Cost analysis of enhanced recovery after surgery in microvascular breast reconstruction

ARTICLE IN PRESS Journal of Plastic, Reconstructive & Aesthetic Surgery (2018) ■■, ■■–■■

Cost analysis of enhanced recovery after surgery in microvascular breast reconstruction Christine Oh a, James Moriarty b, Bijan J. Borah b,c, Kristin C. Mara d, William S. Harmsen d, Michel Saint-Cyr e, Valerie Lemaine a,* a

Division of Plastic and Reconstructive Surgery, Department of Surgery, Mayo Clinic, Rochester, MN, 55905, USA Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, MN, 55905, USA c Division of Health Care Policy and Research, Mayo Clinic, Rochester, MN, 55905, USA d Department of Biomedical Statistics and Informatics, Mayo Clinic, Rochester, MN, 55905, USA e Division of Plastic and Reconstructive Surgery, Department of Surgery, Baylor Scott & White Health, Temple, TX, 76508, USA b

Received 29 November 2017; accepted 18 February 2018

KEYWORDS DIEP flap; Enhanced Recovery After Surgery; ERAS; Muscle-sparing free TRAM; Outcomes; Economic analysis

Summary Background: Enhanced recovery after surgery (ERAS) pathways have been shown in multiple surgical specialties to decrease hospital length of stay (LOS) after surgery. ERAS in breast reconstruction has been found to decrease hospital LOS and inpatient opioid use. ERAS protocols can facilitate a patient’s recovery and can potentially increase the quality of care while decreasing costs. Methods: A standardized ERAS pathway was developed through multidisciplinary collaboration. It addressed all phases of surgical care for patients undergoing free-flap breast reconstruction utilizing an abdominal donor site. In this retrospective cohort study, clinical variables associated with hospitalization costs for patients who underwent free-flap breast reconstruction with the ERAS pathway were compared with those of historical controls, termed traditional recovery after surgery (TRAS). All patients included in the study underwent surgery between September 2010 and September 2014. Predicted costs of the study groups were compared using generalized linear modeling. Results: A total of 200 patients were analyzed: 82 in the ERAS cohort and 118 in the TRAS cohort. Clinical variables that were identified to potentially affect costs were found to have a statistically significant difference between groups and included unilateral versus bilateral procedures (p = 0.04) and the need for postoperative blood transfusion (p = 0.03). The cost

Disclosures: The authors have no financial interests in any of the products or techniques mentioned and have received no external support related to this study. * Corresponding author. Mayo Clinic, Division of Plastic Surgery, 200 First Street SW, Rochester, MN 55905. E-mail address: [email protected] (V. Lemaine). https://doi.org/10.1016/j.bjps.2018.02.018 1748-6815/© 2018 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.

ARTICLE IN PRESS 2

C. Oh et al. regression analysis on the two cohorts was adjusted for these significant variables. Adjusted mean costs of patients with ERAS were found to be $4,576 lesser than those of the TRAS control group ($38,688 versus $43,264). Conclusions: Implementation of the ERAS pathway was associated with significantly decreased costs when compared to historical controls. There has been a healthcare focus toward prudent resource allocation, which dictates the need for plastic surgeons to recognize economic evaluation of clinical practice. The ERAS pathway can increase healthcare accountability by improving quality of care while simultaneously decreasing the costs associated with autologous breast reconstruction. © 2018 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.

Introduction The implementation of enhanced recovery after surgery (ERAS) protocols aims to reduce postoperative morbidity to allow for shorter recovery times and decreased length of hospital stay (LOS).1–3 Multimodal pain management can also expedite recovery and discharge.4 All these factors have an economic advantage in terms of resource allocation and cost– benefit ratio. Results from a cohort of free flap patients comparing the ERAS and traditional recovery after surgery (TRAS) pathways at the Mayo Clinic, Rochester, MN, concluded a shorter mean hospital LOS and decreased postoperative opioid usage.5 This study used the same cohort to perform a cost comparison analysis of the ERAS versus TRAS pathways.

Methods Development of the ERAS pathway A multidisciplinary team developed the ERAS pathway. ERAS diverges from TRAS beginning in the preoperative holding area where patients are administered acetaminophen, celecoxib, and gabapentin for pre-emptive pain management. The anesthesia team administers antiemetics upon induction and maintains euvolemia. Intraoperative local anesthesia for the ERAS pathway is administered in the form of liposomal bupivacaine (Exparel; Pacira Pharmaceuticals, Inc.) diluted with normal saline as a subfascial transversus abdominis block, as well as in the rectus sheath and lower abdominal subcutaneous tissues. Patients recover in the postanesthesia care unit and are transferred to the general postsurgical floor under the care of a plastic surgery floor nurse familiar with flap monitoring. Patients are administered scheduled acetaminophen and celecoxib and are given oral opioids as needed and also parenteral agents for breakthrough pain. Patients are immediately started on a general diet and encouraged to ambulate as tolerated. Intravenous fluids are discontinued as soon as the patient has 600 ml of oral intake. Urinary catheters are removed on postoperative day 1. Discharge planning begins the day after surgery, with a goal for discharge on postoperative day 3 or 4. The ERAS pathway was implemented in November 2012, including incorporation into the electronic medical record as a standardized order set. In the TRAS cohort, the perioperative course was not standardized except for postoperative intensive care unit

(ICU) admission and use of patient-controlled analgesia (PCA). Postoperative pain control and fluid balance were managed by the intensivist on-call.

Study design A database was developed using REDCap (Vanderbilt University, Nashville, TN). Consecutive patients underwent immediate or delayed abdominally based microsurgical breast reconstruction between September 2010 and September 2014. The patient cohort was expanded from Batdorf et al.’s study by two additional years.5 In this retrospective cohort study, women in the ERAS cohort were compared with consecutive historical controls in the TRAS cohort. Specific operative technique was not standardized among the surgeons. The Mayo Clinic Institutional Review Board approved this study. All patients declining research authorization were excluded from analysis.

Data sources Direct medical costs of all services and procedures billed were estimated using a standardized costing approach. These billed services were obtained from the Mayo Clinic Cost Data Warehouse (MCCDW).6 This database provides estimated facility (hospital) costs from line item billed charges using department-level cost-to-charge ratios from the Medicare cost reports. Medical professional costs are estimated using the appropriate Medicare physician fee schedule for each Current Procedural Terminology (CPT) fourth edition code or Healthcare Common Procedure Coding System (HCPCS) tag. To account for potential differences over time in costto-charge ratios and Medicare reimbursement rates, services were valued using 2014 dollars.7 Services that could not be mapped to identical services in 2014 were adjusted to 2014 US dollars using the Gross Domestic Product implicit price deflator.8 Costs were categorized into the BerensonEggers Type of Service categories (BETOS).9,10 The BETOS coding system analyzes growth in Medicare expenditures and allows for clinical categorization that is relatively immune to minor changes in technology or practice patterns, providing objective cost categories.

Statistical analysis Descriptive statistics were reported as mean (SD) or median (IQR) for continuous variables and as the number of patients

ARTICLE IN PRESS Enhanced recovery cost analysis Table 1

3

Patient demographics and comorbidities*. ERAS (N = 82)

Age at surgery Mean (SD) Median Q1, Q3 Insurance Commercial Government BMI Mean (SD) Median Q1, Q3 On Chronic Opioids at time of surgery Fibromyalgia Coagulopathy Factor V Leiden Previous DVT or PE Tobacco Use Never Smoker Former Current Previous Chest Wall Radiation Any Comorbidities COPD or Asthma History of Prior MI, valve replacement Hypertension Diabetes Mellitus Chronic Kidney Disease

TRAS (N = 118)

Total (N = 200)

p value 0.88

49.2 (9.4) 49.4 41.1, 57.2

49.4 (8.7) 49.0 43.5, 54.0

49.3 (9.0) 49.0 43.1, 55.5 0.84

69 (84.1%) 13 (15.9%)

98 (83.1%) 20 (16.9%)

167 (83.5%) 33 (16.5%) 0.08

28.4 (5.3) 28.0 24.6, 31.0 6 (7.3%) 4 (4.9%)

29.7 (4.7) 29.4 26.4, 32.6 4 (3.4%) 6 (5.1%)

29.2 (5.0) 28.6 25.3, 32.0 10 (5.0%) 10 (5.0%)

1 (1.2%) 2 (2.4%)

0 (0.0%) 2 (1.7%)

1 (0.5%) 4 (2.0%)

61 (74.4%) 19 (23.2%) 2 (2.4%) 46 (56.1%) 23 (28.0%) 3 (3.7%) 4 (4.9%) 12 (14.6%) 2 (2.4%) 0 (0.0%)

74 (62.7%) 42 (35.6%) 2 (1.7%) 65 (55.1%) 45 (38.1%) 9 (7.6%) 3 (2.5%) 24 (20.3%) 7 (5.9%) 2 (1.7%)

135 (67.5%) 61 (30.5%) 4 (2.0%) 111 (55.5%) 68 (34.0%) 12 (6.0%) 7 (3.5%) 36 (18.0%) 9 (4.5%) 2 (1.0%)

0.32 1.00 0.59

0.17

0.89 0.14 0.37 0.45 0.35 0.31 0.51

Abbreviations: BMI, body mass index; COPD, chronic obstructive pulmonary disease; ERAS, enhanced recovery after surgery; MI, myocardial infarction; TRAS, traditional recovery after surgery. * Values are mean (SD), median, Q1 (25th) and Q3 (75th) percentiles, or no. of patients (%).

or flaps (percentage) for discrete variables. Discrete variables were analyzed using either the chi-square or Fisher’s exact test as appropriate, and continuous variables were analyzed using the two-sample t-test or Wilcoxon rank sum test. Clinical variables that may affect costs were identified by the senior author (V.L.) and were analyzed for a cost difference between the ERAS and TRAS groups. Categorical variables were tested using the chi-square test. Continuous variables were tested using the Kruskal–Wallis test. Statistically different variables were controlled for in regression analysis of cost variables. A generalized linear model (GLM) with gamma distribution and log-link was used to account for the highly skewed nature of healthcare cost data.11 To investigate predictors of costs, exponentiated coefficients of the GLM model are reported. Statistical differences in incremental costs were determined using 95% confidence intervals. P < 0.05 was considered statistically significant. SAS statistical software version 9.4 was used for all analyses (SAS Institute, Inc.).

Results Two hundred patients were included in the study: 82 in the ERAS cohort and 118 in the TRAS cohort. No significant difference was found between ERAS and TRAS groups with regard

to patient demographics or comorbidities (Table 1). There were a greater number of bilateral reconstructions performed in the ERAS group (12.2% vs. 27.1%, p = 0.01) (Table 2). Statistically different intraoperative variables included length of surgery, use of liposomal bupivacaine, and intraoperative fluid volume. There were no significant differences between the two groups with respect to 30-day postoperative complications (Table 3). Postoperative outcomes in this expanded patient cohort were found to be analogous to the prior reported results confirming decreased hospital LOS and opioid use.5 Clinical variables that were identified to potentially affect costs are listed in Table 4. The only flap variable analyzed was bilateral versus unilateral reconstruction. Postoperative clinical variables analyzed included hospital readmission, flap loss, reoperation within 30 days, and the need for postoperative blood transfusion. Bilateral versus unilateral reconstruction and postoperative blood transfusion were the only two clinical variables found to have a statistically significant effect on costs (Table 4). The exponentiated coefficients, 95% confidence intervals, and p-values of the GLM model for total and Medicare Part A and B costs are depicted in Table 5. Overall total ERAS cost savings appeared to have materialized through savings in the Medicare Part A (facility services) costs. For total costs, ERAS was associated with an 8% reduction in overall costs (exp(b) = 0.92, 95% CI: 0.85 – 0.99). Bilateral reconstruction (vs. unilateral) was

ARTICLE IN PRESS 4

C. Oh et al. Table 2

Reconstruction and intraoperative data*.

Reconstruction Unilateral Bilateral Operation time (h) (unilateral or bilateral)b Mean (SD) Median Q1, Q3 Unilateral – Operation time (h) N Mean (SD) Median Q1, Q3 Bilateral – Operation time (h) N Mean (SD) Median Q1, Q3 Intraoperative Local Analgesia Liposomal bupivacaine Pain pump catheter Intraoperative IV fluid (mL) Intraoperative Transfusion

ERAS (N = 82)

TRAS (N = 118)

Total (N = 200)

10 (12.2%) 72 (87.8%)

32 (27.1%) 86 (72.9%)

42 (21.0%) 158 (79.0%)

10.7 (2.9) 10.7 8.5, 12.5

11.8 (2.9) 11.7 9.7, 13.5

11.3 (2.9) 11.4 9.3, 12.9

10 7.3 (2.5) 6.5 5.5, 8.5

32 9.9 (2.6) 9.4 8.2, 12.0

42 9.3 (2.8) 9.2 7.7, 10.8

72 11.1 (2.6) 11.2 9.1, 12.7

86 12.5 (2.6) 12.3 10.7, 14.1

158 11.9 (2.7) 11.7 10.0, 13.4

71 (86.6%) 0 (0.0%) 2698.7 (3737.0) 5 (6.1%)

6 (5.2%) 12 (10.3%) 4489.8 (4174.6) 8 (6.8%)

77 (38.9%) 12 (6.1%) 3755.4 (4087.8) 13 (6.5%)

p value 0.011

0.007

0.011

0.002

<0.001

<0.001 0.85

Abbreviations: ERAS, enhanced recovery after surgery; TRAS, traditional recovery after surgery. * Values are mean (SD), median, Q1 (25th) and Q3 (75th) percentiles, or no. of patients (%). b Operation time is defined as the time from skin incision to skin closure.

Table 3

Thirty-day postoperative complications*.

Hospital readmission Flap loss within 30 days Partial flap loss Total flap Loss Breast hematoma/seroma Breast cellulitis Mastectomy flap necrosis Deep vein thrombosis Pulmonary embolus Pneumonia treated with antibiotics Cardiac complications (a fib, MI, CHF)

ERAS (N = 82)

TRAS (N = 118)

Total (N = 200)

p value

15 (18.3%)

11 (9.3%)

26 (13.0%)

0.06 0.23

3 (3.7%) 2 (2.4%) 7 (8.5%) 11 (13.4%) 2 (2.4%) 0 (0.0%) 0 (0.0%) 1 (1.2%) 0 (0.0%)

1 (0.8%) 1 (0.8%) 10 (8.5%) 8 (6.8%) 7 (6.0%) 1 (0.9%) 3 (2.6%) 2 (1.7%) 2 (1.7%)

4 (2.0%) 3 (1.5%) 17 (8.5%) 19 (9.5%) 9 (4.5%) 1 (0.5%) 3 (1.5%) 3 (1.5%) 2 (1.0%)

0.99 0.11 0.31 1.00 0.27 1.00 0.51

Abbreviations: A fib, atrial fibrillation; CHF, congestive heart failure; ERAS, enhanced recovery after surgery; MI, myocardial infarct; TRAS, traditional recovery after surgery. * Values are no. of patients (%).

associated with 22% higher costs (exp(b) = 1.22, 95% CI: 1.11 – 1.35). Postoperative blood transfusion was found to be associated with 52% higher costs (exp(b) = 1.52; 95% CI: 1.34 – 1.72). Body mass index was not a predictor of higher total costs or Part A or B costs. Mean predicted total costs and the associated 95% confidence intervals of the ERAS and TRAS groups are presented in Table 6. Mean predicted costs of the ERAS group were $4,576 lesser than those of the TRAS group ($38,688 vs. $43,264). Stratifying by hospital costs versus physician services showed differing results. Physician services were actually found to be greater on average for the ERAS group by $842 ($10,686 vs. $9,844).

The mean predicted total costs and cost differences with associated 95% confidence intervals were then categorized according to BETOS components, as shown in Table 7. BETOS categories are used to analyze Medicare costs, with six separate categories including evaluation/management, procedures, imaging, tests, durable medical equipment, and other. The “other” category includes costs for room, board, and pharmacy. The BETOS coding system was initially developed for analyzing growth in Medicare expenditures.9 In this study, the ERAS group was found to consistently have lower costs than controls in all categories, with the greatest difference associated with the room and board ($12,266 vs. $8,395).

ARTICLE IN PRESS Enhanced recovery cost analysis Table 4

5

Clinical variables that may affect costs*.

BMI Mean (SD) Median Q1, Q3 Tobacco Use Never Former smoker Current smoker COPD or Asthma Prior MI/valve replacement Hypertension Diabetes Mellitus Chronic Kidney Disease Chest Wall Radiation Bilateral Reconstruction 30-Day Hospital Readmission Flap Loss Partial flap loss Total flap loss Reoperation within 30 days Postoperative blood transfusion

ERAS (N = 82)

TRAS (N = 118)

Total (N = 200)

28 (5) 28 25, 31

30 (5) 29 26, 33

61 (74.4%) 19 (23.2%) 2 (2.4%) 3 (3.7%) 4 (4.9%) 12 (14.6%) 2 (2.4%) 0 (0.0%) 46 (56.1%) 72 (87.8%) 15 (18.3%)

74 (62.7%) 42 (35.6%) 2 (1.7%) 9 (7.6%) 3 (2.5%) 24 (20.3%) 7 (5.9%) 2 (1.7%) 65 (55.1%) 86 (72.9%) 11 (9.3%)

135 (67.5%) 61 (30.5%) 4 (2.0%) 12 (6.0%) 7 (3.5%) 36 (18.0%) 9 (4.5%) 2 (1.0%) 111 (55.5%) 158 (79.0%) 26 (13.0%)

3 (3.7%) 2 (2.4%) 14 (17.1%) 4 (4.9%)

1 (0.8%) 1 (0.8%) 17 (14.4%) 17 (14.4%)

4 (2.0%) 3 (1.5%) 31 (15.5%) 21 (10.5%)

p value 0.03

29 (5) 29 25, 32 0.17

0.25 0.38 0.30 0.24 0.24 0.89 0.04 0.06 0.24

0.61 0.03

Abbreviations: BMI, body mass index; COPD, chronic obstructive pulmonary disease; ERAS, enhanced recovery after surgery; MI, myocardial infarction; TRAS, traditional recovery after surgery. * Values are mean (SD), median, Q1 (25th) and Q3 (75th) percentiles, or no. of patients (%).

Table 5 Exponentiated coefficients associated with costs for Medicare Part A and Part B services. Covariate

Exponentiated Coefficient

Overall Costs ERAS 0.919 BMI 1.006 Bilateral 1.222 Postoperative 1.517 Transfusion Medicare Part A Costs ERAS 0.872 BMI 1.007 Bilateral 1.174 Postoperative 1.604 Transfusion Medicare Part B Costs ERAS 1.071 BMI 1.005 Bilateral 1.385 Postoperative 1.254 Transfusion

95% CI

p-value

0.848 – 0.997 0.998 – 1.015 1.108 – 1.348 1.336 – 1.722

0.04 0.13 <0.001 <0.001

0.792 – 0.959 0.997 – 1.017 1.046 – 1.318 1.381 – 1.862

<0.01 0.17 <0.01 <0.001

0.978 – 1.172 0.996 – 1.014 1.242 – 1.545 1.090 – 1.444

0.14 0.30 <0.001 <0.01

Abbreviations: BMI, body mass index; ERAS, enhanced recovery after surgery; TRAS, traditional recovery after surgery.

Discussion ERAS pathways are being adopted by a vast array of surgical specialities. 3,12 Kehlet described how postoperative

complications occur regardless of surgical skill and advances in anesthesia and developed multimodal interventions to address perioperative stress physiology.2 This study cohort was the first to publish outcomes of an ERAS pathway in plastic and reconstructive surgery, specifically microvascular breast reconstruction.5 This study group reported a decrease in LOS from 5.5 days to 3.9 days with no significant increase in complication rates or hospital readmissions.5 ERAS patients had significantly decreased opioid use, had lower pain scores at 24 hours postoperatively and ambulated earlier, which was confirmed in this analysis, which doubled the sample size from the initial study.5 Of note, the TRAS cohort was found to have longer operative times regardless of unilateral versus bilateral procedures when compared to the ERAS group (unilateral 9.9 h v 7.3 h; p = 0.01 and bilateral 12.5 h v 11.1; p = 0.002), which in turn can be a component of the greater intraoperative IV fluid volumes in the TRAS group as well. The change in operative time can be attributed to faster surgeon operative times in addition to a possible improvement in preoperative efficiency given the specific protocol dictating the flow of the case. Longer operative times can potentially lead to increased administering of fluids and hence longer time to postoperative recovery and higher costs attributable to longer LOS. Although there is a focus on decreasing LOS as a costsaving outcome measure, one study showed that the incremental costs incurred by patients on their last full inpatient day was only $420 per day on average, or 2.4% of the $17,734 mean total cost of stay.13 Nursing represents a majority of end-of-stay costs, but another analysis reported a cost analysis showing that a majority of cost of hospital service is associated with buildings, equipment, salaried labor, and

ARTICLE IN PRESS 6

C. Oh et al. Table 6

Mean predicted costs and cost difference of ERAS and TRAS*.

Cost Variable

Mean Predicted Costs for TRAS (95% CI)

Mean Predicted Costs for ERAS (95% CI)

Mean Predicted Cost Difference (95% CI)

Total Costs Medicare Part A (Hospital Costs) Medicare Part B (Physician Services)

$43,264 ($41,611 to $44,889) $33,439 ($32,052 to $34,810)

$38,688 ($37,664 to $39,994) $27,974 ($27,164 to $29,005)

−$4,576 (−$6,411 to −$2,536) −$5,465 (−$7,014 to −$3,765)

$9,844 ($9,540 to $10,150)

$10,686 ($10,423 to $10,957)

$842 ($484 to $1,267)

Abbreviations: ERAS, enhanced recovery after surgery; TRAS, traditional recovery after surgery. * All cost predictions controlled for ERAS/TRAS, BMI, bilateral/unilateral, and postoperative blood transfusion.

Table 7

BETOS analysis of mean predicted costs and cost difference of ERAS and TRAS *.

Cost Variable

Mean Predicted Costs for TRAS (95% CI)

Evaluation and Management Procedure Imaging Tests Other Pharmacyb Room and Boardb

$513 ($480 to $599) $17,787 ($17,188 to $18,354) $488 ($470 to $500) $2,287 ($2,186 to $2,390) $19,291 ($18,415 to $20,174) $3,264 ($3,073 to $3,463) $12,266 ($11,709 to $12,844)

Mean Predicted Costs for ERAS (95% CI) $279 ($259 to $329) $17,642 ($17,210 to $18,101) $465 ($452 to $486) $1,385 ($1,340 to $1,437) $14,101 ($13,664 to $14,671) $2,412 ($2,319 to $2,538) $8,395 ($8,147 to $8,734)

Mean Predicted Cost Difference (95% CI) −$234 (−$319 to −$176) −$145 (−$890 to $614) −$23 (−$39 to $8) −$902 (−$1,015 to −$785) −$5,190 (−$6,138 to −$4,153) −$852 (−$1,063 to −$620) −$3,871 (−$4,474 to −$3,202)

Abbreviations: BETOS, Berenson-Eggers Type of Service; ERAS, enhanced recovery after surgery; TRAS, traditional recovery after surgery. * All cost predictions controlled for ERAS/TRAS, BMI, bilateral/unilateral, and postoperative blood transfusion. b Denotes subcategories of the BETOS “Other” category.

overheads, which are fixed costs in the short term.14 This emphasizes the focus of the ERAS pathway not in decreasing hospital LOS but in enhancing overall perioperative recovery. Opioid use is a major factor in the nature of perioperative recovery. Celecoxib and gabapentin afford significant reductions in postoperative opioid use with faster recovery of bowel function and earlier mobilization.15,16 Oral opioids were available as necessary for breakthrough pain with the addition of parenteral opioids as a last resort. The use of transversus abdominis plane blocks has been shown to decrease opioid consumption in the postoperative period.17 Decreased opioid may allow for earlier patient mobility and indirectly contribute to a decreased LOS. Owing to the manner in which the costing algorithm is designed and the fact that the patients in the control and treatment groups span different years, an alternative approach was used to evaluate the data. Typically, hospital services are evaluated with department-level cost-tocharge ratios and physician services with Medicare reimbursement rates. This approach uses year-specific values for both the cost-to-charge ratios, as well as Medicare reimbursement rates, which can vary year to year. As the TRAS patients are in different years than ERAS patients, there is the potential that a cost difference is driven by changes in the costto-charge ratios or Medicare reimbursement rates and not from the differing protocols. For this reason, all values were based on 2014 data. However, this is not without its own limitations, namely, services that were rendered in the very early part of the study period may no longer have the same CPT codes in 2014. Thus, costing prior years’ utilizations using the most recent year prices becomes challenging when the study period is very long. Consequently, this study

restricted the analysis to patients in 2010 and onward to account for costing variations. The adjusted mean costs of ERAS patients were $4,576 lesser than those of the historical TRAS control cohort. Hospital costs were much higher in the TRAS group and can be primarily attributable to the ICU stay. The TRAS pathway involved admission to a closed ICU under the care of the critical care team. Free flap reconstruction dictates a critical need for close monitoring, especially during the first 24 to 48 hours when most thrombotic complications occur, but it may result in ICU overutilization.18 Head and neck reconstruction requires close airway monitoring and warrants an ICU stay,19 but breast reconstruction can be safely monitored with trained nursing to check the flap on an every 1-to-2-hour basis regardless of the setting.18 A study of surveys collected from 26 plastic surgery centers performing microvascular free flap reconstruction concluded that 45% of responding centers had patients admitted to the ICU owing to a lack of adequately trained nurses.20 This equated to a comparative increase in cost of $2878 to $3345 per day for ICU care.20 This supports the benefits of the ERAS protocol by incorporating improved nursing training, as well as the use of standardized postoperative protocols to reduce costs and improve efficiency. A major limitation of this study is that the costs attributable to the ICU stay were not separated from the other protocol changes associated with ERAS. Patients with TRAS were admitted to the ICU on postoperative day 0 and transferred to the general surgical floor on postoperative day 1. The collected data do not identify ICU-specific costs because the line-item list is not entered chronologically, thus deeming a logistic regression model not possible. ICU costs are incorporated into Medicare Part A and B costs as

ARTICLE IN PRESS Enhanced recovery cost analysis well as throughout the BETOS categories thus cannot be identified separately. In an ideal study design, we would have aimed for a prospective randomized controlled trial. This study was conducted with such practice changes incorporated in the most optimal design possible to control for the ICU variable but remains a weakness of the cost analysis. A patient variable found to have a significant effect on total costs was the need for postoperative blood transfusion. Seventeen of the 118 patients (14.4%) in the TRAS cohort received postoperative blood transfusions compared to 4 of 82 patients (4.9%) in the ERAS cohort. Red blood cell transfusion has been independently associated with significantly higher hospital cost and patient morbidity.21 Increased blood loss and use of perioperative blood transfusion correlates with an increased risk for infections, primarily because of the content of leukocytes and noncellular transfusion components.22,23 The ERAS pathway resulted in fewer postoperative blood transfusions. One possibility is a lower threshold for blood transfusion by ICU intensivists for the TRAS patient cohort. Another possibility is the change in intraoperative fluid management. As patients with ERAS had an aim of strict euvolemic fluid balance, hemoglobin levels postoperatively may have been closer to true physiology without hemodilution. There was no difference in intraoperative transfusion rates between the ERAS and TRAS groups.5 Although BMI was not found to have a statistically significant effect on overall total costs, part A (facility) costs, or part B (professional) costs, it was found to have a statistically significant effect on pharmacy costs. This can likely be explained by the higher absolute dosing of medications required for patients with a higher BMI. Overall physician fees were higher for the ERAS group. This could be explained by the greater number of bilateral breast reconstructions for the ERAS group. The TRAS group incorporates an additional compensation for ICU intensivists. Anesthesiologist fees would be expected to be equivalent between the two groups. Bilateral versus unilateral procedure was found to have a statistically significant effect on cost. This finding does not require much explanation on why costs are higher given longer operating room times, as well as higher direct variable costs. In the ERAS group, there is also a greater proportion of deep inferior epigastric perforator (DIEP) and muscle-sparing transverse rectus abdominis (TRAM) flaps, representing the trend toward perforatorbased flaps with muscle preservation. A 1999 study by Allen et al. published a cost-based comparison between free TRAM and DIEP flaps, concluding twice the cost of TRAM flaps due to longer recovery and LOS compared to DIEP reconstruction.24 There may be a more significant donor-site morbidity with TRAM flap patients (pain and abdominal wall manipulation), which attributes to a much longer LOS.24 For immediate reconstruction, full TRAM flaps accounted for 1% and 2% of cases in ERAS and TRAS cohorts, respectively.5 For delayed reconstruction, full TRAM flaps accounted for 3% and 7% in ERAS and TRAS cohorts, respectively, and a statistically significant difference was found between them.5 There was a slightly higher incidence of full TRAM flaps in the TRAS group, which may have contributed to the higher total costs of the TRAS group. This finding was not considered a significant variable in affecting costs because a majority of the flaps in this study were perforator or muscle sparing with minimal full TRAM flaps in both groups.

7 There were no significant differences between the two groups with respect to 30-day postoperative complications. A greater number of hospital readmissions were noted in the ERAS group (18.3% vs. 9.3% p = 0.06), but most cases were overnight hospitalizations for breast cellulitis initially requiring antibiotics administered IV in the hospital. This correlates with the higher percentage of breast cellulitis cases in the ERAS group (13.4% vs. 6.8%; p = 0.11). It may be possible that the difference in hospital readmission may be a result of earlier hospital discharge in the ERAS group; but given that most cases were of cellulitis, treated initially with IV antibiotics and transitioned to oral antibiotics, and were categorized as a minor complication, it could be argued that an increased hospital LOS can increase the risk of major complications such as deep venous thrombosis (DVT) and pulmonary embolus (PE). The present cost analysis was based on index hospitalization only and thus does not include costs associated with subsequent readmission, which may be considered a weakness of this study. A limitation in the study to be mentioned includes the clinical characteristics determined to affect costs. The clinical variables studied were generalized categories, but each independent variable was not analyzed for a cost difference. It was assumed that certain data points collected, such as recipient vessel selected or number of ambulations per patient, would likely not reveal a statistically significant difference in costs, yet it is possible that such variables may influence cost.

Conclusion Autologous reconstruction is known to provide greater longterm patient satisfaction for the patient compared to implantbased reconstruction, yet it has the least favorable reimbursement-to-operative time ratio.25–29 Implementation of the ERAS pathway can decrease patient morbidity and deliver fiscally responsible care in an evolving healthcare environment in breast reconstruction.

Conflict of interest None.

Funding None.

References 1. McDonald R. Enhanced recovery clinical education programme improves quality of post-operative care. BMJ Qual Improv Rep 2015;4(1). 2. Kehlet H. Multimodal approach to control postoperative pathophysiology and rehabilitation. Br J Anaesth 1997;78(5):606–17. 3. Kehlet H. Multimodal approach to postoperative recovery. Curr Opin Crit Care 2009;15(4):355–8. 4. Davidge KM, Brown M, Morgan P, Semple JL. Processes of care in autogenous breast reconstruction with pedicled TRAM flaps: expediting postoperative discharge in an ambulatory setting. Plast Reconstr Surg 2013;132(3):339e–44e.

ARTICLE IN PRESS 8 5. Batdorf NJ, Lemaine V, Lovely JK, et al. Enhanced recovery after surgery in microvascular breast reconstruction. J Plast Reconstr Aesthet Surg 2015;68(3):395–402. 6. Visscher SL, Naessens JM, Yawn BP, Reinalda MS, Anderson SS, Borah BJ. Developing a standardized healthcare cost data warehouse. BMC Health Serv Res 2017;17(1):396. 7. Suri RM, Thompson JE, Burkhart HM, et al. Improving affordability through innovation in the surgical treatment of mitral valve disease. Mayo Clin Proc 2013;88(10):1075–84. 8. Health Economics Resource Center. Washington DDoVA, Veterans Health Administration. How do I adjust for the effects of inflation? 2013 [Accessed 11 January 2016]. 9. Agency for Healthcare Research and Quality. Using appropriate price indices for analyses of health care expenditures or income across multiple years (MEPS: Medical Expenditure Panel Survey). 2014. http://meps.ahrq.gov/mepsweb/about_meps/Price _Index.shtml. [Accessed 11 January 2016]. 10. Berenson-Eggers Type of Service (BETOS). Centers for Medicare & Medicaid Services. 2016. https://www.cms.gov/Research -Statistics-Data-and-Systems/Statistics-Trends-and-Reports/ MedicareFeeforSvcPartsAB/downloads/betosdesccodes.pdf [Accessed 11 January 2016]. 11. Manning WG, Basu A, Mullahy J. Generalized modeling approaches to risk adjustment of skewed outcomes data. J Health Econ 2005;24(3):465–88. 12. Adamina M, Kehlet H, Tomlinson GA, Senagore AJ, Delaney CP. Enhanced recovery pathways optimize health outcomes and resource utilization: a meta-analysis of randomized controlled trials in colorectal surgery. Surgery 2011;149(6):830–40. 13. Taheri PA, Butz DA, Greenfield LJ. Length of stay has minimal impact on the cost of hospital admission. J Am Coll Surg 2000; 191(2):123–30. 14. Roberts RR, Frutos PW, Ciavarella GG, et al. Distribution of variable vs fixed costs of hospital care. JAMA 1999;281(7):644–9. 15. Sun T, Sacan O, White PF, Coleman J, Rohrich RJ, Kenkel JM. Perioperative versus postoperative celecoxib on patient outcomes after major plastic surgery procedures. Anesth Analg 2008;106(3):950–8. table of contents. 16. Parsa AA, Sprouse-Blum AS, Jackowe DJ, Lee M, Oyama J, Parsa FD. Combined preoperative use of celecoxib and gabapentin in the management of postoperative pain. Aesthetic Plast Surg 2009;33(1):98–103. 17. Zhong T, Ojha M, Bagher S, et al. Transversus abdominis plane block reduces morphine consumption in the early postoperative

C. Oh et al.

18.

19.

20.

21.

22.

23. 24.

25.

26.

27.

28.

29.

period following microsurgical abdominal tissue breast reconstruction: a double-blind, placebo-controlled, randomized trial. Plast Reconstr Surg 2014;134(5):870–8. Cornejo A, Ivatury S, Crane CN, Myers JG, Wang HT. Analysis of free flap complications and utilization of intensive care unit monitoring. J Reconstr Microsurg 2013;29(7):473–9. Bhama PK, Davis GE, Bhrany AD, Lam DJ, Futran ND. The effects of intensive care unit staffing on patient outcomes following microvascular free flap reconstruction of the head and neck: a pilot study. JAMA Otolaryngol Head Neck Surg 2013;139(1): 37–42. Haddock NT, Gobble RM, Levine JP. More consistent postoperative care and monitoring can reduce costs following microvascular free flap reconstruction. J Reconstr Microsurg 2010;26(7): 435–9. Trentino KM, Farmer SL, Swain SG, et al. Increased hospital costs associated with red blood cell transfusion. Transfusion 2015;55(5):1082–9. Nielsen HJ, Reimert CM, Pedersen AN, et al. Time-dependent, spontaneous release of white cell- and platelet-derived bioactive substances from stored human blood. Transfusion 1996;36(11– 12):960–5. Nielsen HJ. Detrimental effects of perioperative blood transfusion. Br J Surg 1995;82(5):582–7. Kaplan JL, Allen RJ. Cost-based comparison between perforator flaps and TRAM flaps for breast reconstruction. Plast Reconstr Surg 2000;105(3):943–8. Patel A, Clune JE, Forte A, Chang CC. The impact of the Medicare sustainable growth rate formula on reconstructive plastic surgery. Plast Reconstr Surg 2010;126(5):270e–1e. Sando IC, Chung KC, Kidwell KM, Kozlow JH, Malay S, Momoh AO. Comprehensive breast reconstruction in an academic surgical practice: an evaluation of the financial impact. Plast Reconstr Surg 2014;134(6):1131–9. Alderman AK, Storey AF, Nair NS, Chung KC. Financial impact of breast reconstruction on an academic surgical practice. Plast Reconstr Surg 2009;123(5):1408–13. Yueh JH, Slavin SA, Adesiyun T, et al. Patient satisfaction in postmastectomy breast reconstruction: a comparative evaluation of DIEP, TRAM, latissimus flap, and implant techniques. Plast Reconstr Surg 2010;125(6):1585–95. Hu ES, Pusic AL, Waljee JF, et al. Patient-reported aesthetic satisfaction with breast reconstruction during the long-term survivorship Period. Plast Reconstr Surg 2009;124(1):1–8.