Accuracy of recorded tumor, node, and metastasis stage in a comprehensive cancer center

Purpose: The benefits of recording the tumor, node, and metastasis (TNM) stages of cancer patients are well accepted, but little is known about how accurately this is performed. An audit was performed to determine the accuracy of recorded stage and to act as a baseline before the implementation of an education program. Patients and Methods: All new patient referrals to Princess Margaret Hospital between July I and August 31, 1997, were reviewed. An audit panel composed of! five health record technicians (HRTs) and 10 doctors was assembled. Each auditor reviewed 10% of the health record. If there was a discrepancy between the stage in the health record and the auditor stage, then the final stage was determined by the audit committee. Analysis of the agreement between the health record, the physician auditor, the HRT auditor, and the final! stage was performed. Results: A total of 855 patients were referred with a new diagnosis of a malignancy for which there was a TNM stage system; 833 patients (97.4%) had a stage assigned. There was agreement between the health record stage and final stage in 80% (95% confidence interval [CI], 77% to 82%) of cases for clinical stage, compared with 90% (95% CI, 87% to 92%) for pathologic stage. Of the major site groups, lung was the least accurately recorded. The most common major discrepancies were due to the recording of X when a definite category could be assigned. Conclusion: This audit demonstrates the importance of staging and provides impetus to develop staging guidelines and education programs. (C) 2002 by American Society of Clinical Oncology.