Data matter: the creation of an acute care surgery registry

Background: Patients presenting to hospital with emergency general surgery (EGS) conditions such as appendicitis, biliary disease, and hernias represent 11% of hospital admissions in Canada. This study evaluated the establishment of a local registry of EGS patients by auditing the accuracy of data abstracted from administrative records compared with manual chart review. Methods: Patients admitted to and consulted by the Acute Care Surgery (ACS) service at London Health Sciences Centre during a 1-year period were entered into a prospective registry. Data on demographics, comorbidities, imaging, diagnosis, treatment, and outcomes were collected from both manual chart review of patient electronic health records and routinely collected administrative data. Accuracy of data from administrative sources compared with those from manual chart review was audited for a 6-month period. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated for comorbidities, imaging, and complications to determine which variables can be accurately collected using administrative data. Results: A total of 2159 patients were included in the analysis. Overall accuracy was good for most comorbidities, ranging from 83.8% to 100%; however, sensitivity, specificity, NPV, and PPV ranged drastically, suggesting that data on some comorbidities may not be accurately collected through health records. Imaging procedures were consistent between the 2 data collection methods, with good overall accuracy of 97.6%. For assessed complications, accuracy was excellent overall, with an average accuracy of 99.6% (range 98.3% to 100%). Conclusion: This study shows that a substantial portion of data necessary for an ACS registry can be reliably extracted from routinely collected administrative data.