Table 4.
Mixed effects multivariable logistic regression model including covariates associated with the probability of isolating Arcobacter spp. from cattle faeces on Lancashire dairy farms. Farm is considered as a random effect (n = 15)
| Covariate | Estimate β | 95% CI | OR | 95% CI | Wald test P value |
|---|---|---|---|---|---|
| Proportion of fresh cows (quintiles) | 0·13 | 0·06 to 0·20 | 1·14 | 1·07–1·21 | <0·001 |
| Average herd daily milk yield (quintiles) | −0·17 | −0·30 to −0·50 | 0·84 | 0·74–0·95 | 0·006 |
| Campylobacter isolated | −1·67 | −1·86 to −1·48 | 0·19 | 0·16–0·23 | <0·001 |
| Faeces: loose vs. firm | 0·23 | 0·06 to 0·39 | 1·25 | 1·07–1·48 | 0·006 |
| Management system* | 0·81 | 0·45 to 1·16 | 2·24 | 1·57–3·20 | <0·001 |
| HAY vs. PSHW | |||||
| Time x1 | 0·51 | 0·37 to 0·64 | 1·66 | 1·44–1·91 | <0·001 |
| Time x2 | 0·13 | 0·003 to −0·26 | 1·14 | 1·00–1·23 | 0·045 |
| Time x3 | −0·21 | −0·34 to −0·08 | 0·81 | 0·71–0·92 | 0·001 |
| Time x4 | −0·14 | −0·28 to −0·01 | 0·87 | 0·76–0·98 | 0·028 |
| Management system × x1 | −0·72 | −1·00 to −0·44 | 0·49 | 0·37–0·64 | <0·001 |
| Management system × x2 | −0·38 | −0·64 to −0·13 | 0·68 | 0·53–0·88 | 0·003 |
| Management system × x3 | 0·16 | −0·10 to 0·43 | 1·18 | 0·91–1·53 | 0·217 |
| Management system × x4 | 0·06 | −0·21 to 0·34 | 1·07 | 0·81–1·40 | 0·649 |
| Baseline (PSHW, lowest level of all other covariates) | 0·002 | −0·51 to 0·52 | |||
OR, Odds ratio; CI, confidence interval.
Between-farm variance 0·033 (95% CI 0·013–0·080).
*
Management system: housed all year (HAY) vs. pasture in summer, housed in winter (PSHW).