Table 2.
Final multilevel logistic regression model of factors associated with seropositivity to Rift Valley fever virus in cattle, sheep and goats in central South Africa, 2015–2016, adjusted for clustering within farms.
| Variable and level | OR | 95% CI | P-value |
|---|---|---|---|
| Species | |||
| Cattle | 1* | ||
| Sheep | 0.48 | 0.32–0.72 | <0.001 |
| Goats | 0.30 | 0.16–0.56 | <0.001 |
| Age (years) | |||
| <2 | 1* | ||
| 2–4 | 2.82 | 1.79–4.44 | <0.001 |
| >4 | 17.08 | 11.29–25.85 | <0.001 |
| Type of farm | |||
| Communal | 1* | ||
| Private | 4.78 | 2.02–11.29 | <0.001 |
| Access to perennial river | |||
| No | 1* | ||
| Yes | 1.53 | 1.02–2.31 | 0.042 |
| Access to seasonal pan | |||
| No | 1* | ||
| Yes | 1.75 | 1.17–2.61 | 0.006 |
| Year of last RVF vaccination on farm | |||
| Never | 1* | ||
| 2009–2011 | 0.96 | 0.56–1.65 | 0.893 |
| 2012–2013 | 0.90 | 0.45–1.78 | 0.760 |
| 2014–2016 | 1.86 | 1.03–3.36 | 0.040 |
| Unknown | 2.71 | 0.26–28.89 | 0.408 |
Variance of random effect for farm nested within district = 1.48 (95% CI: 1.06–2.07; P < 0.001).
Model log-likelihood = −1,114.935; AIC = 2,255.87; n = 2,971.
*Reference category.
OR = odds ratio, CI = confidence interval, AIC = Akaike’s information criterion.