Original Investigation | Gastroenterology and Hepatology

Cost-effectiveness Analysis of Bariatric Surgery for PatientsWith Nonalcoholic Steatohepatitis CirrhosisMatthew J. Klebanoff, BA; Kathleen E. Corey, MD, MPH, MMSc; Sumeyye Samur, PhD; Jin G. Choi, BS; Lee M. Kaplan, MD, PhD;Jagpreet Chhatwal, PhD; Chin Hur, MD, MPH

Abstract

IMPORTANCE Obesity is the most common risk factor for nonalcoholic steatohepatitis (NASH), theprogressive form of nonalcoholic fatty liver disease that can lead to cirrhosis and hepatocellularcarcinoma. Weight loss can be an effective treatment for obesity and may slow the progression ofadvanced liver disease.

OBJECTIVE To assess the cost-effectiveness of bariatric surgery in patients with NASH andcompensated cirrhosis.

DESIGN, SETTING, AND PARTICIPANTS This economic evaluation study used a Markov-basedstate-transition model to simulate the benefits and risks of laparoscopic sleeve gastrectomy (SG),laparoscopic Roux-en-Y gastric bypass (GB), and intensive lifestyle intervention (ILI) compared withusual care in patients with NASH and compensated cirrhosis and varying baseline weight(overweight, mild obesity, moderate obesity, and severe obesity). Patients faced varied risks ofperioperative mortality and complications depending on the type of surgery they underwent. Datawere collected on March 22, 2017.

MAIN OUTCOMES AND MEASURES Life-years, quality-adjusted life-years (QALYs), costs (in 2017$US), and incremental cost-effectiveness ratios (ICERs) were calculated.

RESULTS Demographic characteristics of the patient population were based on a previouslypublished prospective study (n = 161). Patients in the model were 41.0% female, and the base caseage was 54 years. Compared with usual care, SG was associated with an increase in QALYs of 0.263 to1.180 (bounds of ranges represent overweight to severe obesity); GB, 0.263 to 1.207; and ILI, 0.004to 0.216. Sleeve gastrectomy was also associated with an increase in life-years of 0.693 to 1.930; GB,0.694 to 1.947; and ILI, 0.012 to 0.114. With usual care, expected life-years in overweight, mildobesity, moderate obesity, and severe obesity were 12.939, 11.949, 10.976, and 10.095, respectively.With usual care, QALY in overweight was 6.418; mild obesity, 5.790; moderate obesity, 5.186; andsevere obesity, 4.577. Sleeve gastrectomy was the most cost-effective option for patients across allweight classes assessed: ICER for SG in patients with overweight was $66 119 per QALY; mild obesity,$18 716 per QALY; moderate obesity, $10 274 per QALY; and severe obesity, $6563 per QALY. Athreshold analysis on the procedure cost of GB found that for GB to be cost-effective, the cost of thesurgery must be decreased from its baseline value of $28 734 by $4889 for mild obesity, by $3189for moderate obesity, and by $2289 for severe obesity. In overweight patients, GB involved fewerQALYs than SG, and thus decreasing the cost of surgery would not result in cost-effectiveness.

CONCLUSIONS AND RELEVANCE Bariatric surgery could be highly cost-effective in patients withNASH compensated cirrhosis and obesity or overweight. The findings from this analysis suggest that

(continued)

Key PointsQuestion Is bariatric surgery cost-

effective in patients with nonalcoholic

steatohepatitis and compensated

cirrhosis?

Findings In this simulation model study,

laparoscopic sleeve gastrectomy had an

incremental cost-effectiveness ratio of

$66 119 per quality-adjusted life-year in

overweight patients, $18 716 per

quality-adjusted life-year in patients

with mild obesity, $10 274 per quality-

adjusted life-year in patients with

moderate obesity, and $6563 per

quality-adjusted life-year in patients

with severe obesity.

Meaning Bariatric surgery could be

highly cost-effective in patients with

nonalcoholic steatohepatitis and

compensated cirrhosis, even in those

with a lower baseline body mass index.

+ Invited Commentary

+ Supplemental content

Author affiliations and article information arelisted at the end of this article.

Open Access. This is an open access article distributed under the terms of the CC-BY License.

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 1/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

Abstract (continued)

it can inform clinical trials evaluating the effect of bariatric procedures in patients with NASHcirrhosis, including those with a lower body mass index.

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047

Introduction

The prevalence of adult obesity in the United States is expected to reach approximately 50% by2030.1 Obesity is the most common risk factor for nonalcoholic fatty liver disease and the progressivesubtype of the disease, nonalcoholic steatohepatitis (NASH), which can give rise to cirrhosis andhepatocellular carcinoma (HCC). In recent prospective studies, more than 70% of patients withcompensated cirrhosis were overweight or obese.2 In addition to contributing to the development ofadvanced liver disease, obesity also leads to worse outcomes among patients with compensatedcirrhosis. Compared with their normal-weight counterparts, obese patients with compensatedcirrhosis face a nearly 3-fold increased risk of decompensation.2

Weight loss can be an effective treatment for NASH.3 In a recent 12-month trial of lifestyleinterventions, fibrosis improved or stabilized in patients with NASH who lost at least 5% of total bodyweight.4 Weight loss through diet and exercise also decreases portal pressure in patients with NASHand compensated cirrhosis, which likely reduces the likelihood of decompensation.5 In addition,weight loss may also improve the likelihood of receiving a liver transplant for patients on the waitlist.6 Although diet and exercise may improve outcomes in patients with NASH, long-termmaintenance of weight loss is often inadequate. In a trial of lifestyle intervention in patients withNASH, only 50% achieved 7% total body weight loss after 1 year.4

In contrast with lifestyle interventions, bariatric surgery often induces excellent long-termweight loss outcomes and may also have the potential to halt or reverse liver damage in cirrhosis.7,8

However, bariatric surgery carries heightened perioperative risks in patients with cirrhosis. Ananalysis of patients undergoing bariatric surgery in the Nationwide Inpatient Sample showed thatmortality was higher in those with compensated (0.9%) and decompensated cirrhosis (16.3%) thanin those without cirrhosis (0.3%).9 In addition, the cost of bariatric surgery may pose a potentialbarrier for some patients; payers typically deny coverage of these procedures to individuals withbody mass index (BMI; calculated as weight in kilograms divided by height in meters squared) of lessthan 35.0, making surgery inaccessible to many who might benefit from it.

Despite the benefits of weight loss in NASH cirrhosis and the increasing burden of thiscondition, no randomized clinical trial has been performed, to date, to assess the effect of bariatricsurgery and nonsurgical weight loss interventions on the progression of advanced liver disease.Under these circumstances, mathematical modeling provides a platform to integrate the bestavailable data to help guide medical decision making. The aim of our study was to perform acomparative cost-effectiveness analysis of bariatric surgery in patients with NASH cirrhosis who wereoverweight or obese.

Methods

Model StructureThis report follows the Consolidated Health Economic Evaluation Reporting Standards (CHEERS)reporting guideline for economic evaluations. We developed a Markov-based state-transition modelusing TreeAge Pro software (version 2017; TreeAge) to assess cost-effectiveness of the following 4different weight loss strategies: usual care, intensive lifestyle intervention (ILI), laparoscopic sleevegastrectomy (SG), and laparoscopic Roux-en-Y gastric bypass (GB) (Figure). Institutional review boardapproval was not required for this study that did not use human participants.

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 2/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

The baseline patient characteristics of the simulated population were based on a previouslypublished prospective study2 (n = 161). Patients in the model were 41.0% female, and the base caseage was 54 years. Note that this population represents patients with cirrhosis and thus includes ahigher proportion of older and male individuals than the overall population of patients undergoingbariatric surgery. Our analysis was conducted for patients with overweight, mild obesity, moderateobesity, and severe obesity. The model cycle length, or time between state transitions, was 1 year.Each year, patients could remain in the same health state, progress to another, or die. Patients withcompensated cirrhosis could progress to the decompensated cirrhosis state. Based on an analysisthat associated obesity with clinical decompensation,2 we adjusted the probability ofdecompensation according to patients’ BMI. Patients with compensated or decompensated cirrhosiscould progress to HCC.10-14 In the model, liver transplant was a possibility only for individuals withdecompensated cirrhosis or HCC (Figure).15-18 After transplant, these patients faced a risk of deathbased on mortality rates among transplant recipients.19 Possible causes of death included surgicalmortality, liver-related mortality, and background mortality, which was based on patients’ age, sex,and BMI, using data from the US Third National Health and Nutritional Examination Survey.9,20-23

Patients with compensated and decompensated cirrhosis had excess liver-related mortality of 2.1%and 13.0%, respectively.11,12,22 Simulated patients were followed up until death, to capture the long-term benefits of surgery.

Competing StrategiesIn the usual-care strategy, patients did not undergo surgery or other weight loss interventions, andtheir liver disease progressed according to probabilities derived from the literature. We assumedthat, without intervention, patients’ weight remained constant during their lifetime.

In SG and GB strategies, patients faced a risk of 30-day surgical mortality after the operation,which varied depending on the type of procedure, based on the American College of SurgeonsNational Quality Improvement Survey.23 They also encountered the possibility of major and minorcomplications after surgery.23 Major complications required a second operation; minor complicationsrequired readmission but no additional operation. The estimates from the American College ofSurgeons National Quality Improvement Survey were adjusted for our population using data from astudy that compared mortality in patients with compensated cirrhosis and patients withoutcirrhosis.9

After surgery, patients underwent weight loss, which could slow the progression of liver diseasein our model by decreasing the probability of decompensation. We incorporated the association ofweight loss with survival and quality of life for patients in the following 4 weight classes: overweight(BMI, 25.0-29.9), mild obesity (BMI, 30.0-34.9), moderate obesity (BMI, 35.0-39.9), and severeobesity (BMI, �40.0). Postsurgical weight loss was derived from a recent meta-analysis.24 Weightloss in the ILI strategy was derived from the Look AHEAD (Action for Health in Diabetes) trial.25 Long-term trends in weight were extrapolated using data from the Swedish Obese Subjects study.26 We

Figure. Simplified Model Schematic

Patient characteristicsTransition probabilitiesQuality-of-life weightsTreatment efficacySurgical complicationsand mortality

Life yearsQALYs

Costs and ICER

Incidence of clinicalevents (decompensationor transplant)

Hepatocellularcarcinoma

Compensatedcirrhosis

Inputs Outputs

Backgroundmortality

Liver-relateddeath

Decompensatedcirrhosis

Livertransplant

The dashed line between the compensated cirrhosisand decompensated cirrhosis states indicates that theprobability of decompensating varies according tobody mass index and thus decreases with weight loss.ICER indicates incremental cost-effectiveness ratio;QALYs, quality-adjusted life-years.

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 3/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

applied our estimates of percentage of excess weight loss (eTable 1 in the Supplement) to all of theinitial weight categories in our analysis.

Costs and Quality-of-Life AdjustmentsCosts associated with treatment strategies, surgical complications, and health states were derivedfrom the literature.27-32 The initial cost of GB was based on a prior analysis,27 and the cost of SG wascalculated using cost data from the Healthcare Cost and Utilization Project Nationwide InpatientSample.28 All costs from prior years were converted to 2017 $US using the Consumer Price Index.33

Costs were evaluated from the perspective of a third-party payer.Utility estimates were derived from studies on cirrhosis due to other causes because no quality-

of-life studies for NASH cirrhosis were available to inform our model.34 This approach has been usedin a previous cost-effectiveness analysis studying NASH.10 All patients who underwent surgery had adecrement in their quality of life, which was applied for 6 weeks. An additional quality-of-lifedecrement was applied for 6 weeks to patients who experienced major complications, and a 4-weekdecrement was applied after mild complications. Complications also incurred additional costs.27 Weadjusted quality of life based on patients’ age, sex, and weight class (eTable 2 in the Supplement).Costs and utilities were discounted at an annual rate of 3%.35

OutcomesCompared with usual care, we estimated the gain in life-years, quality-adjusted life-years (QALYs),and total costs for patients undergoing ILI, SG, and GB. We assessed cost-effectiveness for eachtreatment strategy by estimating incremental cost-effectiveness ratios (ICERs) for patients withoverweight, mild obesity, moderate obesity, and severe obesity. The ICERs were calculated on anefficiency frontier. The willingness-to-pay threshold to determine cost-effectiveness was $100 000per QALY.

Sensitivity AnalysisData were collected on March 22, 2017. We conducted 1-way sensitivity analyses to examine theassociation of model parameter uncertainty on the cost-effectiveness results for patients in eachweight class. One-way sensitivity analyses were performed for the cost-effective strategy in eachweight class (ie, SG). We also performed probabilistic sensitivity analysis (PSA) to assess how modelparameter uncertainty could affect our results. In PSA, all parameter values were variedsimultaneously, using statistical distributions derived from the literature (Table 1). We conducted PSAusing second-order sampling for 10 000 iterations for each weight class.

Results

Base-Case ResultsAll weight loss strategies involved a gain in life-years and QALYs in patients with compensatedcirrhosis (41% female and 59% male; base case age, 54 years). Compared with usual care, SG wasassociated with an increase in QALYs of 0.263 to 1.180 (bounds of ranges represent overweight tosevere obesity); GB, 0.263 to 1.207; and ILI, 0.004 to 0.216. Sleeve gastrectomy was also associatedwith an increase in life-years of 0.693 to 1.930; GB, 0.694 to 1.947; and ILI, 0.012 to 0.114 (Table 2).With usual care, expected life-years in overweight, mild obesity, moderate obesity, and severeobesity were 12.939, 11.949, 10.976, and 10.095, respectively. These results are consistent withprospective studies that found median survival of greater than 12 years in patients with compensatedcirrhosis.36 With usual care, QALY in overweight was 6.418; mild obesity, 5.790; moderate obesity,5.186; and severe obesity, 4.577. For patients with obesity (mild, moderate, and severe), GB involvedthe greatest increase in life-years and QALYs, followed by SG and then ILI. In contrast, for overweightpatients, GB had the largest increase in life-years; however, SG had the largest increase in QALYs,followed very closely by GB and then by ILI. Note that these differences between GB and SG for

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 4/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

Table 1. Model Inputs: Baseline Values, Ranges, and Parameters for Distributions Used in Deterministic and Probabilistic Sensitivity Analyses

Parameter SourceBase Case(Range or Variation) Distribution Parameter 1a Parameter 2b

Age, y Berzigotti et al,2 2011 54 (34.4 to 73.6) NA NA NA

Female, % Berzigotti et al,2 2011 41 (35 to 49) β 59.122 85.078

Annual discount rate, % Weinstein et al,35 1996 3 (0 to 5) β 8.354 270.122

Transition probabilities

HR per 1-U increase in BMI Berzigotti et al,2 2011 1.06 (1.01 to 1.12) Log normal 4.907 86.682

CC to DC (BMI, 27.2), % NA 7.21 (3.21 to 10.77) β 13.0128 167.469

CC/DC to HCC, %10-14 Mahady et al,10 2012; Sanyal et al,11 2006; Ratziuet al,12 2002; Ascha et al,13 2010; Yatsuji et al,14

2009

0.69 (0.50 to 16.8) β 1.03 147.97

CC to liver-related death, % Sanyal et al,11 2006; Hui et al,22 2003 2.1 (2.0 to 4.0) β 4.483 210.003

DC to liver-related death, % Ratziu et al,12 2002 13.0 (10.0 to 38.0) β 0.774 5.178

HCC to liver-related death, % Fattovich et al,21 2002 42.7 (33.0 to 86.0) β 2.14 2.87

DC to transplantation, % Thuluvath et al,15 2010; Davis et al,16 2011 2.3 (1.0 to 6.2) β 1.282 54.473

HCC to transplantation, % Lang et al,17 2009; Saab et al,18 2010 4.0 (0.0 to 14.0) β 0.59 14.16

Posttransplant to liver-related deathin year 1, %

Wolfe et al,19 2010 11.6 (6.0 to 42.0) β 0.38 2.88

Posttransplant to liver-related deathin year 2 or later, %19

Wolfe et al,19 2010 4.4 (2.4 to 11.0) β 1.59 34.51

30-Day mortality for GB, % Mosko and Nguyen,9 2011; Young et al,23 2015 0.3255 (0.10 to 1.00) β 2.000 612.491

30-Day mortality for SG, % Mosko and Nguyen,9 2011; Young et al,23 2015 0.217 (0.041 to 1.14) β 0.596 273.886

Minor complications for GB, % Mosko and Nguyen,9 2011; Young et al,23 2015 7.86 (3.73 to 16.47) β 5.311 62.255

Major complications for GB, % Mosko and Nguyen,9 2011; Young et al,23 2015 5.34 (2.53 to 11.19) β 5.477 97.095

Minor complications for SG, % Mosko and Nguyen,9 2011; Young et al,23 2015 5.32 (2.52 to 11.15) β 5.476 97.449

Major complications for SG, % Mosko and Nguyen,9 2011; Young et al,23 2015 3.47 (1.65 to 7.28) β 5.600 155.785

Health-related quality-of-lifeweights

CC Chhatwal et al,34 2015 0.90 (0.81 to 0.99) Β 37.52 4.17

DC Chhatwal et al,34 2015 0.80 (0.57 to 0.99) β 8.50 2.12

HCC Chhatwal et al,34 2015 0.79 (0.54 to 0.99) β 7.27 1.93

Liver transplant at year 1 Chhatwal et al,34 2015 0.84 (0.77 to 0.93) β 52.70 10.04

Liver transplant at year 2 or later Chhatwal et al,34 2015 0.93 (0.84 to 0.99) β 27.78 2.09

Treatment-related quality-of-lifeweights

Initial surgery Campbell et al,27 2010 –0.22 (–0.24 to –0.20) β 300.96 1067.03

Minor complications Campbell et al,27 2010 –0.11 (–0.12 to –0.10) β 369.31 2988.03

Major complications Campbell et al,27 2010 –0.36 (–0.40 to –0.32) β 258.87 460.21

Weight-related quality-of-lifeweights

BMI <30.0 Campbell et al,27 2010 0.88 (0.79 to 0.97) β 48.22 6.58

BMI 30.0-34.9 Campbell et al,27 2010 0.85 (0.77 to 0.93) β 60.27 10.64

BMI 35.0-39.9 Campbell et al,27 2010 0.82 (0.74 to 0.90) β 72.33 15.88

BMI ≥40.0 Campbell et al,27 2010 0.78 (0.70 to 0.86) β 88.40 24.93

Health state costs, 2017 $US

CC Saab et al,30 2014; Gordon et al,31 2012;McAdam-Marx et al,32 2011

$5886 (±25%) γ 61.466 0.010

DC Saab et al,30 2014; Gordon et al,31 2012;McAdam-Marx et al,32 2011

$41 082 (±25%) γ 61.466 0.001

HCC Saab et al,30 2014; Gordon et al,31 2012;McAdam-Marx et al,32 2011

$90 344 (±25%) γ 61.466 6.804

Liver transplant year 1 Saab et al,30 2014; Gordon et al,31 2012;McAdam-Marx et al,32 2011

$183 279 (±25%) γ 61.466 3.354

Liver transplant year 2 or later Saab et al,30 2014; Gordon et al,31 2012;McAdam-Marx et al,32 2011

$45 107 (±25%) γ 61.466 0.001

(continued)

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 5/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

overweight patients were very small. When rounded to the nearest thousandth, SG and GB bothinvolved an equal number of QALYs, and GB involved 0.001 additional life-years compared with SG.

Cost-effectiveness analysis found that SG was the optimal strategy in patients with NASHcirrhosis in all weight categories. Compared with the next nondominated strategy, the ICER for SGwas $66 119 per QALY in overweight, $18 716 per QALY in mild obesity, $10 274 per QALY in moderate

Table 1. Model Inputs: Baseline Values, Ranges, and Parameters for Distributions Used in Deterministic and Probabilistic Sensitivity Analyses (continued)

Parameter SourceBase Case(Range or Variation) Distribution Parameter 1a Parameter 2b

Treatment costs, 2017 $US

GB Campbell et al,27 2010; AHRQ,28 2014 $28 734 (±25%) γ 61.466 0.002

SG Campbell et al,27 2010; AHRQ,28 2014 $23 660 (±25%) γ 61.466 0.003

Minor complications Campbell et al,27 2010 $1414 (±25%) γ 61.466 0.043

Major complications Campbell et al,27 2010 $46 091 (±25%) γ 61.466 0.001

ILI in year 1 Rushing et al,29 2017 $1410 (±25%) γ 61.466 0.044

ILI in year 2 Rushing et al,29 2017 $1087 (±25%) γ 61.466 0.057

ILI in year 3 Rushing et al,29 2017 $932 (±25%) γ 61.466 0.067

ILI in year 4 Rushing et al,29 2017 $750 (±25%) γ 61.466 0.083

ILI in years 5-8 Rushing et al,29 2017 $564 (±25%) γ 61.466 0.111

Abbreviations: AHRQ, Agency for Healthcare Research and Quality; BMI, body massindex (calculated as weight in kilograms divided by height in meters squared); CC,compensated cirrhosis; DC, decompensated cirrhosis; GB, laparoscopic Roux-en-Ygastric bypass; HCC, hepatocellular carcinoma; HR, hazard ratio; ILI, intensive lifestyleintervention; NA, not applicable; SG, laparoscopic sleeve gastrectomy.

a Corresponds to the α parameter for β distribution, the k (shape) parameter for γdistribution, and μ for log normal distribution.

b Corresponds to the β parameter for β distribution, θ (scale) parameter for γdistribution, and α for log normal distribution.

Table 2. Results of Cost-effectiveness Analysesa

Strategy Cost, 2017 $USIncrementalCost, 2017 $US QALYs

IncrementalQALYs Life-Years

IncrementalLife-Years ICER, $/QALY

Severe obesity (BMI ≥40.0)

Usual care 214 412 NA 4.577 NA 10.095 NA NA

ILI 223 087 934 4.793 −0.964 10.209 −1.815 Absolutely dominated

SG 222 153 7741 5.757 1.179 12.025 1.930 6563

GB 228 369 6216 5.784 0.027 12.042 0.017 229 919

Moderate obesity (BMI 35.0-39.9)

Usual care 206 809 NA 5.186 NA 10.976 NA NA

ILI 214 953 8145 5.259 0.073 11.063 0.087 Extendedly dominated

SG 216 075 1122 6.088 0.829 12.593 1.530 10 274

GB 222 174 6099 6.106 0.019 12.604 0.011 329 002

Mild obesity (BMI 30.0-34.9)

Usual care 197 486 NA 5.790 NA 11.949 NA NA

ILI 205 128 7642 5.809 0.020 12.001 0.052 Extendedly dominated

SG 209 976 4848 6.457 0.648 13.131 1.130 18 716

GB 215 990 6014 6.458 0.002 13.137 0.006 3 667 701

Overweight (BMI 25.0-29.9)

Usual care 186 264 NA 6.418 NA 12.939 NA NA

ILI 193 506 7242 6.422 0.004 12.951 0.012 Extendedly dominated

SG 203 660 10 153 6.681 0.258 13.632 0.681 66 119

GB 209 606 5946 6.681 0 13.633 0.001 Absolutely dominated

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided byheight in meters squared); GB, laparoscopic Roux-en-Y gastric bypass; ICER, incrementalcost-effectiveness ratio; ILI, intensive lifestyle intervention; NA, not applicable; QALYs,quality-adjusted life-years; SG, laparoscopic sleeve gastrectomy.a An extendedly dominated strategy has an ICER that is higher than that of the next most

effective strategy. An absolutely dominated strategy is more expensive and less

effective than other strategies. Note that the ILI strategy has negative incrementalQALYs and life-years for the severe obesity category; as the ILI strategy was absolutelydominated, the incremental QALYs, life-years, and costs for this group are calculatedagainst the SG strategy.

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 6/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

obesity, and $6563 per QALY in severe obesity (Table 2). Although GB involved a greater increase inQALYs than SG in obese patients, GB was also more expensive, and the ICER for GB ($229 919 perQALY for severe, $329 002 per QALY for moderate, and $3 667 701 per QALY for mild obesity)exceeded the commonly accepted willingness-to-pay threshold of $100 000 per QALY.

A threshold analysis on the procedure cost of GB found that for GB to be cost-effective, the costof the surgery must be decreased from its baseline value of $28 734 by $4889 for mild obesity, by$3189 for moderate obesity, and by $2289 for severe obesity (Table 3). In overweight patients, GBinvolved fewer QALYs than SG, and thus decreasing the cost of surgery would not result in cost-effectiveness.

Sensitivity AnalysisWe conducted 1-way sensitivity analysis only on the cost-effective interventions identified in the basecase analyses (ie, SG). eFigures 1 to 4 in the Supplement show the 10 model parameters for eachweight class that had the largest association with the ICER, using tornado diagrams. We found that inpatients with mild, moderate, or severe obesity, SG remained cost-effective despite variations in allmodel inputs. In overweight patients, the ICER for SG remained below the willingness-to-paythreshold ($100 000/QALY), despite modifications to all model inputs except for age and the hazardratio for decompensation associated with a 1-U increase in BMI. The ICER rose above $100 000 perQALY (vs usual care) when age was varied to its high extreme (73.6 years) and when the hazard ratiofor decompensation associated with a 1-U increase in BMI was varied to its low extreme (1.01).

In the PSA, model parameters were varied simultaneously 10 000 times, and cost-effectivenesswas calculated. According to our PSA results, SG was the most cost-effective option for patients withoverweight or obesity in most iterations. In patients with severe obesity, SG was the optimal strategyin 79% of iterations, using a willingness-to-pay threshold of $100 000 per QALY (eFigure 5 in theSupplement). In those with moderate obesity, SG was the most cost-effective option in 64% ofiterations (eFigure 6 in the Supplement). In patients with mild obesity, it was the most cost-effectiveoption in 76% of iterations (eFigure 7 in the Supplement). In the overweight population, SG was theoptimal strategy in 53% of model iterations (eFigure 8 in the Supplement).

Discussion

In this modeling-based study, we examined whether the benefits of bariatric surgery outweighed theheightened risks of mortality and complications in surgical candidates with NASH cirrhosis. Ourresults projected that surgery would outperform usual care and ILI across all weight classes assessed,including overweight, in terms of life expectancy and QALY. Furthermore, we found that bariatricsurgery was cost-effective for patients with obesity (mild, moderate, and severe) and those in theoverweight range.

Although studies demonstrate that decompensation is more likely at higher BMIs2 and thatweight loss leads to decreased portal hypertension in patients with cirrhosis and obesity,5 less isknown regarding the effect of bariatric surgery on clinical outcomes in NASH cirrhosis.7,8 In aprevious analysis, Klebanoff et al37 examined the long-term cost-effectiveness of bariatric surgery inpatients with NASH and varying degrees of fibrosis, but that study did not include patients withNASH and cirrhosis. However, given the continued rise in the prevalence of NASH cirrhosis,38 weshould explore the potential role of weight loss in treating this disease. Therefore, our analysis

Table 3. Threshold Analysis: Procedure Cost That Could Make GB Cost-effectivea

Patient ProfileReduced Cost of GB at Which GB Is the MostCost-effective Intervention $

Cost Reduction Relative to BaselineCost of GB, 2017 $US

Severe obesity (BMI ≥40.0) 25 965 2289

Moderate obesity(BMI 35.0-39.9)

25 065 3189

Mild obesity (BMI 30.0-34.9) 23 365 4889

Abbreviations: BMI, body mass index (calculated asweight in kilograms divided by height in meterssquared); GB, laparoscopic Roux-en-Y gastric bypass.a Baseline cost of GB was $28 734. Overweight is not

included because GB was absolutely dominated forthis BMI class (ie, GB resulted in fewer quality-adjusted life-years than sleeve gastrectomy).

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 7/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

underscores a real and urgent need for clinical studies of bariatric surgery and other weight losstherapies in NASH cirrhosis.

For patients with obesity, bariatric surgery remained cost-effective in our 1-way sensitivityanalyses, which included varying the probability of surgical mortality and complications. Thesefindings are important because the weighing of risks and benefits of bariatric surgery in the settingof cirrhosis may not appear obvious to patients or physicians. Surgery may involve a 3-fold increasedrisk of death in patients with compensated cirrhosis, compared with those without cirrhosis.9 Inaddition, individuals with cirrhosis have a diminished life expectancy, and thus may not derive long-term health benefits from surgery, including improvements in comorbidities such as hypertension,dyslipidemia, and diabetes.39,40 Despite these considerations, our analysis suggests that the benefitsof surgery may outweigh the risks. In sensitivity analyses, surgery remained cost-effective in patientswith obesity, including those with mild obesity (BMI, 30.0-35.0), even when we assumed a smallbenefit of weight loss on liver disease progression and an increase in the the risk of surgical mortality.In addition, a recent analysis found no association between compensated cirrhosis and increasedmortality in patients undergoing bariatric surgery, suggesting that our analysis may even overstatethe dangers of bariatric surgery.41 Our findings are consistent with the available evidence, suggestingthat bariatric surgery offers an acceptably low risk in patients with compensated cirrhosis, potentiallyeven in those with portal hypertension.42,43

To date, no clinical studies have compared outcomes of SG and GB in patients with NASHcirrhosis. We chose to include both procedures in our model, because they are the 2 most commonbariatric procedures performed in the United States.44 Of the treatment options in this analysis, GBinvolved the largest increase in life expectancy and quality of life for patients with obesity, but SGwas the most cost-effective. Although GB involved a slightly greater increase in QALYs, it was alsomarginally more expensive than SG. Our findings suggest that the slightly superior weight lossattributable to GB might not be worth the higher cost of the procedure. In addition to the extra costof GB, it also entails a higher risk of complications, which may limit its cost-effectiveness. Data frompatients without cirrhosis show that GB induces more weight loss and more successful resolution ofcomorbidities compared with SG, but also involves a higher rate of deep wound infections, seriousmorbidity, and 30-day reoperation.23 Despite the overall lower rate of complications for SG, it isassociated with an increased risk of portal vein thrombosis, a rare but serious complication withmortality exceeding 40% among affected patients.45 Although this risk poses an important clinicalconsideration, patients and physicians are increasingly opting for SG, owing to various potentialadvantages. For example, SG preserves endoscopic access to the gastric tube in the event of varicealbleeding, as well as access to the biliary system.46 While our model is informed by limited data andshould not be used to guide selection of a particular bariatric procedure, our findings support theidea that SG may provide a more favorable combination of risk, benefit, and cost compared with GB.Given the rapidly increasing popularity of SG at the expense of GB, our results may reflect a growingconventional wisdom concerning these procedures among bariatric surgeons.

Strengths and LimitationsOur study has several strengths. This model is, to our knowledge, the first cost-effectiveness analysisof bariatric surgery and lifestyle intervention among patients with NASH cirrhosis. In addition,although 1 previous modeling study explored weight loss in patients with cirrhosis,47 it did not assesscost-effectiveness and did not include SG, which is now the most popular bariatric procedure in theUnited States.48 Compared with prior analyses, ours also incorporated not only individuals withsevere obesity, but also those with lower baseline BMI.

As with any simulation model, our analysis must be viewed in the context of some limitations.First, we used weight loss data from studies of patients without cirrhosis; weight loss after bariatricsurgery may differ between patients with and without cirrhosis. This limitation is not major, becausethe available data suggest excellent weight loss outcomes for patients with and without cirrhosis.7

We also extrapolated long-term trends in BMI using data from the Swedish Obese Subjects study,26

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 8/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

given a lack of long-term data surrounding trends in weight after SG and GB. In addition, we assumedthat weight loss would slow the progression of compensated cirrhosis to decompensated cirrhosis.This decision gains support from an analysis that described an increased risk of decompensationassociated with high BMI,2 as well as a trial that found decreased portal hypertension after weightloss.5 Evidence also suggests that bariatric surgery induces resolution of noncirrhotic NASH,49 andsome limited data show that surgery may even cause regression of fibrosis in patients withcirrhosis.8,50 Future studies that explore the effects of bariatric surgery in this population will confirmor refute our modeling assumptions. We also assumed that weight loss would benefit patients in ourmodel by leading to a decrease in background mortality, a decrease in the probability ofdecompensation, and an increase in weight-related quality of life. Thus, we may have overestimatedthe benefit of surgery on survival and quality of life. However, we also did not explicitly model othercomorbidities that resolve or improve with weight loss, such as diabetes, hypertension, anddyslipidemia, which may counterbalance any potential overestimation of clinical benefit. In addition,to inform the effect of weight loss on the probability of decompensation, our analysis used data froma study that included a large portion of patients with cirrhosis secondary to viral hepatitis or alcoholuse.2 Given the association between NASH and obesity, we suspect that weight loss might have aneven greater benefit in individuals with NASH. Accordingly, our estimates of the benefits of bariatricsurgery in this population may be conservative.

Our results for patients with a BMI of less than 35.0 should also be viewed with some caution.The available evidence suggests that bariatric surgery is safe and does not lead to excessive weightloss in patients with mild obesity. Based on the current literature, the American Society for Metabolicand Bariatric Surgery51 and the International Federation for the Surgery of Obesity and MetabolicDisorders52 recommend bariatric surgery as an option for suitable patients with a BMI of 30.0 to35.0. For patients with a BMI of less than 30.0, scant data are available to guide clinical decisionmaking. The lack of data to inform modeling for patients with a BMI of less than 30.0 makes ouranalysis of the overweight patients an exploratory one, although we can reasonably expect similarbeneficial results of surgery in the lower BMI group.

Conclusions

Our results underscore the promise that bariatric procedures may hold for patients with NASHcirrhosis, including those with lower BMI. Our analysis affirms that the benefits of surgery likelyoutweigh the risks in otherwise eligible individuals, and surgery could be a cost-effectiveintervention. Given the increasing burden of NASH cirrhosis and the lack of effective treatments, itwould appear that future studies must explore the effects of bariatric surgery on disease progressionin individuals with NASH cirrhosis.

ARTICLE INFORMATIONAccepted for Publication: January 2, 2019.

Published: February 22, 2019. doi:10.1001/jamanetworkopen.2019.0047

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2019 Klebanoff MJet al. JAMA Network Open.

Corresponding Author: Jagpreet Chhatwal, PhD, Massachusetts General Hospital Institute for TechnologyAssessment, 101 Merrimac St, 10th Floor, Boston, MA 02114 (jagchhatwal@mgh.harvard.edu).

Author Affiliations: Currently a medical student at Yale School of Medicine, New Haven, Connecticut (Klebanoff);Gastroenterology Division, Massachusetts General Hospital, Boston (Corey, Kaplan, Chhatwal); Harvard MedicalSchool, Boston, Massachusetts (Corey, Kaplan, Chhatwal); Institute for Clinical and Economic Review, Boston,Massachusetts (Samur); Massachusetts General Hospital Institute for Technology Assessment, Boston (Samur,Choi, Chhatwal); Division of Digestive and Liver Diseases, Department of Medicine, Columbia University MedicalCenter, New York, New York (Hur).

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 9/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

Author Contributions: Drs Chhatwal and Hur were co–senior authors. Drs Chhatwal and Hur had full access to allthe data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Klebanoff, Corey, Kaplan, Chhatwal, Hur.

Acquisition, analysis, or interpretation of data: Klebanoff, Corey, Samur, Choi, Hur.

Drafting of the manuscript: Klebanoff, Hur.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Klebanoff, Chhatwal, Hur.

Administrative, technical, or material support: Klebanoff, Samur, Choi, Kaplan, Hur.

Supervision: Corey, Kaplan, Chhatwal, Hur.

Conflict of Interest Disclosures: Dr Corey reported receiving personal fees from Novo Nordisk and Gilead outsidethe submitted work. Mr Choi reported receiving personal fees from Cambridge Biomedical and EconomicConsulting Group, LLC, outside the submitted work. Dr Chhatwal reported receiving research grants from andserving of advisory panels of Merck and Gilead outside the submitted work. Dr Hur reported receiving consultingfees and personal fees from Novo Nordisk outside the submitted work. No other disclosures were reported.

REFERENCES1. Wang YC, McPherson K, Marsh T, Gortmaker SL, Brown M. Health and economic burden of the projected obesitytrends in the USA and the UK. Lancet. 2011;378(9793):815-825. doi:10.1016/S0140-6736(11)60814-3

2. Berzigotti A, Garcia-Tsao G, Bosch J, et al; Portal Hypertension Collaborative Group. Obesity is an independentrisk factor for clinical decompensation in patients with cirrhosis. Hepatology. 2011;54(2):555-561. doi:10.1002/hep.24418

3. Musso G, Cassader M, Rosina F, Gambino R. Impact of current treatments on liver disease, glucose metabolismand cardiovascular risk in non-alcoholic fatty liver disease (NAFLD): a systematic review and meta-analysis ofrandomised trials. Diabetologia. 2012;55(4):885-904. doi:10.1007/s00125-011-2446-4

4. Vilar-Gomez E, Martinez-Perez Y, Calzadilla-Bertot L, et al. Weight loss through lifestyle modificationsignificantly reduces features of nonalcoholic steatohepatitis. Gastroenterology. 2015;149(2):367-78.e5. doi:10.1053/j.gastro.2015.04.005

5. Berzigotti A, Albillos A, Villanueva C, et al; Ciberehd SportDiet Collaborative Group. Effects of an intensivelifestyle intervention program on portal hypertension in patients with cirrhosis and obesity: the SportDiet Study.Hepatology. 2017;65(4):1293-1305. doi:10.1002/hep.28992

6. Segev DL, Thompson RE, Locke JE, et al. Prolonged waiting times for liver transplantation in obese patients.Ann Surg. 2008;248(5):863-870. doi:10.1097/SLA.0b013e31818a01ef

7. Jan A, Narwaria M, Mahawar KK. A systematic review of bariatric surgery in patients with liver cirrhosis. ObesSurg. 2015;25(8):1518-1526. doi:10.1007/s11695-015-1727-2

8. Kral JG, Thung SN, Biron S, et al. Effects of surgical treatment of the metabolic syndrome on liver fibrosis andcirrhosis. Surgery. 2004;135(1):48-58. doi:10.1016/j.surg.2003.10.003

9. Mosko JD, Nguyen GC. Increased perioperative mortality following bariatric surgery among patients withcirrhosis. Clin Gastroenterol Hepatol. 2011;9(10):897-901. doi:10.1016/j.cgh.2011.07.007

10. Mahady SE, Wong G, Craig JC, George J. Pioglitazone and vitamin E for nonalcoholic steatohepatitis: a costutility analysis. Hepatology. 2012;56(6):2172-2179. doi:10.1002/hep.25887

11. Sanyal AJ, Banas C, Sargeant C, et al. Similarities and differences in outcomes of cirrhosis due to nonalcoholicsteatohepatitis and hepatitis C. Hepatology. 2006;43(4):682-689. doi:10.1002/hep.21103

12. Ratziu V, Bonyhay L, Di Martino V, et al. Survival, liver failure, and hepatocellular carcinoma in obesity-relatedcryptogenic cirrhosis. Hepatology. 2002;35(6):1485-1493. doi:10.1053/jhep.2002.33324

13. Ascha MS, Hanouneh IA, Lopez R, Tamimi TA, Feldstein AF, Zein NN. The incidence and risk factors ofhepatocellular carcinoma in patients with nonalcoholic steatohepatitis. Hepatology. 2010;51(6):1972-1978. doi:10.1002/hep.23527

14. Yatsuji S, Hashimoto E, Tobari M, Taniai M, Tokushige K, Shiratori K. Clinical features and outcomes of cirrhosisdue to non-alcoholic steatohepatitis compared with cirrhosis caused by chronic hepatitis C. J GastroenterolHepatol. 2009;24(2):248-254. doi:10.1111/j.1440-1746.2008.05640.x

15. Thuluvath PJ, Guidinger MK, Fung JJ, Johnson LB, Rayhill SC, Pelletier SJ. Liver transplantation in the UnitedStates, 1999-2008. Am J Transplant. 2010;10(4, pt 2):1003-1019. doi:10.1111/j.1600-6143.2010.03037.x

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 10/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

16. Davis KL, Mitra D, Medjedovic J, Beam C, Rustgi V. Direct economic burden of chronic hepatitis C virus in aUnited States managed care population. J Clin Gastroenterol. 2011;45(2):e17-e24. doi:10.1097/MCG.0b013e3181e12c09

17. Lang K, Danchenko N, Gondek K, Shah S, Thompson D. The burden of illness associated with hepatocellularcarcinoma in the United States. J Hepatol. 2009;50(1):89-99. doi:10.1016/j.jhep.2008.07.029

18. Saab S, Hunt DR, Stone MA, McClune A, Tong MJ. Timing of hepatitis C antiviral therapy in patients withadvanced liver disease: a decision analysis model. Liver Transpl. 2010;16(6):748-759.

19. Wolfe RA, Roys EC, Merion RM. Trends in organ donation and transplantation in the United States, 1999-2008.Am J Transplant. 2010;10(4, pt 2):961-972. doi:10.1111/j.1600-6143.2010.03021.x

20. Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA. 2003;289(2):187-193. doi:10.1001/jama.289.2.187

21. Fattovich G, Pantalena M, Zagni I, Realdi G, Schalm SW, Christensen E; European Concerted Action on ViralHepatitis (EUROHEP). Effect of hepatitis B and C virus infections on the natural history of compensated cirrhosis:a cohort study of 297 patients. Am J Gastroenterol. 2002;97(11):2886-2895. doi:10.1111/j.1572-0241.2002.07057.x

22. Hui JM, Kench JG, Chitturi S, et al. Long-term outcomes of cirrhosis in nonalcoholic steatohepatitis comparedwith hepatitis C. Hepatology. 2003;38(2):420-427. doi:10.1053/jhep.2003.50320

23. Young MT, Gebhart A, Phelan MJ, Nguyen NT. Use and outcomes of laparoscopic sleeve gastrectomy vslaparoscopic gastric bypass: analysis of the American College of Surgeons NSQIP. J Am Coll Surg. 2015;220(5):880-885. doi:10.1016/j.jamcollsurg.2015.01.059

24. Chang SH, Stoll CR, Song J, Varela JE, Eagon CJ, Colditz GA. The effectiveness and risks of bariatric surgery: anupdated systematic review and meta-analysis, 2003-2012. JAMA Surg. 2014;149(3):275-287. doi:10.1001/jamasurg.2013.3654

25. Look AHEAD Research Group. Eight-year weight losses with an intensive lifestyle intervention: the LookAHEAD study. Obesity (Silver Spring). 2014;22(1):5-13. doi:10.1002/oby.20662

26. Sjöström L, Peltonen M, Jacobson P, et al. Bariatric surgery and long-term cardiovascular events. JAMA. 2012;307(1):56-65. doi:10.1001/jama.2011.1914

27. Campbell J, McGarry LA, Shikora SA, Hale BC, Lee JT, Weinstein MC. Cost-effectiveness of laparoscopic gastricbanding and bypass for morbid obesity. Am J Manag Care. 2010;16(7):e174-e187.

28. Agency for Healthcare Research and Quality. Healthcare Cost and Utilization Project National Inpatient Sample(NIS). https://hcupnet.ahrq.gov/. Modified August 13, 2018. Accessed March 22, 2017.

29. Rushing J, Wing R, Wadden TA, et al; Look AHEAD Research Group. Cost of intervention delivery in a lifestyleweight loss trial in type 2 diabetes: results from the Look AHEAD clinical trial. Obes Sci Pract. 2017;3(1):15-24. doi:10.1002/osp4.92

30. Saab S, Gordon SC, Park H, Sulkowski M, Ahmed A, Younossi Z. Cost-effectiveness analysis of sofosbuvir pluspeginterferon/ribavirin in the treatment of chronic hepatitis C virus genotype 1 infection. Aliment Pharmacol Ther.2014;40(6):657-675. doi:10.1111/apt.12871

31. Gordon SC, Pockros PJ, Terrault NA, et al. Impact of disease severity on healthcare costs in patients withchronic hepatitis C (CHC) virus infection. Hepatology. 2012;56(5):1651-1660. doi:10.1002/hep.25842

32. McAdam-Marx C, McGarry LJ, Hane CA, Biskupiak J, Deniz B, Brixner DI. All-cause and incremental per patientper year cost associated with chronic hepatitis C virus and associated liver complications in the United States:a managed care perspective. J Manag Care Pharm. 2011;17(7):531-546.

33. United States Bureau of Labor Statistics. Consumer Price Index Inflation Calculator. 2015. https://data.bls.gov/cgi-bin/cpicalc.pl. Accessed July 25, 2017.

34. Chhatwal J, Kanwal F, Roberts MS, Dunn MA. Cost-effectiveness and budget impact of hepatitis C virustreatment with sofosbuvir and ledipasvir in the United States. Ann Intern Med. 2015;162(6):397-406. doi:10.7326/M14-1336

35. Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russell LB. Recommendations of the Panel on Cost-effectiveness in Health and Medicine. JAMA. 1996;276(15):1253-1258. doi:10.1001/jama.1996.03540150055031

36. D’Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis:a systematic review of 118 studies. J Hepatol. 2006;44(1):217-231. doi:10.1016/j.jhep.2005.10.013

37. Klebanoff MJ, Corey KE, Chhatwal J, Kaplan LM, Chung RT, Hur C. Bariatric surgery for nonalcoholicsteatohepatitis: a clinical and cost-effectiveness analysis. Hepatology. 2017;65(4):1156-1164. doi:10.1002/hep.28958

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 11/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022

38. Estes C, Razavi H, Loomba R, Younossi Z, Sanyal AJ. Modeling the epidemic of nonalcoholic fatty liver diseasedemonstrates an exponential increase in burden of disease. Hepatology. 2018;67(1):123-133.

39. Batsis JA, Romero-Corral A, Collazo-Clavell ML, Sarr MG, Somers VK, Lopez-Jimenez F. Effect of bariatricsurgery on the metabolic syndrome: a population-based, long-term controlled study. Mayo Clin Proc. 2008;83(8):897-907. doi:10.1016/S0025-6196(11)60766-0

40. Schauer PR, Bhatt DL, Kirwan JP, et al; STAMPEDE Investigators. Bariatric surgery versus intensive medicaltherapy for diabetes: 5-year outcomes. N Engl J Med. 2017;376(7):641-651. doi:10.1056/NEJMoa1600869

41. Lipshultz H, Issak A, Porter A, et al. 3–Bariatric surgery is safe in patients with compensated cirrhosis: anupdate on outcomes from 2008-2013. Gastroenterology. 2018;154(6)(suppl 1):S1. doi:10.1016/S0016-5085(18)30492-X

42. Pestana L, Swain J, Dierkhising R, Kendrick ML, Kamath PS, Watt KD. Bariatric surgery in patients with cirrhosiswith and without portal hypertension: a single-center experience. Mayo Clin Proc. 2015;90(2):209-215. doi:10.1016/j.mayocp.2014.11.012

43. Hanipah ZN, Punchai S, McCullough A, et al. Bariatric surgery in patients with cirrhosis and portalhypertension. Obes Surg. 2018;28(11):3431-3438. doi:10.1007/s11695-018-3372-z

44. Chhatwal J, Samur S, Kues B, et al. Optimal timing of hepatitis C treatment for patients on the liver transplantwaiting list. Hepatology. 2017;65(3):777-788. doi:10.1002/hep.28926

45. Belnap L, Rodgers GM, Cottam D, Zaveri H, Drury C, Surve A. Portal vein thrombosis after laparoscopic sleevegastrectomy: presentation and management. Surg Obes Relat Dis. 2016;12(10):1787-1794. doi:10.1016/j.soard.2016.03.005

46. Takata MC, Campos GM, Ciovica R, et al. Laparoscopic bariatric surgery improves candidacy in morbidly obesepatients awaiting transplantation. Surg Obes Relat Dis. 2008;4(2):159-164. doi:10.1016/j.soard.2007.12.009

47. Bromberger B, Porrett P, Choudhury R, Dumon K, Murayama KM. Weight loss interventions for morbidlyobese patients with compensated cirrhosis: a Markov decision analysis model. J Gastrointest Surg. 2014;18(2):321-327. doi:10.1007/s11605-013-2298-y

48. American Society for Metabolic and Bariatric Surgery. Estimate of bariatric surgery numbers, 2011-2017. https://asmbs.org/resources/estimate-of-bariatric-surgery-numbers. Accessed December 8, 2017.

49. Lassailly G, Caiazzo R, Buob D, et al. Bariatric surgery reduces features of nonalcoholic steatohepatitis inmorbidly obese patients. Gastroenterology. 2015;149(2):379-388. doi:10.1053/j.gastro.2015.04.014

50. Woodford RM, Burton PR, O’Brien PE, Laurie C, Brown WA. Laparoscopic adjustable gastric banding inpatients with unexpected cirrhosis: safety and outcomes. Obes Surg. 2015;25(10):1858-1862. doi:10.1007/s11695-015-1623-9

51. ASMBS Clinical Issues Committee. Bariatric surgery in class I obesity (body mass index 30-35 kg/m2). Surg ObesRelat Dis. 2013;9(1):e1-e10. doi:10.1016/j.soard.2012.09.002

52. Busetto L, Dixon J, De Luca M, Shikora S, Pories W, Angrisani L. Bariatric surgery in class I obesity: a PositionStatement from the International Federation for the Surgery of Obesity and Metabolic Disorders (IFSO). ObesSurg. 2014;24(4):487-519. doi:10.1007/s11695-014-1214-1

SUPPLEMENT.eTable 1. Percentage of Excess Weight Loss After Bariatric Surgery and Lifestyle InterventioneTable 2. Health-Related Quality-of-Life Utilities of the United States PopulationeFigure 1. One-Way Sensitivity Analysis for SG in Severe ObesityeFigure 2. One-Way Sensitivity Analysis for SG in Moderate ObesityeFigure 3. One-Way Sensitivity Analysis for SG in Mild ObesityeFigure 4. One-Way Sensitivity Analysis for SG in OverweighteFigure 5. Probabilistic Sensitivity Analysis for Severe ObesityeFigure 6. Probabilistic Sensitivity Analysis for Moderate ObesityeFigure 7. Probabilistic Sensitivity Analysis for Mild ObesityeFigure 8. Probabilistic Sensitivity Analysis for OverweighteReferences.

JAMA Network Open | Gastroenterology and Hepatology Cost-effectiveness of Bariatric Surgery for Patients With NASH Cirrhosis

JAMA Network Open. 2019;2(2):e190047. doi:10.1001/jamanetworkopen.2019.0047 (Reprinted) February 22, 2019 12/12

Downloaded From: https://jamanetwork.com/ on 05/06/2022