[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.
Variation in Rates of Serious Complications Across Bariatric Centers of Excellence in 12 States
Variation in Rates of Serious Complications Across Bariatric Centers of Excellence in 12 States

Risk-adjusted and reliability-adjusted outcomes in 165 centers across 12 states. Data from Healthcare Cost and Utilization Project’s State Inpatient Database.9

Figure 2.
Variation in Rates of Serious Complications Across Low-Volume, Medium-Volume, and High-Volume Bariatric Centers of Excellence
Variation in Rates of Serious Complications Across Low-Volume, Medium-Volume, and High-Volume Bariatric Centers of Excellence

Mean rate of complications is based on risk-adjusted data prior to reliability adjustment. Data are from the Healthcare Cost and Utilization Project’s State Inpatient Database, 2010-2013.9

Table 1.  
Patient Characteristicsa
Patient Characteristicsa
Table 2.  
Hospital-Based Centers of Excellence Characteristicsa
Hospital-Based Centers of Excellence Characteristicsa
Table 3.  
Variation in Rates of Serious Complications at Bariatric Centers of Excellence by Statea
Variation in Rates of Serious Complications at Bariatric Centers of Excellence by Statea
1.
American Society for Metabolic and Bariatric Surgery. Estimate of bariatric surgery numbers, 2011-2015. https://asmbs.org/resources/estimate-of-bariatric-surgery-numbers. Accessed September 20, 2016.
2.
Flum  DR, Salem  L, Elrod  JA, Dellinger  EP, Cheadle  A, Chan  L.  Early mortality among Medicare beneficiaries undergoing bariatric surgical procedures.  JAMA. 2005;294(15):1903-1908.PubMedGoogle ScholarCrossref
3.
Hollenbeak  CS, Rogers  AM, Barrus  B, Wadiwala  I, Cooney  RN.  Surgical volume impacts bariatric surgery mortality: a case for centers of excellence.  Surgery. 2008;144(5):736-743.PubMedGoogle ScholarCrossref
4.
Pratt  GM, McLees  B, Pories  WJ.  The ASBS Bariatric Surgery Centers of Excellence program: a blueprint for quality improvement.  Surg Obes Relat Dis. 2006;2(5):497-503.PubMedGoogle ScholarCrossref
5.
Schirmer  B, Jones  DB.  The American College of Surgeons Bariatric Surgery Center Network: establishing standards.  Bull Am Coll Surg. 2007;92(8):21-27.PubMedGoogle Scholar
6.
Gebhart  A, Young  M, Phelan  M, Nguyen  NT.  Impact of accreditation in bariatric surgery.  Surg Obes Relat Dis. 2014;10(5):767-773.PubMedGoogle ScholarCrossref
7.
Dimick  JB, Osborne  NH, Nicholas  L, Birkmeyer  JD.  Identifying high-quality bariatric surgery centers: hospital volume or risk-adjusted outcomes?  J Am Coll Surg. 2009;209(6):702-706.PubMedGoogle ScholarCrossref
8.
Chhabra  KR, Dimick  JB.  Hospital networks and value-based payment: fertile ground for regionalizing high-risk surgery.  JAMA. 2015;314(13):1335-1336.PubMedGoogle ScholarCrossref
9.
Healthcare Cost and Utilization Project. Overview of the State Inpatient Databases (SID). http://www.hcup-us.ahrq.gov/sidoverview.jsp. Accessed January 11, 2016.
10.
Santry  HP, Gillen  DL, Lauderdale  DS.  Trends in bariatric surgical procedures.  JAMA. 2005;294(15):1909-1917.PubMedGoogle ScholarCrossref
11.
Krell  RW, Finks  JF, English  WJ, Dimick  JB.  Profiling hospitals on bariatric surgery quality: which outcomes are most reliable ?  J Am Coll Surg. 2014;219(4):725-734.e3.PubMedGoogle ScholarCrossref
12.
Scally  CP, Thumma  JR, Birkmeyer  JD, Dimick  JB.  Impact of surgical quality improvement on payments in Medicare patients.  Ann Surg. 2015;262(2):249-252.PubMedGoogle ScholarCrossref
13.
Osborne  NH, Nicholas  LH, Ryan  AM, Thumma  JR, Dimick  JB.  Association of hospital participation in a quality reporting program with surgical outcomes and expenditures for Medicare beneficiaries.  JAMA. 2015;313(5):496-504.PubMedGoogle ScholarCrossref
14.
Livingston  EH.  Procedure incidence and in-hospital complication rates of bariatric surgery in the United States.  Am J Surg. 2004;188(2):105-110.PubMedGoogle ScholarCrossref
15.
Ibrahim  AM, Hughes  TG, Thumma  JR, Dimick  JB.  Association of hospital critical access status with surgical outcomes and expenditures among Medicare beneficiaries.  JAMA. 2016;315(19):2095-2103.PubMedGoogle ScholarCrossref
16.
Dartmouth Institute for Health Policy and Clinical Practice. Research methods. In: The Dartmouth Atlas of Health Care. Lebanon, NH: Trustees of Dartmouth College; 2017. http://www.dartmouthatlas.org/downloads/methods/research_methods.pdf. Accessed August 12, 2016.
17.
Elixhauser  A, Steiner  C, Harris  DR, Coffey  RM.  Comorbidity measures for use with administrative data.  Med Care. 1998;36(1):8-27.PubMedGoogle ScholarCrossref
18.
Southern  DA, Quan  H, Ghali  WA.  Comparison of the Elixhauser and Charlson/Deyo methods of comorbidity measurement in administrative data.  Med Care. 2004;42(4):355-360.PubMedGoogle ScholarCrossref
19.
Jones  HE, Spiegelhalter  DJ.  The identification of “unusual” health-care providers from a hierarchical model.  Am Stat. 2011;65(3):154-163.Google ScholarCrossref
20.
Dimick  JB, Staiger  DO, Birkmeyer  JD.  Ranking hospitals on surgical mortality: the importance of reliability adjustment.  Health Serv Res. 2010;45(6, pt 1):1614-1629.PubMedGoogle ScholarCrossref
21.
Grenda  TR, Krell  RW, Dimick  JB.  Reliability of hospital cost profiles in inpatient surgery.  Surgery. 2015;159(2):375-380.PubMedGoogle Scholar
22.
Elixhauser Comorbidity Software, version 3.7. Healthcare Cost and Utilization Project website. https://www.hcup-us.ahrq.gov/toolssoftware/comorbidity/comorbidity.jsp. Accessed March 17, 2017.
23.
American Hospital Association. AHA Annual Survey Database. http://www.ahadataviewer.com/book-cd-products/AHA-Survey. Accessed December 6, 2016.
24.
Saleh  F, Doumouras  AG, Gmora  S, Anvari  M, Hong  D.  Outcomes the Ontario Bariatric Network: a cohort study.  CMAJ Open. 2016;4(3):E383-E389.PubMedGoogle ScholarCrossref
25.
Morton  JM, Garg  T, Nguyen  N.  Does hospital accreditation impact bariatric surgery safety?  Ann Surg. 2014;260(3):504-508.PubMedGoogle ScholarCrossref
26.
O’Brien  PE.  Controversies in bariatric surgery.  Br J Surg. 2015;102(6):611-618.PubMedGoogle ScholarCrossref
27.
Berger  ER, Clements  RH, Morton  JM,  et al.  The impact of different surgical techniques on outcomes in laparoscopic sleeve gastrectomies: the first report from the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP).  Ann Surg. 2016;264(3):464-473.PubMedGoogle ScholarCrossref
28.
Birkmeyer  JD, Siewers  AE, Finlayson  EV,  et al.  Hospital volume and surgical mortality in the United States.  N Engl J Med. 2002;346(15):1128-1137.PubMedGoogle ScholarCrossref
29.
Birkmeyer  JD, Finks  JF, O’Reilly  A,  et al; Michigan Bariatric Surgery Collaborative.  Surgical skill and complication rates after bariatric surgery.  N Engl J Med. 2013;369(15):1434-1442.PubMedGoogle ScholarCrossref
30.
Sinha  A, Jayaraman  L, Punhani  D, Chowbey  P.  Enhanced recovery after bariatric surgery in the severely obese, morbidly obese, super-morbidly obese and super-super morbidly obese using evidence-based clinical pathways: a comparative study.  Obes Surg. 2017;27(3):560-568.PubMedGoogle ScholarCrossref
31.
Telem  DA, Majid  SF, Powers  K, DeMaria  E, Morton  J, Jones  DB.  Assessing national provision of care: variability in bariatric clinical care pathways.  Surg Obes Relat Dis. 2017;13(2):281-284.PubMedGoogle ScholarCrossref
32.
Yeats  M, Wedergren  S, Fox  N, Thompson  JS.  The use and modification of clinical pathways to achieve specific outcomes in bariatric surgery.  Am Surg. 2005;71(2):152-154.PubMedGoogle Scholar
33.
Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program. Decreasing Readmissions Through Opportunities Provided (D.R.O.P) [video]. http://reports.nsqip.facs.org/MBSAQIPDropVideo/. Accessed August 12, 2016.
34.
Morton  J.  The first metabolic and bariatric surgery accreditation and quality improvement program quality initiative: decreasing readmissions through opportunities provided.  Surg Obes Relat Dis. 2014;10(3):377-378.PubMedGoogle ScholarCrossref
35.
Iezzoni  LI, Daley  J, Heeren  T,  et al.  Identifying complications of care using administrative data.  Med Care. 1994;32(7):700-715.PubMedGoogle ScholarCrossref
36.
Weingart  SN, Iezzoni  LI, Davis  RB,  et al.  Use of administrative data to find substandard care: validation of the complications screening program.  Med Care. 2000;38(8):796-806.PubMedGoogle ScholarCrossref
Original Investigation
July 2017

Variation in Outcomes at Bariatric Surgery Centers of Excellence

Author Affiliations
  • 1Center for Healthcare Outcomes and Policy, University of Michigan, Ann Arbor
  • 2Surgical Innovation Editor, JAMA Surgery
JAMA Surg. 2017;152(7):629-636. doi:10.1001/jamasurg.2017.0542
Key Points

Question  How much variation in surgical outcomes exists among bariatric centers of excellence in the United States?

Findings  In this analysis of claims data from 145 527 patients who underwent bariatric procedures, there was a 17-fold variation in rates of serious complications across accredited bariatric centers nationally and up to 9.5-fold variation across individual states. Significant variation was also seen in low-volume, medium-volume, and high-volume bariatric centers.

Meaning  Accreditation alone does not ensure uniform high-quality care for bariatric procedures, and further quality improvement efforts should be considered.

Abstract

Importance  In the United States, reports about perioperative complications associated with bariatric surgery led to the establishment of accreditation criteria for bariatric centers of excellence and many bariatric centers obtaining accreditation. Currently, most bariatric procedures occur at these centers, but to what extent they uniformly provide high-quality care remains unknown.

Objective  To describe the variation in surgical outcomes across bariatric centers of excellence and the geographic availability of high-quality centers.

Design, Setting, and Participants  This retrospective review analyzed the claims data of 145 527 patients who underwent bariatric surgery at bariatric centers of excellence between January 1, 2010, and December 31, 2013. Data were obtained from the Healthcare Cost and Utilization Project’s State Inpatient Database. This database included unique hospital identification numbers in 12 states (Arkansas, Arizona, Florida, Iowa, Massachusetts, Maryland, North Carolina, Nebraska, New Jersey, New York, Washington, and Wisconsin), allowing comparisons among 165 centers of excellence located in those states. Participants were identified using International Classification of Diseases, Ninth Revision, Clinical Modification codes. Those included in the study cohort were patients with a primary diagnosis of morbid obesity and who underwent laparoscopic Roux-en-Y gastric bypass, open Roux-en-Y gastric bypass, laparoscopic gastric band placement, or laparoscopic sleeve gastrectomy. Excluded from the cohort were patients younger than 18 years or who had an abdominal malignant neoplasm. Data were analyzed July 1, 2016, through January 10, 2017.

Main Outcomes and Measures  Risk-adjusted and reliability-adjusted serious complication rates within 30 days of the index operation were calculated for each center. Centers were stratified by geographic location and operative volume.

Results  In this analysis of claims data from 145 527 patients, wide variation in quality was found across 165 bariatric centers of excellence, both nationwide and statewide. At the national level, the risk-adjusted and reliability-adjusted serious complication rates at each center varied 17-fold, ranging from 0.6% to 10.3%. At the state level, variation ranged from 2.1-fold (Wisconsin decile range, 1.5%-3.3%) to 9.5-fold (Nebraska decile range, 1.0%-10.3%). After dividing hospitals into quintiles of quality on the basis of their adjusted complication rates, 38 of 132 (28.8%) had a center in a higher quintile of quality located within the same hospital service area. Variation in rates of complications existed at centers with low volume (annual mean [SD] procedure volume, 156 [20] patients; complication range, 0.6%-6.4%; 9.8-fold variation), medium volume (annual mean [SD] procedure volume, 239 [27] patients; complication range, 0.6%-10.3%; 17.5-fold variation), and high volume (annual mean [SD] procedure volume, 448 [131] patients; complication range, 0.6%-4.9%; 7.5-fold variation).

Conclusions and Relevance  Even among accredited bariatric surgery centers, wide variation exists in rates of postoperative serious complications across geographic location and operative volumes. Given that a large proportion of centers are geographically located near higher-performing centers, opportunities for improvement through regional collaboratives or selective referral should be considered.

Introduction

Improving safety for bariatric surgery has emerged as a national clinical and policy priority because it is both a common procedure and associated with substantial perioperative risk. More than 196 000 bariatric operations are performed annually in the United States,1 and initial reports prior to any bariatric accreditation process revealed 30-day mortality rates as high as 9% in low-volume centers.2 In response, 2 leading surgical organizations separately established criteria for bariatric centers of excellence to help payers guide patients to high-quality centers of care.3-5 Subsequently, many centers underwent the necessary facility and personnel changes to obtain accreditation; today, most (>88%) bariatric procedures are performed at accredited centers.6 In 2014, the 2 separate efforts merged to form the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP), which not only grants accreditation but also facilitates outcome tracking and feedback for bariatric surgeons and centers within the MBSAQIP.

Quiz Ref IDTo what extent accredited bariatric centers uniformly provide high-quality care remains unknown. The accreditation process has no direct assessment of risk-adjusted outcomes to determine center eligibility. Instead, the criteria to become an accredited center focus on structural requirements, such as bariatric operating tables, longer laparoscopic instruments, and minimum annual case volumes. Although important to the optimal care of patients obtaining care at bariatric centers, such structural elements are only weakly associated with patient outcomes.7 Because accredited centers care for most of the patients seeking bariatric surgery, any variation among these centers may identify opportunities for improvement that can be addressed through current quality improvement programs. Moreover, understanding the geographic correlation between high- and low-quality centers could inform the feasibility of quality improvement strategies (eg, selective referral, in-network regionalization) that have been proposed for other surgical procedures.8

In this context, we performed a retrospective review of claims data of patients who underwent bariatric surgery at accredited centers to identify the variations in postoperative complications across these centers. Our findings, based on data obtained prior to the implementation of MBSAQIP, may help inform how continued progress in quality improvement for bariatric surgery can be achieved.

Methods
Data Source

We obtained claims data from the Healthcare Cost and Utilization Project’s State Inpatient Database, which was created by the Agency for Healthcare Research and Quality.9 This federal-state database provides patient-level data for inpatient discharges from short-term, nonfederal, acute care, general, and specialty hospitals in participating US states. For this specific study, data were available from 165 bariatric centers of excellence in 12 states (Arkansas, Arizona, Florida, Iowa, Massachusetts, Maryland, North Carolina, Nebraska, New Jersey, New York, Washington, and Wisconsin). The other 38 states either did not participate in the Healthcare Cost and Utilization Project or did not have 3 consecutive years of data available during the study period—between January 1, 2010, and December 31, 2013. This study was approved by the University of Michigan Institutional Review Board and was deemed exempt from review because of the use of secondary data.

Identification of Procedures and Study Cohort

Patients who underwent bariatric procedures between January 1, 2010, and December 31, 2013, were identified using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes. Included in the study cohort were patients with a primary diagnosis of morbid obesity (ICD-9-CM codes 278.0, 278.00, and 278.01) who underwent laparoscopic Roux-en-Y gastric bypass, open Roux-en-Y gastric bypass, laparoscopic gastric band placement, or laparoscopic sleeve gastrectomy (ICD-9-CM codes 43.89, 44.3, 44.31, 44.38, 44.39, 44.68, 44.95, 44.96, 44.97, 44.99, 44.5, 45.51, and 45.9). Excluded from the cohort were patients younger than 18 years or who had an abdominal malignant neoplasm (ICD-9-CM codes 150.0-159.9 or 230.1-230.9). A summary of the cohort selection following STROBE criteria is shown in the eFigure in the Supplement.

Outcome Variables

In this study, the primary outcome of interest was the rate of serious complications, and ICD-9-CM codes for bariatric surgery populations were used to identify complications.10,11 These complications fell into the following categories: anastomotic leak, cardiac, genitourinary, hemorrhagic, neurologic, obstruction, postoperative shock, pulmonary, splenic injury, thromboembolic, wound infection, and reoperation. Serious complications were defined as any of the above complications requiring a prolonged length of stay greater than the 75th percentile for the specific procedure performed. This length-of-stay criteria has been applied in multiple previous studies to give clinical face validity (ie, that the complication had a meaningful clinical effect) to the rate of complications and to make it more specific.12-15

A secondary interest in this study was to assess the geographic proximity between high-quality and low-quality centers of excellence. Definitions from the Darthmouth Atlas of Healthcare16 were applied to assign each center to a hospital service area. Hospital service areas are local health care markets for hospital care and are determined by a collection of zip codes surrounding a hospital that represents the location where most people in that zip code would obtain their care.

Finally, we wanted to understand the extent of variation that exists across different operative volumes. Hospitals were divided into terciles on the basis of their annual volume of bariatric procedures and were labeled “low volume,” “medium volume,” or “high volume.”

Statistical Analysis

The first aim of this study was to determine variation in quality among centers of excellence for bariatric surgery. To do so, centers were ranked by their risk-adjusted and reliability-adjusted rates of serious complications. To accommodate differences in case mix (the type of patients and procedures performed), a multivariable logistic regression model that accounted for patient age, sex, race, comorbidities (as described by Elixhauser et al17 and Southern et al18), and operation type (eg, sleeve gastrectomy, Roux-en-Y gastric bypass) was applied to calculate a risk-adjusted rate of serious complications for each center. To address possible secular trends, the year of operation was also included in the regression model. Because operative volume at each center varies and may contribute to random noise in the observed rates of complications, a reliability adjustment was also applied. This approach has been previously described and applied in comparing surgical centers.11,19-21 Briefly, the approach uses hierarchical modeling to “shrink” lower-volume hospitals toward the overall population mean in proportion to the strength of the hospitals’ statistical signal (ie, surgical volume). Thus, the final rates of complications used to rank hospital quality were based on risk-adjusted and reliability-adjusted outcomes, making the ranking a conservative estimate of variation across the centers.

The second aim of this study was to describe the geographic availability of higher-quality centers of excellence. To do so, each center was placed in rank order according to the center’s risk-adjusted and reliability-adjusted rate of serious complications. Centers in the highest complication quintile were labeled “very low quality” and then ordinally as “low quality,” “medium quality,” or “high quality” until those in the lowest complication quintile were labeled “very high quality.” Availability of a higher-quality bariatric center was determined by assessing if the center had another facility in the same hospital service area labeled in a higher quintile of quality.

All reported P values were 2 sided, with P < .05 as a threshold for significance, calculated using a 2-tailed t test. To adjust for clustering within hospitals, robust SEs were applied to all models. All statistical analyses were completed with Stata software, version 14 (StataCorp).

Results

This study included 145 527 patients who underwent a bariatric procedure; this cohort had a mean (SD) age of 44.6 (12.1) years and were predominantly women (113 101 [77.7%]) (Table 1). The most common of these procedures was the laparoscopic Roux-en-Y gastric bypass (78 193 [53.7%]), followed by laparoscopic sleeve gastrectomy (44 887 [30.8%]). In addition to a diagnosis of morbid obesity, the most common comorbidity conditions were hypertension (80 807 [55.5%]) and diabetes (44 190 [30.4%]).

Also included in the study were 165 accredited bariatric centers, nearly all of which (164 [99.4%]) were located in an urban setting and most of which (117 [70.9%]) were within a teaching hospital (Table 2). The centers were geographically diverse, with 78 (47.3%) located in the Northeast and 52 (31.5%) in the South. The mean (SD) number of operating rooms in these centers was 25 (20), and the mean (SD) nursing staff ratio was 7.8 (2.5). Of the 145 527 patients who underwent bariatric procedures, 6970 (4.8%) had a complication and 2367 (1.6%) had a serious complication. Quiz Ref IDThe most common complication was postoperative bleeding (2796 [1.9%]). Seventy-two patients (0.05%) had an in-hospital mortality. Risk factors for developing a complication are summarized in the eTable in the Supplement.

Wide variation in rates of complications at individual centers of excellence was observed. Quiz Ref IDThe risk-adjusted and reliability-adjusted serious complication rates at each of the 165 centers varied 17-fold, ranging from 0.6% to 10.3% (Figure 1). The top and bottom deciles varied 3.7-fold (top decile, 0.9%; bottom decile, 3.3%; P < .005). Quiz Ref IDSimilar variation was seen at the state level (Table 3), ranging from 2.1-fold variation (Wisconsin decile range, 1.5%-3.3%) to 9.5-fold variation (Nebraska decile range, 1.0%-10.3%). Variation in rates of complications existed at centers with low volume (annual mean [SD] procedure volume, 156 [20] patients; complication range, 0.6%-6.4%; 9.8-fold variation), medium volume (annual mean [SD] procedure volume, 239 [27] patients; complication range, 0.6%-10.3%; 17.5-fold variation), and high volume (annual mean [SD] procedure volume, 448 [131] patients; complication range, 0.6%-4.9%; 7.5-fold variation) (Figure 2). Mean (SD) complication rates were 1.9% (0.9%) at low-volume centers, 2.0% (1.0%) at medium-volume centers, and 1.5% (0.3%) at high-volume centers (P = .31).

The geographic availability of a higher-quality center was assessed at the state and hospital service area levels. Quiz Ref IDAfter dividing hospitals into quintiles of quality on the basis of their adjusted complication rates, all states in the study had both low-quality hospitals and high-quality hospitals. Of the 132 hospitals in the lowest 4 quartiles, 38 hospitals (28.8%) had a bariatric center in a higher quintile of quality located within the same hospital service area.

Discussion

Our study presents 2 principle findings that improve our understanding of perioperative safety for bariatric surgery at centers of excellence. First, there was widespread variation in quality—up to 17-fold between top and bottom performers—across bariatric centers of excellence in the United States. Similar to findings by bariatric accreditation programs in other countries,24 this finding suggests that participation alone in the Center of Excellence Program did not ensure uniform high-quality care. Second, a large proportion of lower-performing centers were often located in the same hospital service area as higher-performing centers. Such geographic proximity suggests that patients living in most health care markets would benefit from information about the local bariatric centers in their area. Given that most bariatric procedures are now performed at accredited centers, wide variation among these centers suggests that accreditation alone does not discriminate enough to guide patients to the best centers for care.

Whether bariatric centers of excellence have improved safety for bariatric procedures has been a matter of debate. Advocates for centers of excellence note that hospitals with accreditation have improved their outcomes over time.3,6,25 In contrast, a later study demonstrated that similar improvements were made in non–centers of excellence, and, when taken together, there are no significant differences in outcomes between the 2 types of bariatric centers.26 Nonetheless, establishing minimum standards and systematically adhering to best practices are an excellent place to start improving outcomes across all centers performing bariatric procedures. The accredited centers model has been widely adopted, with more than 720 centers having now obtained accreditation,27 and major private insurers still restrict payments to these centers. This study helps us move past the debate by identifying an important opportunity for improvement across accredited centers, where nearly all bariatric procedures are performed.

The variation in quality seen at bariatric centers may have multiple explanations, which we attempted to directly address in our study. Because all centers have to meet a minimum annual volume requirement (>125 operations per year), differences in quality are unlikely related to the volume-outcome association seen in other surgical procedures.28 Moreover, by stratifying centers by their operative volume, we found wide variation among low-volume, medium-volume, and high-volume centers. An alternative explanation for variation in complication rates may be the differences in case mix across different centers (ie, proportion of sleeve gastrectomies vs proportion of gastric bypass). We accounted for procedure type in our modeling and still found considerable variation, suggesting that additional mechanisms not captured in our administrative data are at work.

Two other potential mechanisms not captured in our data may explain the variation in serious complications. The first may be the skill of the surgeon performing the procedure. Pilot data of direct video observation of experienced bariatric surgeons found that their technical skill levels varied widely.29 Moreover, this skill variation was strongly associated with rates of postoperative outcomes. The second possible explanation for the variation in serious complications may be inconsistent adherence to bariatric care pathways. Currently, there is conflicting evidence as to what extent these pathways are being adopted and if these pathways correlate with improved outcomes.30-32 Both surgical skill and care pathway adherence are active areas of research that may identify approaches to reduce variation after bariatric operations.

Our findings have multiple implications for both practice and policy to improve the safety of bariatric surgery. First, hospital accreditation alone does not ensure uniform high-quality care. Specifically, because most bariatric procedures are now performed at accredited centers, patients and payers can no longer use this designation to identify the highest-performing hospitals. Second, all the resources and infrastructure are already in place to identify true centers of excellence. Because the MBSAQIP does collect outcomes, including serious complications, from all accredited centers, the program could readily stratify nearly all bariatric centers into low-quality, medium-quality, and high-quality performers. Such information could be valuable for patients who are choosing a center for bariatric surgery, for surgeons who are identifying quality improvement targets, and for administrators of large hospital systems who are optimizing their network service lines. Third, now that nearly all bariatric centers are enrolled in MBSAQIP, they can use the program as a platform for widespread quality improvement. The MBSAQIP has launched multiple efforts to reduce readmissions and enhance recovery pathways.33,34 Given that the population eligible for bariatric procedures far outnumbers the available surgeon workforce, this general improvement approach may be favorable to selective referral. Whether such widespread quality improvement efforts can reduce the observed variation in serious complication rates remains to be seen.

Limitations

These findings should be interpreted in the context of the study’s multiple limitations. First, we used administrative data that are inherently limited in identifying complication codes. We used multiple strategies to make our coding more sensitive and specific, including using outcomes from the Complication Screening Project.35,36 In addition, we added a length of stay criterion to give the complication outcomes clinical face validity.12,13 Second, because of state privacy laws and the nature of the data available through the Healthcare Cost and Utilization Project’s State Inpatient Database, we were able to examine centers in only 12 states. These states, however, were geographically diverse with a broad case mix and provided a representative sample of bariatric centers across the country. Third, our data set did not allow evaluation of surgeon-level outcomes. Future studies with this level of specificity may further help identify sources and offer explanations for the variation we observed. Finally, using claims data did not allow us to measure other important outcomes associated with bariatric procedures, such as patient satisfaction and weight loss. While these measures are important to studying quality, they are secondary to the patient safety concerns raised by the perioperative complications mentioned here. In contrast, administrative claims data do carry unique benefits that are central to this study. Mostly notably, claims data allowed us to generate a large enough sample to detect meaningful differences across centers, which other data sources may not be powered to identify.

Conclusions

This study found widespread variation in postoperative outcomes among bariatric centers of excellence. As centers and payers look toward improving care for the bariatric population, focus should be aimed at reducing variation where most procedures are performed. Given that a large proportion of low-quality centers are geographically located near higher performing centers, opportunities for improvement through selective referral or regional collaboration should be considered.

Back to top
Article Information

Corresponding Author: Andrew M. Ibrahim, MD, MSc, Center for Healthcare Outcomes and Policy, University of Michigan, 2800 Plymouth Ave, Bldg 10-G016, Ann Arbor, MI 48109 (iandrew@umich.edu).

Accepted for Publication: January 14, 2017.

Published Online: April 26, 2017. doi:10.1001/jamasurg.2017.0542

Author Contributions: Dr Ibrahim and and Mr Thumma had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Ibrahim, Ghaferi, Dimick.

Acquisition, analysis, or interpretation of data: Ibrahim, Thumma, Dimick.

Drafting of the manuscript: Ibrahim, Thumma, Dimick.

Critical revision of the manuscript for important intellectual content: Ibrahim, Ghaferi, Dimick.

Statistical analysis: Thumma, Dimick.

Obtained funding: Dimick.

Administrative, technical, or material support: Ghaferi, Dimick.

Study supervision: Ghaferi, Dimick.

Conflict of Interest Disclosures: Dr Dimick has a financial interest in ArborMetrix, Inc, which had no role in the analysis herein. No other disclosures were reported.

Funding/Support: This study was supported in part by funding from the Robert Wood Johnson Foundation and the US Department of Veterans Affairs (Dr Ibrahim); by grants 5K08HS02362 and P30HS024403 from the Agency for Healthcare Research and Quality and Patient Centered Outcomes Research Institute Award CE-1304-6596 (Dr Ghaferi); and by grant R01AG039434-04 from the National Institute on Aging of the National Institutes of Health (Dr Dimick).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: Dr Dimick is the Surgical Innovation Editor of JAMA Surgery but was not involved in the editorial review or decision to accept the manuscript for publication.

References
1.
American Society for Metabolic and Bariatric Surgery. Estimate of bariatric surgery numbers, 2011-2015. https://asmbs.org/resources/estimate-of-bariatric-surgery-numbers. Accessed September 20, 2016.
2.
Flum  DR, Salem  L, Elrod  JA, Dellinger  EP, Cheadle  A, Chan  L.  Early mortality among Medicare beneficiaries undergoing bariatric surgical procedures.  JAMA. 2005;294(15):1903-1908.PubMedGoogle ScholarCrossref
3.
Hollenbeak  CS, Rogers  AM, Barrus  B, Wadiwala  I, Cooney  RN.  Surgical volume impacts bariatric surgery mortality: a case for centers of excellence.  Surgery. 2008;144(5):736-743.PubMedGoogle ScholarCrossref
4.
Pratt  GM, McLees  B, Pories  WJ.  The ASBS Bariatric Surgery Centers of Excellence program: a blueprint for quality improvement.  Surg Obes Relat Dis. 2006;2(5):497-503.PubMedGoogle ScholarCrossref
5.
Schirmer  B, Jones  DB.  The American College of Surgeons Bariatric Surgery Center Network: establishing standards.  Bull Am Coll Surg. 2007;92(8):21-27.PubMedGoogle Scholar
6.
Gebhart  A, Young  M, Phelan  M, Nguyen  NT.  Impact of accreditation in bariatric surgery.  Surg Obes Relat Dis. 2014;10(5):767-773.PubMedGoogle ScholarCrossref
7.
Dimick  JB, Osborne  NH, Nicholas  L, Birkmeyer  JD.  Identifying high-quality bariatric surgery centers: hospital volume or risk-adjusted outcomes?  J Am Coll Surg. 2009;209(6):702-706.PubMedGoogle ScholarCrossref
8.
Chhabra  KR, Dimick  JB.  Hospital networks and value-based payment: fertile ground for regionalizing high-risk surgery.  JAMA. 2015;314(13):1335-1336.PubMedGoogle ScholarCrossref
9.
Healthcare Cost and Utilization Project. Overview of the State Inpatient Databases (SID). http://www.hcup-us.ahrq.gov/sidoverview.jsp. Accessed January 11, 2016.
10.
Santry  HP, Gillen  DL, Lauderdale  DS.  Trends in bariatric surgical procedures.  JAMA. 2005;294(15):1909-1917.PubMedGoogle ScholarCrossref
11.
Krell  RW, Finks  JF, English  WJ, Dimick  JB.  Profiling hospitals on bariatric surgery quality: which outcomes are most reliable ?  J Am Coll Surg. 2014;219(4):725-734.e3.PubMedGoogle ScholarCrossref
12.
Scally  CP, Thumma  JR, Birkmeyer  JD, Dimick  JB.  Impact of surgical quality improvement on payments in Medicare patients.  Ann Surg. 2015;262(2):249-252.PubMedGoogle ScholarCrossref
13.
Osborne  NH, Nicholas  LH, Ryan  AM, Thumma  JR, Dimick  JB.  Association of hospital participation in a quality reporting program with surgical outcomes and expenditures for Medicare beneficiaries.  JAMA. 2015;313(5):496-504.PubMedGoogle ScholarCrossref
14.
Livingston  EH.  Procedure incidence and in-hospital complication rates of bariatric surgery in the United States.  Am J Surg. 2004;188(2):105-110.PubMedGoogle ScholarCrossref
15.
Ibrahim  AM, Hughes  TG, Thumma  JR, Dimick  JB.  Association of hospital critical access status with surgical outcomes and expenditures among Medicare beneficiaries.  JAMA. 2016;315(19):2095-2103.PubMedGoogle ScholarCrossref
16.
Dartmouth Institute for Health Policy and Clinical Practice. Research methods. In: The Dartmouth Atlas of Health Care. Lebanon, NH: Trustees of Dartmouth College; 2017. http://www.dartmouthatlas.org/downloads/methods/research_methods.pdf. Accessed August 12, 2016.
17.
Elixhauser  A, Steiner  C, Harris  DR, Coffey  RM.  Comorbidity measures for use with administrative data.  Med Care. 1998;36(1):8-27.PubMedGoogle ScholarCrossref
18.
Southern  DA, Quan  H, Ghali  WA.  Comparison of the Elixhauser and Charlson/Deyo methods of comorbidity measurement in administrative data.  Med Care. 2004;42(4):355-360.PubMedGoogle ScholarCrossref
19.
Jones  HE, Spiegelhalter  DJ.  The identification of “unusual” health-care providers from a hierarchical model.  Am Stat. 2011;65(3):154-163.Google ScholarCrossref
20.
Dimick  JB, Staiger  DO, Birkmeyer  JD.  Ranking hospitals on surgical mortality: the importance of reliability adjustment.  Health Serv Res. 2010;45(6, pt 1):1614-1629.PubMedGoogle ScholarCrossref
21.
Grenda  TR, Krell  RW, Dimick  JB.  Reliability of hospital cost profiles in inpatient surgery.  Surgery. 2015;159(2):375-380.PubMedGoogle Scholar
22.
Elixhauser Comorbidity Software, version 3.7. Healthcare Cost and Utilization Project website. https://www.hcup-us.ahrq.gov/toolssoftware/comorbidity/comorbidity.jsp. Accessed March 17, 2017.
23.
American Hospital Association. AHA Annual Survey Database. http://www.ahadataviewer.com/book-cd-products/AHA-Survey. Accessed December 6, 2016.
24.
Saleh  F, Doumouras  AG, Gmora  S, Anvari  M, Hong  D.  Outcomes the Ontario Bariatric Network: a cohort study.  CMAJ Open. 2016;4(3):E383-E389.PubMedGoogle ScholarCrossref
25.
Morton  JM, Garg  T, Nguyen  N.  Does hospital accreditation impact bariatric surgery safety?  Ann Surg. 2014;260(3):504-508.PubMedGoogle ScholarCrossref
26.
O’Brien  PE.  Controversies in bariatric surgery.  Br J Surg. 2015;102(6):611-618.PubMedGoogle ScholarCrossref
27.
Berger  ER, Clements  RH, Morton  JM,  et al.  The impact of different surgical techniques on outcomes in laparoscopic sleeve gastrectomies: the first report from the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP).  Ann Surg. 2016;264(3):464-473.PubMedGoogle ScholarCrossref
28.
Birkmeyer  JD, Siewers  AE, Finlayson  EV,  et al.  Hospital volume and surgical mortality in the United States.  N Engl J Med. 2002;346(15):1128-1137.PubMedGoogle ScholarCrossref
29.
Birkmeyer  JD, Finks  JF, O’Reilly  A,  et al; Michigan Bariatric Surgery Collaborative.  Surgical skill and complication rates after bariatric surgery.  N Engl J Med. 2013;369(15):1434-1442.PubMedGoogle ScholarCrossref
30.
Sinha  A, Jayaraman  L, Punhani  D, Chowbey  P.  Enhanced recovery after bariatric surgery in the severely obese, morbidly obese, super-morbidly obese and super-super morbidly obese using evidence-based clinical pathways: a comparative study.  Obes Surg. 2017;27(3):560-568.PubMedGoogle ScholarCrossref
31.
Telem  DA, Majid  SF, Powers  K, DeMaria  E, Morton  J, Jones  DB.  Assessing national provision of care: variability in bariatric clinical care pathways.  Surg Obes Relat Dis. 2017;13(2):281-284.PubMedGoogle ScholarCrossref
32.
Yeats  M, Wedergren  S, Fox  N, Thompson  JS.  The use and modification of clinical pathways to achieve specific outcomes in bariatric surgery.  Am Surg. 2005;71(2):152-154.PubMedGoogle Scholar
33.
Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program. Decreasing Readmissions Through Opportunities Provided (D.R.O.P) [video]. http://reports.nsqip.facs.org/MBSAQIPDropVideo/. Accessed August 12, 2016.
34.
Morton  J.  The first metabolic and bariatric surgery accreditation and quality improvement program quality initiative: decreasing readmissions through opportunities provided.  Surg Obes Relat Dis. 2014;10(3):377-378.PubMedGoogle ScholarCrossref
35.
Iezzoni  LI, Daley  J, Heeren  T,  et al.  Identifying complications of care using administrative data.  Med Care. 1994;32(7):700-715.PubMedGoogle ScholarCrossref
36.
Weingart  SN, Iezzoni  LI, Davis  RB,  et al.  Use of administrative data to find substandard care: validation of the complications screening program.  Med Care. 2000;38(8):796-806.PubMedGoogle ScholarCrossref
×