- Email: [email protected]
Patient-Reported Outcomes Accurately Measure the Value of an Enhanced Recovery Program in Liver Surgery
Patient-Reported Outcomes Accurately Measure the Value of an Enhanced Recovery Program in Liver Surgery Ryan W Day, MD, Charles S Cleeland, PhD, Xin S Wang, MD, MPH, Sharon Fielder, John Calhoun, MS, Claudius Conrad, MD, PhD, Jean-Nicolas Vauthey, MD, FACS, Vijaya Gottumukkala, MD, Thomas A Aloia, MD, FACS
APN,
Enhanced recovery (ER) pathways have become increasingly integrated into surgical practice. Studies that compare ER and traditional pathways often focus on outcomes confined to inpatient hospitalization and rarely assess a patient’s functional recovery. The aim of this study was to compare functional outcomes for patients treated on an Enhanced Recovery in Liver Surgery (ERLS) pathway vs a traditional pathway. STUDY DESIGN: One hundred and eighteen hepatectomy patients rated symptom severity and life interference using the validated MD Anderson Symptom Inventory preoperatively and postoperatively at every outpatient visit until 31 days after surgery. The ERLS protocol included patient education, narcotic-sparing anesthesia and analgesia, diet advancement, restrictive fluid administration, early ambulation, and avoidance of drains and tubes. RESULTS: Seventy-five ERLS pathway patients were clinically comparable with 43 patients simultaneously treated on a traditional pathway. The ERLS patients reported lower immediate postoperative pain scores and experienced fewer complications and decreased length of stay. As measured by symptom burden on life interference, ERLS patients were more likely to return to baseline functional status in a shorter time interval. The only independent predictor of faster return to baseline interference levels was treatment on an ERLS pathway (p ¼ 0.021; odds ratio ¼ 2.62). In addition, ERLS pathway patients were more likely to return to intended oncologic therapy (95% vs 87%) at a shorter time interval compared to patients on the traditional pathway (44.7 vs 60.2 days). CONCLUSIONS: In oncologic liver surgery, enhanced recovery’s primary mechanism of action is reduction in life interference by postoperative surgical symptoms, allowing patients to return sooner to normal function and adjuvant cancer therapies. (J Am Coll Surg 2015;221:1023e1030. 2015 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved)
BACKGROUND:
Disclosure Information: Nothing to disclose. Support: This research was supported in part by National Institutes of Health Grant, CA016672. Drs Gottumukkala and Aloia contributed equally to this work. Poster abstract presented at the 123rd Annual Western Surgical Association Meeting, Napa Valley, CA, November 2015. Received August 10, 2015; Revised September 9, 2015; Accepted September 9, 2015. From the Department of Surgical Oncology (Day, Fielder, Conrad, Vauthey, Aloia), Department of Symptom Research (Cleeland, Wang), Institute for Cancer Care Innovation (Calhoun), and Department of Anesthesiology (Gottumukkala), The University of Texas MD Anderson Cancer Center, Houston, TX. Correspondence address: Thomas A Aloia, MD, FACS, Department of Surgical Oncology, The University of Texas MD Anderson Cancer Center, 1400 Herman Pressler, Unit 1484, Houston, TX 77030. email: taaloia@ mdanderson.org
ª 2015 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved.
The ultimate goal of any surgery is to return the patient to at least their baseline functional status, if not an improved functional status compared with their preoperative state, as rapidly as possible and with the least amount of intercurrent disability. Although fast-track surgical protocols have been reported for decades,1,2 it is only recently that the enhanced recovery (ER) movement in perioperative care has significantly penetrated North American surgical practice. Multiple recently published meta-analyses have clearly demonstrated that patients are benefiting from these changes in philosophy and practice.3-5 Most commonly, the included studies focused on outcomes confined to inpatient hospitalization using concrete primary end points, including early return of bowel function, lower complication rates, and/or shorter length of inpatient stay.6-8
1023
http://dx.doi.org/10.1016/j.jamcollsurg.2015.09.011 ISSN 1072-7515/15
1024
Day et al
Patient-Centered Value of Enhanced Recovery
Abbreviations and Acronyms
ER ERLS GI MDASI-GI PRO RIOT
¼ ¼ ¼ ¼
enhanced recovery Enhanced Recovery in Liver Surgery gastrointestinal MD Anderson Symptom Inventory, gastrointestinal version ¼ patient-reported outcomes ¼ return to intended oncologic therapy
In addition, many of the published studies that have examined the impact of ER protocols on short-term outcomes have focused on patients undergoing nononcologic procedures.9-11 When surgical oncology patients have been included, they have tended to have early-stage disease amenable to minimally invasive surgical approaches.12,13 For the majority of oncologic operations, however, postoperative recovery carries the additional demand of returning the patient to adjuvant oncologic therapies. Failure to return to intended oncologic therapy (RIOT)14 after cancer surgery due to complications and lingering poor performance status is strongly associated with worse oncologic outcomes, including shortened overall survival.15,16 Therefore, the end point of length of stay does not adequately measure the oncologic value of surgical recovery to patients with cancer. In surgical practice, and particularly in the field of surgical oncology, tools for the measurement of functional recovery are lacking.17 To address this knowledge gap, this study was designed to assess the ability of an ER program to deliver rapid functional recovery after surgical oncology procedures. To measure the recovery process, the study used a validated patient-reported outcomes (PRO) assessment tool to compare the quality of recovery between an Enhanced Recovery in Liver Surgery (ERLS) pathway and a traditional recovery pathway.
METHODS After study approval by the University of Texas MD Anderson Cancer Center IRB (protocol PA14-1079), hepatectomy patient data entered into a prospectively maintained hepatobiliary surgery database were assessed. The PRO tool used in this study was the gastrointestinal version of the MD Anderson Symptom Inventory (MDASI-GI).18 The MDASI-GI is a validated instrument in cancer patients18 composed of 24 questions that the patient evaluates on Likert scales (0 to 10) (Appendix). The tool has 3 sections, including 13 core symptom questions, which are common to all formats of the MDASI. The GI symptom-specific module comprises 5 questions unique to the patients with GI cancer (eg, constipation,
J Am Coll Surg
diarrhea/watery stools via stoma, swallowing, change in taste, feeling bloated). Lastly, the interferences portion is composed of 6 questions that assess the impact that symptoms are having on the patient’s function and well being (eg, general activity, mood, work including housework, relations with other people, walking, and enjoyment of life). A score of 0 for an item signifies that the patient was not experiencing the symptom of interest or that they were fully functional without life interference from symptoms, respectively. A score of 10 indicates that a symptom was graded as “the worst possible experience” or that interference was graded as “completely interfering” with daily functioning, respectively. The symptom inventory was administered to patients undergoing hepatectomy for any diagnosis between September 2013 and January 2015. Initially baseline data were collected on the last preoperative visit before the operation (typically 1 to 3 days before surgery). Postoperatively, as an inpatient, the survey was administered on postoperative days 1, 3, and 5. In addition, midterm recovery was measured by collecting MDASI-GI at each postsurgical outpatient visit until 31 days had passed since the operation (typically postoperative weeks 1 and 4). All survey data were used for time to recovery and event analyses. The last survey administered within the 31-day postsurgical time period was used for quantitative binary statistical analyses. Demographic, operative, hospitalization, and complication data were obtained from the hepatobiliary surgical database and supplemented by retrospective review of the electronic medical record. For each patient, the preoperative and postoperative MDASI-GI questionnaires were compared across several domains. First, the magnitude and direction of score change from preoperative survey to postoperative survey for the total score and for each component score of the MDASI-GI (core, GI, and interference) were calculated. Individual MDASI item, component scores, and total score were then converted to the categorical value of “returned to baseline,” defined as a postoperative score that was no more than 2 points higher than the preoperative score. Demographic and clinical information included age, sex, preoperative systemic therapy, and American Society of Anesthesiologists score. Operative data included type of operation, use of minimally invasive vs open approach, use of epidural analgesia, length of operation, magnitude of liver resection, and use of the ERLS pathway. There were 2 patients who were initially in the minimally invasive group that required conversion to an open operation due to dense adhesions; for the purpose of this analysis, these were considered open and not minimally invasive procedures. Major hepatectomy was defined as nonanatomic resection
Vol. 221, No. 6, December 2015
Day et al
of 4 or more segments,19 right hepatectomy, left hepatectomy, extended right hepatectomy, or extended left hepatectomy as defined by The Brisbane 2000 terminology.20 Major complications were defined as any adverse event requiring reintervention or ICU admission. Outcomes assessed included length of stay, patient-reported pain scores on arrival to the post-anesthesia care unit (anesthesia goal for initial pain control was a patient-reported pain score of <4 on a 0 to 10 scale), return to baseline MDASI-GI composite scores, RIOT rate, and time to RIOT. The ERLS pathway was composed of preoperative education regarding the principles of ER and expected processes, in addition to routine liver surgery education presented to traditional pathway patients. Patients on the ERLS pathway received multimodality pain control. Preoperatively, this cohort received 3 oral medications: tramadol extended release, pregabalin, and celecoxib. Intraoperatively, a narcotic limiting perioperative inflammatory modulation and immunoprotective anesthetic technique was used. Postoperatively, patients were reinitiated on the 3-drug non-narcotic oral protocol. In addition, for open procedures, patients received either epidural catheters or IV dilaudid patientcontrolled analgesia on an as-needed basis. For minimally invasive procedures, epidural catheters were not used, and specimen extraction incisions and port sites were infiltrated with long-acting local anesthetic (liposomal bupivacaine). Nasogastric tubes were not used. The ERLS patients were placed on a clear liquid diet on the same day as the operation. Intravenous fluids were restricted and all IV fluids were discontinued after 600 mL po intake, followed by rapid diet advancement. In addition to continuing the 3 oral drug, narcotic-sparing regimen, postoperative breakthrough pain was managed with a defined stair step as needed algorithm (acetaminophen/tramadol/low-dose hydromorphone) (Table 1). To regulate compliance, surgeons entered the ERLS protocol in a staggered fashion. Once entered, all hepatectomy patients were eligible for ERLS with no restrictions based on case type, surgical approach, or age. Compliance audits determined near complete (>90%) compliance with all phases of the ER protocol, save for patient allergy, comorbidity, or complication, where safety indicated a departure from protocol. Statistics All univariate statistical analyses were performed using the chi-square test or Fisher’s exact test for categorical variables and the Mann-Whitney U test for continuous variables. Continuous variables are reported as median value and interquartile range or mean SD. A 2-tailed univariate p < 0.05 was considered significant and all variables
Patient-Centered Value of Enhanced Recovery
1025
that were found to have a univariate p < 0.10 were included in multivariate analysis. The multivariate analysis was performed using binary logistic regression and the backwards stepwise variable elimination method. Time-to-event analysis was performed using one-minus event Kaplan-Meier analysis and the log rank test. All statistical analyses were performed using SPSS software (version 22.0, IBM Corp).
RESULTS Study patient demographic and clinical profile For the 118 patients in this study, median age was 59 years (interquartile range 48 to 67 years), 65 (55%) were male, and mean American Society of Anesthesiologists score was 3. Thirty-one patients (26.3%) underwent a minimally invasive approach to hepatectomy and 37 patients (31.4%) underwent major hepatectomy. Final pathology determined that operative indications included 23 primary hepatic malignancies (19.5%), 93 secondary metastatic malignancies (78.8%), and 2 benign lesions (1.7%), which were suspicious for malignancy on preoperative imaging. The ERLS pathway was used in 75 patients (63.6%). Compared with patients on a traditional recovery pathway, patients who were treated on the ERLS pathway had similar median age (59 vs 60 years; p ¼ 0.700) and similar rates of American Society of Anesthesiologists 3/ 4 classification (97% vs 93%; p ¼ 0.353). The proportion of ERLS patients receiving preoperative chemotherapy was also similar to the traditional pathway (75% vs 79%; p ¼ 0.588). The ERLS patients underwent a similar proportion of major hepatectomies (32% vs 30%; p ¼ 0.513), but were more likely to have a minimally invasive approach (35% vs 12%; p ¼ 0.006) (Table 2). Study patient clinical outcomes The ERLS and traditional pathways had similar procedure duration (268 vs 286 minutes; p ¼ 0.557), perioperative transfusions (9% vs 18%; p ¼ 0.146), major complications (12% vs 16%; p ¼ 0.513), and readmission rates (9% vs 12%; p ¼ 0.691). The ERLS patients were less likely to have a complication of any severity (37% vs 61%; p ¼ 0.015). For patient-reported immediate postoperative pain scores, ERLS patients were more likely to report <4 out of 10 pain (76% vs 56%; p ¼ 0.023). Hospital length of stay for the index hospitalization was lower for ERLS patients (4.8 vs 6.1 days; p ¼ 0.027). There were no ICU admission and no 30-day mortalities in either group. Patient-reported functional outcomes With regard to MDASI-GI scores, during the initial 31 days of follow-up, the ERLS patients were more likely
Enhanced Recovery in Liver Surgery vs Traditional Pathway
Factors
Preoperative Education
Intraoperative Perioperative steroids Opioid sparing anesthesia Intravenous/inhalational anesthetics Fluid management Regional analgesia Drains Postoperative Baseline analgesia
prn analgesia
Tubes Early ambulation
Fluid management Early oral intake Ready for discharge criteria
No mechanical bowel preparation required Celecoxib 400 mg po, pregabalin 75 mg po (avoid if older than 65 years), tramadol extended release 300 mg po morning of surgery, anxiolytics and anti-nausea medication as needed
Open and MIS liver surgery patient education material provided KVO IV Clear liquids after lunch day before surgery, npo post midnight Mechanical bowel preparation used selectively Anxiolytics and anti-nausea medication as needed
Dexamethasone 10 mg IV on induction of anesthesia
No
Yes Propofol as main anesthetic agent; IV dexmedetomidine, IV ketamine, IV lidocaine infusions titrated by anesthesia provider Goal directed: monitor stroke volume MIS: local anesthetic wound infiltration with long-acting liposomal bupivacaine; Open: epidural preferred Minimized
No Combined protocol with narcotics and inhalational agents Goal directed: unmonitored MIS: local wound infiltration with bupivacaine; Open: epidural preferred Selective
Pregabalin 75 mg po bid, start PM POD0 48 hours; acetaminophen 500 mg po 1, start PM POD0; celecoxib 200 mg po bid, start POD1; tramadol 50 mg po q6h, start POD1 48 hours; epidural basal rate if placed in operating room Epidural patient: titrated per pain service; nonepidural patient: mild pain (1 to 3): acetaminophen 500 mg po q6h (limit 2 gm/day); moderate pain (4 to 6): tramadol 50 mg po q6h; severe pain (7 to 10): hydromorphone 0.5 mg q15min 2 No NG tube Day of surgery: sit on edge of bed; POD1 out of bed to chair and ambulation at least 4 times daily
Epidural, PCA hydromorphone, po hydrocodone
Hepatobiliary fluid protocol, minimize IV fluid rate, SL after 600 mL po POD0 patients allowed clear liquids, POD1 regular diet Formalized: independently ambulatory, good pain control, tolerating diet, bowel function, no infections, comorbidities under control
Patient-Centered Value of Enhanced Recovery
Bowel state Preventive analgesia
Open and MIS liver surgery patient education material provided, as well as ER-specific patient education material, including information about ER principles, patient and caregiver expectations, and pain management Saline lock IV in preoperative holding Solids up to 6 hours before surgery, clear liquids permissible up to 2 hours before surgery
Traditional
Day et al
Fluid management Preoperative fasting
Enhanced recovery
1026
Table 1.
IV hydromorphone, po tramadol, po hydrocodone, IV/po acetaminophen, IV ketorolac
ADAT, advance diet as tolerated; ER, enhanced recovery; KVO, keep vein open; MIS, minimally invasive surgery; PCA, patient-controlled analgesia; POD, postoperative day; prn, as needed; SL, saline lock.
J Am Coll Surg
Selective NG tube Day of surgery: sit on edge of bed; POD1 out of bed to chair and selective ambulation at least 4 times daily Hepatobiliary fluid protocol, minimize IV fluid rate POD0 npo with ice, POD1 clears, ADAT Formalized: independently ambulatory, good pain control, tolerating diet, bowel function, no infections, comorbidities under control
Vol. 221, No. 6, December 2015
Table 2.
Day et al
Patient-Centered Value of Enhanced Recovery
1027
Enhanced Recovery in Liver Surgery vs Traditional Pathway Univariate Analysis
Factor
Age, years, median (IQR)* Sex, male Preoperative chemotherapy ASA 3/4y Minimally invasive Major hepatectomy Epidural Operative time, minutes, mean SD* Operative time 300 minutes Perioperative transfusion PACU pain <4/10 Any complication Major complication LOS, days, mean SD* RTB total RTB core RTB GI RTB interference Difference in total, mean SD* Difference in core, mean SD* Difference in GI, mean SD* Difference in interference, mean SD* Failure to RIOT Days to RIOT, mean SD*
Traditional (n ¼ 43 [36.4%])
ERLS (n ¼ 75 [63.6%])
p Value
60 (4767) 18 (42) 34 (79) 40 (93) 5 (12) 13 (30) 26 (61) 286 123 17 (40) 8 (18) 24 (56) 26 (61) 7 (16) 6.1 2.6 8 (19) 14 (33) 27 (63) 12 (28) 15.1 17.3 6.3 9.3 2.4 4.3 6.6 7.7 3 (13)z 60.2 35.3z
59 (4897) 47 (63) 56 (75) 73 (97) 26 (35) 24 (32) 37 (49) 268 126 29 (39) 7 (9) 57 (76) 28 (37) 9 (12) 4.8 2.0 27 (36) 32 (43) 47 (63) 37 (49) 12.1 18.9 6.0 11.8 1.6 5.8 4.4 7.3 2 (5)x 44.7 17.52x
0.700 0.029 0.588 0.353 0.006 0.842 0.243 0.557 0.926 0.146 0.023 0.015 0.513 0.027 0.046 0.279 0.989 0.023 0.179 0.290 0.299 0.037 0.373 0.134
Values are n (%) unless otherwise indicated. ASA, American Society of Anesthesiologists; GI, gastrointestinal; LOS, length of stay; PACU, postanesthesia care unit; RIOT, return to intended oncologic therapy; RTB, return to baseline. *Mann-Whitney U test. y Fisher’s exact test. z n ¼ 24 eligible. x n ¼ 38 eligible.
to report a return to baseline in total score (36% vs 19%; p ¼ 0.046) and total interference score (49% vs 28%; p ¼ 0.023). The ERLS and traditional pathway patients had a similar likelihood of experiencing return to baseline of core symptoms (43% vs 33%; p ¼ 0.279) and GI scores (63% vs 63%; p ¼ 0.989). A difference between absolute preoperative and postoperative scores was only significant for the composite interference scores (mean score difference: 4.4 vs 6.6; p ¼ 0.037). The ERLS patients also returned to baseline interference faster than patients treated on a traditional pathway (p ¼ 0.025) (Fig. 1). The largest impact that ERLS had on interference scores was observed in patients who were undergoing open major hepatectomy. Compared with 8% return to baseline interference with traditional recovery in open major hepatectomy, 46% of open major hepatectomy patients on the ERLS protocol returned to baseline interference within 31 days of surgery (p ¼ 0.020).
Analysis of risk factors for recovery Univariate analysis determined that the following factors were associated with return to baseline of the interference composite scores: minimally invasive approach, minor hepatectomy, and ERLS pathway. However, multivariate analysis determined that the only independent predictor of return to baseline interference score was being on an ERLS pathway (76% vs 55%; p ¼ 0.021, odds ratio ¼ 2.62) (Table 3). Assessment of return to intended oncologic therapy rates Longer-term follow-up was used to determine the RIOT rate and time to RIOT for both study groups. A total of 52 patients were intended to receive adjuvant therapy, 24 non-ERLS patients and 38 ERLS patients. Although nearly all patients initiated RIOT in a timely fashion, patients treated on the ERLS pathway tended to be more
1028
Day et al
Figure 1. Comparison of return to baseline interference between enhanced recovery in liver surgery and traditional pathway.
likely to initiate RIOT (95% vs 87%) with a shorter time to RIOT compared to the traditional recovery pathway (44.7 vs 60.2 days).
DISCUSSION The findings of this analysis are novel across several domains. First, these data suggest that the central impact of ER protocols on patient outcomes is the ability to lessen the degree to which postoperative symptoms interfere with early return to life function and enjoyment. Similar to previous reports, patients in the ERLS pathway experienced fewer overall complications.9,21,22 Table 3.
J Am Coll Surg
Patient-Centered Value of Enhanced Recovery
The shortened length of stay observed in this study is likely a derivative benefit of complication reduction, but does not completely account for the broader functional benefits of ER, particularly in the setting of oncologic surgery. Patients on both pathways reported similar levels of symptom burden. However, it appears that by minimizing the interference of postoperative symptom burden on life activities and enjoyment, ER protocols steepen the slope of recovery from largest deficit to preoperative baseline, returning patients more rapidly to their desired functional status. Second, these data demonstrate the value of a PRO tool that assesses not just symptom scores, but also measures the degree to which postoperative symptoms interfere with function and life experience. Typical PRO tools ask the patient to rate a series of symptoms on Likert scales. Many, such as the SF-36 and the FACT-GI are validated and can reliably reflect the presence of complications after surgery.23 Unfortunately, these tools only indirectly measure aggregate functional recovery that correlates with return to activity norms, enjoyment of life, and initiation of adjuvant therapies.17 In contrast, the MDASI-GI is a validated PRO instrument that catalogs both the symptoms and, potentially more importantly, the level of interference that the aggregate symptom burden has on the patient’s life and functional capacity.18 As such, it is an ideal tool to measure functional recovery after surgery. Many surgeons have had the experience of seeing a patient in postoperative clinic who subjectively reports severe postoperative symptoms (eg, pain or fatigue), yet is highly functional (eg, returned to driving and work). Indeed, many of the patients in the current study reported postoperative pain, fatigue, and other significant symptoms.
Factors Contributing to Return to Baseline Interference
Factor
Age 65 years or older Male Preoperative chemotherapy Minimally invasive Major hepatectomy Operation time 300 minutes Epidural ERLS LOS >5 days Any complication Major complication
No RTB interference (n ¼ 69) n %
22 35 51 14 26 30 35 38 34 35 9
32 51 74 20 38 45 51 55 49 51 13
RTB interference (n ¼ 49) n %
17 30 39 17 11 16 28 37 19 19 7
35 61 80 34 22 33 57 76 39 39 14
ERLS, enhanced recovery in liver surgery; LOS, length of stay; RTB, return to baseline.
Univariate p Value
0.749 0.259 0.475 0.080 0.079 0.235 0.491 0.023 0.259 0.199 0.846
Multivariate p Value
Odds ratio (95% CI)
0.530 0.069
0.021
2.62 (1.155.94)
Vol. 221, No. 6, December 2015
Day et al
However, it was only the interference score that correlated with their functional outcomes and their overall sense of well being. Certainly from a clinical perspective, surgeons need to closely monitor and compassionately respond to subjective symptom burden, but from a research perspective, this study emphasizes that symptom scores may be less accurate measures of functional recovery compared with interference scores. Third, the inclusion of patients across the spectrum of liver surgical oncology procedures allowed critical comparison of the influence of surgical approach, magnitude of hepatectomy, epidural use, and perioperative recovery programs on functional recovery. Although overall patient numbers did not allow for a full statistical evaluation of every factor of significance in liver surgery, the study was powered to evaluate the impact of major factors historically associated with post-hepatectomy outcomes. After controlling for the potential confounders of surgical approach, magnitude of hepatectomy and epidural use, the ERLS protocol was the only independent factor associated with return to baseline interference scores. The ERLS patients also achieved return to baseline interference scores more rapidly than patients on a traditional pathway. It is important to note that “return to baseline” was chosen instead of “return to normal” as the primary dependent variable in this analysis. As with many surgical oncology practices, and hepatobiliary surgery practice in particular, this was a highly pretreated population, with 76% of patients receiving preoperative systemic chemotherapy. Given this clinical setting, preoperative baseline MDASIGI scores were frequently elevated, reflecting residual chemotherapy-associated toxicities. Despite these deficits, it is remarkable that the use of ERLS could ameliorate the “double functional hit” of systemic therapy followed by surgery, providing functional recovery to more than half of the patients, at a median of 17 days after surgery. Fourth, there was a strong association between return to baseline interference scores and RIOT. This is a critically important finding, as it provides a potential mechanistic link between ERLS and improved long-term oncologic outcomes in cancer patients. Studies across multiple cancer types have demonstrated that timely RIOT leads to improved disease-free and overall survival.24-26 Nearfuture studies are needed to confirm the links between ERLS pathways, higher rates of and more rapid RIOT, and improved long-term oncologic outcomes.
CONCLUSIONS Using a validated patient-reported outcomes tool that discriminated life interference scores from symptom
Patient-Centered Value of Enhanced Recovery
1029
burden, this analysis identified a strong association between an enhanced recovery in liver surgery pathway and rapid functional recovery of a heavily pretreated cohort of hepatic surgery patients, independent of operative approach and magnitude of hepatectomy. Author Contributions Study conception and design: Day, Cleeland, Wang, Fielder, Calhoun, Gottumukkala, Aloia Acquisition of data: Day, Fielder, Conrad, Gottumukkala, Aloia Analysis and interpretation of data: Day, Vauthey, Gottumukkala, Aloia Drafting of manuscript: Day, Gottumukkala, Aloia Critical revision: Day, Cleeland, Wang, Fielder, Calhoun, Conrad, Vauthey, Gottumukkala, Aloia REFERENCES 1. Kehlet H, Wilmore DW. Fast-track surgery. Br J Surg 2005; 92:3e4. 2. Kehlet H. Future perspectives and research initiatives in fasttrack surgery. Langenbecks Arch Surg 2006;391:495e498. 3. Hughes MJ, McNally S, Wigmore SJ. Enhanced recovery following liver surgery: a systematic review and meta-analysis. HPB (Oxford) 2014;16:699e706. 4. Coolsen MM, van Dam RM, van der Wilt AA, et al. Systematic review and meta-analysis of enhanced recovery after pancreatic surgery with particular emphasis on pancreaticoduodenectomies. World J Surg 2013;37:1909e1918. 5. Greco M, Capretti G, Beretta L, et al. Enhanced recovery program in colorectal surgery: a meta-analysis of randomized controlled trials. World J Surg 2014;38:1531e1541. 6. Connor S, Cross A, Sakowska M, et al. Effects of introducing an enhanced recovery after surgery programme for patients undergoing open hepatic resection. HPB (Oxford) 2013;15:294e301. 7. di Sebastiano P, Festa L, De Bonis A, et al. A modified fast-track program for pancreatic surgery: a prospective single-center experience. Langenbecks Arch Surg 2011;396:345e351. 8. Serclova Z, Dytrych P, Marvan J, et al. Fast-track in open intestinal surgery: prospective randomized study (Clinical Trials Gov Identifier no. NCT00123456). Clin Nutr 2009;28:618e624. 9. Lin DX, Li X, Ye QW, et al. Implementation of a fast-track clinical pathway decreases postoperative length of stay and hospital charges for liver resection. Cell Biochem Biophys 2011; 61:413e419. 10. Takamoto T, Hashimoto T, Inoue K, et al. Applicability of enhanced recovery program for advanced liver surgery. World J Surg 2014;38:2676e2682. 11. McNicol FJ, Kennedy RH, Phillips RK, Clark SK. Laparoscopic total colectomy and ileorectal anastomosis (IRA), supported by an enhanced recovery programme in cases of familial adenomatous polyposis. Colorectal Dis 2012;14: 458e462. 12. Schultz NA, Larsen PN, Klarskov B, et al. Evaluation of a fast-track programme for patients undergoing liver resection. Br J Surg 2013;100:138e143. 13. Alcantara-Moral M, Serra-Aracil X, Gil-Egea MJ, et al. Observational cross-sectional study of compliance with the fast track
1030
14.
15. 16. 17. 18.
19.
Day et al
Patient-Centered Value of Enhanced Recovery
protocol in elective surgery for colon cancer in Spain. Int J Colorectal Dis 2014;29:477e483. Aloia TA, Zimmitti G, Conrad C, et al. Return to intended oncologic treatment (RIOT): a novel metric for evaluating the quality of oncosurgical therapy for malignancy. J Surg Oncol 2014;110:107e114. Merkow RP, Bilimoria KY, Tomlinson JS, et al. Postoperative complications reduce adjuvant chemotherapy use in resectable pancreatic cancer. Ann Surg 2014;260:372e377. Merkow RP, Bentrem DJ, Mulcahy MF, et al. Effect of postoperative complications on adjuvant chemotherapy use for stage III colon cancer. Ann Surg 2013;258:847e853. Lee L, Dumitra T, Fiore JF Jr, Mayo NE, Feldman LS. How well are we measuring postoperative “recovery” after abdominal surgery? Qual Life Res 2015;24:2583e2590. Wang XS, Williams LA, Eng C, et al. Validation and application of a module of the M. D. Anderson Symptom Inventory for measuring multiple symptoms in patients with gastrointestinal cancer (the MDASI-GI). Cancer 2010;116: 2053e2063. Andreou A, Vauthey JN, Cherqui D, et al. Improved long-term survival after major resection for hepatocellular carcinoma: a multicenter analysis based on a new definition of major hepatectomy. J Gastrointest Surg 2013;17:66e77; discussion 77.
J Am Coll Surg
20. Pang YY. The Brisbane 2000 terminology of liver anatomy and resections. HPB 2000;2:33339. HPB (Oxford) 2002;4:99. author reply 100. 21. van Dam RM, Hendry PO, Coolsen MM, et al. Initial experience with a multimodal enhanced recovery programme in patients undergoing liver resection. Br J Surg 2008;95:969e975. 22. Stoot JH, van Dam RM, Busch OR, et al. The effect of a multimodal fast-track programme on outcomes in laparoscopic liver surgery: a multicentre pilot study. HPB (Oxford) 2009;11: 140e144. 23. Awdeh H, Kassak K, Sfeir P, et al. The SF-36 and 6-minute walk test are significant predictors of complications after major surgery. World J Surg 2015;39:1406e1412. 24. Ahmed S, Iqbal N, Yadav S, et al. Time to adjuvant therapy and other variables in localized gastric and gastroesophageal junction (GEJ) cancer (IJGC-D-13-00162). J Gastrointest Cancer 2014; 45:284e290. 25. Gagliato Dde M, Gonzalez-Angulo AM, Lei X, et al. Clinical impact of delaying initiation of adjuvant chemotherapy in patients with breast cancer. J Clin Oncol 2014;32:735e744. 26. Biagi JJ, Raphael MJ, Mackillop WJ, et al. Association between time to initiation of adjuvant chemotherapy and survival in colorectal cancer: a systematic review and meta-analysis. JAMA 2011;305:2335e2342.
Vol. 221, No. 6, December 2015
APPENDIX
Day et al
Patient-Centered Value of Enhanced Recovery
1030.e1
1030.e2
Day et al
Patient-Centered Value of Enhanced Recovery
J Am Coll Surg
APN,
Enhanced recovery (ER) pathways have become increasingly integrated into surgical practice. Studies that compare ER and traditional pathways often focus on outcomes confined to inpatient hospitalization and rarely assess a patient’s functional recovery. The aim of this study was to compare functional outcomes for patients treated on an Enhanced Recovery in Liver Surgery (ERLS) pathway vs a traditional pathway. STUDY DESIGN: One hundred and eighteen hepatectomy patients rated symptom severity and life interference using the validated MD Anderson Symptom Inventory preoperatively and postoperatively at every outpatient visit until 31 days after surgery. The ERLS protocol included patient education, narcotic-sparing anesthesia and analgesia, diet advancement, restrictive fluid administration, early ambulation, and avoidance of drains and tubes. RESULTS: Seventy-five ERLS pathway patients were clinically comparable with 43 patients simultaneously treated on a traditional pathway. The ERLS patients reported lower immediate postoperative pain scores and experienced fewer complications and decreased length of stay. As measured by symptom burden on life interference, ERLS patients were more likely to return to baseline functional status in a shorter time interval. The only independent predictor of faster return to baseline interference levels was treatment on an ERLS pathway (p ¼ 0.021; odds ratio ¼ 2.62). In addition, ERLS pathway patients were more likely to return to intended oncologic therapy (95% vs 87%) at a shorter time interval compared to patients on the traditional pathway (44.7 vs 60.2 days). CONCLUSIONS: In oncologic liver surgery, enhanced recovery’s primary mechanism of action is reduction in life interference by postoperative surgical symptoms, allowing patients to return sooner to normal function and adjuvant cancer therapies. (J Am Coll Surg 2015;221:1023e1030. 2015 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved)
BACKGROUND:
Disclosure Information: Nothing to disclose. Support: This research was supported in part by National Institutes of Health Grant, CA016672. Drs Gottumukkala and Aloia contributed equally to this work. Poster abstract presented at the 123rd Annual Western Surgical Association Meeting, Napa Valley, CA, November 2015. Received August 10, 2015; Revised September 9, 2015; Accepted September 9, 2015. From the Department of Surgical Oncology (Day, Fielder, Conrad, Vauthey, Aloia), Department of Symptom Research (Cleeland, Wang), Institute for Cancer Care Innovation (Calhoun), and Department of Anesthesiology (Gottumukkala), The University of Texas MD Anderson Cancer Center, Houston, TX. Correspondence address: Thomas A Aloia, MD, FACS, Department of Surgical Oncology, The University of Texas MD Anderson Cancer Center, 1400 Herman Pressler, Unit 1484, Houston, TX 77030. email: taaloia@ mdanderson.org
ª 2015 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved.
The ultimate goal of any surgery is to return the patient to at least their baseline functional status, if not an improved functional status compared with their preoperative state, as rapidly as possible and with the least amount of intercurrent disability. Although fast-track surgical protocols have been reported for decades,1,2 it is only recently that the enhanced recovery (ER) movement in perioperative care has significantly penetrated North American surgical practice. Multiple recently published meta-analyses have clearly demonstrated that patients are benefiting from these changes in philosophy and practice.3-5 Most commonly, the included studies focused on outcomes confined to inpatient hospitalization using concrete primary end points, including early return of bowel function, lower complication rates, and/or shorter length of inpatient stay.6-8
1023
http://dx.doi.org/10.1016/j.jamcollsurg.2015.09.011 ISSN 1072-7515/15
1024
Day et al
Patient-Centered Value of Enhanced Recovery
Abbreviations and Acronyms
ER ERLS GI MDASI-GI PRO RIOT
¼ ¼ ¼ ¼
enhanced recovery Enhanced Recovery in Liver Surgery gastrointestinal MD Anderson Symptom Inventory, gastrointestinal version ¼ patient-reported outcomes ¼ return to intended oncologic therapy
In addition, many of the published studies that have examined the impact of ER protocols on short-term outcomes have focused on patients undergoing nononcologic procedures.9-11 When surgical oncology patients have been included, they have tended to have early-stage disease amenable to minimally invasive surgical approaches.12,13 For the majority of oncologic operations, however, postoperative recovery carries the additional demand of returning the patient to adjuvant oncologic therapies. Failure to return to intended oncologic therapy (RIOT)14 after cancer surgery due to complications and lingering poor performance status is strongly associated with worse oncologic outcomes, including shortened overall survival.15,16 Therefore, the end point of length of stay does not adequately measure the oncologic value of surgical recovery to patients with cancer. In surgical practice, and particularly in the field of surgical oncology, tools for the measurement of functional recovery are lacking.17 To address this knowledge gap, this study was designed to assess the ability of an ER program to deliver rapid functional recovery after surgical oncology procedures. To measure the recovery process, the study used a validated patient-reported outcomes (PRO) assessment tool to compare the quality of recovery between an Enhanced Recovery in Liver Surgery (ERLS) pathway and a traditional recovery pathway.
METHODS After study approval by the University of Texas MD Anderson Cancer Center IRB (protocol PA14-1079), hepatectomy patient data entered into a prospectively maintained hepatobiliary surgery database were assessed. The PRO tool used in this study was the gastrointestinal version of the MD Anderson Symptom Inventory (MDASI-GI).18 The MDASI-GI is a validated instrument in cancer patients18 composed of 24 questions that the patient evaluates on Likert scales (0 to 10) (Appendix). The tool has 3 sections, including 13 core symptom questions, which are common to all formats of the MDASI. The GI symptom-specific module comprises 5 questions unique to the patients with GI cancer (eg, constipation,
J Am Coll Surg
diarrhea/watery stools via stoma, swallowing, change in taste, feeling bloated). Lastly, the interferences portion is composed of 6 questions that assess the impact that symptoms are having on the patient’s function and well being (eg, general activity, mood, work including housework, relations with other people, walking, and enjoyment of life). A score of 0 for an item signifies that the patient was not experiencing the symptom of interest or that they were fully functional without life interference from symptoms, respectively. A score of 10 indicates that a symptom was graded as “the worst possible experience” or that interference was graded as “completely interfering” with daily functioning, respectively. The symptom inventory was administered to patients undergoing hepatectomy for any diagnosis between September 2013 and January 2015. Initially baseline data were collected on the last preoperative visit before the operation (typically 1 to 3 days before surgery). Postoperatively, as an inpatient, the survey was administered on postoperative days 1, 3, and 5. In addition, midterm recovery was measured by collecting MDASI-GI at each postsurgical outpatient visit until 31 days had passed since the operation (typically postoperative weeks 1 and 4). All survey data were used for time to recovery and event analyses. The last survey administered within the 31-day postsurgical time period was used for quantitative binary statistical analyses. Demographic, operative, hospitalization, and complication data were obtained from the hepatobiliary surgical database and supplemented by retrospective review of the electronic medical record. For each patient, the preoperative and postoperative MDASI-GI questionnaires were compared across several domains. First, the magnitude and direction of score change from preoperative survey to postoperative survey for the total score and for each component score of the MDASI-GI (core, GI, and interference) were calculated. Individual MDASI item, component scores, and total score were then converted to the categorical value of “returned to baseline,” defined as a postoperative score that was no more than 2 points higher than the preoperative score. Demographic and clinical information included age, sex, preoperative systemic therapy, and American Society of Anesthesiologists score. Operative data included type of operation, use of minimally invasive vs open approach, use of epidural analgesia, length of operation, magnitude of liver resection, and use of the ERLS pathway. There were 2 patients who were initially in the minimally invasive group that required conversion to an open operation due to dense adhesions; for the purpose of this analysis, these were considered open and not minimally invasive procedures. Major hepatectomy was defined as nonanatomic resection
Vol. 221, No. 6, December 2015
Day et al
of 4 or more segments,19 right hepatectomy, left hepatectomy, extended right hepatectomy, or extended left hepatectomy as defined by The Brisbane 2000 terminology.20 Major complications were defined as any adverse event requiring reintervention or ICU admission. Outcomes assessed included length of stay, patient-reported pain scores on arrival to the post-anesthesia care unit (anesthesia goal for initial pain control was a patient-reported pain score of <4 on a 0 to 10 scale), return to baseline MDASI-GI composite scores, RIOT rate, and time to RIOT. The ERLS pathway was composed of preoperative education regarding the principles of ER and expected processes, in addition to routine liver surgery education presented to traditional pathway patients. Patients on the ERLS pathway received multimodality pain control. Preoperatively, this cohort received 3 oral medications: tramadol extended release, pregabalin, and celecoxib. Intraoperatively, a narcotic limiting perioperative inflammatory modulation and immunoprotective anesthetic technique was used. Postoperatively, patients were reinitiated on the 3-drug non-narcotic oral protocol. In addition, for open procedures, patients received either epidural catheters or IV dilaudid patientcontrolled analgesia on an as-needed basis. For minimally invasive procedures, epidural catheters were not used, and specimen extraction incisions and port sites were infiltrated with long-acting local anesthetic (liposomal bupivacaine). Nasogastric tubes were not used. The ERLS patients were placed on a clear liquid diet on the same day as the operation. Intravenous fluids were restricted and all IV fluids were discontinued after 600 mL po intake, followed by rapid diet advancement. In addition to continuing the 3 oral drug, narcotic-sparing regimen, postoperative breakthrough pain was managed with a defined stair step as needed algorithm (acetaminophen/tramadol/low-dose hydromorphone) (Table 1). To regulate compliance, surgeons entered the ERLS protocol in a staggered fashion. Once entered, all hepatectomy patients were eligible for ERLS with no restrictions based on case type, surgical approach, or age. Compliance audits determined near complete (>90%) compliance with all phases of the ER protocol, save for patient allergy, comorbidity, or complication, where safety indicated a departure from protocol. Statistics All univariate statistical analyses were performed using the chi-square test or Fisher’s exact test for categorical variables and the Mann-Whitney U test for continuous variables. Continuous variables are reported as median value and interquartile range or mean SD. A 2-tailed univariate p < 0.05 was considered significant and all variables
Patient-Centered Value of Enhanced Recovery
1025
that were found to have a univariate p < 0.10 were included in multivariate analysis. The multivariate analysis was performed using binary logistic regression and the backwards stepwise variable elimination method. Time-to-event analysis was performed using one-minus event Kaplan-Meier analysis and the log rank test. All statistical analyses were performed using SPSS software (version 22.0, IBM Corp).
RESULTS Study patient demographic and clinical profile For the 118 patients in this study, median age was 59 years (interquartile range 48 to 67 years), 65 (55%) were male, and mean American Society of Anesthesiologists score was 3. Thirty-one patients (26.3%) underwent a minimally invasive approach to hepatectomy and 37 patients (31.4%) underwent major hepatectomy. Final pathology determined that operative indications included 23 primary hepatic malignancies (19.5%), 93 secondary metastatic malignancies (78.8%), and 2 benign lesions (1.7%), which were suspicious for malignancy on preoperative imaging. The ERLS pathway was used in 75 patients (63.6%). Compared with patients on a traditional recovery pathway, patients who were treated on the ERLS pathway had similar median age (59 vs 60 years; p ¼ 0.700) and similar rates of American Society of Anesthesiologists 3/ 4 classification (97% vs 93%; p ¼ 0.353). The proportion of ERLS patients receiving preoperative chemotherapy was also similar to the traditional pathway (75% vs 79%; p ¼ 0.588). The ERLS patients underwent a similar proportion of major hepatectomies (32% vs 30%; p ¼ 0.513), but were more likely to have a minimally invasive approach (35% vs 12%; p ¼ 0.006) (Table 2). Study patient clinical outcomes The ERLS and traditional pathways had similar procedure duration (268 vs 286 minutes; p ¼ 0.557), perioperative transfusions (9% vs 18%; p ¼ 0.146), major complications (12% vs 16%; p ¼ 0.513), and readmission rates (9% vs 12%; p ¼ 0.691). The ERLS patients were less likely to have a complication of any severity (37% vs 61%; p ¼ 0.015). For patient-reported immediate postoperative pain scores, ERLS patients were more likely to report <4 out of 10 pain (76% vs 56%; p ¼ 0.023). Hospital length of stay for the index hospitalization was lower for ERLS patients (4.8 vs 6.1 days; p ¼ 0.027). There were no ICU admission and no 30-day mortalities in either group. Patient-reported functional outcomes With regard to MDASI-GI scores, during the initial 31 days of follow-up, the ERLS patients were more likely
Enhanced Recovery in Liver Surgery vs Traditional Pathway
Factors
Preoperative Education
Intraoperative Perioperative steroids Opioid sparing anesthesia Intravenous/inhalational anesthetics Fluid management Regional analgesia Drains Postoperative Baseline analgesia
prn analgesia
Tubes Early ambulation
Fluid management Early oral intake Ready for discharge criteria
No mechanical bowel preparation required Celecoxib 400 mg po, pregabalin 75 mg po (avoid if older than 65 years), tramadol extended release 300 mg po morning of surgery, anxiolytics and anti-nausea medication as needed
Open and MIS liver surgery patient education material provided KVO IV Clear liquids after lunch day before surgery, npo post midnight Mechanical bowel preparation used selectively Anxiolytics and anti-nausea medication as needed
Dexamethasone 10 mg IV on induction of anesthesia
No
Yes Propofol as main anesthetic agent; IV dexmedetomidine, IV ketamine, IV lidocaine infusions titrated by anesthesia provider Goal directed: monitor stroke volume MIS: local anesthetic wound infiltration with long-acting liposomal bupivacaine; Open: epidural preferred Minimized
No Combined protocol with narcotics and inhalational agents Goal directed: unmonitored MIS: local wound infiltration with bupivacaine; Open: epidural preferred Selective
Pregabalin 75 mg po bid, start PM POD0 48 hours; acetaminophen 500 mg po 1, start PM POD0; celecoxib 200 mg po bid, start POD1; tramadol 50 mg po q6h, start POD1 48 hours; epidural basal rate if placed in operating room Epidural patient: titrated per pain service; nonepidural patient: mild pain (1 to 3): acetaminophen 500 mg po q6h (limit 2 gm/day); moderate pain (4 to 6): tramadol 50 mg po q6h; severe pain (7 to 10): hydromorphone 0.5 mg q15min 2 No NG tube Day of surgery: sit on edge of bed; POD1 out of bed to chair and ambulation at least 4 times daily
Epidural, PCA hydromorphone, po hydrocodone
Hepatobiliary fluid protocol, minimize IV fluid rate, SL after 600 mL po POD0 patients allowed clear liquids, POD1 regular diet Formalized: independently ambulatory, good pain control, tolerating diet, bowel function, no infections, comorbidities under control
Patient-Centered Value of Enhanced Recovery
Bowel state Preventive analgesia
Open and MIS liver surgery patient education material provided, as well as ER-specific patient education material, including information about ER principles, patient and caregiver expectations, and pain management Saline lock IV in preoperative holding Solids up to 6 hours before surgery, clear liquids permissible up to 2 hours before surgery
Traditional
Day et al
Fluid management Preoperative fasting
Enhanced recovery
1026
Table 1.
IV hydromorphone, po tramadol, po hydrocodone, IV/po acetaminophen, IV ketorolac
ADAT, advance diet as tolerated; ER, enhanced recovery; KVO, keep vein open; MIS, minimally invasive surgery; PCA, patient-controlled analgesia; POD, postoperative day; prn, as needed; SL, saline lock.
J Am Coll Surg
Selective NG tube Day of surgery: sit on edge of bed; POD1 out of bed to chair and selective ambulation at least 4 times daily Hepatobiliary fluid protocol, minimize IV fluid rate POD0 npo with ice, POD1 clears, ADAT Formalized: independently ambulatory, good pain control, tolerating diet, bowel function, no infections, comorbidities under control
Vol. 221, No. 6, December 2015
Table 2.
Day et al
Patient-Centered Value of Enhanced Recovery
1027
Enhanced Recovery in Liver Surgery vs Traditional Pathway Univariate Analysis
Factor
Age, years, median (IQR)* Sex, male Preoperative chemotherapy ASA 3/4y Minimally invasive Major hepatectomy Epidural Operative time, minutes, mean SD* Operative time 300 minutes Perioperative transfusion PACU pain <4/10 Any complication Major complication LOS, days, mean SD* RTB total RTB core RTB GI RTB interference Difference in total, mean SD* Difference in core, mean SD* Difference in GI, mean SD* Difference in interference, mean SD* Failure to RIOT Days to RIOT, mean SD*
Traditional (n ¼ 43 [36.4%])
ERLS (n ¼ 75 [63.6%])
p Value
60 (4767) 18 (42) 34 (79) 40 (93) 5 (12) 13 (30) 26 (61) 286 123 17 (40) 8 (18) 24 (56) 26 (61) 7 (16) 6.1 2.6 8 (19) 14 (33) 27 (63) 12 (28) 15.1 17.3 6.3 9.3 2.4 4.3 6.6 7.7 3 (13)z 60.2 35.3z
59 (4897) 47 (63) 56 (75) 73 (97) 26 (35) 24 (32) 37 (49) 268 126 29 (39) 7 (9) 57 (76) 28 (37) 9 (12) 4.8 2.0 27 (36) 32 (43) 47 (63) 37 (49) 12.1 18.9 6.0 11.8 1.6 5.8 4.4 7.3 2 (5)x 44.7 17.52x
0.700 0.029 0.588 0.353 0.006 0.842 0.243 0.557 0.926 0.146 0.023 0.015 0.513 0.027 0.046 0.279 0.989 0.023 0.179 0.290 0.299 0.037 0.373 0.134
Values are n (%) unless otherwise indicated. ASA, American Society of Anesthesiologists; GI, gastrointestinal; LOS, length of stay; PACU, postanesthesia care unit; RIOT, return to intended oncologic therapy; RTB, return to baseline. *Mann-Whitney U test. y Fisher’s exact test. z n ¼ 24 eligible. x n ¼ 38 eligible.
to report a return to baseline in total score (36% vs 19%; p ¼ 0.046) and total interference score (49% vs 28%; p ¼ 0.023). The ERLS and traditional pathway patients had a similar likelihood of experiencing return to baseline of core symptoms (43% vs 33%; p ¼ 0.279) and GI scores (63% vs 63%; p ¼ 0.989). A difference between absolute preoperative and postoperative scores was only significant for the composite interference scores (mean score difference: 4.4 vs 6.6; p ¼ 0.037). The ERLS patients also returned to baseline interference faster than patients treated on a traditional pathway (p ¼ 0.025) (Fig. 1). The largest impact that ERLS had on interference scores was observed in patients who were undergoing open major hepatectomy. Compared with 8% return to baseline interference with traditional recovery in open major hepatectomy, 46% of open major hepatectomy patients on the ERLS protocol returned to baseline interference within 31 days of surgery (p ¼ 0.020).
Analysis of risk factors for recovery Univariate analysis determined that the following factors were associated with return to baseline of the interference composite scores: minimally invasive approach, minor hepatectomy, and ERLS pathway. However, multivariate analysis determined that the only independent predictor of return to baseline interference score was being on an ERLS pathway (76% vs 55%; p ¼ 0.021, odds ratio ¼ 2.62) (Table 3). Assessment of return to intended oncologic therapy rates Longer-term follow-up was used to determine the RIOT rate and time to RIOT for both study groups. A total of 52 patients were intended to receive adjuvant therapy, 24 non-ERLS patients and 38 ERLS patients. Although nearly all patients initiated RIOT in a timely fashion, patients treated on the ERLS pathway tended to be more
1028
Day et al
Figure 1. Comparison of return to baseline interference between enhanced recovery in liver surgery and traditional pathway.
likely to initiate RIOT (95% vs 87%) with a shorter time to RIOT compared to the traditional recovery pathway (44.7 vs 60.2 days).
DISCUSSION The findings of this analysis are novel across several domains. First, these data suggest that the central impact of ER protocols on patient outcomes is the ability to lessen the degree to which postoperative symptoms interfere with early return to life function and enjoyment. Similar to previous reports, patients in the ERLS pathway experienced fewer overall complications.9,21,22 Table 3.
J Am Coll Surg
Patient-Centered Value of Enhanced Recovery
The shortened length of stay observed in this study is likely a derivative benefit of complication reduction, but does not completely account for the broader functional benefits of ER, particularly in the setting of oncologic surgery. Patients on both pathways reported similar levels of symptom burden. However, it appears that by minimizing the interference of postoperative symptom burden on life activities and enjoyment, ER protocols steepen the slope of recovery from largest deficit to preoperative baseline, returning patients more rapidly to their desired functional status. Second, these data demonstrate the value of a PRO tool that assesses not just symptom scores, but also measures the degree to which postoperative symptoms interfere with function and life experience. Typical PRO tools ask the patient to rate a series of symptoms on Likert scales. Many, such as the SF-36 and the FACT-GI are validated and can reliably reflect the presence of complications after surgery.23 Unfortunately, these tools only indirectly measure aggregate functional recovery that correlates with return to activity norms, enjoyment of life, and initiation of adjuvant therapies.17 In contrast, the MDASI-GI is a validated PRO instrument that catalogs both the symptoms and, potentially more importantly, the level of interference that the aggregate symptom burden has on the patient’s life and functional capacity.18 As such, it is an ideal tool to measure functional recovery after surgery. Many surgeons have had the experience of seeing a patient in postoperative clinic who subjectively reports severe postoperative symptoms (eg, pain or fatigue), yet is highly functional (eg, returned to driving and work). Indeed, many of the patients in the current study reported postoperative pain, fatigue, and other significant symptoms.
Factors Contributing to Return to Baseline Interference
Factor
Age 65 years or older Male Preoperative chemotherapy Minimally invasive Major hepatectomy Operation time 300 minutes Epidural ERLS LOS >5 days Any complication Major complication
No RTB interference (n ¼ 69) n %
22 35 51 14 26 30 35 38 34 35 9
32 51 74 20 38 45 51 55 49 51 13
RTB interference (n ¼ 49) n %
17 30 39 17 11 16 28 37 19 19 7
35 61 80 34 22 33 57 76 39 39 14
ERLS, enhanced recovery in liver surgery; LOS, length of stay; RTB, return to baseline.
Univariate p Value
0.749 0.259 0.475 0.080 0.079 0.235 0.491 0.023 0.259 0.199 0.846
Multivariate p Value
Odds ratio (95% CI)
0.530 0.069
0.021
2.62 (1.155.94)
Vol. 221, No. 6, December 2015
Day et al
However, it was only the interference score that correlated with their functional outcomes and their overall sense of well being. Certainly from a clinical perspective, surgeons need to closely monitor and compassionately respond to subjective symptom burden, but from a research perspective, this study emphasizes that symptom scores may be less accurate measures of functional recovery compared with interference scores. Third, the inclusion of patients across the spectrum of liver surgical oncology procedures allowed critical comparison of the influence of surgical approach, magnitude of hepatectomy, epidural use, and perioperative recovery programs on functional recovery. Although overall patient numbers did not allow for a full statistical evaluation of every factor of significance in liver surgery, the study was powered to evaluate the impact of major factors historically associated with post-hepatectomy outcomes. After controlling for the potential confounders of surgical approach, magnitude of hepatectomy and epidural use, the ERLS protocol was the only independent factor associated with return to baseline interference scores. The ERLS patients also achieved return to baseline interference scores more rapidly than patients on a traditional pathway. It is important to note that “return to baseline” was chosen instead of “return to normal” as the primary dependent variable in this analysis. As with many surgical oncology practices, and hepatobiliary surgery practice in particular, this was a highly pretreated population, with 76% of patients receiving preoperative systemic chemotherapy. Given this clinical setting, preoperative baseline MDASIGI scores were frequently elevated, reflecting residual chemotherapy-associated toxicities. Despite these deficits, it is remarkable that the use of ERLS could ameliorate the “double functional hit” of systemic therapy followed by surgery, providing functional recovery to more than half of the patients, at a median of 17 days after surgery. Fourth, there was a strong association between return to baseline interference scores and RIOT. This is a critically important finding, as it provides a potential mechanistic link between ERLS and improved long-term oncologic outcomes in cancer patients. Studies across multiple cancer types have demonstrated that timely RIOT leads to improved disease-free and overall survival.24-26 Nearfuture studies are needed to confirm the links between ERLS pathways, higher rates of and more rapid RIOT, and improved long-term oncologic outcomes.
CONCLUSIONS Using a validated patient-reported outcomes tool that discriminated life interference scores from symptom
Patient-Centered Value of Enhanced Recovery
1029
burden, this analysis identified a strong association between an enhanced recovery in liver surgery pathway and rapid functional recovery of a heavily pretreated cohort of hepatic surgery patients, independent of operative approach and magnitude of hepatectomy. Author Contributions Study conception and design: Day, Cleeland, Wang, Fielder, Calhoun, Gottumukkala, Aloia Acquisition of data: Day, Fielder, Conrad, Gottumukkala, Aloia Analysis and interpretation of data: Day, Vauthey, Gottumukkala, Aloia Drafting of manuscript: Day, Gottumukkala, Aloia Critical revision: Day, Cleeland, Wang, Fielder, Calhoun, Conrad, Vauthey, Gottumukkala, Aloia REFERENCES 1. Kehlet H, Wilmore DW. Fast-track surgery. Br J Surg 2005; 92:3e4. 2. Kehlet H. Future perspectives and research initiatives in fasttrack surgery. Langenbecks Arch Surg 2006;391:495e498. 3. Hughes MJ, McNally S, Wigmore SJ. Enhanced recovery following liver surgery: a systematic review and meta-analysis. HPB (Oxford) 2014;16:699e706. 4. Coolsen MM, van Dam RM, van der Wilt AA, et al. Systematic review and meta-analysis of enhanced recovery after pancreatic surgery with particular emphasis on pancreaticoduodenectomies. World J Surg 2013;37:1909e1918. 5. Greco M, Capretti G, Beretta L, et al. Enhanced recovery program in colorectal surgery: a meta-analysis of randomized controlled trials. World J Surg 2014;38:1531e1541. 6. Connor S, Cross A, Sakowska M, et al. Effects of introducing an enhanced recovery after surgery programme for patients undergoing open hepatic resection. HPB (Oxford) 2013;15:294e301. 7. di Sebastiano P, Festa L, De Bonis A, et al. A modified fast-track program for pancreatic surgery: a prospective single-center experience. Langenbecks Arch Surg 2011;396:345e351. 8. Serclova Z, Dytrych P, Marvan J, et al. Fast-track in open intestinal surgery: prospective randomized study (Clinical Trials Gov Identifier no. NCT00123456). Clin Nutr 2009;28:618e624. 9. Lin DX, Li X, Ye QW, et al. Implementation of a fast-track clinical pathway decreases postoperative length of stay and hospital charges for liver resection. Cell Biochem Biophys 2011; 61:413e419. 10. Takamoto T, Hashimoto T, Inoue K, et al. Applicability of enhanced recovery program for advanced liver surgery. World J Surg 2014;38:2676e2682. 11. McNicol FJ, Kennedy RH, Phillips RK, Clark SK. Laparoscopic total colectomy and ileorectal anastomosis (IRA), supported by an enhanced recovery programme in cases of familial adenomatous polyposis. Colorectal Dis 2012;14: 458e462. 12. Schultz NA, Larsen PN, Klarskov B, et al. Evaluation of a fast-track programme for patients undergoing liver resection. Br J Surg 2013;100:138e143. 13. Alcantara-Moral M, Serra-Aracil X, Gil-Egea MJ, et al. Observational cross-sectional study of compliance with the fast track
1030
14.
15. 16. 17. 18.
19.
Day et al
Patient-Centered Value of Enhanced Recovery
protocol in elective surgery for colon cancer in Spain. Int J Colorectal Dis 2014;29:477e483. Aloia TA, Zimmitti G, Conrad C, et al. Return to intended oncologic treatment (RIOT): a novel metric for evaluating the quality of oncosurgical therapy for malignancy. J Surg Oncol 2014;110:107e114. Merkow RP, Bilimoria KY, Tomlinson JS, et al. Postoperative complications reduce adjuvant chemotherapy use in resectable pancreatic cancer. Ann Surg 2014;260:372e377. Merkow RP, Bentrem DJ, Mulcahy MF, et al. Effect of postoperative complications on adjuvant chemotherapy use for stage III colon cancer. Ann Surg 2013;258:847e853. Lee L, Dumitra T, Fiore JF Jr, Mayo NE, Feldman LS. How well are we measuring postoperative “recovery” after abdominal surgery? Qual Life Res 2015;24:2583e2590. Wang XS, Williams LA, Eng C, et al. Validation and application of a module of the M. D. Anderson Symptom Inventory for measuring multiple symptoms in patients with gastrointestinal cancer (the MDASI-GI). Cancer 2010;116: 2053e2063. Andreou A, Vauthey JN, Cherqui D, et al. Improved long-term survival after major resection for hepatocellular carcinoma: a multicenter analysis based on a new definition of major hepatectomy. J Gastrointest Surg 2013;17:66e77; discussion 77.
J Am Coll Surg
20. Pang YY. The Brisbane 2000 terminology of liver anatomy and resections. HPB 2000;2:33339. HPB (Oxford) 2002;4:99. author reply 100. 21. van Dam RM, Hendry PO, Coolsen MM, et al. Initial experience with a multimodal enhanced recovery programme in patients undergoing liver resection. Br J Surg 2008;95:969e975. 22. Stoot JH, van Dam RM, Busch OR, et al. The effect of a multimodal fast-track programme on outcomes in laparoscopic liver surgery: a multicentre pilot study. HPB (Oxford) 2009;11: 140e144. 23. Awdeh H, Kassak K, Sfeir P, et al. The SF-36 and 6-minute walk test are significant predictors of complications after major surgery. World J Surg 2015;39:1406e1412. 24. Ahmed S, Iqbal N, Yadav S, et al. Time to adjuvant therapy and other variables in localized gastric and gastroesophageal junction (GEJ) cancer (IJGC-D-13-00162). J Gastrointest Cancer 2014; 45:284e290. 25. Gagliato Dde M, Gonzalez-Angulo AM, Lei X, et al. Clinical impact of delaying initiation of adjuvant chemotherapy in patients with breast cancer. J Clin Oncol 2014;32:735e744. 26. Biagi JJ, Raphael MJ, Mackillop WJ, et al. Association between time to initiation of adjuvant chemotherapy and survival in colorectal cancer: a systematic review and meta-analysis. JAMA 2011;305:2335e2342.
Vol. 221, No. 6, December 2015
APPENDIX
Day et al
Patient-Centered Value of Enhanced Recovery
1030.e1
1030.e2
Day et al
Patient-Centered Value of Enhanced Recovery
J Am Coll Surg