Injury recurrence among untreated and medically treated victims of violence in the USA

Studies on the risk of intentional re-injury tend to use samples of admissions to hospitals and conclude that prior intentional injury is predictive of future such admissions. Limiting samples to persons receiving medical attention misses the equally important population of those intentionally injured but not receiving medical attention. Approximately 2 million violence-related injuries go medically unevaluated each year in the United States. A significant portion of these are severe. Including this population captures the dark figure of untreated intentional injury and allows for the estimation of the effects of medical intervention on future risk. This paper tests the hypothesis that those going without medical evaluation will have higher odds of injury recurrence than the traditionally studied population of those receiving health care. Data are drawn from the National Crime Victimization Survey (NCVS), an ongoing and nationally representative longitudinal survey of households in the United States. The NCVS collects information at the household and person-level, and when victimization is identified through screener questions, at the incident-level. In addition to capturing crime and violence reported to the police and medical professionals, this methodology captures incidents that go unreported and untreated. Controlling for correlates of violence and reinjury, persons injured in violent events receiving medical attention alone, or in combination with police notification, are at substantially lower risk of incurring future violence and violence-related injuries than persons injured but not receiving medical or police attention.