MedStar Georgetown University Hospital Washington, DC
Wei Yan, MD1, Shandiz Shahbazi, MD1, Disha Sharma, MBBS2, Ilan Vavilin, MD1, Bryan S. Stone, DO1, Cory Higley, DO, MPH1, Albert C. Shu, MD1, Rabin Neupane, MBBS2, Andrew Stevens, MD1, Nadim G. Haddad, MD3, Sandeep Nadella, MBBS1; 1MedStar Georgetown University Hospital, Washington, DC; 2MedStar Washington Hospital Center, Washington, DC; 3MedStar Georgetown University Hospital, Potomac, MD
Introduction: Acute pancreatitis (AP) has a broad clinical spectrum: from mild, self-limited disease to fulminant illness resulting in multi-organ failure. Studies have shown certain comorbidities and laboratory findings are associated with more severe disease requiring intensive care unit (ICU) level of care. We characterized real-world diagnostic trends and clinical characteristics in patients admitted to the ICU with AP in a large multi-state health system. Methods: ICD-10 codes for AP were used to identify adult patients presenting to emergency departments (ED) within the MedStar Health system between March 2015-June 2019. These included ED visits and inpatient admissions. Manual review yielded 2542 charts with complete clinical information, 1903 of which had AP as defined by Revised Atlanta Classification (RAC). Patients were put into 1) ICU level-of-care and 2) Non-ICU groups. The following were compared: laboratory values including lipase, clinical symptoms, imaging, workup of etiology, comorbidities, complication rate, treatment variables, readmission rate and length of stay. Statistical analysis was performed using PRISM (GraphPad). Results: A total of 116 patients required ICU level care and were more likely to have AP defined by RAC (OR 2.46, p< 0.0005) compared to those not admitted to the ICU. Evidence of AP on imaging was significantly higher in the ICU group (OR 3.34, p< 0.0001). The ICU group was more likely to receive diagnostic workup (OR 1.67, p= 0.02) and the most common etiology was alcohol-induced. Underlying pulmonary disease (OR 1.66, p=0.02), hypertension (HTN) together with coronary artery disease (CAD) (OR 2.02, p=0.02) and cirrhosis (OR 2.91,p=0.005) were independent risk factors associated with ICU admission, while CAD or HTN alone, congestive heart failure, chronic kidney disease and cancer were not. Complication rates (OR 9.8, p< 0.0001) and length of stay (13.5 vs. 3.6 days,p< 0.0001) were significantly higher in the ICU group compared to the non-ICU group. Average age (51 vs. 52 years,p=0.21) and readmission rates (8.1% vs. 9.7%,p=0.25) between the two groups were similar. Discussion: As expected, patients with AP requiring ICU level care had higher incidence of comorbidities and experienced longer hospital stays. Specifically, patients who have lung disease, CAD, HTN, or cirrhosis may benefit from closer monitoring. Our study provides valuable insights into patient and treatment characteristics that may predispose to ICU admissions in AP.
Disclosures: Wei Yan indicated no relevant financial relationships. Shandiz Shahbazi indicated no relevant financial relationships. Disha Sharma indicated no relevant financial relationships. Ilan Vavilin indicated no relevant financial relationships. Bryan Stone indicated no relevant financial relationships. Cory Higley indicated no relevant financial relationships. Albert Shu indicated no relevant financial relationships. Rabin Neupane indicated no relevant financial relationships. Andrew Stevens indicated no relevant financial relationships. Nadim Haddad indicated no relevant financial relationships. Sandeep Nadella indicated no relevant financial relationships.