Managed health care providers, governmental agencies, lawmakers, and the general public are now relying on informational databases to influence health care decisions. This information is also used to derive efficiency profiles for both the individual physician and the institution. However, database reporting procedures are not standardized or monitored. Further, the ICD-9-CM coding system lacks specificity. Consequently, the accuracy of the information in a typical informational database is questionable. As a result, credentialing of physicians and institutions may be based on erroneous information from inaccurate databases. The objective of our study was to document the accuracy of a local informational database. We did a retrospective review of 40 randomly selected charts with surgical procedure codes (ICD-9-CM) for decompression or diskectomy. We compared the diagnosis, procedures, comorbid events, and complications as originally recorded by the surgeon with results of our chart review. We also compared the procedure codes as originally recorded by a hospital recorder with results of our chart review. Results indicated that the information actually reported to the hospital database was lacking or inaccurate in 70% of cases. Discrepancies were not the fault of hospital coders; the coding accuracy rate was 95%. Discrepancies were largely the result of two problems. First, the primary procedure code (ie, the first code on the list recorded by the hospital medical records personnel) did not always match the actual primary procedure (ie, the most complex procedure) done by the surgeon. Second, the ICD-9-CM code was not specific enough to accurately describe the actual surgery done. We conclude that critical discrepancies were recorded into the informational database.