Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

Health Services Research & Development

Go to the ORD website
Go to the QUERI website

HSR&D Citation Abstract

Search | Search by Center | Search by Source | Keywords in Title

Stop the Scanners! External Validation of Algorithms Predicting Complications in Patients with Crohn’s Disease Presenting to the Emergency Department

Govani S, Waljee AK, Swoger J, Saul M, Higgins P. Stop the Scanners! External Validation of Algorithms Predicting Complications in Patients with Crohn’s Disease Presenting to the Emergency Department. Poster session presented at: Crohn's and Colitis Foundation of America Advances In Inflammatory Bowel Diseases Clinical and Research Conference; 2014 Dec 6; Orlando, FL.




Abstract:

Stop the Scanners! External Validation of Algorithms Predicting Complications in Patients With Crohn's Disease Presenting to the Emergency Department Govani Shail1, Waljee Akbar1, Swoger Jason2, Saul Melissa2, Higgins Peter1 1University of Michigan, Ann Arbor, Michigan, 2University of Pittsburgh, Pittsburgh, Pennsylvania BACKGROUND: Patients with Crohn's disease (CD) are frequently over-exposed to radiation from computed tomography (CT), leading to increased cancer risk in this young population. We have published predictive equations using data from a single tertiary care center to identify CD patients in the emergency department (ED) who obtain little benefit from CT scans in the ED. These two equations using CRP and ESR predict the presence of complications (perforation, abscess or other serious outcome) in CD patients presenting to the ED. The first equation, created using logistic regression, had a sensitivity of 93%, a negative predictive value (NPV) of 98.1% and potentially saved 43% of patients from unnecessary CT scan radiation. The second, simplified model, ESR + 5 CRP (mg/dL) #10, had similar sensitivity but lower specificity but only saved 18.5% of the patients from CT. We aimed to validate these two algorithms in an external cohort at a second tertiary care center in a different state. METHODS: The electronic medical records of a large tertiary care medical system were queried for patients with CD presenting to the ED who underwent a CT of the abdomen/pelvis between 2009 and 2012. De-identified medical records of these visits and labs were obtained and manually reviewed. Labs were from within 24 hours of the ED visit. Visits due to trauma, and CTs without oral and intravenous contrast were excluded. The logistic regression model and the simple equation for complications were tested to determine sensitivity, NPV, and scan avoidance rate. RESULTS: A total of 919 ED visits were reviewed for 559 individuals. Of these, 455 had CD with 797 visits. Applying exclusion criteria, 501 visits with scans were analyzed. ESR and CRP data were complete in only 211 visits, which were the basis of this external validation analysis. The logistic regression model had a sensitivity of 80% and NPV of 96.0%, saving 46.9% of the patients from unnecessary CT scans. The complication miss rate was 1.9% (4/211). The simple model had a sensitivity of 90%, NPV of 93.1% and saved 13.7% of the patients from CT. The miss rate was 1.0% (2/211). CONCLUSIONS: Prediction models for the need for ED CT scans in CD patients were developed on a dataset from one tertiary care center and have now been validated at a second, external tertiary care center. Models predicting complications in CD using CRP and ESR perform well and have the potential to save these patients from unnecessary CT scans and radiation in 45% of cases, if a miss rate of 1-2% is acceptable. These models require further study of generalizability beyond tertiary care centers, and prospective studies of pragmatic implementation in the ED setting.





Questions about the HSR&D website? Email the Web Team.

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.