Clinical outcomes of acute pancreatitis in patients with cirrhosis

Journal Pre-proof Clinical outcomes of acute pancreatitis in patients with cirrhosis C. Roberto Simons-Linares, Carlos Romero-Marrero, Sunguk Jang, Amit Bhatt, Rocio Lopez, John Vargo, Tyler Stevens, William Carey, Prabhleen Chahal PII:

S1424-3903(19)30765-3

DOI:

https://doi.org/10.1016/j.pan.2019.11.002

Reference:

PAN 1112

To appear in:

Pancreatology

Received Date: 31 May 2019 Revised Date:

2 November 2019

Accepted Date: 6 November 2019

Please cite this article as: Simons-Linares CR, Romero-Marrero C, Jang S, Bhatt A, Lopez R, Vargo J, Stevens T, Carey W, Chahal P, Clinical outcomes of acute pancreatitis in patients with cirrhosis, Pancreatology (2019), doi: https://doi.org/10.1016/j.pan.2019.11.002. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier B.V. on behalf of IAP and EPC.

Running title: Clinical Outcomes of Acute Pancreatitis in Patients with Cirrhosis.

Short Title: Outcomes of acute pancreatitis in cirrhosis. cirrhosis.

Authors: C. Roberto Simons-Linares MD MSc, Carlos Romero-Marrero MD, Sunguk Jang MD, Amit Bhatt MD, Rocio Lopez, John Vargo MD MPH AGAF FACG FASGE, Tyler Stevens MD, William Carey, MD MACG, Prabhleen Chahal MD FACG FASGE. Affiliations: Gastroenterology and Hepatology department, Digestive Disease Institute, Cleveland Clinic Foundation, Cleveland, Ohio USA.

Corresponding Author: Author Prabhleen Chahal, MD, FACG. Postal and email: 9500 Euclid Avenue, Cleveland, OH 44195 [email protected]

Keywords: cirrhosis outcomes, acute pancreatitis, liver cirrhosis acute pancreatitis, cirrhosis pancreatitis.

Institutional Review Board: Approved.

Conflict of Interest: C. Roberto Simons-Linares, Carlos Romero-Marrero, Sunguk Jang, Amit Bhatt, Rocio Lopez, John Vargo, William Carey, and Prabhleen Chahal have no conflicts of interest to disclose.

Word count: 2450 words including abstract 289 words. Does not include references and tables.

Abbreviations

Acute pancreatitis (AP), National Inpatient Sample (NIS), Healthcare Cost and Utilization Project (HCUP), Health Insurance Portability and Accountability Act (HIPAA), United States (US), Charlson comorbidity index (CCI), odds ratios (OR), Acute kidney injury (AKI), acute respiratory distress syndrome (ARDS), systemic inflammatory response syndrome (SIRS), computed tomography (CT), magnetic resonance cholangiopancreatography (MRCP), endoscopic retrograde cholangiopancreatography (ERCP), esophagogastroduodenoscopy (EGD), gastrointestinal bleeding (GIB), Cholecystectomy (CCY), percutaneous cholecystostomy (PC), percutaneous transhepatic cholangiogram (PTHC), portal vein thrombosis (PVT), myocardial infarction (MI), Model of End-stage Liver Disease (MELD), US dollars (US$), International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CMN).

Page 2 of 20

ABSTRACT

Background: Background AP outcomes in cirrhotic patients have not yet been studied. We aim to investigate the outcomes of cirrhotics patients with acute pancreatitis. Methods: Methods The National Inpatient Sample (NIS) database (2003-2013) was queried for patients with a discharge diagnosis of AP and liver cirrhosis. Cirrhosis was further classified as compensated and decompensated using the validated Baveno IV criteria. Primary outcome was inpatient mortality. The analysis was adjusted for age, gender, race, Charlson comorbidity index (CCI), median income quartile, and hospital characteristics. Results: Results Over 2.8 million patients with acute pancreatitis were analyzed. Cirrhosis prevalence was 2.8% (80,093). Both compensated and decompensated cirrhosis subjects had significantly higher mortality. Highest odds ratios (OR) were: inpatient mortality (OR 3.4, P<0.001), Shock (OR 1.5, P=0.02), Ileus (OR: 1.3, p=0.02, ARDS (OR 1.2, p=0.03), upper endoscopy performed (OR 2.0, p<0.001), blood transfusions (OR 3.1, p<0.001), gastrointestinal bleed (OR 5.5, p<0.001), sepsis (OR 1.3, p=0.005), portal vein thrombosis (PVT) (OR 7.2, p<0.001), acute cholecystitis (OR 1.3, p<0.001). Interestingly, cirrhosis patients had lower hospital length of stay, (OR 0.16, p<0.001), AKI (OR 0.93, p=0.06), myocardial infarction (OR 0.31, p<0.001), SIRS (OR 0.62, p<0.001), parenteral nutrition requirement (OR 0.84, p=0.002). Decompensated cirrhosis had higher inflation-adjusted hospital charges (+$3,896.60; p<0.001). Conclusion: Conclusion AP patients with cirrhosis have higher inpatient mortality, but it is unlikely to be due to AP severity as patients had lower incidence of SIRS and AKI. Higher mortality is possibly related to complications of cirrhosis and portal hypertension itself such as GI bleed, shock, PVT, AC and sepsis.

Page 3 of 20

INTRODUCTION

Acute pancreatitis is an acute inflammation of the pancreas that involves unique and complex pathophysiological changes. It particularly increases the vascular permeability that leads to capillary leakage and third-spacing of fluid; leading to intravascular volume depletion, hypotension, shock and multi-organ failure. The main therapy for AP is early aggressive intravenous fluid resuscitation, which indeed can be very challenging in patients with liver cirrhosis and especially in those with decompensated cirrhosis that leads to fluid overload (ascites, edema or anasarca)[1-3].

The incidence of acute pancreatitis in the United States is increasing and recent data suggest that hospitalizations due to acute pancreatitis have had a 13% increase in the last decade [4]. However, the mortality, hospital length of stay (LOS) and health care cost have decreased. Nevertheless, AP related hospitalizations remain a significant burden for our health care system with an annual cost of $2.6 billion for inpatient cases reported in 2012 [5]. Liver cirrhosis is a common chronic disease that is major public health problem. Worldwide is estimated that over 800 million people suffer from liver cirrhosis with a mortality rate of 2 million deaths per year. It is also one of the most common chronic GI disease and decompensated cases represent a frequent hospital discharge diagnosis [6]. Ascites, fluid overload and low intravascular volume are common complication of cirrhotic portal hypertension.

A 20-year epidemiologic study in Finland identified that the incidence of both acute alcoholic pancreatitis and cirrhosis increased during the study period which followed the increasing trend in alcohol consumption in Finland. Interestingly, there was a temporary decrease in hospitalizations for acute alcoholic pancreatitis and liver cirrhosis that correlated with the diminished alcohol consumption in the country during the economic recession in the mid 1990s [7]. This study demonstrates a clear epidemiologic association between the two diseases. Surprisingly, there is very limited literature on the clinical outcomes of AP among patients with liver cirrhosis.

The aim of our study was to investigate the prevalence of cirrhosis among patients admitted with acute pancreatitis and to investigate their outcomes of mortality, morbidity and hospitalization cost.

METHODS

Study setting: The National Inpatient Sample (NIS) database is a publicly available database provided by the Healthcare Cost and Utilization Project (HCUP) of the Agency for Healthcare Research and Quality. It contains clinical and resource utilization data and it is compliance with the Health Insurance Portability and Accountability Act (HIPAA). It protects the privacy of patients, physicians and hospitals. The NIS is the largest inpatient database in the United States (US) and contains data from hospital discharges in the US and it represents over 40

million hospitalizations each year that are generalizable to the American population. It contains over 100 clinical outcomes and non-clinical outcome data for each hospital stay with primary and secondary diagnosis as well as baseline population characteristics, patient’s comorbidity measures, total charges (US$) and hospital characteristics.

Subjects: The NIS from 2003 to 2013 was queried according to the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CMN) for patients with a discharge diagnosis of AP (ICD-9 577.0) and liver cirrhosis (ICD-9 codes 571.2, 571.5, 571.6). Patients were further classified in cirrhotic and non-cirrhotic. Subgroup analysis of the cirrhotic group was performed to compare decompensated and compensated cirrhosis as per the Baveno IV classification of cirrhosis severity. The patients with compensated cirrhosis were defined as those without ascites, variceal bleeding or hepatic encephalopathy; corresponding to Baveno IV stages 1 and 2. The patients with decompensated liver disease were defined as those with ascites (ICD-9 789.5), portal hypertension (ICD-9 572.3), hepatic encephalopathy (ICD-9 572.2) or variceal bleed (ICD-9 456.0, 456.2); corresponding to Baveno IV stages 3 and 4.

Outcomes: The primary outcome was inpatient mortality. Secondary outcomes included organ failure such as acute kidney injury (AKI), acute respiratory distress syndrome (ARDS) or need for mechanical ventilation, ileus, myocardial infarction (MI), stroke, shock, sepsis, any bacterial infection, systemic inflammatory response syndrome (SIRS), venous thromboembolism (including pulmonary embolism, deep vein thrombosis, portal vein thrombosis), parenteral nutrition requirement, gastrointestinal bleed (GIB), red blood cell transfusion requirement, esophagogastroduodenoscopy (EGD), imaging test utilization (Computed tomography, magnetic resonance cholangiopancreatography), pancreatic fluid collection drainage, paracentesis requirement, length of hospital stay (LOS), US$ inflation-adjusted hospital charges. Results were adjusted for age, gender, race, Charlson comorbidity index (CCI), median income quartile, and hospital characteristics.

Analysis: Data are presented as population weighted mean ± standard error for continuous variables or population weighted percent ± standard error for categorical factors. A univariate analysis was performed to assess differences between subjects with and without cirrhosis; continuous variables were compared using t-tests and categorical variables were compared using Rao-Scott chi-square tests. In addition, separate univariate and multivariable logistic regression analyses were performed to assess the association between cirrhosis and in-hospital death and other binary outcomes of interest. Also, univariate and multivariable linear regression analyses were performed to assess the

association of cirrhosis with length of hospital stay and hospital charges. All multivariable models included age, gender, race, median income quartile, CCI, and hospital characteristics. The Deyo modification of the Charlson Comorbidity Index (CCI) was used to define severity of co-morbid conditions.

Inflation adjustment for cost analysis: Charges were adjusted for inflation to 2013 dollars using the gross domestic product deflator (Table 1.1.4 Price Indexes for Gross Domestic Product, US Department of Commerce, Bureau of Economic Analysis). The temporal trend in inflation-adjusted hospital charges in subjects with AP was assessed by linear and polynomial linear regression. The most appropriate functional form for the trend was assessed by examination of regression diagnostic plots. The interaction between cirrhosis and year was evaluated to see differences in trends by group.

All analyses were performed using SAS (version 9.4, The SAS Institute, Cary, NC), which account for the complex sampling design of NIS and appropriately weight participants in statistical models.

RESULTS

Over 2.8 million patients with acute pancreatitis were analyzed. Cirrhosis prevalence was 2.8% (80,093) and 69% (55,509) of those patients were classified as compensated cirrhosis. Overall, cirrhotics were mostly males (62% vs. 51%; <0.001) and the age of the study groups were grossly comparable, but decompensated cirrhotics were younger (53.3

vs. 52.5 years old; <0.001). Interestingly, Hispanics were more commonly found in the decompensated cirrhosis group when compared to compensated cirrhosis (16.7% vs. 13.7%; p<0.001). Inversely, African Americans were found more often to have compensated cirrhosis than decompensated cirrhosis (15.3% vs. 11.8%; p<0.001). Cirrhotic patients were less likely to be obese when compared to non-cirrhotics (9.5% vs. 5.5%; p<0.001), probably due to malnutrition (table 1 and 2).

After adjusting for age, gender, race, CCI, median income quartile, and hospital characteristics; both compensated (OR 1.7, P<0.001; CI: 1.4-2.0) and decompensated cirrhosis patients (OR 2.4, P<0.001; CI: 2.0-2.9) had significantly higher inpatient mortality when compared to patients without cirrhosis. Moreover, when comparing among cirrhotic patients; decompensated patients had higher inpatient mortality than compensated cirrhotic patients (OR 1.4, P<0.001; CI: 1.1-1.8) (Table 3). In addition, patients with decompensated cirrhosis had higher inflation-adjusted hospital charges than non-cirrhosis patients (+$3,896.6; p<0.001) (Figure 1).

The most common organ failure in AP patients is by far AKI. In our study, AKI was more common among cirrhotic patients but this statistical difference disappears after adjusting for all factors (table 2). A similar finding was seen with small bowel ileus and after adjusting the data, there was no statistical difference between groups (Table 2). Patients with decompensated cirrhosis had significantly higher occurrence of ARDS and mechanical ventilation requirement (OR 1.2, p=0.003; CI 1.08-1.4), as well as hemodynamic shock (OR 1.5, P<0.001; CI: 1.2-1.9). In addition, venous thromboembolism (VTE) was more common

in patients with cirrhosis; but when we analyzed the different types of VTE independently, we found that this significant difference was only true for portal vein thrombosis (PVT) and was present across compensated and decompensated cirrhotics. No statistical difference was found when testing for pulmonary embolism (PE) and deep venous thrombosis (DVT) (Table 2).

Interestingly, patients with decompensated cirrhosis who had biliary AP were more likely to have concomitant acute cholecystitis (AC) (OR 1.8, P<0.001; CI: 1.1-2.6). Moreover, no difference in biliary AP frequency and/or occurrence of ascending cholangitis was found. Both compensated (OR 0.81, P<0.001; CI: 0.76-0.87) and decompensated (OR 0.55, P<0.001; CI: 0.48-0.62) cirrhotics were less likely to undergo cholecystectomy. No difference was found between cirrhotic and non-cirrhotic patients undergoing ERCP.

Not surprisingly, cirrhotic patients had higher rates of GIB (OR 2.8, P<0.001; CI: 2.5-3.0) and these odds were highest in decompensated cirrhotics (OR 5.5, P<0.001; CI: 5.0-6.2). Red blood cell transfusions requirement was higher in both, compensated (OR 2.0, P<0.001; CI: 1.9-2.2) and decompensated cirrhotic patients (OR 3.1, P<0.001; CI: 2.9-3.4). EGD requirement was more frequently performed only in decompensated cirrhotics (OR 2.0, P<0.001; CI: 1.8-2.2). Furthermore, decompensated cirrhotics had almost 20 times higher odds of undergoing paracentesis.

Interestingly, compensated (OR 0.16, p<0.001; CI 0.03-0.30) and decompensated cirrhotic patients (OR 0.46, p<0.001; CI 0.26-0.66) had shorter hospital length of stay and had lower

odds of having inpatient myocardial infarction (decompensated cirrhotics OR 0.31, p<0.001 and compensated cirrhotics OR 0.63, p=0.008). Additionally, compensated cirrhotics had lower rates of SIRS (OR 0.62, p<0.001; CI 0.48-0.80) and parenteral nutrition requirement (OR 0.84, p=0.002; CI 0.75-0.94).

A slight trend towards higher occurrence of sepsis was found only in decompensated cirrhotics (OR 1.3, p=0.005; CI 1.08-1.5). Moreover, cirrhotic patients had lower odds of having a pancreatic fluid collection drained, although it was not statistically significant. Noteworthy, both compensated and decompensated cirrhotics had higher likelihood of undergoing exploratory laparotomy, with the latter group having almost 16 times higher odds (OR 15.7, p<0.001; CI 14.2-17.4).

DISCUSSION

Although our study shows that among AP patients admitted to the hospital, cirrhotics had higher inpatient mortality, this seems to be from complications of cirrhosis and portal hypertension such as GIB, shock, and sepsis; rather than acute pancreatitis inflammation as evidenced by the lower incidence of SIRS, AKI and ileus in patients with cirrhosis. These findings have important clinical implications as cirrhotic patients admitted with AP may require a different approach in order to decrease the higher mortality demonstrated by this study. Furthermore, cirrhotic patients are often fluid overloaded and AP aggressive IV fluid hydration may be challenging or not doable in this population. Studies are needed in this

area to determine if the aggressive IV fluid therapy (250-500 cc/hr) has beneficial or detrimental effect in the liver cirrhosis patients.

Acute pancreatitis inflammation, especially when severe, increases the vascular permeability that leads to capillary leakage and sequestration of fluid. This third-spacing of fluid phenomena then leads to severe intravascular volume depletion, hypotension, hypoperfusion, pancreatic ischemia and necrosis. Ultimately, leading to shock and multi-organ failure. Most of the AP cases are mild but 10-20% of cases can be severe and characterized by persistent organ failure and pancreatic necrosis. Acute pancreatitis causes SIRS, either transient or persistent (>48h)[8]. In the presented study, SIRS was less likely to be present in patients with cirrhosis and AP. This could indicate that cirrhotic patient’s higher mortality is not due to pancreatic inflammation but rather to cirrhosis and portal hypertension complications. Importantly, AKI is the most common organ failure during an AP attack, but it could also be a complication of cirrhotic portal hypertension. In AP, kidney injury is most likely from a pre-renal etiology due to intravascular volume depletion from sequestration of fluid. Differently, in liver cirrhosis, kidney injury can be multifactorial. It can happen as a consequence of systemic and splanchnic arterial vasodilatation and consecutive reduction in effective circulating blood volume, and renal perfusion may be critically impaired in patients with cirrhosis and portal hypertension. But the etiology of AKI in cirrhosis can be: pre-renal (i.e. hypovolemia due to GIB, lactulose-induced diarrhea), hepatorenal syndrome due to renal vasoconstriction in the absence of alternative identifiable causes (unresponsive to volume resuscitation), and/or intrinsic causes such as acute tubular necrosis. In the presented study, it seems that most patients with liver cirrhosis had mild pancreatitis as evidenced by the lack of

worsening of kidney injury that one could expect from a more severe pancreatitis in the settings of a cirrhotic physiology [9, 10].

GI bleed is a common complication of portal hypertension seen in cirrhosis [11]. Moreover, GI bleed can be very severe when is due to esophageal varices requiring endoscopic intervention[12, 13]. Our study found that cirrhotic patients were more likely to have GI bleed, require RBC transfusions and EGD intervention. These findings could potentially explain the higher mortality in cirrhotics in our study. Another potential explanation for increased mortality in cirrhotics could be infections and sepsis, which are also a well-known complication of liver cirrhosis. The immune dysfunction in liver cirrhosis is due to an increase in anti-inflammatory cytokines and suppression of proinflammatory cytokines. Furthermore, there is an increased of the gut permeability and an alteration in the gut flora, which can lead to bacterial translocation, infection and bacteremia [14].

In our study, sepsis and infections were more frequent in cirrhotics, which can be explained by immunosuppressed state in cirrhotics. Also, cirrhotics were diagnosed with cholecystitis more frequently than non-cirrhotics, this could be explained by a) cirrhotics are more likely to have right upper quadrant ultrasound and ascites may also impact or mimic pericholecystic fluid, b) cirrhotics have a higher risk for cholecystitis. Gallstones prevalence in the general population is between 10-15%, in contrast cirrhotics have a higher prevalence (25-30%)[15, 16]. Another surprising finding was the shorter hospital stay of cirrhotic patients (both compensated and decompensated). This could be due to a premature hospital discharge in this population, as there is growing literature on premature hospital discharge of cirrhotic patients in the US. A study found that this

may be the leading factor of high readmission rates among cirrhotics, hence roughly 2/3 of admitted patients are a readmission and only 1/3 are new admissions[17-19].

One of the limitations that we had to overcome was the lack of laboratory data to calculate Child-Pugh and Model of End-stage Liver Disease (MELD) scores. Hence, we used the Baveno IV criteria, in an effort to stratify compensated from decompensated patients. The Baveno IV consensus purpose was to standardize the definitions in variceal bleed cirrhotics for further use in clinical trials; but also to be used in other studies that require stratification of cirrhotic patients with portal hypertension (decompensated) from those without (compensated)[20]. In clinical practice, the Baveno IV criteria aligns with the objectives of our study and we applied this well validated score to stratify our cirrhotic subgroup analysis (compensated vs. decompensated cirrhosis). For the mentioned reasons we decided to use the Baveno IV score rather than using our own definitions that would not be validated and would be a limitation.

In addition, the current NIS database study findings correlate with our own center’s data and results. We carried out a retrospective study in hospitalized AP patients at large tertiary hospital. Propensity score matching was performed to match patients with and without cirrhosis on a 1:2 basis. In addition, subgroup analysis of cirrhotic patients according to Child-Pugh and Model of End-stage Liver Disease (MELD) scores was performed. 819 patients with AP were analyzed. Cirrhosis prevalence was 6.1%. Mean Child-Pugh score was 8.2, mean MELD score was 16. Etiology of AP was similar in both groups. Importantly, cirrhosis subjects had significantly higher inpatient mortality (7.5% vs. 1.3%, P=0.1), severe AP (17.5% vs. 7.5%), shock (7.9% vs. 3%), ARDS (10% vs. 3.8%), intensive care unit (ICU)

requirement (15% vs. 6.3%). Although it did not reach statistical significance, the trend was present. Surprisingly, cirrhotics had lower prevalence of admission SIRS (22.5% vs. 32.5%), persistent SIRS (25% vs. 15%). Moreover, cirrhotics had similar rates of pancreatic necrosis, ileus, BISAP score, Marshall score, admission hematocrit, blood urea nitrogen (BUN) and LOS. Finally, cirrhotic patients who had severe AP and/or required ICU and/or died in-hospital were mainly Child-C and had high MELD score (17-38) but they had low BISAP and Marshall scores. The conclusion of the study was that cirrhotics who are hospitalized with AP have higher inpatient mortality, most likely related to decompensated cirrhosis (GI bleed, sepsis)– as they appear to have lower prevalence of SIRS, low BISAP and Marshall scores and similar rates of pancreatic necrosis compared to non-cirrhotics. Of note, cirrhotic patients who died were mostly Child-Pugh C and had very high MELD score[21].

Finally, optimal management of AP is still a challenge and multiple factors can affect its multidisciplinary decision-making approach [22]. It is important to recognize risk factors or populations that are at higher risk of death during an AP attack. To the best of our knowledge, the presented study is the first to evaluate the clinical outcomes of acute pancreatitis in patients with liver cirrhosis. More studies are needed to corroborates our findings. Studies focusing on aggressive IV fluid therapy outcomes in cirrhosis patients are also needed.

Our study has several strengths, it is the first and biggest patient cohort to date to be reported. We evaluated patients with compensated and decompensated cirrhosis in an

attempt to determine the true impact or complications of cirrhosis/portal hypertension. Moreover, the current NIS study correlates with our own center data on the same topic and we have provided a summary of this comparison. The study was also adjusted for age, gender, race, CCI, median income quartile, and hospital characteristics. Nevertheless, our study has several limitations such as NIS data accounts for inpatient data only and misses post-discharge outcomes. Also, there is reliance on administrative data and billing codes for ascertaining outcomes and exposures. The latter is important because there might be a degree of diagnosis codes misclassification. Although our study attempted to separate outcomes between compensated and decompensated cirrhosis with the Baveno IV classification; the NIS database lacks of laboratory data to adjust for Child-Pugh and MELD scores. Finally, the database has 17.64% of data missing for the race variables, which is a significant number.

CONCLUSION

Cirrhotic patients that are hospitalized with AP have higher inpatient mortality in the US. This is unlikely to be due to AP severity as cirrhotics had lower incidence of SIRS and AKI. We propose that the higher mortality is possibly related to complications of cirrhosis and portal hypertension itself, as evidenced by higher incidence of GI bleed, shock and sepsis in cirrhotics. Our study findings correlate with another single center study where cirrhotics with AP had low BISAP and Marshall scores, less SIRS and pancreatic necrosis; but higher GI bleed and sepsis that led the authors to conclude that the higher mortality was due to decompensated cirrhosis rather than AP. Finally, our study has important clinical findings

in two very common GI conditions, and further studies are needed to corroborate our findings.

REFERENCES 1. 2.

3. 4.

5. 6.

7.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Banks, P.A., et al., Classification of acute pancreatitis--2012: revision of the Atlanta classification and definitions by international consensus. Gut, 2013. 62(1): p. 102-11. Kadiyala, V., et al., The Atlanta Classification, Revised Atlanta Classification, and Determinant-Based Classification of Acute Pancreatitis: Which Is Best at Stratifying Outcomes? Pancreas, 2016. 45(4): p. 510-5. Gardner, T.B., et al., Fluid resuscitation in acute pancreatitis. Clin Gastroenterol Hepatol, 2008. 6(10): p. 1070-6. Krishna, S.G., et al., The Changing Epidemiology of Acute Pancreatitis Hospitalizations: A Decade of Trends and the Impact of Chronic Pancreatitis. Pancreas, 2017. 46(4): p. 482-488. Peery, A.F., et al., Burden of gastrointestinal disease in the United States: 2012 update. Gastroenterology, 2012. 143(5): p. 1179-1187.e3. Marcellin, P. and B.K. Kutala, Liver diseases: A major, neglected global public health problem requiring urgent actions and large-scale screening. Liver Int, 2018. 38 Suppl 1: p. 2-6. Sand, J., A. Valikoski, and I. Nordback, Alcohol consumption in the country and hospitalizations for acute alcohol pancreatitis and liver cirrhosis during a 20-year period. Alcohol Alcohol, 2009. 44(3): p. 321-5. John, B.J., et al., Persistent Systemic Inflammatory Response Syndrome predicts the need for tertiary care in Acute Pancreatitis. Acta Gastroenterol Belg, 2017. 80(3): p. 377-380. Wu, C.C., et al., Incidence and factors predictive of acute renal failure in patients with advanced liver cirrhosis. Clin Nephrol, 2006. 65(1): p. 28-33. Bucsics, T. and E. Krones, Renal dysfunction in cirrhosis: acute kidney injury and the hepatorenal syndrome. Gastroenterol Rep (Oxf), 2017. 5(2): p. 127-137. Ibrahim, M., I. Mostafa, and J. Deviere, New Developments in Managing Variceal Bleeding. Gastroenterology, 2018. 154(7): p. 1964-1969. Long, B. and A. Koyfman, The emergency medicine evaluation and management of the patient with cirrhosis. Am J Emerg Med, 2018. 36(4): p. 689-698. Tayyem, O., et al., Evaluation and management of variceal bleeding. Dis Mon, 2018. 64(7): p. 312-320. Noor, M.T. and P. Manoria, Immune Dysfunction in Cirrhosis. J Clin Transl Hepatol, 2017. 5(1): p. 50-58. Acalovschi, M., Gallstones in patients with liver cirrhosis: incidence, etiology, clinical and therapeutical aspects. World J Gastroenterol, 2014. 20(23): p. 7277-85. Li, X., et al., Liver cirrhosis: a risk factor for gallstone disease in chronic hepatitis C patients in China. Medicine (Baltimore), 2017. 96(26): p. e7427. Chirapongsathorn, S., et al., Incidence and cost analysis of hospital admission and 30-day readmission among patients with cirrhosis. Hepatol Commun, 2018. 2(2): p. 188-198. Fagan, K.J., et al., Burden of decompensated cirrhosis and ascites on hospital services in a tertiary care facility: time for change? Intern Med J, 2014. 44(9): p. 865-72. Wei, M., et al., Hospital Cirrhosis Volume and Readmission in Patients with Cirrhosis in California. Dig Dis Sci, 2018. 63(9): p. 2267-2274. Thabut, D., et al., Multicenter prospective validation of the Baveno IV and Baveno II/III criteria in cirrhosis patients with variceal bleeding. Hepatology, 2015. 61(3): p. 1024-32.

21.

22.

Abushamma, S., et al., Clinical Outcomes of Acute Pancreatitis in Patients with Liver Cirrhosis - A Single-Center Experience: 32. American Journal of Gastroenterology, 2018. 113: p. S18. Lankisch, P.G., M. Apte, and P.A. Banks, Acute pancreatitis. Lancet, 2015. 386(9988): p. 85-96.

Figure Legends: Figure 1. Trends in charges

Table 1. Patient Characteristics Factor Age (years), mean ± SE Gender . Male . Female Race . White . Black . Hispanic . Other . Unknown Obese Primary expected payer . Medicare . Medicaid . Private Insurance . Other Median household income quartile for patient's ZIP Code . 1 . 2 . 3 . 4 Bed size of hospital . Small . Medium . Large Region of hospital . Northeast . Midwest . South . West Urban hospital Teaching hospital CCI, mean ± SE

No Cirrhosis

Compensated Cirrhosis

Decompensated Cirrhosis

52.5±0.05 3

52.7±0.13 3

53.3±0.18 1, 2

51.3% ± 0.09% 48.7% ± 0.09% 2, 3

62.1% ± 0.50% 37.9% ± 0.50% 1, 3

64.5% ± 0.72% 35.5% ± 0.72% 1, 2

2, 3

1, 3

1, 2

p-value <0.001 <0.001

<0.001 53.6% ± 0.40% 13.9% ± 0.22% 10.3% ± 0.22% 4.5% ± 0.10% 17.8% ± 0.44% 9.5% ± 0.07% 2, 3

51.1% ± 0.65% 15.3% ± 0.46% 13.7% ± 0.47% 4.3% ± 0.23% 15.5% ± 0.59% 5.3% ± 0.22% 1

54.5% ± 0.85% 11.8% ± 0.54% 16.7% ± 0.67% 6.1% ± 0.37% 10.9% ± 0.61% 5.6% ± 0.34% 1

32.2% ± 0.12% 2, 3 15.0% ± 0.13% 34.5% ± 0.18% 18.4% ± 0.16%

30.6% ± 0.49% 1 27.7% ± 0.53% 20.8% ± 0.45% 20.9% ± 0.48%

30.8% ± 0.69% 1 27.8% ± 0.70% 22.3% ± 0.65% 19.1% ± 0.61%

<0.001 <0.001

<0.001 31.6% ± 0.31% 2, 3 26.8% ± 0.24% 23.2% ± 0.21% 18.5% ± 0.30%

37.4% ± 0.64% 1, 3 26.6% ± 0.53% 21.5% ± 0.48% 14.5% ± 0.45%

33.8% ± 0.84% 1, 2 25.7% ± 0.72% 23.0% ± 0.68% 17.4% ± 0.66% <0.001

2, 3

14.9% ± 0.20% 26.9% ± 0.25% 58.2% ± 0.30%

1

13.1% ± 0.42% 26.3% ± 0.58% 60.6% ± 0.66%

1

12.6% ± 0.57% 24.4% ± 0.74% 63.1% ± 0.86%

<0.001 2, 3

16.9% ± 0.24% 22.7% ± 0.25% 40.6% ± 0.31% 19.8% ± 0.24% 83.5% ± 0.18% 2, 3 37.8% ± 0.33% 2, 3 0.78±0.00 2, 3

Data presented as Percent ± Standard Error (SE) unless otherwise stated. Charlson comorbidity index (CCI), Standard deviation (SE).

1, 3

17.0% ± 0.58% 18.6% ± 0.51% 41.1% ± 0.68% 23.2% ± 0.57% 86.6% ± 0.43% 1, 3 40.6% ± 0.70% 1, 3 1.9±0.01 1, 3

1, 2

16.8% ± 0.73% 18.2% ± 0.67% 38.4% ± 0.89% 26.5% ± 0.80% 89.5% ± 0.52% 1, 2 43.7% ± 0.93% 1, 2 3.3±0.03 1, 2

<0.001 <0.001 <0.001

Table 2. Association Between Cirrhosis and Binary Outcomes of AP Compensated vs. No Cirrhosis Outcome

Decompensated vs. No Cirrhosis

Decompensated vs. Compensated Cirrhosis

OR (95% CI)

p-value

OR (95% CI)

p-value

OR (95% CI)

p-value

1.3 (1.2, 1.4) 0.93 (0.87, 1.00)

<0.001 0.062

2.0 (1.8, 2.1) 0.96 (0.87, 1.06)

<0.001 0.4

1.5 (1.4, 1.7) 1.03 (0.92, 1.2)

<0.001 0.63

0.82 (0.73, 0.92) 0.85 (0.75, 0.95)

<0.001 0.005

1.06 (0.91, 1.2) 1.09 (0.93, 1.3)

0.43 0.29

1.3 (1.07, 1.6) 1.3 (1.06, 1.6)

0.007 0.009

Unadjusted

0.76 (0.55, 1.06)

0.1

0.71 (0.43, 1.2)

0.19

0.93 (0.51, 1.7)

0.83

Adjusted

0.63 (0.45, 0.89)

0.008

0.31 (0.18, 0.54)

<0.001

0.49 (0.25, 0.94)

0.031

1.2 (0.97, 1.5) 1.04 (0.83, 1.3)

0.089 0.74

2.3 (1.8, 2.8) 1.5 (1.2, 1.9)

<0.001 <0.001

1.9 (1.4, 2.5) 1.5 (1.06, 2.0)

<0.001 0.02

0.68 (0.53, 0.87) 0.62 (0.48, 0.80)

0.003 <0.001

1.1 (0.84, 1.5) 0.87 (0.64, 1.2)

0.42 0.39

1.7 (1.1, 2.5) 1.4 (0.96, 2.1)

0.01 0.083

1.1 (0.99, 1.3) 0.96 (0.82, 1.1)

0.072 0.61

1.9 (1.6, 2.3) 1.3 (1.08, 1.5)

<0.001 0.005

1.7 (1.3, 2.1) 1.3 (1.06, 1.7)

<0.001 0.013

1.6 (1.4, 2.0) 1.4 (1.1, 1.6)

<0.001 <0.001

4.0 (3.4, 4.7) 2.7 (2.2, 3.2)

<0.001 <0.001

2.4 (1.9, 3.1) 2.0 (1.5, 2.5)

<0.001 <0.001

AKI Unadjusted Adjusted Ileus Unadjusted Adjusted Myocardial Infarction(MI)

Shock Unadjusted Adjusted SIRS Unadjusted Adjusted Sepsis Unadjusted Adjusted Any venous thrombosis Unadjusted Adjusted

Deep vein thrombosis (DVT) Unadjusted Adjusted Pulmonary embolism (PE) Unadjusted Adjusted Portal vein thrombosis (PVT) Unadjusted Adjusted EGD Unadjusted Adjusted Transfusion Unadjusted Adjusted GIB Unadjusted Adjusted Parenteral Nutrition Unadjusted Adjusted Abdominal CT Unadjusted Adjusted MRCP Unadjusted

1.03 (0.70, 1.5) 0.84 (0.57, 1.2)

0.9 0.39

0.92 (0.51, 1.7) 0.54 (0.29, 1.02)

0.79 0.057

0.90 (0.45, 1.8) 0.64 (0.31, 1.3)

0.77 0.23

0.88 (0.58, 1.3) 0.75 (0.48, 1.2)

0.54 0.19

0.81 (0.42, 1.6) 0.57 (0.29, 1.1)

0.52 0.098

0.92 (0.42, 2.0) 0.76 (0.35, 1.7)

0.83 0.5

2.8 (2.3, 3.5) 2.5 (2.0, 3.2)

<0.001 <0.001

9.3 (7.7, 11.2) 7.2 (5.8, 9.0)

<0.001 <0.001

3.3 (2.5, 4.4) 2.9 (2.1, 3.9)

<0.001 <0.001

1.10 (1.02, 1.2) 1.00 (0.94, 1.08)

0.009 0.93

2.5 (2.3, 2.7) 2.0 (1.8, 2.2)

<0.001 <0.001

2.2 (2.0, 2.5) 2.0 (1.8, 2.2)

<0.001 <0.001

2.5 (2.3, 2.7) 2.0 (1.9, 2.2)

<0.001 <0.001

5.2 (4.8, 5.7) 3.1 (2.9, 3.4)

<0.001 <0.001

2.1 (1.9, 2.3) 1.6 (1.4, 1.8)

<0.001 <0.001

3.2 (2.9, 3.5) 2.8 (2.5, 3.0)

<0.001 <0.001

7.1 (6.4, 7.9) 5.5 (5.0, 6.2)

<0.001 <0.001

2.2 (2.0, 2.5) 2.0 (1.8, 2.3)

<0.001 <0.001

0.93 (0.83, 1.03) 0.84 (0.75, 0.94)

0.18 0.002

1.07 (0.91, 1.2) 0.84 (0.71, 0.98)

0.42 0.031

1.1 (0.95, 1.4) 1.00 (0.82, 1.2)

0.14 0.98

0.94 (0.82, 1.08) 0.88 (0.77, 1.01)

0.37 0.067

1.00 (0.84, 1.2) 0.93 (0.78, 1.1)

0.96 0.46

1.07 (0.87, 1.3) 1.06 (0.86, 1.3)

0.52 0.58

0.83 (0.67, 1.03)

0.097

1.3 (1.04, 1.7)

0.023

1.6 (1.2, 2.2)

0.004

Adjusted

0.91 (0.73, 1.1)

0.39

1.4 (1.08, 1.8)

0.011

1.5 (1.1, 2.1)

0.011

ERCP Unadjusted

0.83 (0.72, 0.96)

0.011

0.78 (0.63, 0.96)

0.021

0.94 (0.73, 1.2)

0.62

Adjusted CCY

0.93 (0.80, 1.07)

0.31

0.94 (0.75, 1.2)

0.57

Unadjusted

0.64 (0.60, 0.69)

<0.001

0.38 (0.33, 0.43)

<0.001

Adjusted

0.81 (0.76, 0.87)

<0.001

0.55 (0.48, 0.62)

<0.001

0.79 (0.45, 1.4) 0.75 (0.43, 1.3)

0.41 0.33

1.3 (0.69, 2.6) 0.90 (0.44, 1.8)

0.39 0.76

1.7 (0.71, 4.0) 1.2 (0.48, 2.9)

0.23 0.71

0.63 (0.31, 1.3) 0.59 (0.29, 1.2)

0.2 0.14

1.07 (0.48, 2.4) 0.55 (0.20, 1.5)

0.86 0.25

1.7 (0.59, 4.9) 0.94 (0.28, 3.2)

0.32 0.92

0.72 (0.51, 1.00) 0.70 (0.49, 0.99)

0.052 0.041

0.73 (0.46, 1.2) 0.71 (0.43, 1.2)

0.21 0.18

1.02 (0.59, 1.8) 1.02 (0.58, 1.8)

0.93 0.95

4.1 (3.6, 4.6) 3.6 (3.2, 4.0)

<0.001 <0.001

26.2 (24.0, 28.6) 19.5 (17.6, 21.7)

<0.001 <0.001

6.4 (5.6, 7.4) 5.5 (4.7, 6.3)

<0.001 <0.001

1.3 (1.1, 1.4) 1.07 (0.95, 1.2)

<0.001 0.28

1.8 (1.6, 2.1) 1.2 (1.08, 1.4)

<0.001 0.003

1.4 (1.2, 1.7) 1.2 (0.97, 1.4)

<0.001 0.098

1.8 (1.6, 2.1)

<0.001

3.4 (2.9, 4.0)

<0.001

1.9 (1.5, 2.3)

<0.001

1.7 (1.4, 2.0)

<0.001

2.4 (2.0, 2.9)

<0.001

1.4 (1.1, 1.8)

0.002

PC Unadjusted Adjusted PTHC Unadjusted Adjusted Pancreatic fluid collection drainage Unadjusted Adjusted Paracentesis Unadjusted Adjusted ARDS/Mechanical Ventilation Unadjusted Adjusted In-hospital death Unadjusted Adjusted

1.01 (0.78, 1.3) 0.58 (0.51, 0.67) 0.67 (0.58, 0.77)

0.93 <0.001 <0.001

*Adjusted for age, gender, race, median income quartile, CCI and hospital characteristics Abbreviations: Acute kidney injury (AKI), acute respiratory distress syndrome (ARDS), systemic inflammatory response syndrome (SIRS), computed tomography (CT), magnetic resonance cholangiopancreatography (MRCP), endoscopic retrograde cholangiopancreatography (ERCP), esophagogastroduodenoscopy (EGD), gastrointestinal bleeding (GIB), Cholecystectomy (CCY), percutaneous cholecystostomy (PC), percutaneous transhepatic cholangiogram (PTHC).

Table 3. Association Between Cirrhosis Decompensated and Continuous Outcomes of AP Compensated vs. No Cirrhosis Outcome Length of stay (days) Unadjusted Adjusted Inflation-adjusted Charges ($) Unadjusted Adjusted

Decompensated vs. No Cirrhosis

Decompensated vs. Compensated Cirrhosis

OR (95% CI)

p-value

OR (95% CI)

p-value

OR (95% CI)

p-value

0.53 (0.40, 0.66) 0.16 (0.03, 0.30)

<0.001 0.054

1.3 (1.07, 1.5) 0.46 (0.26, 0.66)

<0.001 <0.001

0.74 (0.50, 0.98) 0.30 (0.06, 0.54)

<0.001 0.038

2571.5 (1657.0, 3485.9) -1258.0 (-2184.0, -331.9)

<0.001 0.023

*Adjusted for age, gender, race, median income quartile, CCI and hospital characteristics

12271.6 (10437.2, 14106.1) 3896.6 (2061.4, 5731.7)

<0.001 <0.001

9700.2 (7661.6, 11738.7) 5154.5 (3136.6, 7172.5)

<0.001 <0.001