Identifying Algorithms to Improve the Accuracy of Unverified Diagnosis Codes for Birth Defects

We found that surveillance programs that rely on unverified diagnosis codes can use algorithms to dramatically increase the accuracy of case finding, without having to review medical records. This can be important for etiologic studies. However, the use of increasingly restrictive case definition algorithms led to a decrease in completeness and the disproportionate exclusion of less severe cases, which could limit the widespread use of these approaches.