Table 2.
Multivariable Logistic Regression Model Assessing Factors Associated With Receipt of Echinocandin Versus Fluconazole for Initial Treatment of Adult Candidemia Patients (N = 1801)—9 Surveillance Sites, United States, 2017–2018
| Characteristic | Adjusted Odds Ratio | 95% Confidence Interval | PValue |
|---|---|---|---|
| Age group, y | .4062 | ||
| 18–44 (referent) | 1.00 | … | |
| 45–64 | .92 | .69–1.21 | |
| ≥65 | .83 | .63–1.10 | |
| Cirrhosis | 2.06 | 1.29–3.29 | .0026 |
| Hospitalized within 90 d before diagnosis | .80 | .65–.98 | .0332 |
| Surveillance site a | <.0001 | ||
| California | 2.08 | 1.31–3.30 | |
| Colorado | 2.06 | 1.35–3.15 | |
| Georgia | 1.46 | 1.07–2.01 | |
| Maryland | 3.09 | 2.09–4.57 | |
| Minnesota | 2.11 | 1.42–3.14 | |
| New Mexico | .96 | .48–1.89 | |
| New York | 1.64 | .95–2.83 | |
| Oregon | 1.15 | .72–1.84 | |
| Tennessee (referent)b | 1.00 | … |
a
The Emerging Infections Program conducted surveillance in select counties in each of these states.
b
Tennessee was chosen as the referent group because patients from this site had the lowest odds of receiving an echinocandin versus fluconazole in unadjusted odds ratios.