Elizaveta Ragulin-Coyne

Perioperative mortality after pancreatectomy: A risk score to aid decision-making

BACKGROUND Undergoing a pancreatectomy obligates the patient to risks and benefits. For complex operations such as pancreatectomy, the objective assessment of baseline risks may be useful in decision-making. We developed an integer-based risk score estimating in-hospital mortality after pancreatectomy, incorporating institution-specific mortality rates to enhance its use. METHODS Pancreatic resections were identified from the Nationwide Inpatient Sample (1998-2006), and categorized as proximal, distal, or nonspecified by the International Classification of Diseases, 9th edition. Logistic regression and bootstrap methods were used to estimate in-hospital mortality using demographics, diagnosis, comorbidities (Charlson index), procedure, and hospital volume; 80% of this cohort was selected randomly to create the score and 20% was used for validation. Score assignments were subsequently individually fitted to risk distributions around specific mortality rates. RESULTS Sixteen thousand one hundred sixteen patient discharges were identified. Nationwide in-hospital mortality was 5.3%. Integers were assigned to predictors (age group, Charlson index, sex, diagnosis, pancreatectomy type, and hospital volume) and applied to an additive score. Three score groups were defined to stratify in-hospital mortality (national mortality, 1.3%, 4.9%, and 14.3%; P < .0001), with sufficient discrimination of derivation and validation sets (C statistics, 0.72 and 0.74). Score groups were shifted algorithmically to calculate risk based on institutional data (eg, with institutional mortality of 2.0%, low-, medium-, and high-risk patient groups had 0.5%, 1.9%, and 5.4% mortality, respectively). A web-based tool was developed and is available online (http://www.umassmed.edu/surgery/panc_mortality_custom.aspx). CONCLUSION To maximize patient benefit, objective assessment of risk for major procedures is necessary. We developed a Surgical Outcomes Analysis and Research risk score predicting pancreatectomy mortality that combines national and institution-specific data to enhance decision-making. This type of risk stratification tool may identify opportunities to improve care for patients undergoing specific operative procedures.

National trends in pancreaticoduodenal trauma: interventions and outcomes

OBJECTIVES Pancreaticoduodenal trauma (PDT) is associated with substantial mortality and morbidity. In this study, contemporary trends were analysed using national data. METHODS The Nationwide Inpatient Sample for 1998-2009 was queried for patients with PDT. Interventions including any operation (Any-Op) and pancreas-specific surgery (PSURG) were identified. Trends in treatment and outcomes were determined [complications, length of stay (LoS), mortality] for the Any-Op, PSURG and non-operative (Non-Op) groups. Analyses included chi-squared tests, Cochran-Armitage trend tests and logistic regression. RESULTS A total of 27 216 patients (nationally weighted) with PDT were identified. Over time, the frequency of PDT increased by 8.3%, whereas the proportion of patients submitted to PSURG declined (from 21.7% to 19.8%; P = 0.0004) and the percentage of patients submitted to non-operative management increased (from 56.7% to 59.1%; P = 0.01). In the Non-Op group, mortality decreased from 9.7% to 8.6% (P < 0.001); morbidity and LoS remained unchanged at ∼40% and ∼12 days, respectively. In the PSURG group, mortality remained stable at ∼15%, complications increased from 50.2% to 71.8% (P < 0.0001) and LoS remained stable at ∼21 days. For all PDT patients, significant independent predictors of mortality included: the presence of combined pancreatic and duodenal injuries; penetrating trauma, and age >50 years. Having any operation (Any-Op) was associated with mortality, but PSURG was not a predictor of death. CONCLUSIONS The utilization of operations for PDT has declined without affecting mortality, but operative morbidity increased significantly over the 12 years to 2009. The development of an evidence-based approach to invasive manoeuvres and an early multidisciplinary approach involving pancreatic surgeons may improve outcomes in patients with these morbid injuries.

Utilization of laparoscopy for resections of stomach and esophagus cancers: Is hospital the deciding factor?

86 Background: Foregut surgery is technically complex. Outcomes for such high-stakes operations receive increasing scrutiny and the use of minimally invasive approaches has been further adopted. This study aims to determine national trends in laparoscopy utilization and patient outcomes for potentially curative cancer resections of the esophagus and stomach. Methods: Retrospective review of all esophageal and gastric cancer resections in the Nationwide Inpatient Sample during 1998 to 2011. Univariate analyses of sex, race, admission status, Elixhauser comorbidity score, year, insurance, hospital characteristics, procedure, and center volume were performed by chi-square. Cochran-Armitage test was used for trends. Logistic regressions were used to model inpatient mortality, complications and laparoscopy. Results: From 1998 to 2011, 120,527 and 25,540 patients (nationally-weighted records) underwent gastrectomies and esophagectomies for cancer. From early (1998-2002) to late (2008-2011) study years, inpatien...

528 Is Routine Intraoperative Cholangiogram Necessary in the 21st Century? a National View

The video presents technical steps for Laparoscopic Reversal of Roux-en-Y Gastric Bypass (RYGB) to normal anatomy, and results of pre and post-op metabolic testing. The patient underwent RYGB at another institution (BMI 46). Two years after RYGB (BMI 25), presented with recalcitrant hyperinsulinemic hypoglycemia episodes and hypoglycemia unawareness. A laparoscopic gastrostomy was placed in the excluded stomach. A meal test (MT) administered orally and through the gastrostomy showed altered insulin and glycemic responses through the RYGB. Laparoscopic reversal was performed. Patient symptoms subsided; and 6 months later (BMI 29), a MT showed normal insulin and glycemic profiles.

Surgery for pancreatic cancer: Unnecessary or insufficient?

e14596 Background: All potential cures for pancreatic cancer include resection. However, some patients with resectable disease do not undergo surgery, while others undergo operation but succumb quickly to recurrent disease or postoperative complications. We attempted to construct an optimal cohort of patients for resection by identifying two outlying groups who either 1) might have benefited from resection; or 2) underwent non-beneficial surgery. METHODS Patients ≥65 years old with pancreatic adenocarcinoma were identified in SEER-Medicare 1991-2007. Potential suboptimal utilization was defined as patients with locoregional disease who did not undergo resection, or patients who received resections yet died <6 months from diagnosis. Univariate and multivariate analyses were performed to identify predictors of suboptimal outcome. Survival was examined by Cox. RESULTS Among 17,450 patients, 9612 (55.1%) had metastases at diagnosis. For metastatic patients, 5,574 (58%) visited a surgeon, 828 (8.6%) were recommended for surgery, and 185 (1.9%) were resected. 7,604 (79.1%) of metastatic patients died within 6 months of diagnosis, including 100 (54.1%) of the highly-selected resection cohort. Younger metastatic patients were more likely to receive surgery (p=0.007). For patients with locoregional disease, only 6,001 (76.6%) visited a surgeon, 3,135 (40%) were recommended for resection, and 2,528 (32.3%) underwent operation. In locoregional disease, odds of surviving >6 months were independently increased by receipt of surgery (OR 3.9, 3.5-4.2) and chemotherapy (OR 4.9, 4.4-5.4). Resection also improved overall survival (HR 0.4, 0.38-0.42). However, of the total 2,713 all-stage resected patients, 654 (24.1%) died within 6 months of diagnosis. CONCLUSIONS Only one-third of patients with locoregional pancreatic cancer undergo resection, suggesting potential underutilization of surgery; however, one-quarter of resected patients die within 6 months. These results suggest that surgical therapy may be imperfectly utilized in this lethal disease. Careful consideration of resection as well as more precise stratification of perioperative risk and potential for residual/recurrent disease may help optimize the cohort on whom we operate.

Potential predictors of pancreatic cancer: A population-based screen.

164 Background: Pancreatic cancer (PC) is the fourth leading cause of cancer-related death in the U.S. Unfortunately, PC is usually diagnosed at late stages. We hypothesized that certain diagnoses may precede PC diagnosis and assist in early identification of pancreatic cancer patients. METHODS SEER-Medicare 1991-2005 was used to identify PC patients. PC and prePC diagnoses were identified using ICD9 codes. We then examined pre-pancreatic cancer (prePC) diagnoses and compared the frequency of those diagnoses by PC stage at time of cancer diagnosis. Stepwise logistic regression was used to assess potential PrePC diagnoses. Risk factors were compared by stage at diagnosis using Kruskal-Wallis test, stratified univariate analysis and logistic regression. Time to diagnosis was calculated for each PrePC diagnosis. RESULTS 19,801 PC patients were identified. Significant (p<0.05) prePC diagnoses included acute pancreatitis, chronic pancreatitis, cyst/pseudocyst, other pancreatic disease, bile duct obstruction, diabetes, weight loss, jaundice, abdominal pain, hepatomegaly. Median time (months) and interquartile range percentiles (25th-75th) before PC diagnosis were as follows: acute pancreatitis 0.97 (0.33-8.6), chronic pancreatitis 1.56 (0.37-11), cyst/pseudocyst 0.83 (0.3-3.5), other pancreatic disease 0.47 (0.2-1.2), bile duct obstruction 0.4 (0.17- 0.83), diabetes 30.6 (11.3-59.8), weight loss 1.16 (0.43-5.1), jaundice 0.43 (0.2-0.8), abdominal pain 16 (1.07-55.5), hepatomegaly 1.06 (0.33-1.07). Patients diagnosed at AJCC stage 0 had a mean of 3.53 prePC diagnoses (±SD 1.42); stage IA, 2.80 (1.68); stage IB, 2.42 (1.57); stage IIA, 2.44 (1.63); stage IIB, 2.46 (1.64); stage III, 2.33 (1.59); and Stage IV, 1.79 (1.40) (p<0.001). CONCLUSIONS PC patients who presented at later stages were less likely to have prePC claims identified prior to PC diagnosis compared with patients diagnosed at earlier stages. This analysis of potential prePC diagnoses suggests that access to care and earlier identification of PC related conditions may factor into the stage at which this lethal disease is identified. Further studies need to be conducted to identify and analyze additional predictors of PC and better screen individuals at risk. No significant financial relationships to disclose.