Table 2

Predictors of CJ Leverages

Arrest: Drop Charges(n = 867)Conviction: Avoid Jail/Prison(n = 847)Probation/Parole Mandate(n = 845)
OR95% CIOR95% CIOR95% CI
Age above median (>44 years)0.57**(0.38–0.85)0.54**(0.35–0.84)0.50**(0.32–0.78)
Male1.60**(1.07–2.38)1.78**(1.15–2.75)2.89***(1.82–4.57)
African American1.75(0.87–3.50)1.67(0.89–3.39)1.70(0.84–3.46)
White1.66(0.88–3.15)1.11(0.57–2.13)1.34(0.69–2.58)
Substance abuse2.17***(1.41–3.33)2.13**(1.35–3.38)
BPRS score above median (>30)1.26(0.84–1.89)1.26(0.82–1.93)1.82**(1.15–2.87)
GAF score above median (>47)0.85(0.55–1.32)0.76(0.47–1.23)
Insight score above median (>18)1.070.71–1.60)1.00(0.65–1.54)1.40(0.89–2.19)
Past hospitalizations above median (>3)1.96**(1.29–2.98)2.82***(1.75–4.55)1.72*(1.10–2.72)
Time in treatment above median (>20 years)1.76**(1.14–2.71)1.77*(1.12–2.81)1.580.99–2.52)
Model significanceLikelihood ratio = 56.7; df = 10***Likelihood ratio = 62.71; df = 9***Likelihood ratio = 61.39; df = 9***
  • * p < .05

  • ** p < .01

  • *** p < .001

  • Excluded from the pooled model because sites' odds ratios differed significantly, according to Zelen's test of homogeneity.