Table 3. Model-Based Adjusted Rates Within Trial Groups and Adjusted Relative Rates Between Intervention to Control Communities of Outcomes Using the Intention-to-Treat Population.
| Outcome by group | Adjusted rate (95% CI)a | Adjusted rate ratio (95% CI)b | P valuec | |
|---|---|---|---|---|
| Intervention communities | Control communities | |||
| Drug overdose deaths | ||||
| Overall | 55.99 (50.00-62.71) | 61.11 (54.25-68.83) | 0.92 (0.78-1.07) | .26 |
| Research sited | ||||
| Kentucky | 69.98 (53.48-91.57) | 64.75 (42.22-99.30) | 1.08 (0.66-1.76) | NR |
| Massachusetts | 55.20 (36.14-84.32) | 57.32 (40.70-80.72) | 0.96 (0.56-1.65) | |
| New York | 54.05 (38.51-75.87) | 61.99 (48.03-80.01) | 0.87 (0.58-1.30) | |
| Ohio | 48.01 (41.01-56.19) | 58.73 (46.22-74.62) | 0.82 (0.62-1.09) | |
| Urban or rural classificationd | ||||
| Urban | 57.05 (49.63-65.58) | 60.82 (51.80-71.41) | 0.94 (0.76-1.15) | NR |
| Rural | 54.37 (43.50-67.94) | 62.11 (49.06-78.63) | 0.88 (0.63-1.21) | |
| Aged | ||||
| 18-34 y | 51.73 (44.80-59.72) | 57.81 (48.55-68.85) | 0.89 (0.72-1.11) | NR |
| 35-54 y | 78.54 (65.64-93.98) | 87.03 (74.49-101.7) | 0.90 (0.74-1.11) | |
| ≥55 y | 42.61 (35.72-50.83) | 48.22 (36.28-64.10) | 0.88 (0.64-1.22) | |
| Sexd | ||||
| Male | 70.84 (59.40-84.49) | 80.87 (70.51-92.75) | 0.88 (0.73-1.05) | NR |
| Female | 39.22 (33.88-45.40) | 44.53 (37.04-53.53) | 0.88 (0.71-1.09) | |
| Race and ethnicityd,h | ||||
| Hispanic | 50.69 (33.37-76.99) | 53.89 (40.43-71.85) | 0.94 (0.57-1.55) | |
| Non-Hispanic Black | 76.83 (60.34-97.83) | 93.80 (74.50-118.1) | 0.82 (0.59-1.14) | |
| Non-Hispanic White | 52.59 (44.43-62.25) | 57.01 (48.13-67.52) | 0.92 (0.76-1.11) | NR |
| Non-Hispanic other | 22.37 (11.09-45.12) | 29.98 (13.93-64.52) | 0.75 (0.27-2.04) | |
| Overdose deaths involving any opioid and any psychostimulant (other than cocaine) | ||||
| Overall | 6.68 (5.14-8.69) | 10.55 (7.75-14.36) | 0.63 (0.44-0.91) | .02 |
| Research sited | ||||
| Kentucky | 18.27 (12.57-26.56) | 24.61 (14.72-41.16) | 0.74 (0.40-1.36) | NR |
| Massachusetts | 2.94 (0.65-13.26) | 3.45 (1.03-11.61) | 0.85 (0.12-5.87) | |
| New York | 4.26 (2.07-8.76) | 10.39 (3.81-28.37) | 0.41 (0.11-1.53) | |
| Ohio | 8.86 (6.00-13.10) | 12.60 (6.58-24.11) | 0.70 (0.34-1.45) | |
| Urban/rural classificationd | ||||
| Urban | 5.83 (4.06-8.39) | 8.75 (4.88-15.71) | 0.67 (0.37-1.22) | NR |
| Rural | 7.53 (4.97-11.42) | 12.88 (8.45-19.61) | 0.59 (0.33-1.04) | |
| Overdose deaths involving any opioid and cocaine | ||||
| Overalle,f | 10.32 (8.32-12.80) | 10.94 (8.28-14.47) | 0.94 (0.69-1.30) | .72 |
| Research sited | ||||
| Kentucky | 7.51 (3.44-16.38) | 6.77 (1.94-23.64) | 1.11 (0.26-4.64) | NR |
| Massachusetts | 12.15 (6.44-22.92) | 12.52 (4.75-33.01) | 0.97 (0.33-2.81) | |
| New York | 15.68 (9.08-27.07) | 16.27 (10.40-25.48) | 0.96 (0.48-1.92) | |
| Ohio | 8.14 (5.70-11.63) | 9.97 (5.88-16.89) | 0.82 (0.44-1.53) | |
| Urban or rural classificationd | ||||
| Urban | 13.81 (10.42-18.29) | 13.13 (9.15-18.85) | 1.05 (0.69-1.61) | |
| Rural | 6.83 (4.44-10.51) | 10.10 (6.08-16.80) | 0.68 (0.35-1.29) | |
| Overdose deaths involving any opioid and any benzodiazepine | ||||
| Overalle,g | 6.59 (5.25-8.27) | 6.64 (5.00-8.82) | 0.99 (0.73-1.35) | .96 |
| Research sited | ||||
| Kentucky | 7.73 (4.90-12.19) | 8.30 (4.13-16.67) | 0.93 (0.42-2.07) | NR |
| Massachusetts | 7.28 (3.65-14.52) | 4.95 (1.41-17.37) | 1.47 (0.42-5.09) | |
| New York | 7.31 (4.48-11.94) | 7.24 (4.34-12.08) | 1.01 (0.48-2.12) | |
| Ohio | 4.95 (2.71-9.05) | 6.18 (4.08-9.34) | 0.80 (0.40-1.63) | |
| Urban or rural classificationd | ||||
| Urban | 7.07 (5.54-9.01) | 6.44 (4.49-9.22) | 1.10 (0.72-1.68) | NR |
| Rural | 5.46 (3.26-9.14) | 7.60 (4.48-12.89) | 0.72 (0.35-1.46) | |
Abbreviation: NR, not reported.
Adjusted rate per 100 000 residents aged 18 years and older.
Adjusted relative rate of outcome between intervention and control communities. For subgroups (research site, urban or rural classification, age, sex, race and ethnicity), the 95% CIs reported used a Bonferroni correction.
For subgroups (research site, urban or rural classification, age, sex, race and ethnicity), false discovery rate–adjusted P value for the interaction between stratification variable and trial group were not significant.
A separate negative binomial model was fit for each test of effect modification. Each model included the following fixed effects: trial group, research site (Kentucky, Massachusetts, New York, Ohio), urban or rural classification, baseline opioid overdose death rate, baseline rate of outcome, stratification variable (age, sex, race and ethnicity, if needed), and a 2-way interaction between trial group and the stratification variable.
Results obtained from a negative binomial model adjusting for research site (Kentucky, Massachusetts, New York , Ohio), urban or rural classification, baseline opioid overdose death rate, and baseline rate of the outcome. The natural log of the observed community population size of individuals 18 years of age or older measured from the 2020 Bridged-Race Population Estimates or the 2017-2021 American Community Survey 5-Year Averages is used as an offset.
Estimated dispersion parameter: k, 0.1472; 95% CI, 0.0774-0.2798.
Estimated dispersion parameter: k, 0.0925; 95% CI, 0.0298-0.2866.
Other included American Indian, Alaskan Native, Asian, Native Hawaiian, other Pacific Islander, and other race not previously mentioned.