Validation of Lipase and Systemic Inflammatory Response Syndrome as Prognostic Indicators in Pediatric Acute Pancreatitis: A Retrospective Analysis.

INTRODUCTION: Acute pancreatitis (AP) is understudied in the pediatric population despite increasing incidence. Although many cases are mild and resolve with supportive care, severe acute pancreatitis (SAP) can be associated with significant morbidity and mortality. There is a lack of pediatric-specific predictive tools to help stratify risk of SAP in children. METHODS: A retrospective cohort study of patients with AP or recurrent AP at Cohen Children's Medical Center between 2011 and 2016 was performed. Lipase level and the presence of pediatric systemic inflammatory response syndrome (SIRS) on admission were examined as potential predictors of SAP and length of stay (LOS). A multivariate logistic regression or analysis of covariance was used to conduct the multivariate analysis. RESULTS: Seventy-nine pediatric patients met inclusion criteria. Approximately 37% (29/79) had SIRS on admission, 22% (17/79) developed SAP, and there were no mortalities. In both the univariate and multivariate models, SIRS was a predictor of SAP. Mean (SD) LOS for patients with SIRS compared with without SIRS was 9.6 ±â€Š8.3 compared with 6.3 ±â€Š6.9 days (P < 0.05). The mean LOS of patients with one or more comorbidity (48%, 38/79) was 10.0 ±â€Š9.5 compared with 5.2 ±â€Š4.0 days (P < 0.01) for those patients without any comorbidities. Only the presence of comorbidities predicted length of time spent nil per os (NPO; P = 0.0022). Patients with comorbidities stayed an average of 5.6 ±â€Š7.6 days NPO, whereas those without comorbidities spent 2.8 ±â€Š2.4 days NPO. Lipase was not predictive of SAP, LOS, or length of time spent NPO. CONCLUSIONS: These results support the use of SIRS as a simple screening tool on admission to identify children at risk for the development of SAP. The presence of any comorbidity was predictive of LOS and length of NPO in the multivariate model. This may reflect that comorbidities prolong pancreatitis or influence disposition planning.

Predictors of 30-day readmission following pancreatic surgery: A retrospective review.

BACKGROUND: Pancreatectomies have been identified as procedures with an increased risk of readmission. In surgical patients, readmissions within 30 days of discharge are usually procedure-related. We sought to determine predictors of 30-day readmission following pancreatic resections in a large healthcare system. METHODS: We retrospectively collected information from the records of 383 patients who underwent pancreatic resections from 2004-2013. To find the predictors of readmission in the 30 days after discharge, we performed a univariate screen of possible variables using the Fisher's exact test for categorical variables and the Mann-Whitney U test for continuous variables. Multivariate analysis was used to determine the independent factors. RESULTS: Fifty-eight (15.1%) patients were readmitted within 30 days of discharge. Of the patients readmitted, the most common diagnoses at readmission were sepsis (17.2%), and dehydration (8.6%). Multivariate logistic regression found that the development of intra-abdominal fluid collections (OR = 5.32, P < 0.0001), new thromboembolic events (OR = 4.08, P = 0.016), and pre-operative BMI (OR = 1.06, P = 0.040) were independent risk factors of readmission within 30 days of discharge. CONCLUSION: Our data demonstrate that factors predictive of 30-day readmission are a combination of patient characteristics and the development of post-operative complications. Targeted interventions may be used to reduce the risk of readmission.

Outcomes following operative vs. non-operative management of blunt traumatic pancreatic injuries: a retrospective multi-institutional study.

BACKGROUND: Traumatic pancreatic injuries are rare, and guidelines specifying management are controversial and difficult to apply in the acute clinical setting. Due to sparse data on these injuries, we carried out a retrospective review to determine outcomes following surgical or non-surgical management of traumatic pancreatic injuries. We hypothesize a higher morbidity and mortality rate in patients treated surgically when compared to patients treated non-surgically. METHODS: We performed a retrospective review of data from four trauma centers in New York from 1990-2014, comparing patients who had blunt traumatic pancreatic injuries who were managed operatively to those managed non-operatively. We compared continuous variables using the Mann-Whitney U test and categorical variables using the chi-square and Fisher's exact tests. Univariate analysis was performed to determine the possible confounding factors associated with mortality in both treatment groups. RESULTS: Twenty nine patients were managed operatively and 32 non-operatively. There was a significant difference between the operative and non-operative groups in median age (37.0 vs. 16.2 years, P = 0.016), grade of pancreatic injury (grade I; 30.8 vs. 85.2%, P value for all comparisons <0.0001), median injury severity score (ISS) (16.0 vs. 4.0, P = 0.002), blood transfusion (55.2 vs. 15.6%, P = 0.0012), other abdominal injuries (79.3 vs. 38.7%, P = 0.0014), pelvic fractures (17.2 vs. 0.00%, P = 0.020), intensive care unit (ICU) admission (86.2 vs. 50.0%, P = 0.003), median length of stay (LOS) (16.0 vs. 4.0 days, P <0.0001), and mortality (27.6 vs. 3.1%, P = 0.010). CONCLUSIONS: Patients with traumatic pancreatic injuries treated operatively were more severely injured and suffered greater complications than those treated non-operatively. The greater morbidity and mortality associated with these patients warrants further study to determine optimal triage strategies and which subset of patients is likely to benefit from surgery.