Rates and Characteristics of Violent Death Victims Among Persons Released From Prison

Background: Persons released from prison face multiple challenges to successful re-entry, including high risk of mortality by natural causes and violence.  Little is known about violent fatalities, such as homicide and suicide, among releasees.  Additional study is needed to ensure safe and successful re-entry of releasees and decrease violence in the communities to which they return.  This study examines characteristics of violent death victims among prison releasees in North Carolina, asking: Do rates of violent death among releasees differ significantly from the general population? What characteristics are associated with releasees who died violently?

Methods: This is a retrospective cohort study of 48,228 inmates released from NC Division of Corrections (NCDC) from 7/06 through 6/07.  This cohort was predominantly male (87.1%), nonwhite (55.4%), never married (60.9%), with less than 12^th grade education (69.5%), and mean age of 35 (SD=10.3).  The NCDC provided inmate records for match with the NC National Violent Death Reporting System, and releasees who died violently within 12 months were identified and data merged. Probabilistic matching was used, with SSN, name, and DOB creating single variable strings, with subsequent manual review of matches and non-matches.  We identified 88 violent deaths.

Analysis involved computation of rate of violent deaths per 100,000 population and use of Chi-square and logistic regression to identify significant characteristics associated with risk of violent death and compute adjusted odds ratios.

Results: Rates of violent death per 100,000 were higher for releasees than the general population overall (182.5, 24.8), for homicide (107.8, 8.5) and suicide (62.2, 4.0). The proportion of homicides to suicides differed significantly between releasees (59.1%, 34.1%) and general population (32.9%, 63%), X² (2,N=3,436) = 7248.2, p < .001. Controlling for demographic and prison-related characteristics revealed for all violent deaths significantly higher odds of violent death for males, (adjusted OR=2.7,p = .039), mentally ill in prison (adjusted OR=2.2, p=.002), and younger age (adjusted OR=0.97,p =.033).   For homicides significantly higher odds occurred for nonwhites (2.9  p=.001), >2 incarcerations (2.1,p =.019), and younger age (0.9,p =.001). For suicides significantly lower odds occurred in nonwhite (0.4,p = .016), without supervised release (0.4,p =.047), and receiving substance abuse treatment in prison (0.3,p =.010).  

Conclusions: Violent death rates are significantly higher among releasees than the general population, with males, young adults, and mentally ill at higher risk of overall violent death; young adults, nonwhites, and persons with multiple incarcerations at higher risk of homicide; and whites, supervised post-release, and untreated for substance abuse in prison at higher risk of suicide. Though absolute numbers are not high, violent deaths among releasees involve not simply loss of an individual with rehabilitation potential, but also significant impact on family and community. Re-entry practices must address violent death risks in pre-release interventions, particularly targeting high risk groups such as the mentally ill or untreated substance abusing individuals.  More research is needed to better understand the predictors and socio-environmental contexts in which violent deaths occur post-release, so that releasees have maximal opportunity to be safe as they work to reintegrate into the community.