Jordan Holmes

Decreased Severity in Recurrent Versus Initial Episodes of Acute Pancreatitis.

Objectives The comparative outcomes of initial versus recurrent acute pancreatitis (AP) have not been clearly established. Aim The aim was to compare the clinical outcomes of those with an initial episode of AP to those with recurrent AP stratified by the number of prior episodes. Methods This retrospective cohort study included consecutive patients with AP admitted to the Cleveland Clinic between 2008 and 2011. The odds of severe AP, multisystem organ failure, ICU admission, new local complications, elevated blood urea nitrogen and bedside index for severity in acute pancreatitis score, systemic inflammatory response syndrome, and mortality were compared using univariable and multivariable logistic regression. Results Two hundred and ninety two patients were included, of which 213 (72%) were admitted on their initial AP episode. Mortality in patients experiencing first episode was 4.7%, compared to 0% in patients with recurrent attack of pancreatitis (P = 0.047). Prior episodes of AP were found to be protective against multisystem organ failure (odds ratio, 0.14 for each prior episode; confidence interval, 0.01–0.76) and intensive care unit admission (0.24, confidence interval, 0.06–0.91), adjusting for potential confounding factors such as transfer status and obesity. Conclusions Patients presenting with recurrent AP may be at decreased risk of a clinically severe course and incur decreased mortality.

Continuing Medical Education Questions: April 2014

A. Endoscopic therapy provides initial and long-term bleeding control in the majority of patients with GIAD-related bleeding. B. Endoscopic therapy provides only initial bleeding control in the majority of patients with GIAD-related bleeding. C. Isolated small bowel GIAD has lower rebleeding risk following endoscopic therapy compared with gastric or colonic GIAD. D. Intraoperative enteroscopy with segmental small bowel resection of GIAD-aff ected bowel remains the fi rst-line option for treatment of small bowel GIAD. 3. Which one of the following statements is correct regarding the role of Heyde’s syndrome on GIAD?

Positive predictive value of ICD-9 discharge diagnosis of acute pancreatitis

Acute pancreatitis (AP) is a disease with a rising incidence that causes considerable morbidity and, in its most severe form, high mortality. Research to investigate risk factors and treatment outcomes in AP use administrative databases for case finding, usually based on the International Classification of Diseases, Ninth Revision (ICD-9) 577.0; however, in a recent systematic review, the positive predictive value (PPV) for ICD-9 code 577.0 ranged from 40% to 75%, revealing low to moderate PPV and specificity. A single-center retrospective cohort study was conducted in our institution to determine the PPVof ICD-9 code (577.0) for primary discharge diagnosis of AP to derive an algorithm for improved PPVand specificity. We also compared the accuracy of the discharge diagnosis based on the admitting service (internal medicine vs gastroenterology vs surgery). A manual chart review was conducted of consecutive patients admitted through the emergency department between January 1, 2010, and December 31, 2011, with a primary discharge diagnosis of ICD-9 577.0 assigned by the admitting physician. The ICD-9 code discharge diagnosis was compared with the American College of Gastroenterology guidelines definition, which is made based on 2 of the following criteria: (1) abdominal pain characteristic of AP, (2) serum amylase and/or lipase level 3 times the upper limit of normal, and (3) characteristic findings of AP on computed tomographic scan or ultrasound. For encounters that did not fulfill the criteria for AP, the data were collected on alternative diagnoses. Finally, a new algorithm was tested that included ICD-9 code of 577.0 plus the presence of a serum lipase level 3 times the upper limit of normal. The PPV of ICD-9 code (577.0) for primary discharge diagnosis of AP was calculated as the percentage of true AP cases out of all patients with a diagnosis of ICD-9 577.0. The PPV of the new algorithm was calculated as the percentage of true AP cases out of all patients with primary discharge diagnosis of ICD-9 577.0 and serum lipase level 3 times the upper limit of normal. The results of our study are shown in Table 1. Low to moderate PPVof 67% for ICD-9 577.0 was observed for the discharge diagnosis of AP. No significant difference was found in the accuracy of discharge diagnosis among the different admitting services. The implication is that studies relying on ICD-9 discharge diagnosis for case finding will include a large number of patients who did not truly have AP. This misclassification will result in difficulty in interpreting the studies and applying them to the care of patients with AP. We tested an improved method for identifying ‘‘true’’ AP cases for research purposes, which was to filter the billinggenerated list using elevated lipase level. This was accomplished by cross-matching the billing and laboratory databases, preserving the automated and quick case finding needed for large epidemiological studies. The new algorithm demonstrated improved PPV and specificity. The tradeoff was a lower sensitivity, which we find acceptable. Based on limited data, it seems that ICD-10 may improve the identification of AP. A nationwide study validating the ICD-10 coding for AP in the Swedish Patient Register found a PPV of 83%. In conclusion, our validation of the ICD-9 577.0 primary discharge diagnosis of AP showed a low to moderate PPV, which improved with a new and improved algorithm of ICD-9 code plus lipase levels. The authors declare no conflict of interest.