Table 2. Results of multilevel Poisson regression analyses of risk for nontuberculous mycobacterial infection, Missouri, USA, 2008–2019.
| Variable | Rate ratio (95% CI) |
|---|---|
| Fixed effects | |
| County-level flooding | |
| No flooding | Referent |
| 1–3 times | 1.19 (1.11–1.29) |
| 4–5 times | 1.29 (1.17–1.43) |
| >5 times | 1.38 (1.26–1.52) |
| County rural–urban context | |
| Rural | 2.82 (1.90–4.19) |
| Urban | 2.08 (1.53–2.82) |
| Metro | Referent |
| County-level poverty levels | |
| 1st quartile (lowest) | Referent |
| 2nd quartile | 0.88 (0.62–1.25) |
| 3rd quartile | 0.71 (0.49–1.03) |
| 4th quartile (highest) | 0.78 (0.54–1.13) |
| County-level ratio non-Hispanic Black | |
| 1st quartile (lowest) | Referent |
| 2nd quartile | 1.17 (0.81–1.69) |
| 3rd quartile | 1.09 (0.76–1.56) |
| 4th quartile (highest) | 0.84 (0.58–1.21) |
| Patient age group, y | |
| <20 | Referent |
| 20–49 | 7.23 (5.11–10.2) |
| 50–64 | 26.7 (18.9–37.7) |
| >65 | 76.8 (54.6–108.2) |
| Sex | |
| M | Referent |
| F | 0.94 (0.89–0.99) |
| Race/ethnicity | |
| Non-Hispanic White | Referent |
| Non-Hispanic Black | 2.62 (2.31–2.98) |
| Others* |
32.2 (30.3–34.2) |
| Random effect | |
| Variance 0.32, p<0.001 | 1.73 |
*Asian (91; 1.72%), native American (9; 0.17%), unknown race (1,976; 37%).