Abstract T P365: Childhood Stroke: ICD-9 Code Positive Predictive Values for Identifying Hemorrhagic and Ischemic Events

Abstract

Introduction: Childhood stroke event rates have been reported both using administrative databases and population-based epidemiological studies. The latter include verification of stroke as a case and categorization of type. Administrative databases allow cheaper, quicker estimation of rates and possible extrapolation of estimated rates to population-based rates. However, these estimations rely on the accuracy and interpretation of the ICD-9 coding. Methods: The Greater Cincinnati/Northern Kentucky Stroke Study measures temporal trends in the incidence rates in a biracial population of 1.3 million. Discharge lists with primary and secondary ICD-9 codes 430-436 from 16 area hospitals for 2010 were obtained; 437-438, 674 and 747 were also included at the Children’s hospital. Detailed information from medical records of potential cases was abstracted by trained research nurses and reviewed by stroke physicians, who determined if the event was a case and, also the event type (hemorrhagic stroke, infarction or transient ischemic attack(TIA)). Results: A total of 89 potential events in children <20 years of age were reviewed, yielding 19 confirmed cases. Positive predictive values (PPV) for the primary ICD-9 codes for specific types varied from 0% to 100%. Primary and secondary ICD-9 codes, event types, and percent correct are presented in the Table. Conclusions: Childhood stroke cases captured through selected ICD-9 codes: 430-432, 434.x1, 434.9 and 435.9 (marked with an asterisk in the table) in the primary position would yield 14 strokes/TIAs, and underestimate the number of events by 26%. However, using both primary and secondary codes they would yield 34 strokes//TIAs, and overestimate the number of events by 79%. Population-based epidemiology studies are essential to monitor the validity of using ICD-9 codes to estimate childhood stroke/TIA incidence.