an anesthesiologist champion. It is a group of like-minded individuals coming together to improve efficiency. I think that is really a key message. You brought up a very controversial point with respect to the chest tube output. We chose 300 mL in 24 hours. As long as there was <300 mL, no chylous drainage, and no air leak, chest tubes were allowed to be removed. Why did we choose 300 mL? It was a very arbitrary number. In fact, it was a compromise amongst the 3 thoracic surgeons in the group. There is 1 surgeon who is usually
910 Madani et al
more conservative than the other 2, so we had to compromise. The existing literature on chest tube removal thresholds suggests that the practice of keeping a chest tube for prolonged periods until low drainage amounts have reached is not unsupported by any strong evidence. When I was a junior resident, it was ingrained in my head that you cannot take a chest tube out until there is <200 mL. We are seeing increasing studies showing you can go up to thresholds as high as 500 mL in 24 hours and the rate of tube reinsertion does not change. We have looked at our results and have actually increased our threshold to 450 mL in 24 hours. Dr Jai Prasad (Detroit, MI): I am talking from my experience in using enhanced recovery for colorectal surgery patients. The 2 things that we found very effective were duration of NPO before surgery and the use of intraoperative intravenous fluids. Were there any specific things that you changed as part of your enhanced recovery program in the preoperative or intraoperative period? I know you made a reference to smoking cessation and air leaks. Were there any other factors that were incorporated as part of the enhanced recovery protocol? Dr Amin Madani: Yes, the preoperative visits were very protocolized. We had the patient meet with a preoperative nurse. They go through an extensive amount of counseling. We teach them all their expected goals on every postoperative day. We really try to make an effort to engage the
Surgery October 2015
patient in their own care. They meet with the physiotherapist. They learn about how to do spirometry. We talked about a smoking cessation program. It is a very systematic approach. Intraoperatively, you talked about limiting fluids. In thoracic patients, although there is no specific number that we use, we usually try to limit fluids to avoid complications such as pulmonary edema. Dr James Madura (Phoenix, AZ): I think you could just as easily say that evidence-based medicine and patient expectations reduce duration of stay and complications, because I think that is what these enhanced recovery programs really are. Lung resections may be a little different than GI surgery, but what I have noticed is that there are populations of patients who undergo colon or small bowel resections in whom we can predict will not meet the enhanced recovery program goals. For example, an 80-year-old, diabetic lady who underwent a large cancer operation. Have you identified any of these factors as predictors of not meeting your goals in the enhanced recovery program? Dr Amin Madani: That’s a very good point. I agree that there are some patients who you cannot really fast track. Although we have these protocolized pathways, the surgeon still reserves the capacity and judgment to not adhere to the pathway if they feel it is appropriate. We did not look into specific patient factors that would predict adherence to the pathway milestones. Perhaps we can do that in the future.