RESUMEN
BACKGROUND: Inaccurate use of International Classification of Diseases, Ninth Revision (ICD-9), codes obfuscates registries used for research, resulting in unreliable data and inaccurate measurement of outcomes, and it may contribute to mismanagement of patients. Thus it is important to understand the prevalence of ICD-9 code misuse. We chose chronic kidney disease (CKD) as a condition of interest after several patients recruited for a previous study indicated they did not have the disease, despite the presence of the ICD-9 code (585.x) in their electronic medical record (EMR). METHODS: Retrospective chart review of patients with the ICD-9 code for CKD stage 3 (585.3; n = 325). Data were collected from EMRs at 3 primary care practices Buffalo, New York (n = 2), and Kansas City, Kansas (n = 1). RESULTS: Across all practices, 47% of patients with the CKD ICD-9 code did not have clinical indicators for the disease, based on Kidney Disease Outcomes Quality Initiative guidelines. CONCLUSIONS: The CKD stage 3 ICD-9 code usage did not accurately reflect the prevalence of disease among this population. This has clinical implications because patients may be treated or receive tests for a disease they do not have. This also presents an important issue for research projects that rely on accurate data from EMRs to identify and recruit patients.