Predicting Recidivism of Older Offenders: The Role of Serious Mental Illness (SMI) and Pharmacy Spending
Contrary to the common belief that older offenders age out of the criminal justice system, an increasing number of older adults have been involved with recurring arrests and incarcerations over the past decades (Lewis et al., 2006). Equally alarming is that many of these older offenders exhibit the signs and symptoms of serious mental illnesses (SMI) but often remain untreated. Considering the significant influence of SMI and its treatment on elders risk for re-arrest, the purpose of the study was to examine how these factors predict re-arrest of older offenders while controlling for demographics and the number of previous arrests.
Methods
Multiple sources of data including Florida Medicaid claims and the Florida Department of Law Enforcement data were used to select 1,100 offenders (age 60 and older) who were residents in the state of Florida. To be included for analyses, subjects had to be arrested at least once during the baseline year from July 2002 to June 2003 and receive Medicaid services at least once from July 2002 to June 2007. Re-arrest during the fiscal year of 2007-2008 was used as a distal outcome. Multivariate logistic regression estimated the likelihood of re-arrest in relation to demographics (race/ethnicity, age, gender), SMI diagnoses (psychosis, bipolar I disorder, major depression, other bipolar disorders), Medicaid spending and yearly numbers of arrests during the five years (2002-2007) were examined.
Results
The average age of the sample was 65.23 (SD = 5.24), 18.3% were female, and 30.5% were blacks. Psychosis (15.7%) was the most frequent SMI diagnoses followed by depression (13.4%), other bipolar disorders (12.5%), and bipolar I disorders (4.6%). The average Medicaid spending in pharmacy for mental health was $4461.73 (SD = 8159.16) over the five year period. The yearly number of arrests during the five years was in the range of 1.36-1.78. Of 1,100 subjects, 11.2% were rearrested in the outcome year of 2007-2008. The results of the multivariate logistic regression show that re-arrest was predicted significantly by Black race (OR = 1.60), younger age (OR = 0.89), psychotic disorder (OR = 1.91), lower Medicaid for psychotropic medication (OR = 0.59), and a higher number of rearrest in 2002-2003 (OR = 1.49) and 2006-2007 (OR = 2.18).
Conclusions and Implications
One finding of particular interest was that higher Medicaid spending on psychotropic medication was associated with a significantly reduced rate of re-arrests among older offenders. Considering for high public costs associated with repeat offenders, there is a need for improving screening and treatment systems for older offenders with mental health care needs. It is important for social workers to recognize and advocate mental health care needs among older offenders. Further studies should systematically examine the relationship of Medicaid spending in pharmacy for mental health to mental health outcomes and criminal behaviors among older offenders.