Do billing codes accurately capture intravenous tissue plasminogen activator treatment rates? Justified concern for clinical performance measures based on billing code assignment

Background: International Classification of Diseases, Ninth Revision, Clinical Modi-fication (ICD-9-CM) codes are commonly used to determine US national stroke vol-ume and intravenous (IV) tissue plasminogen activator (tPA) treatment rates; however, this method is often criticized because of assumed poor validity and reli-ability of coding assignment. We sought to understand the validity of IV tPA ICD-9-CM code assignments within a comprehensive stroke center in the southeastern United States. Methods: Confirmed stroke registry IV tPA cases were retrieved from 2009 to 2011; tPA drip and ship cases were eliminated from the analysis. Retained clinical data included admission National Institutes of Health Stroke Scale (NIHSS) scores, hemorrhagic transformation, diffusion positive magnetic resonance imaging (MRI) results, and discharge modified Rankin Scale (mRS) score. Results: A total of 247 IV tPA cases were assembled, of which 78% were appropriately assigned the IV tPA billing code. ICD-9 code 434.91 (cerebral artery occlusion with infarct) was used for 72% of the sample, 434.11 (cerebral emboli with infarct) was used for 9% of cases, and 433.11 (carotid occlusion with infarct) was assigned to 2% of cases. Interestingly, 435 (transient cerebral ischemia) was assigned to 2%(n 5 7) with all of these cases having NIHSS score more than 2 at time of treatment, diffusion MRI documentation of infarction in 29%, and 43% having a discharge mRS score more than 2. Conclusions: Our findings support the concern that billing codes may significantly underestimate actual IV tPA treatment volume in the United States and suggest the need for regular audit of billing codes by Stroke Center leaders, with provision of feedback and education to coders, aimed at improving code assignment.