Table 3. Clinic-Level Implementation Outcomes of the PROUD Trial Assessed in the 3 Years Postrandomization by Health System and Clinica.
| Outcome | Outcome value by health system and clinic | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Health system 1 | Health system 2 | Health system 3 | Health system 4 | Health system 5 | Health system 6 | |||||||
| PROUD | Usual care | PROUD | Usual care | PROUD | Usual care | PROUD | Usual care | PROUD | Usual care | PROUD | Usual care | |
| Patients, No.b | 30 708 | 52 486 | 20 336 | 41 296 | 21 265 | 21 339 | 30 138 | 20 881 | 29 572 | 27 789 | 39 491 | 42 042 |
| Primary outcome: patient-years of OUD treatment, No.c | ||||||||||||
| Prerandomization | 13.3 | 3.9 | 5.3 | 11.2 | 8.3 | 20.3 | 1.1 | 13.6 | 0.4 | 0 | 12.4 | 14.3 |
| Postrandomization | 68.1 | 13.1 | 22.1 | 17.7 | 33.7 | 56.3 | 6.5 | 23.5 | 9.8 | 0 | 43.4 | 26 |
| Difference (post vs pre)d | 54.8 | 9.2 | 16.7 | 6.5 | 25.5 | 36.0 | 5.5 | 9.8 | 9.5 | 0 | 31.0 | 11.7 |
| PROUD vs usual care difference (95% CI)e,f | 45.6 (31.8 to 58.5) | 10.2 (−1.4 to 20.3) | −10.5 (−28.4 to 9.0) | −4.4 (−12.4 to 4.4) | 9.5 (1.2 to 14.0) | 19.3 (8.5 to 30.7) | ||||||
| Secondary outcomes of patient-years of OUD treatment | ||||||||||||
| Ongoing treatmentg | ||||||||||||
| Prerandomization | 9.2 | 1.1 | 4.7 | 9.3 | 4.9 | 8.3 | 0.3 | 10.1 | 0 | 0 | 8.3 | 8.2 |
| Postrandomization | 18.0 | 7.5 | 4.1 | 11.9 | 11.3 | 17.1 | 0 | 17.4 | 0 | 0 | 9.7 | 10.4 |
| Difference (post vs pre)d | 8.9 | 6.3 | −0.6 | 2.6 | 6.4 | 8.9 | −0.3 | 7.3 | 0 | 0 | 1.5 | 2.2 |
| PROUD vs usual care difference (95% CI)e,h | 2.5 (−5.5 to 10.6) | −3.2 (−8.9 to 2.5) | −2.4 (−13.6 to 8.8) | −7.6 (−15.8 to 0.7) | NA | −0.7 (−8.5 to 7.1) | ||||||
| Newly initiatede,i | ||||||||||||
| Prerandomization | 4.2 | 2.8 | 0.6 | 2.0 | 3.4 | 12.0 | 0.7 | 3.5 | 0.4 | 0 | 4.1 | 6.0 |
| Postrandomization | 50.1 | 5.7 | 18.0 | 5.8 | 22.4 | 39.2 | 6.5 | 6.1 | 9.8 | 0 | 33.7 | 15.6 |
| Difference (post vs pre)d | 45.9 | 2.9 | 17.3 | 3.9 | 19.0 | 27.1 | 5.8 | 2.6 | 9.4 | 0 | 29.6 | 9.6 |
| PROUD vs usual care difference (95% CI)e,f | 43.0 (28 to 55.8) | 13.4 (1.5 to 21.8) | −8.1 (−25 to 8.8) | 3.2 (−2.4 to 9.3) | 9.4 (1.1 to 14.0) | 20.0 (9.8 to 29.6) | ||||||
| Restricted to when nurse was seeing patients | ||||||||||||
| Prerandomization | 13.3 | 3.9 | 5.3 | 11.2 | 8.3 | 20.3 | 1.1 | 13.6 | 0.4 | 0 | 12.4 | 14.3 |
| Postrandomization | 73.9 | 13.5 | 16.9 | 12.6 | 32.0 | 53.1 | 6.2 | 16.5 | 8.0 | 0.0 | 44.6 | 25.3 |
| Difference (post vs pre)d | 60.6 | 9.6 | 11.6 | 1.3 | 23.7 | 32.8 | 5.2 | 2.9 | 7.6 | 0.0 | 32.2 | 11.1 |
| PROUD vs usual care difference (95% CI)e,f | 51.0 (35.9 to 65.2) | 10.2 (0.7 to 18.8) | −9.1 (−26.6 to 11.3) | 2.3 (−5.3 to 9.1) | 7.6 (1.0 to 11.3) | 21.1 (9.5 to 32.8) | ||||||
| Newly initiated and restricted to when nurse was seeing patientsi | ||||||||||||
| Prerandomization | 4.2 | 2.8 | 0.6 | 2.0 | 3.4 | 12.0 | 0.7 | 3.5 | 0.4 | 0 | 4.1 | 6.0 |
| Postrandomization | 56.3 | 6.3 | 14.4 | 4.1 | 23.2 | 40.5 | 6.2 | 5.7 | 8.0 | 0 | 35.7 | 16.1 |
| Difference (post vs pre)d | 52.2 | 3.5 | 13.8 | 2.2 | 19.8 | 28.5 | 5.5 | 2.2 | 7.6 | 0.0 | 31.6 | 10.0 |
| PROUD vs usual care difference (95% CI)e,f | 48.7 (31.9 to 62.9) | 11.6 (1.4 to 18.7) | −8.7 (−25.4 to 8.3) | 3.3 (−2.8 to 9.8) | 7.6 (1.0 to 11.3) | 21.6 (10.9 to 31.8) | ||||||
Abbreviations: NA, not applicable; OUD, opioid use disorder; PROUD, Primary Care OUD Treatment trial.
Results presented per 10 000 primary care patients, which was calculated by dividing by the number of patients seen in the clinic postrandomization and multiplied by 10 000.
Number of eligible primary care patients assigned to the clinics in 3 years prior to randomization and up to 2 years postrandomization.
Treatment was defined as having a medication order or procedure code for buprenorphine formulations that are indicated for OUD or having a medication order or procedure code for extended-release injectable naltrexone.
Calculated by subtracting prerandomization from postrandomization patient-years of OUD treatment for each clinic.
Calculated by subtracting the usual care difference (in prerandomization and postrandomization patient-years of OUD treatment) from the intervention clinic difference for each health system.
95% CI is based on bootstrap t method, described in the statistical analysis plan in Supplement 1.
Ongoing treatment was defined as OUD treatment for patients with less than a 365-day gap in OUD treatment.
95% CI is based on a normal approximation rather than bootstrap t method due to rare outcome events.
Newly initiated was defined as OUD treatment for those with no OUD treatment in the past 365 days.