This study examined the quality of injury-related coding in State hospital data and the usefulness of the data in injury control research and surveillance.
Findings from the study about the quality of injury-related coding in hospital data include the following: about 75 percent of non-fata hospitalized cases were coded with a traditional injury diagnosis in the primary diagnosis field, and 90 percent had a traditional injury diagnosis somewhere in the first six diagnosis fields; for cases that had an injury diagnosis code in the first three diagnosis fields, 88.1 percent were E coded, indicating an external cause of the injury or poisoning; and E coding completeness varied by State, with some States reporting high rates of E coding by using non-specific E codes. This study examined the quality of injury-related coding in State hospital data and the usefulness of the data to injury control research and surveillance. Data for the study were obtained from hospital discharge records from 19 States for the year 1997 that included 17.8 million records. A set of criteria were established to identify which cases could be classified as injury-related, resulting in a sample of 1,218,210 cases identified as likely acute or non-acute injury. The data was analyzed to determine the use of E-codes in initial and subsequent diagnosis fields. The findings indicate that in order to capture all injury-related cases, researchers will need to examine secondary diagnosis fields, in addition to the primary diagnosis field. The findings also indicate that is is possible ot combine data from multiple States if researchers are aware of the difference in State data collection and recordation methods. Recommendations for data administrators are discussed. Tables and references
Downloads
Similar Publications
- Method to the Madness: Tracking and Interviewing Respondents in a Longitudinal Study of Prisoner Reentry
- Spectroscopic Differentiation and Chromatographic Separation of Regioisomeric Indole Aldehydes: Synthetic Cannabinoids Precursors
- Predicting the Origin of Stains From Next Generation Sequencing mRNA Data