Association between Violent Behavior and Longitudinal Patterns of Health Services Use

This study used a person-centered approach to determine the relationship between violent behavior and health services use in a nationally representative cohort. This longitudinal analysis used data from Waves 1–3 (n = 5,114) of The National Longitudinal Study of Adolescent to Adult Health. Data, including social, familial, and behavioral health survey variables. This study used variables indicating the participant had or had not received counseling, substance abuse, and/or sexually transmitted infection (STI) treatment. In addition, respondents were asked if they ever needed care but did not receive it, and if they had ever shot or stabbed someone. Latent Transition Analysis was used to identify specific classes of violent behavior and health service utilization. Latent transition analysis (LTA) also describes changes between specified latent classes, how classes are characterized, and the probability of changing groups. Between 80 and 98 percent of participants recorded not receiving services, and 20–23 percent report needing care but not receiving it. Using Collins and Lanza’s approach to LTA, a three-status model was optimal based upon both model fit and parsimony. This resulted in groups with a Low Lifetime Probability and Low Service Utilization (LPLS), Low Youth Probability and Unmet Service Needs (YPUS), and Low Adult Probability with High Service Engagement (APSE). The probability of engaging in violence was directly associated with need and utilization of health and mental health services. Results suggest that one potential mechanism to prevent or mitigate violent behavior is to ensure adequate, timely health and mental health treatment are received.