We performed a retrospective study using the ACS-NSQIP registry, which has 719 participating sites with 8,581,877 cases [16]. We identified patients undergoing operative pancreatic necrosectomy (Current Procedural Terminology (CPT) code 48,105) by surgeons from 2007 to 2019. This is the only CPT code for pancreatic necrosectomy and includes open necrosectomy as well as minimally invasive necrosectomies such as VARD. This was plotted as the number of cases, deaths and CD4 complications per year in ACS-NSQIP and as a percentage of annual NSQIP records. The preoperative variables collected included age, sex, modified 11-item frailty index (mFI), functional status, alcohol use (defined as 2 or more drinks in the previous 2 weeks), smoking history, history of congestive heart failure, history of myocardial infarction, steroid use, and emergency surgery status).

The mFI is a measure of frailty that has been predictive of operative outcomes in multiple studies and is calculated using the concept of “accumulating deficits” defined by variables available in the NSQIP database [17]. The 11 mFI variables are (1) nonindependent functional status; (2) history of diabetes mellitus; (3) history of chronic obstructive pulmonary disease or pneumonia; (4) history of congestive heart failure; (5) history of myocardial infarction; (6) history of percutaneous coronary intervention, stenting, or angina; (7) history of hypertension requiring medication; (8) history of peripheral vascular disease or ischemic rest pain; (9) history of transient ischemic attack or cerebrovascular event; (10) history of cerebrovascular accident with neurologic deficit; and (11) history of impaired sensorium. Functional status as assigned by the submitting center is defined by the patient’s independence in performing activities of daily living (ADLs) in the 30 days preceding surgery. Functional status 1 refers to an independent patient without the need for assistance to perform ADLs. Functional status 2 is partial dependence on assistance from another person. Functional status 3 is completely dependent on another person for ADLs. Nonindependent status indicates a functional status of 2 or 3.

The preoperative laboratory tests analyzed were white blood cell count (WBC), hematocrit (Hct), platelet count, serum sodium, blood urea nitrogen (BUN), creatinine, albumin, bilirubin, alkaline phosphatase, aspartate aminotransferase (AST), partial thromboplastin time (PTT), international normalized ratio (INR), and prothrombin time (PT).

Primary outcomes were postoperative complications stratified into Clavien-Dindo (CD) class 4 complications requiring ICU care including septic shock, myocardial infarction, cardiac arrest, pulmonary embolism, dialysis, reintubation, on ventilator for more than 48 h, and CD class 5 (death). Frailty, emergency surgery status, functional status and preoperative lab work (were evaluated by univariate and multivariate analyses as predictors of mortality. (NSQIP provides the last available result and duration of preoperative days, in this dataset laboratory results were obtained a mean of 2.4 ± 8.5 days preoperatively.) Post-SPN CPT codes for endoscopic necrosectomy (43,240, 48,999) and interventional radiology drainage procedures (10,030, 49,405, 49,406, and 49,407) were also collected.

Annual cases, deaths, and CD4 complications per year in ACS-NSQIP were reported both as absolute values and also as a percentage of annual NSQIP records. As the ACS-NSQIP registries adds new hospitals each year, including community hospitals as well as tertiary referral centers, a comparator group of pancreatoduodenectomies (Whipples) was identified (CPT codes 4815, 48,152, 48,153 and 48,154) to examine relative time trends in another pancreatic procedure. Whipples were selected as a comparator due to their likelihood of being performed at tertiary referral centers thought also likely to perform surgical necrosectomy.

Statistical analysis was performed using SPSS Version 28 (IBM, Armonk, NY). A p value less than 0.05 was considered statistically significant. Categorical variables were reported as a percentage of the total group and analyzed using Pearson’s chi-square and Fisher’s exact tests. Time trends of mFI, mortality and CD4 complications were performed using Pearson’s chi-square with the Mantel–Haenszel test. Nonparametric testing for significance of ordinal and continuous variables was confirmed with the Wilcoxon rank sum test. Listwise deletion was used if a case had missing data for any of the variables and that case was excluded from analysis. Logistic regression variables for mortality and complications were selected from prior literature, Kolbe et al.’s 2015 study of SPN in NSQIP [15]. ACS-NSQIP was used under the requirements of the UC San Diego Institutional Review Board and followed the ACS-NSQIP data use agreement.