PAPERS OF THE 134TH ASA ANNUAL MEETING
Does Hospital Accreditation Impact Bariatric Surgery Safety? John M. Morton, MD, MPH,∗ Trit Garg, BA,∗ and Ninh Nguyen, MD† Objective: To evaluate the impact of hospital accreditation upon bariatric surgery outcomes. Background: Since 2004, the American College of Surgeons and the American Society of Metabolic and Bariatric Surgery have accredited bariatric hospitals. Few studies have evaluated the impact of hospital accreditation on all bariatric surgery outcomes. Methods: Bariatric surgery hospitalizations were identified using International Classification of Diseases, Ninth Revision (ICD9) codes in the 2010 Nationwide Inpatient Sample (NIS). Hospital names and American Hospital Association (AHA) codes were used to identify accredited bariatric centers. Relevant ICD9 codes were used for identifying demographics, length of stay (LOS), total charges, mortality, complications, and failure to rescue (FTR) events. Results: There were 117,478 weighted bariatric patient discharges corresponding to 235 unique hospitals in the 2010 NIS data set. A total of 72,615 (61.8%) weighted discharges, corresponding to 145 (61.7%) named or AHAidentifiable hospitals were included. Among the 145 hospitals, 66 (45.5%) were unaccredited and 79 (54.5%) accredited. Compared with accredited centers, unaccredited centers had a higher mean LOS (2.25 vs 1.99 days, P < 0.0001), as well as total charges ($51,189 vs $42,212, P < 0.0001). Incidence of any complication was higher at unaccredited centers than at accredited centers (12.3% vs 11.3%, P = 0.001), as was mortality (0.13% vs 0.07%, P = 0.019) and FTR (0.97% vs 0.55%, P = 0.046). Multivariable logistic regression analysis identified unaccredited status as a positive predictor of incidence of complication [odds ratio (OR) = 1.08, P < 0.0001], as well as mortality (OR = 2.13, P = 0.013). Conclusions and Relevance: Hospital accreditation status is associated with safer outcomes, shorter LOS, and lower total charges after bariatric surgery. Keywords: accreditation, bariatric, database, obesity, outcomes, quality, safety, surgery (Ann Surg 2014;260:504–509)
B
ariatric surgery has proven to be the most effective and enduring option in treating the morbidly obese.1,2 Although mortality and other complications are risks associated with bariatric surgery, they have shown to be inversely correlated with the volume of cases performed by the surgeon.3–8 Because of this demonstrated volumeoutcome relationship, the American College of Surgeons (ACS) and the American Society of Metabolic and Bariatric Surgery (ASMBS) initiated an accreditation process in 2004 whereby hospitals are designated as Centers of Excellence (COE) if they perform a high volume of bariatric surgery, defined as more than 125 cases per year, and are equipped with appropriate equipment, experienced surgeons, and a multidisciplinary team to care for patients.9 This policy of facility
From the ∗ Bariatric and Minimally Invasive (BMI) Surgery, Stanford School of Medicine, Stanford, CA; and †Gastrointestinal Surgery, University of California, Irvine. Disclosure: The authors declare no conflicts of interest. Reprints: John M. Morton, MD, MPH, Bariatric and Minimally Invasive (BMI) Surgery, Stanford School of Medicine, 300 Pasteur Drive, H3680, Stanford, CA 94305 ([email protected]). C 2014 by Lippincott Williams & Wilkins Copyright ISSN: 0003-4932/14/26003-0504 DOI: 10.1097/SLA.0000000000000891
504 | www.annalsofsurgery.com
accreditation was established to improve quality of care and patient outcomes. It has previously been shown through utilization of a variety of data sets that accreditation improves outcomes after bariatric surgery. For instance, Kohn et al used the largest, all-payer database in the USA. With 102,069 surgeries analyzed from over 8 years to show that accreditation status was associated with a significant decrease in complications.10 Similarly, Nguyen et al11 used the University Health Consortium database with 35,284 bariatric surgery patients to demonstrate that accredited centers were associated with a significant reduction in mortality, shorter length of stay (LOS), and lower cost. Finally, Kwon et al12 used the national MarketScan Commercial Claims and Encounter Database to show that accredited centers had significant reductions in inpatient mortality, 90-day reoperations, complications, and readmissions. Most recently in 2013, Jafari et al13 used the Nationwide Inpatient Sample (NIS) data set to show that for laparoscopic stapled procedures—Laparoscopic Roux-en-Y gastric bypass (LRYGB) and Laparoscopic Sleeve Gastrectomy (LSG)—high-volume centers (HVCs), defined by an annual volume of laparoscopic stapled procedures of greater than 50, that were also accredited had significantly lower in-patient mortality, and also higher complications. This combination of higher complications but lower mortality for accredited centers may suggest that accredited centers are able to better identify patients with complications and prevent mortality, and thus have a lower rate of failure to rescue (FTR). However, further research is needed to specifically assess differences in FTR between accredited and unaccredited centers. Three studies have been cited against accreditation.14–16 However, the Livingston study predated accreditation and the Michigan study compared accredited versus nonaccredited hospitals in the Michigan Bariatric Surgery Collaborative where participation is predicated on criteria that are indistinguishable from national accreditation criteria. Finally, the Dimick study did not address mortality, FTR, and relied upon a little-used econometric analytic technique that utilized a non-Medicare population as a control group for accreditation. The Dimick study did not account for the fact that the non-Medicare population was also exposed to private payor accreditation also indistinguishable from Centers for Medicare and Medicaid Services (CMS) accreditation. To our knowledge, no study has used the NIS data set to evaluate the impact of accreditation on outcomes after all 3 types of bariatric surgery—LRYGB, LSG, and Laparoscopic Adjustable Gastric Banding (LAGB). Moreover, although previous studies investigating the impact of accreditation have analyzed mortality and complications, no study to our knowledge has investigated patient safety indicators (PSIs), such as FTR, as developed by the Agency for Healthcare Research and Quality (AHRQ). This study aimed to use the most contemporary NIS data set from 2010 to evaluate the impact of accreditation on outcomes after all types of bariatric surgery, including complications, mortality, and FTR.
METHODS Data Source The database used was the 2010 NIS, created as part of the AHRQ’s Healthcare Research Cost and Utilization Project. The NIS Annals of Surgery r Volume 260, Number 3, September 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
is the largest, publicly available, all-payer, inpatient administrative database, containing 5 to 8 million hospitalization records from about 1000 hospitals in more than 37 different states. The data set represents approximately 20% of community hospitals in the United States and contains hospital descriptors such as geographic region, size, rural/urban location, and teaching status. Data elements on the patient-level include primary and secondary diagnoses codes, demographics, procedures performed, LOS, total charges, and discharge status. Also included are sample weights that are representative estimates of national inpatient use.17
Study Sample Patients were identified retrospectively using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure codes for LRYGB (44.38), LAGB (44.95), and LSG (43.82, 43.89, and 44.68), with a confirmatory diagnosis code for morbid obesity (278). Patients were excluded if they were transfers, as per their admission status, or if they were younger than 18 years. Using ICD-9-CM diagnosis codes, patients were excluded if they had malignant neoplasm of digestive organs or peritoneum (150– 159), inflammatory bowel disease (555–556), noninfectious colitis (557–558), or familial adenomatous polyposis (211). Hospitals were excluded if no name was provided in the NIS data set. Using hospital names and American Hospital Association (AHA) identification codes, hospitals were identified as accredited or nonaccredited based on the list published by the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) Web site and their status in 2010 was confirmed.9
Patient and Hospital Characteristics Patient age, sex, race, and insurance status were examined. Using relevant ICD-9-CM diagnosis codes, a modification of the Charlson Comorbidity Index was calculated with possible values of 0, 1, or ≥2.18 Hospitals were classified as HVCs if total bariatric volume was greater than or equal to 125 as this was the volume standard in 2010. Hospital size and teaching status were also recorded.
Outcomes Our outcomes of interest were total hospital bariatric volume, patient complications, mortality, LOS, total charges, and FTR. Complications were identified using relevant ICD-9-CM codes as described previously.19 FTR was defined by the AHRQ’s PSIs. The PSIs are a set of indicators providing information on potential inpatient adverse events after surgeries, medical procedures, and childbirth. The PSIs were developed through a comprehensive literature review, ICD-9-CM code study, review by a physician panel, risk adjustment measures, and empirical analyses. The PSIs screen for problems that patients experience as a result of exposure to the health care system and that are likely amenable to prevention by changes at the system or provider level. Our PSI of interest, FTR, captures death among those patients with potential complications of care defined by a list of ICD-9-CM diagnosis codes available on the AHRQ Web site.20
Statistical Analysis Sample weights designed by the NIS were used to create national representative estimates of all outcomes of interest. Continuous variables were analyzed using t test, and categorical variables were analyzed using χ 2 test. Multivariate, multilevel, mixed-effects logistic regression analyses were utilized for mortality and complications with hospital clustering adjustment to correct for potential confounders. All analyses were performed with SAS software version 9.3 (SAS, Carey, NC). C 2014 Lippincott Williams & Wilkins
Bariatric Accreditation Improves Safety
RESULTS There were 117,478 weighted bariatric patient discharges corresponding to 235 unique hospitals in the 2010 NIS data set. A total of 72,615 (61.8% of total) weighted discharges, corresponding to 145 (61.7% of total) name or AHA-identifiable hospitals were included in the analysis. Among the 145 hospitals included, 66 (45.5%) were identified as being unaccredited, corresponding to a total of 12,366 (17.0%) weighted patient discharges, and 79 (54.5%) as accredited, corresponding to a total of 60,249 (83.0%) weighted discharges. Shown in Table 1 are characteristics of accredited and unaccredited hospitals. Mean volume of bariatric cases performed in 2010 was similar in both unaccredited and accredited centers (279 vs 265, P = 0.909). A lower percentage of unaccredited centers were HVCs (52.8% vs 80.8%, P < 0.0001), whereas a greater proportion of unaccredited centers were teaching hospitals (66.2% vs 58.1%, P < 0.0001). Shown in Table 2 are characteristics of patients undergoing bariatric procedures at unaccredited and accredited centers. The distribution of type of bariatric procedure was significantly different between unaccredited and accredited centers (P < 0.0001), with percent of LRYGB being more frequent in unaccredited centers (69.5% vs 60.3%), percent of LAGB more in accredited centers (15.6% vs 25.7%), and percent of LSG being similar in both (14.8% vs 14.0%). Mean age of patients was similar at both unaccredited and accredited centers (44.7 vs 44.6, P = 0.522), whereas distribution of age groups significantly differed with more elderly patients at accredited centers (P < 0.0001). Percentage of female patients was similar at both unaccredited and accredited centers (78.8% vs 78.3%, P = 0.186). The distribution of insurance status significantly differed between unaccredited and accredited centers (P < 0.0001), with private insurance being more frequent at unaccredited centers (70.3% vs 68.8%), as well as self-pay (10.1% vs 2.57%), whereas Medicare more at accredited centers (8.75% vs 13.5%) as well as Medicaid (6.67% vs 11.6%). Charlson comorbidity scores were distributed differently between unaccredited and accredited centers (P < 0.0001), with accredited centers having a higher percentage of patients with a Charlson comorbidity score of 3 or higher (1.22% vs 3.02%). Shown in Table 3 are differences in LOS between accredited and unaccredited centers. Compared with accredited centers, unaccredited centers had a higher mean LOS (2.25 vs 1.99 days, P < 0.0001). Moreover, the distribution of LOS was significantly different (P < 0.0001) with unaccredited centers having a lesser frequency of 1 day LOS (30.7% vs 39.8%), but greater frequency of 2 days LOS (45.8% vs 38.1%) as well as 5 or more days LOS (4.18% vs 3.05%). Shown in Table 4 are our outcomes of interest. Compared with accredited centers, unaccredited centers total charges were less ($51,189 vs $42,212, P < 0.0001). Incidence of any complication was higher at unaccredited centers than accredited (12.3% vs 11.3%, P = 0.001), as was mortality (0.13% vs 0.07%, P = 0.019) and FTR (0.97% vs 0.55%, P = 0.046). A detailed summary of incidence of individual complications by accreditation status is provided in Table 5. Most notably, unaccredited centers had a significantly higher
TABLE 1. Hospital Characteristics by Accreditation Status, 2010 Hospitals, no. (%) Mean volume HVC, % Teaching hospital, %
Unaccredited
Accredited
P
66 (45.5) 279 52.8 66.2
79 (54.5) 265 80.8 58.1
0.909
Does Hospital Accreditation Impact Bariatric Surgery Safety? John M. Morton, MD, MPH,∗ Trit Garg, BA,∗ and Ninh Nguyen, MD† Objective: To evaluate the impact of hospital accreditation upon bariatric surgery outcomes. Background: Since 2004, the American College of Surgeons and the American Society of Metabolic and Bariatric Surgery have accredited bariatric hospitals. Few studies have evaluated the impact of hospital accreditation on all bariatric surgery outcomes. Methods: Bariatric surgery hospitalizations were identified using International Classification of Diseases, Ninth Revision (ICD9) codes in the 2010 Nationwide Inpatient Sample (NIS). Hospital names and American Hospital Association (AHA) codes were used to identify accredited bariatric centers. Relevant ICD9 codes were used for identifying demographics, length of stay (LOS), total charges, mortality, complications, and failure to rescue (FTR) events. Results: There were 117,478 weighted bariatric patient discharges corresponding to 235 unique hospitals in the 2010 NIS data set. A total of 72,615 (61.8%) weighted discharges, corresponding to 145 (61.7%) named or AHAidentifiable hospitals were included. Among the 145 hospitals, 66 (45.5%) were unaccredited and 79 (54.5%) accredited. Compared with accredited centers, unaccredited centers had a higher mean LOS (2.25 vs 1.99 days, P < 0.0001), as well as total charges ($51,189 vs $42,212, P < 0.0001). Incidence of any complication was higher at unaccredited centers than at accredited centers (12.3% vs 11.3%, P = 0.001), as was mortality (0.13% vs 0.07%, P = 0.019) and FTR (0.97% vs 0.55%, P = 0.046). Multivariable logistic regression analysis identified unaccredited status as a positive predictor of incidence of complication [odds ratio (OR) = 1.08, P < 0.0001], as well as mortality (OR = 2.13, P = 0.013). Conclusions and Relevance: Hospital accreditation status is associated with safer outcomes, shorter LOS, and lower total charges after bariatric surgery. Keywords: accreditation, bariatric, database, obesity, outcomes, quality, safety, surgery (Ann Surg 2014;260:504–509)
B
ariatric surgery has proven to be the most effective and enduring option in treating the morbidly obese.1,2 Although mortality and other complications are risks associated with bariatric surgery, they have shown to be inversely correlated with the volume of cases performed by the surgeon.3–8 Because of this demonstrated volumeoutcome relationship, the American College of Surgeons (ACS) and the American Society of Metabolic and Bariatric Surgery (ASMBS) initiated an accreditation process in 2004 whereby hospitals are designated as Centers of Excellence (COE) if they perform a high volume of bariatric surgery, defined as more than 125 cases per year, and are equipped with appropriate equipment, experienced surgeons, and a multidisciplinary team to care for patients.9 This policy of facility
From the ∗ Bariatric and Minimally Invasive (BMI) Surgery, Stanford School of Medicine, Stanford, CA; and †Gastrointestinal Surgery, University of California, Irvine. Disclosure: The authors declare no conflicts of interest. Reprints: John M. Morton, MD, MPH, Bariatric and Minimally Invasive (BMI) Surgery, Stanford School of Medicine, 300 Pasteur Drive, H3680, Stanford, CA 94305 ([email protected]). C 2014 by Lippincott Williams & Wilkins Copyright ISSN: 0003-4932/14/26003-0504 DOI: 10.1097/SLA.0000000000000891
504 | www.annalsofsurgery.com
accreditation was established to improve quality of care and patient outcomes. It has previously been shown through utilization of a variety of data sets that accreditation improves outcomes after bariatric surgery. For instance, Kohn et al used the largest, all-payer database in the USA. With 102,069 surgeries analyzed from over 8 years to show that accreditation status was associated with a significant decrease in complications.10 Similarly, Nguyen et al11 used the University Health Consortium database with 35,284 bariatric surgery patients to demonstrate that accredited centers were associated with a significant reduction in mortality, shorter length of stay (LOS), and lower cost. Finally, Kwon et al12 used the national MarketScan Commercial Claims and Encounter Database to show that accredited centers had significant reductions in inpatient mortality, 90-day reoperations, complications, and readmissions. Most recently in 2013, Jafari et al13 used the Nationwide Inpatient Sample (NIS) data set to show that for laparoscopic stapled procedures—Laparoscopic Roux-en-Y gastric bypass (LRYGB) and Laparoscopic Sleeve Gastrectomy (LSG)—high-volume centers (HVCs), defined by an annual volume of laparoscopic stapled procedures of greater than 50, that were also accredited had significantly lower in-patient mortality, and also higher complications. This combination of higher complications but lower mortality for accredited centers may suggest that accredited centers are able to better identify patients with complications and prevent mortality, and thus have a lower rate of failure to rescue (FTR). However, further research is needed to specifically assess differences in FTR between accredited and unaccredited centers. Three studies have been cited against accreditation.14–16 However, the Livingston study predated accreditation and the Michigan study compared accredited versus nonaccredited hospitals in the Michigan Bariatric Surgery Collaborative where participation is predicated on criteria that are indistinguishable from national accreditation criteria. Finally, the Dimick study did not address mortality, FTR, and relied upon a little-used econometric analytic technique that utilized a non-Medicare population as a control group for accreditation. The Dimick study did not account for the fact that the non-Medicare population was also exposed to private payor accreditation also indistinguishable from Centers for Medicare and Medicaid Services (CMS) accreditation. To our knowledge, no study has used the NIS data set to evaluate the impact of accreditation on outcomes after all 3 types of bariatric surgery—LRYGB, LSG, and Laparoscopic Adjustable Gastric Banding (LAGB). Moreover, although previous studies investigating the impact of accreditation have analyzed mortality and complications, no study to our knowledge has investigated patient safety indicators (PSIs), such as FTR, as developed by the Agency for Healthcare Research and Quality (AHRQ). This study aimed to use the most contemporary NIS data set from 2010 to evaluate the impact of accreditation on outcomes after all types of bariatric surgery, including complications, mortality, and FTR.
METHODS Data Source The database used was the 2010 NIS, created as part of the AHRQ’s Healthcare Research Cost and Utilization Project. The NIS Annals of Surgery r Volume 260, Number 3, September 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
is the largest, publicly available, all-payer, inpatient administrative database, containing 5 to 8 million hospitalization records from about 1000 hospitals in more than 37 different states. The data set represents approximately 20% of community hospitals in the United States and contains hospital descriptors such as geographic region, size, rural/urban location, and teaching status. Data elements on the patient-level include primary and secondary diagnoses codes, demographics, procedures performed, LOS, total charges, and discharge status. Also included are sample weights that are representative estimates of national inpatient use.17
Study Sample Patients were identified retrospectively using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure codes for LRYGB (44.38), LAGB (44.95), and LSG (43.82, 43.89, and 44.68), with a confirmatory diagnosis code for morbid obesity (278). Patients were excluded if they were transfers, as per their admission status, or if they were younger than 18 years. Using ICD-9-CM diagnosis codes, patients were excluded if they had malignant neoplasm of digestive organs or peritoneum (150– 159), inflammatory bowel disease (555–556), noninfectious colitis (557–558), or familial adenomatous polyposis (211). Hospitals were excluded if no name was provided in the NIS data set. Using hospital names and American Hospital Association (AHA) identification codes, hospitals were identified as accredited or nonaccredited based on the list published by the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) Web site and their status in 2010 was confirmed.9
Patient and Hospital Characteristics Patient age, sex, race, and insurance status were examined. Using relevant ICD-9-CM diagnosis codes, a modification of the Charlson Comorbidity Index was calculated with possible values of 0, 1, or ≥2.18 Hospitals were classified as HVCs if total bariatric volume was greater than or equal to 125 as this was the volume standard in 2010. Hospital size and teaching status were also recorded.
Outcomes Our outcomes of interest were total hospital bariatric volume, patient complications, mortality, LOS, total charges, and FTR. Complications were identified using relevant ICD-9-CM codes as described previously.19 FTR was defined by the AHRQ’s PSIs. The PSIs are a set of indicators providing information on potential inpatient adverse events after surgeries, medical procedures, and childbirth. The PSIs were developed through a comprehensive literature review, ICD-9-CM code study, review by a physician panel, risk adjustment measures, and empirical analyses. The PSIs screen for problems that patients experience as a result of exposure to the health care system and that are likely amenable to prevention by changes at the system or provider level. Our PSI of interest, FTR, captures death among those patients with potential complications of care defined by a list of ICD-9-CM diagnosis codes available on the AHRQ Web site.20
Statistical Analysis Sample weights designed by the NIS were used to create national representative estimates of all outcomes of interest. Continuous variables were analyzed using t test, and categorical variables were analyzed using χ 2 test. Multivariate, multilevel, mixed-effects logistic regression analyses were utilized for mortality and complications with hospital clustering adjustment to correct for potential confounders. All analyses were performed with SAS software version 9.3 (SAS, Carey, NC). C 2014 Lippincott Williams & Wilkins
Bariatric Accreditation Improves Safety
RESULTS There were 117,478 weighted bariatric patient discharges corresponding to 235 unique hospitals in the 2010 NIS data set. A total of 72,615 (61.8% of total) weighted discharges, corresponding to 145 (61.7% of total) name or AHA-identifiable hospitals were included in the analysis. Among the 145 hospitals included, 66 (45.5%) were identified as being unaccredited, corresponding to a total of 12,366 (17.0%) weighted patient discharges, and 79 (54.5%) as accredited, corresponding to a total of 60,249 (83.0%) weighted discharges. Shown in Table 1 are characteristics of accredited and unaccredited hospitals. Mean volume of bariatric cases performed in 2010 was similar in both unaccredited and accredited centers (279 vs 265, P = 0.909). A lower percentage of unaccredited centers were HVCs (52.8% vs 80.8%, P < 0.0001), whereas a greater proportion of unaccredited centers were teaching hospitals (66.2% vs 58.1%, P < 0.0001). Shown in Table 2 are characteristics of patients undergoing bariatric procedures at unaccredited and accredited centers. The distribution of type of bariatric procedure was significantly different between unaccredited and accredited centers (P < 0.0001), with percent of LRYGB being more frequent in unaccredited centers (69.5% vs 60.3%), percent of LAGB more in accredited centers (15.6% vs 25.7%), and percent of LSG being similar in both (14.8% vs 14.0%). Mean age of patients was similar at both unaccredited and accredited centers (44.7 vs 44.6, P = 0.522), whereas distribution of age groups significantly differed with more elderly patients at accredited centers (P < 0.0001). Percentage of female patients was similar at both unaccredited and accredited centers (78.8% vs 78.3%, P = 0.186). The distribution of insurance status significantly differed between unaccredited and accredited centers (P < 0.0001), with private insurance being more frequent at unaccredited centers (70.3% vs 68.8%), as well as self-pay (10.1% vs 2.57%), whereas Medicare more at accredited centers (8.75% vs 13.5%) as well as Medicaid (6.67% vs 11.6%). Charlson comorbidity scores were distributed differently between unaccredited and accredited centers (P < 0.0001), with accredited centers having a higher percentage of patients with a Charlson comorbidity score of 3 or higher (1.22% vs 3.02%). Shown in Table 3 are differences in LOS between accredited and unaccredited centers. Compared with accredited centers, unaccredited centers had a higher mean LOS (2.25 vs 1.99 days, P < 0.0001). Moreover, the distribution of LOS was significantly different (P < 0.0001) with unaccredited centers having a lesser frequency of 1 day LOS (30.7% vs 39.8%), but greater frequency of 2 days LOS (45.8% vs 38.1%) as well as 5 or more days LOS (4.18% vs 3.05%). Shown in Table 4 are our outcomes of interest. Compared with accredited centers, unaccredited centers total charges were less ($51,189 vs $42,212, P < 0.0001). Incidence of any complication was higher at unaccredited centers than accredited (12.3% vs 11.3%, P = 0.001), as was mortality (0.13% vs 0.07%, P = 0.019) and FTR (0.97% vs 0.55%, P = 0.046). A detailed summary of incidence of individual complications by accreditation status is provided in Table 5. Most notably, unaccredited centers had a significantly higher
TABLE 1. Hospital Characteristics by Accreditation Status, 2010 Hospitals, no. (%) Mean volume HVC, % Teaching hospital, %
Unaccredited
Accredited
P
66 (45.5) 279 52.8 66.2
79 (54.5) 265 80.8 58.1
0.909
