Table 4.
Sensitivity Analysis Estimating the Association Between Medication Possession Ratio (MPR) and Violent/Nonviolent Crime (by Death and Duration of Follow-Up)
| Subgroup | MPR | Violent Crime | Nonviolent Crime |
|---|---|---|---|
| ARRa (95% CIb) | ARR (95% CI) | ||
| Surviving participants (n = 10514,c 92%) | |||
| ≤0.19 | 1.35 (1.23, 1.48)d | 1.22 (1.22, 1.42) | |
| 0.20–0.39 | 2.29 (2.03, 2.58) | 1.63 (1.49, 1.79) | |
| 0.40–0.59 | 2.26 (1.99, 2.56) | 1.65 (1.50, 1.81) | |
| 0.60–0.79 | 1.70 (1.51, 1.91) | 1.47 (1.36, 1.60) | |
| ≥0.80 | Reference | Reference | |
| MPR (<0.80) | 1.55 (1.43, 1.68) | 1.41 (1.33, 1.51) | |
| <5 years of follow-up (n = 2530, 22%) | |||
| ≤0.19 | 1.57 (1.18, 2.09) | 1.44 (1.17, 1.78) | |
| 0.20–0.39 | 2.31 (1.55, 3.44) | 1.69 (1.29, 2.22) | |
| 0.40–0.59 | 2.18 (1.46, 3.27) | 1.40 (1.08, 1.80) | |
| 0.60–0.79 | 1.89 (1.20, 2.96) | 1.29 (1.00, 1.65)e | |
| ≥0.80 | Reference | Reference | |
| MPR (<0.80) | 1.77 (1.35, 2.32) | 1.43 (1.19, 1.72) | |
| 5–<10 years of follow-up (n = 3395, 30%) | |||
| ≤0.19 | 1.48 (1.24, 1.76) | 1.28 (1.13, 1.45) | |
| 0.20–0.39 | 2.27 (1.84, 2.81) | 1.51 (1.29, 1.78) | |
| 0.40–0.59 | 2.16 (1.71, 2.72) | 1.59 (1.36, 1.86) | |
| 0.60–0.79 | 1.67 (1.34, 2.08) | 1.41 (1.23, 1.61) | |
| ≥0.80 | Reference | Reference | |
| MPR (<0.80) | 1.64 (1.40, 1.92) | 1.37 (1.23, 1.53) | |
| ≥10 years of follow-up (n = 5537, 48%) | |||
| ≤0.19 | 1.36 (1.22, 1.51) | 1.31 (1.18, 1.44) | |
| 0.20–0.39 | 2.28 (1.97, 2.65) | 1.70 (1.50, 1.91) | |
| 0.40–0.59 | 2.24 (1.93, 2.61) | 1.69 (1.50, 1.91) | |
| 0.60–0.79 | 1.74 (1.52, 2.00) | 1.53 (1.38, 1.69) | |
| ≥0.80 | Reference | Reference | |
| MPR (<0.80) | 1.55 (1.41, 1.71) | 1.43 (1.32, 1.55) |
Note: ARR, adjusted rate ratio; CI: confidence interval; MPR, medication possession ratio.
aEach multivariable GEE model controlled for age at enrollment (centered, age 18), gender (men and women), ethnicity (White, Indigenous, and other); education level (<Gd. 10, Gd. 10/11, Gd. 12, and Vocational/University), use of substance disorder-related services (continuous variable), number of offenses in the previous year (continuous variable), number of 120-day intervals (continuous variable), and duration of follow-up in days (offset variable).
b95% CIs for the adjusted rate ratios were estimated using robust standard errors.
c948 participants (8%) who died during the study period were excluded from the analysis.
dBold indicates significant at P value <.001.
eItalics bold indicates significant at P value ≤.05.