Mental and Physical Healthcare Utilization Among Young Prisoners: A Longitudinal Study

ABSTRACT

The present longitudinal study describes changes in and covariates of healthcare utilization in 75 young males (aged 17– 22 years) during their first year in a Portuguese prison. The prisoners were assessed 1, 3, 6, and 12 months after their arrival in custody. Multilevel negative binomial regression analyses were conducted to examine the effect of time in prison as well as several clinical, socio-demographic, and criminological characteristics on: (a) treatment for mental health problems and (b) treatment for physical health problems. Although healthcare utilization for mental health issues remained stable over 12 months of imprisonment, visits for physical health problems were highest during the first month and then declined significantly. Patterns of healthcare utilization were associated with time spent in prison, mental health symptoms, Portuguese nationality, older age at the onset of imprisonment, criminal history, and severe disciplinary infractions. Strikingly, prisoners who initially reported more mental health symptoms upon arrival to the prison facility received less mental health services for the first six months of imprisonment. The limitations of this study and its implications for correctional research and practice are discussed.

KEYWORDS:

• Young prisoners
• mental and physical healthcare utilization
• longitudinal research

Notes

1. Newly incarcerated prisoners (n = 17) and prisoners who had already been incarcerated for a period of time before admission to the current prison facility (n = 58) were compared on all predictor variables included in the study as observed during the 1st month in prison (i.e., wave 1). Prisoners who had already been incarcerated for a period of time (i.e., 1 month or more) had less visits (p < .001), were more likely to have received a sentence (p < .001), had a more elaborate criminal history (p = .029), and had fewer mental health symptoms (p < .001) than would be expected by chance. These two groups were similar on the other variables. Due to the small sample size, we were not able to run the analyses separate for both groups; however, we did include prior time served in prison as a (continuous) covariate.

2. To examine attrition bias, we developed a probit selection model in which we estimated attrition with the independent variables of the study. This model was not significant, which suggests a random pattern of attrition.

3. We developed an ordered logit regression model for each missing item, and used the missing item as an outcome and the other items of the BSI as predictors (i.e., those more correlated with the missing item). After fitting the best model to the data, marginal effects were calculated to predict the prisoner's score on the missing item based on the score of the items included in the model.