Table 2.
Logistic regression analysis of factors associated with the carriage of multiple-genotype infections (MIs)
| Variable | Odds ratio | Likelihood ratio χ2 | P-value |
|---|---|---|---|
| Sampling year | 1.207 [1.153–1.264] | 67.822 | 1.79 × 10-16 *** |
| Parasite density | 1.257 [1.086–1.454] | 10.656 | 0.001 *** |
| Age | 0.997 [0.989–1.006] | 3.286 | 0.070 |
| Sex: Male | 1.040 [0.835–1.295] | 0.122 | 0.727 |
| Rescue treatment | 0.964 [0.732–1.269] | 0.070 | 0.791 |
| Blood transfusion | 0.834 [0.574–1.211] | 0.904 | 0.341 |
| Location | Compared with Mawker–Thai | 15.832 | 0.001 *** |
| Maela Camp | 0.612 [0.462–0.811] | — | — |
| Wang Pha | 0.672 [0.496–0.909] | — | — |
| Mae Kon Ken | 0.963 [0.596–1.555] | — | — |
| Season | Compared with Jan–Mar | 1.114 | 0.774 |
| April–June | 0.910 [0.671–1.234] | — | — |
| July–September | 1.061 [0.747–1.505] | — | — |
| October–December | 0.950 [0.686–1.317] | — | — |
Figures in square brackets represent 95% CIs. We used the likelihood ratio test to determine whether the exclusion of a particular variable significantly reduces the goodness of fit of the regression model. The likelihood ratio χ2 and its associated P-value are derived from comparing the full model and the constrained model without the variable in question. Significant values are indicated by asterisks (***). Carriage of MIs is associated with lower admission parasite densities and decreases significantly over time.