Table 2. Multivariable Cox Proportional Hazards Regression Models for All-Cause Mortality Using Intent-to-Treat and As-Treated Approaches to MOUD Exposure for IDU-IE–Associated Discharges From July 2011 to June 2015a.
| Characteristic | Hazard Ratio (95% CI) | ||
|---|---|---|---|
| Model 1: MOUD within 3 mo after IDU-IE | Model 2: monthly MOUD receipt after IDU-IE | Model 3: monthly after IDU-IE, including month of discontinuation | |
| MOUD | 1.29 (0.61-2.72) | 0.30 (0.10-0.89) | 0.39 (0.15-1.00) |
| MOUD 3 mo before admission | 0.77 (0.34-1.75) | 1.35 (0.65-2.81) | 1.33 (0.62-2.82) |
| Age, y | |||
| 18-34 | 0.24 (0.11-0.51) | 0.26 (0.12-0.55) | 0.26 (0.12-0.55) |
| 35-49 | 0.60 (0.34-1.06) | 0.62 (0.35-1.10) | 0.62 (0.35-1.1) |
| 50-64 | 1 [Reference] | 1 [Reference] | 1 [Reference] |
| Female | 0.80 (0.45-1.42) | 0.80 (0.45-1.43) | 0.80 (0.42-1.58) |
| Elixhauser Comorbidity Index | 1.03 (0.97-1.09) | 1.03 (0.97-1.09) | 1.01 (0.96-1.08) |
| Mental illness | 0.77 (0.39-1.49) | 0.81 (0.42-1.57) | 0.81 (0.42-1.58) |
| Homelessness | 0.60 (0.32-1.6) | 0.60 (0.31-1.14) | 0.61 (0.32-1.2) |
Abbreviations: IDU-IE, injection drug use–associated infective endocarditis; MOUD, medication for opioid use disorder.
a
Homelessness, major mental illness, and Elixhauser Comorbidity Index covariates were defined as evidence during any part of the study period not only before the index IDU-IE–related hospitalization. There were 61 deaths in each model.