Trends in access to minimally invasive pancreaticoduodenectomy for pancreatic cancers.

BACKGROUND: Minimally invasive pancreaticoduodenectomy (MIPD), including robotic (RPD) and laparoscopy (LPD), is becoming more frequently employed in the management of pancreatic ductal adenocarcinoma (PDAC), though the majority of operations are still performed via open approach (OPD). Access to technologic advances often neglect the underserved. Whether disparities in access to MIPD exist, remain unclear. METHODS: The National Cancer Database (NCDB) was queried (2010-2020) for patients who underwent pancreatoduodenectomy for PDAC. Cochran-Armitage tests assessed for trends over time. Social determinants of health (SDH) were compared between approaches. Multinomial logistic models identified predictors of MIPD. RESULTS: Of 16,468 patients, 80.03 % underwent OPD and 19.97 % underwent MIPD (22.60 % robotic; 77.40 % laparoscopic). Black race negatively predicted LPD (vs white (OR 0.822; 95 % CI 0.701-0.964)). Predictors of RPD included Medicare/other government insurance (vs uninsured or Medicaid (OR 1.660; 95 % CI 1.123-2.454)) and private insurance (vs uninsured or Medicaid (OR 1.597; 95 % CI 1.090-2.340)). Early (2010-2014) vs late (2015-2020) diagnosis, stratified by race, demonstrated an increase in Non-White patients undergoing OPD (13.15 % vs 14.63 %; p = 0.016), but not LPD (11.41 % vs 13.57 %;p = 0.125) or RPD (14.15 % vs 15.23 %; p = 0.774). CONCLUSION: SDH predict surgical approach more than clinical stage, facility type, or comorbidity status. Disparities in race and insurance coverage are different between surgical approaches.

The Evaluation of Gallstone Disease in the Year Before Pancreatic Cancer Diagnosis.

INTRODUCTION: Patients with pancreatic cancer can present with a variety of insidious abdominal symptoms, complicating initial diagnosis. Early symptoms of pancreatic cancer often mirror those associated with gallstone disease, which has been demonstrated to be a risk factor for this malignancy. This study aims to compare the incidence of gallstone disease in the year before diagnosis of pancreatic ductal adenocarcinoma (PDAC) as compared to the general population, and evaluate the association of gallstone disease with stage at diagnosis and surgical intervention. METHODS: Patients with PDAC were identified from SEER-Medicare (2008-2015). The incidence of gallstone disease (defined as cholelithiasis, cholecystitis and/or cholecystectomy) in the 1 year before cancer diagnosis was compared to the annual incidence in an age-matched, sex-matched, and race-matched noncancer Medicare cohort. RESULTS: Among 14,654 patients with PDAC, 4.4% had gallstone disease in the year before cancer diagnosis. Among the noncancer controls (n = 14,654), 1.9% had gallstone disease. Both cohorts had similar age, sex and race distributions. PDAC patients with gallstone disease were diagnosed at an earlier stage (stage 0/I-II, 45.8% versus 38.1%, P < 0.0001) and a higher proportion underwent resection (22.7% versus 17.4%, P = 0.0004) compared to patients without gallstone disease. CONCLUSIONS: In the year before PDAC diagnosis, patients present with gallstone disease more often than the general population. Improving follow-up care and differential diagnosis strategies may help combat the high mortality rate in PDAC by providing an opportunity for earlier stage of diagnosis and earlier intervention.

Disparities in the Receipt of Recommended Curative Treatment for Patients with Early-Stage Hepatocellular Carcinoma.

BACKGROUND: In early-stage hepatocellular carcinoma (HCC), the receipt of recommended care is critical for long-term survival. Unfortunately, not all patients decide to undergo therapy. We sought to identify factors associated with the decision to decline recommended intervention among patients with early-stage HCC. METHODS: The National Cancer Database was queried for patients diagnosed with clinical stages I and II HCC (2004-2017). Cohorts were created based on the receipt or decline of recommended interventions-hepatectomy, liver transplantation, and ablation. Multivariable logistic regression identified predictors for declining intervention, and propensity score analysis was used to calculate the respective odds. Survival analysis was performed using the Kaplan-Meier method. RESULTS: Of 20,863 patients, 856 (4.1%) declined intervention. Patients who were documented as having declined intervention were more often Black (vs. other: OR, 1.3; 95% CI, 1.1-1.6; p = 0.0038), had Medicaid or no insurance (vs. Private, Medicare, or other government insurance): OR, 1.9; 95% CI, 1.6-2.3; p < 0.0001), lived in a low-income area (vs. other: OR, 1.4; 95% CI, 1.2-1.7; p < 0.0001), and received treatment at a non-academic center (vs. academic: OR, 2.1; 95% CI, 1.9-2.5; p < 0.0001). Patients who declined recommended interventions had worse survival compared to those who received treatment (22.9 vs. 59.2 months; p < 0.0001, respectively). CONCLUSIONS: Racial and socioeconomic disparities persist in the decision to undergo recommended treatment. Underutilization of treatment acts as a barrier to addressing racial and socioeconomic disparities in early-stage HCC outcomes.

Surveillance Patterns for Hepatocellular Carcinoma among Screening-Eligible Patients in the Medicare Population.

INTRODUCTION: Routine screening plays a critical role in the diagnosis of hepatocellular carcinoma (HCC), but not all patients undergo consistent surveillance. This study aims to evaluate surveillance patterns and their association with diagnosis stage and survival among Medicare patients at risk for HCC. PATIENTS AND METHODS: Patients with HCC and guideline-based screening eligibility who underwent imaging with ultrasound or abdominal magnetic resonance imaging (MRI) in the 2 years prior to diagnosis were identified from SEER-Medicare (2008-2015). Three surveillance cohorts were created: diagnostic (imaging only within 3 months prior), intermittent (imaging only once within 2 years prior, excluding diagnostic), and routine (at least two imaging encounters within 2 years prior, excluding diagnostic). Multivariable logistic regression was used to predict early-stage diagnosis (stage I-II), and 5-year survival was evaluated using the accelerated failure time method with Weibull distribution. RESULTS: Among 2261 eligible patients, 26.1% were classified as diagnostic, 15.8% as intermittent, and 58.1% as routine surveillance. The median age was 74 years (IQR 70-78 years). The majority of patients had a preexisting cirrhosis diagnosis (81.5%). Routine and intermittent, compared with diagnostic, surveillance were predictive of early-stage disease (routine: OR 2.05, 95% CI 1.64-2.56; intermittent: OR 1.43, 95% CI 1.07-1.90). Patients who underwent routine surveillance had significantly lower risk of mortality (HR 0.84, 95% CI 0.75-0.94) compared with the diagnostic group. CONCLUSIONS: A large proportion of screening-eligible patients do not undergo routine surveillance, which is associated with late-stage diagnosis and higher risk of mortality. These findings demonstrate the impact of timely and consistent healthcare access and can guide interventions for promoting surveillance among these patients.

Exclusion criteria: Evaluating the impact of enrollment requirements in SEER-Medicare research.

BACKGROUND: Enrollment criteria are routinely utilized in patient selection in SEER-Medicare but little is known about how this may be impacting research outcomes. This study evaluated demographics and survival among pancreatic cancer patients who are included and excluded from SEER-Medicare analyses. METHODS: Patients ≥66 years old with pancreatic cancer were identified (SEER-Medicare, 2008-2015). Two patient cohorts were compared: included (continuous enrollment in Medicare Parts A and B and no enrollment in Medicare Advantage), and excluded. Mortality was assessed using a Standardized Mortality Ratio. RESULTS: Among 49,017 patients with pancreatic cancer, 59.5% were in the included cohort. The excluded cohort was younger (median age 74 vs. 77) with more male (49.9% vs. 47.8%), non-white (33.0% vs. 21.3%) and urban-dwelling patients (91.0% vs. 85.0%). Those excluded had a higher mortality risk (SMR 1.06, 95%CI 1.04-1.07). CONCLUSIONS: There are significant differences in patient demographics and mortality among those who are and are not routinely included in SEER-Medicare analyses and our study provides a critical opportunity to quantify this potential bias.

Beyond insurance status: the impact of Medicaid expansion on the diagnosis of Hepatocellular Carcinoma.

BACKGROUND: Medicaid expansion has led to earlier stage diagnoses in several cancers but has not been studied in hepatocellular carcinoma (HCC), a disease with complex risk factors. We examined the effect of Medicaid expansion on the diagnosis of HCC and associations with county-level social vulnerability. METHODS: Patients with HCC <65 years of age were identified from the SEER database (2010-2016). County-level social vulnerability factors were obtained from the CDC SVI and BRFSS. A Difference-in-Difference analysis evaluated change in early-stage diagnoses (stage I-II) between expansion and non-expansion states. A Difference-in-Difference-in-Difference analysis evaluated expansion impact among counties with higher proportions of social vulnerability. RESULTS: Of 19,751 patients identified, 81.5% were in expansion states. Uninsured status decreased in expansion states (6.3%-2.4%, p < 0.0001) and remained unchanged in non-expansion states (12.7%-14.8%, p = 0.43). There was no significant difference in the incidence of early-stage diagnoses between expansion states and non-expansion states. Results were consistent when accounting for social vulnerability. CONCLUSION: Medicaid expansion was not associated with earlier stage diagnoses in patients with HCC, including those with higher social vulnerability. Unlike other cancers, expanded access did not translate into higher utilization of care in HCC, suggesting barriers on a multitude of levels.

The Effect of Hospital Versus Surgeon Volume on Short-Term Patient Outcomes After Pancreaticoduodenectomy: a SEER-Medicare Analysis.

BACKGROUND: The volume-outcome relationship has been well-established for pancreaticoduodenectomy (PD). It remains unclear if this is primarily driven by hospital volume or individual surgeon experience. OBJECTIVE: This study aimed to determine the relationship of hospital and surgeon volume on short-term outcomes of patients with pancreatic adenocarcinoma undergoing PD. METHODS: Patients >65 years of age who underwent PD for pancreatic adenocarcinoma were identified from the Surveillance, Epidemiology, and End Results (SEER)-Medicare database (2008-2015). Analyses were stratified by hospital volume and then surgeon volume, creating four volume cohorts: low-low (low hospital, low surgeon), low-high (low hospital, high surgeon), high-low (high hospital, low surgeon), high-high (high hospital, high surgeon). Propensity scores were created for the odds of undergoing surgery with high-volume surgeons. Following matching, multivariable analysis was used to assess the impact of surgeon volume on outcomes within each hospital volume cohort. RESULTS: In total, 2450 patients were identified: 54.3% were treated at high-volume hospitals (27.0% low-volume surgeons, 73.0% high-volume surgeons) and 45.7% were treated at low-volume hospitals (76.9% low-volume surgeons, 23.1% high-volume surgeons). On matched multivariable analysis, there were no significant differences in the risk of major complications, 90-day mortality, and 30-day readmission based on surgeon volume within the low and high hospital volume cohorts. CONCLUSION: Compared with surgeon volume, hospital volume is a more significant factor in predicting short-term outcomes after PD. This suggests that a focus on resources and care pathways, in combination with volume metrics, is more likely to achieve high-quality care for patients undergoing PD across all hospitals.

The impact of upper gastrointestinal surgical volume on short term pancreaticoduodenectomy outcomes for pancreatic adenocarcinoma in the SEER-Medicare population.

BACKGROUND: Patients undergoing pancreaticoduodenectomy (PD) at low volume PD hospitals with high volume for other complex operations have comparable outcomes to high volume PD centers. We evaluated the impact of upper gastrointestinal operations (UGI) hospital volume on the outcomes of elderly, high risk patients undergoing PD. METHODS: Patients >65 years old who underwent PD for pancreatic adenocarcinoma were identified from SEER-Medicare (2008-2015). Four volume cohorts were created using PD tertiles and UGI median: low (1st tertile PD), mixed-low (2nd tertile PD, low UGI), mixed-high (2nd tertile PD, high UGI) and high (3rd tertile PD). Multivariable logistic and negative binomial regression assessed short-term complications. RESULTS: In total, 2717 patients were identified with a median age of 74.5 years. Patients treated at low, mixed-low and mixed-high volume hospitals, versus high volume, had higher risk of short-term complications, including major complications (low: OR 1.441, 95%CI 1.165-1.783; mixed-low: OR 1.374, 95%CI 1.085-1.740; mixed-high: OR 1.418, 95%CI 1.098-1.832) and 90-day mortality (low: OR 2.16, 95%CI 1.454-3.209; mixed-low: OR 2.068, 95%CI 1.347-3.175; mixed-high: OR 1.96, 95%CI 1.245-3.086). CONCLUSION: Patients with pancreatic adenocarcinoma who are older and more medically complex benefit from undergoing surgery at high volume PD centers, independent of the operative experience of that center.

Unique predictors and economic burden of superficial and deep/organ space surgical site infections following pancreatectomy.

BACKGROUND: Surgical site infections (SSIs) are common following pancreatectomy and associated with significant morbidity and economic burden. We sought to identify distinct predictors for superficial versus deep/organ space SSIs and their effects on surgical outcomes. METHODS: ACS-NSQIP targeted pancreatectomy 2014 and 2015 databases were queried. Univariate and multivariate models were developed for both types of SSI, length of stay (LOS), and readmission. Costs were estimated based on Centers for Medicare & Medicaid Services (CMS) recommendations. RESULTS: Of 8093 patients, there were 422 (5.2%) superficial and 1005 (12.4%) deep/organ space SSIs. On multivariate analyses, preoperative biliary stenting was predictive only for superficial SSI (OR: 2.21), while BMI of 25-29.9 (OR: 1.25) and BMI ≥30 kg/m2 (OR: 1.53), pancreatic duct size <3 mm (OR: 1.30), and intermediate (OR: 1.67) versus hard gland texture were predictors of deep/organ-space SSI. Superficial and deep/organ space SSIs were independent predictors of prolonged LOS (OR: 1.74 vs 1.80) and readmission (OR: 2.59 vs 6.57). Additional readmission costs per patient secondary to superficial SSI and deep/organ space SSI were $7661.37 and $18,409.42, respectively. CONCLUSION: Deep/organ space SSI contributes more profoundly to prolonged hospital stay, readmission, and additional costs, suggesting that strategies should focus on preferential prevention of deep/organ space infections.

A nationwide assessment of outcomes after bile duct reconstruction.

BACKGROUND: Bile duct reconstruction (BDR) is used to manage benign and malignant neoplasms, congenital anomalies, bile duct injuries and other non-malignant diseases. BDR outcomes overall, by year, and by indication were compared. METHODS: Retrospective analysis of Nationwide Inpatient Sample discharges (2004-2011) including ICD-9 codes for BDR. All statistical testing was performed using survey weighting. Univariate analysis of admission characteristics by chi square testing. Multivariate modelling for inpatient complications and inpatient death by logistic regression. RESULTS: Identified 67 160 weighted patient admissions: 2.5% congenital anomaly, 37.4% malignant neoplasm, 2.3% benign neoplasm, 9.9% biliary injury, 47.9% other non-malignant disease. Most BDRs were performed in teaching hospitals (69.6%) but only 25% at centres with a BDR volume more than 35/year. 32.3% involved ≥ 1 complication, and 84.7% were discharges home. There was a 4.2% inpatient death rate. The complication rate increased but the inpatient death rate decreased over time. The rates of acute renal failure increased. Significant multivariate predictors of inpatient death include indication of biliary injury or malignancy, and predictors of any complication include public insurance and non-elective admission. CONCLUSION: This is the first national description of BDRs using a large database. In this diverse sampling, both procedure indication and patient characteristics influence morbidity and mortality.

Patient selection and the volume effect in pancreatic surgery: unequal benefits?

BACKGROUND: The volume effect in pancreatic surgery is well established. Regionalization to high-volume centres has been proposed. The effect of this proposal on practice patterns is unknown. METHODS: Retrospective review of pancreatectomy patients in the Nationwide Inpatient Sample 2004-2011. Inpatient mortality and complication rates were calculated. Patients were stratified by annual centre pancreatic resection volume (low <5, medium 5-18, high >18). Multivariable regression model evaluated predictors of resection at a high-volume centre. RESULTS: In total, 129,609 patients underwent a pancreatectomy. The crude inpatient mortality rate was 4.3%. 36.0% experienced complications. 66.5% underwent a resection at high-volume centres. In 2004, low-, medium- and high-volume centres resected 16.3%, 24.5% and 59.2% of patients, compared with 7.6%, 19.3% and 73.1% in 2011. High-volume centres had lower mortality (P < 0.001), fewer complications (P < 0.001) and a shorter median length of stay (P < 0.001). Patients at non-high-volume centres had more comorbidities (P = 0.001), lower rates of private insurance (P < 0.001) and more non-elective admissions (P < 0.001). DISCUSSION: In spite of a shift to high-volume hospitals, a substantial cohort still receives a resection outside of these centres. Patients receiving non-high-volume care demonstrate less favourable comorbidities, insurance and urgency of operation. The implications are twofold: already disadvantaged patients may not benefit from the high-volume effect; and patients predisposed to do well may contribute to observed superior outcomes at high-volume centres.

Does increasing insurance improve outcomes for US cancer patients?

BACKGROUND: Although debate continues on US healthcare and insurance reform, data are lacking on the effect of insurance on community-level cancer outcomes. Therefore, the objective of the present study was to examine the association of insurance and cancer outcomes. MATERIALS AND METHODS: The US Census Bureau Current Population Survey, Small Area Health Insurance Estimates (2000) were used for the rates of uninsurance. Counties were divided into tertiles according to the uninsurance rates. The data were compared with the cancer incidence and survival for patients residing in counties captured by the Surveillance, Epidemiology, and End Results database (2000-2006). Aggregate patient data were collected of US adults (aged ≥18 y) diagnosed with the following cancers: pancreatic, esophageal, liver or bile duct, lung or bronchial, ovarian, colorectal, breast, prostate, melanoma, and thyroid. The outcomes included the stage at diagnosis, surgery, and survival. Univariate tests and proportional hazards were calculated. RESULTS: The US uninsurance rate was 14.2%, and the range for the Surveillance, Epidemiology, and End Results counties was 8.3%-24.1%. Overall, patients from lower uninsurance rate counties demonstrated longer median survival. Adjusting for patient characteristics and cancer stage (for each cancer), the patients in the higher uninsurance rate counties demonstrated greater mortality (8%-15% increased risk on proportional hazards). The county uninsurance rate was associated with the stage at diagnosis for all cancers, except pancreatic and esophageal, and was also associated with the likelihood of being recommended for cancer-directed surgery (for all cancers). CONCLUSIONS: Health insurance coverage at a community level appears to influence survival for patients with cancer. Additional investigations are needed to examine whether individual versus community associations exist and how best to surmount barriers to cancer care.

Is routine intraoperative cholangiogram necessary in the twenty-first century? A national view.

INTRODUCTION: Intraoperative cholangiogram (IOC) can define biliary ductal anatomy. Routine IOC has been proposed previously. However, current surgeon IOC utilization practice patterns and outcomes are unclear. METHODS: Nationwide Inpatient Sample 2004-2009 was queried for patients with acute biliary disease undergoing cholecystectomy (CCY). Analyses only included surgeons performing ≥10 CCY/year. We dichotomized surgeons into a routine IOC group vs. selective. Outcomes included bile duct injury, complications, mortality, length of stay, and cost. RESULTS: Of the nonweighted patients, 111,815 underwent CCY. A total of 4,740 actual surgeon yearly volumes were examined. On average, each surgeon performed 23.6 CCYs and 7.9 IOCs annually, using IOC in 33 % of cases. The routine IOC group used IOC for 96 % of cases, whereas selective IOC group used IOC ∼25 % of the time. Routine IOC surgeons had no difference in mortality (0.4 %) or rate of bile duct injury (0.25 vs. 0.26 %), but higher overall complications (7.3 vs. 6.8 %, p = 0.04). Patients of routine IOC surgeons received more additional procedures and incurred higher costs. CONCLUSION: Routine IOC does not decrease the rate of bile duct injury, but is associated with significant added cost. Surgeons' routine use of IOC is correlated with increased rates of postsurgical procedures, and is associated with increased overall complications. These data suggest routine IOC may not improve outcomes.

Is it worth looking? Abdominal imaging after pancreatic cancer resection: a national study.

INTRODUCTION: Abdominal imaging is often performed after pancreatic cancer resection. We attempted to quantify the volume and estimate the cost of complex imaging after pancreatectomy nationwide, and to determine whether their frequent use confers benefit. METHODS: Patients with pancreatic adenocarcinoma who underwent resection were identified in Surveillance, Epidemiology and End Results-Medicare (1991-2005). Claims for abdominal imaging ≤5 years after resection were analyzed. Patients receiving annual CT scans were identified. Univariate and multivariate analyses were performed. To assess frequency of annual CT scanning in patients with superior survival, the top decile was further analyzed. RESULTS: Eleven thousand eight hundred fifty studies were performed on 2,217 patients. Ten thousand five hundred forty-two (89%) were CT scans. The median number of scans doubled from three in 1991 to six in 2005 (p < 0.0001). Among patients with sufficient survival to allow for analysis, 51.3% received annual CT scans, while only 32.4% of top-performing patients received annual scans. Univariate analysis of the 10% of patients with superior survival did not reveal any significant benefit associated with annual imaging. CONCLUSION: Utilization of complex imaging after pancreatic cancer resection has increased substantially among Medicare beneficiaries, driven primarily by an increasing number of CT scans. Our study demonstrated no significant survival benefit among patients who received scans on a routine basis.

Underutilization of therapy for hepatocellular carcinoma in the medicare population.

BACKGROUND: The incidence of hepatocellular carcinoma (HCC) is increasing in the United States, and the care of these patients remains highly specialized and complex. Multiple treatment options are available for HCC but their use and effectiveness remain unknown. METHODS: Using Surveillance, Epidemiology, and End Results (SEER)-Medicare linked data, 8730 patients who were diagnosed with HCC between 1991 and 2005 were identified. Therapy included surgical resection (8.7%), liver transplantation (1.4%), ablation (3.6%), or transarterial chemoembolization (16%). Patients who received no or palliative-only treatment were grouped together (NoTx; 70.3%). Patient, disease, and tumor factors were examined as determinants of therapy. RESULTS: HCC is increasing in the Medicare population. The median age at diagnosis was 75.1 years and 73.6% of patients were coded as white, 17.2% as Asian, 8.3% as black, and 0.9% as other race. The rate of therapy increased over time, but only 29.7% of patients overall underwent therapy. In patients with early stage HCC, only 43.1% underwent therapy. In the NoTx group, 49.4% did not have cirrhosis, 36.0% had tumors that measured <5 cm, and 39.8% were diagnosed with stage I or II disease when variables were complete. The use of therapy for all HCC patients increased over time, correlating with a commensurate increase in median survival. In multivariate regression analysis, patients who received any modality of treatment achieved significant benefit compared with the NoTx group (odds ratio, 0.41; 95% confidence interval, 0.39-0.43). CONCLUSIONS: In the Medicare population, HCC patients who received therapy experienced a substantial survival advantage over their nonoperative peers (NoTx). Despite evidence that many patients had favorable biological characteristics, <30% of patients diagnosed with HCC received any treatment.

Progress for resectable pancreatic [corrected] cancer?: a population-based assessment of US practices.

BACKGROUND: : Pancreatic adenocarcinoma is a deadly disease; however, recent studies have suggested improved outcomes in patients with locoregional cancer. Progress was evaluated at a national level in resected patients, as measured by the proportion who received guideline-directed treatment and trends in survival. METHODS: : The linked Surveillance, Epidemiology, and End Results and Medicare databases were queried to identify resections for pancreatic adenocarcinoma performed between 1991 and 2002. Receipt and timing of chemotherapy and radiation with respect to time-trend were assessed. Using logistic regression, factors associated with adjuvant combination chemoradiotherapy were identified. Kaplan-Meier curves stratified by year and treatment were used to assess survival. RESULTS: : Of the 1910 patients, 47.9% (n = 915) received some form of adjuvant therapy within the first 6 months postoperatively; 34.4% (n = 658) received combination chemoradiotherapy (chemoRT). ChemoRT demonstrated a significant increase, from 29.2% to 37.5% (P < .0001). Neoadjuvant therapy was used in 5.7% (n = 108) of patients; no trend was observed during the study (P = .1275). The in-hospital mortality rate was 8.0% (n = 153 patients); no significant trend was noted (P = .3116). Kaplan-Meier survival, stratified by year group of diagnosis, did not change significantly over time (log-rank test, P = .4381), even with comparisons of the first 3 years with the last 3 years of the study (log-rank test, P = .3579). CONCLUSIONS: : Adherence to guideline-directed care isimproving in the United States; however, the pace is slow, and overall survival has yet to be impacted significantly. Both increased use of adjuvant therapy and the development of more promising systemic treatments are necessary to improve survival for patients with resectable pancreatic cancer. Cancer 2010. (c) 2010 American Cancer Society.

Admission volume determines outcome for patients with acute pancreatitis.

BACKGROUND & AIMS: There is controversy over the optimal management strategy for patients with acute pancreatitis (AP). Studies have shown a hospital volume benefit for in-hospital mortality after surgery, and we examined whether a similar mortality benefit exists for patients admitted with AP. METHODS: Using the Nationwide Inpatient Sample, discharge records for all adult admissions with a primary diagnosis of AP (n = 416,489) from 1998 to 2006 were examined. Hospitals were categorized based on number of patients with AP; the highest third were defined as high volume (HV, >or=118 cases/year) and the lower two thirds as low volume (LV, <118 cases/year). A matched cohort based on propensity scores (n = 43,108 in each group) eliminated all demographic differences to create a case-controlled analysis. Adjusted mortality was the primary outcome measure. RESULTS: In-hospital mortality for patients with AP was 1.6%. Hospital admissions for AP increased over the study period (P < .0001). HV hospitals tended to be large (82%), urban (99%), academic centers (59%) that cared for patients with greater comorbidities (P < .001). Adjusted length of stay was lower at HV compared with LV hospitals (odds ratio, 0.86; 95% confidence interval, 0.82-0.90). After adjusting for patient and hospital factors, the mortality rate was significantly lower for patients treated at HV hospitals (hazard ratio, 0.74; 95% confidence interval, 0.67-0.83). CONCLUSIONS: The rates of admissions for AP in the United States are increasing. At hospitals that admit the most patients with AP, patients had a shorter length of stay, lower hospital charges, and lower mortality rates than controls in this matched analysis.