Table 2.

Crude and adjusted odds ratios (OR) with their 95% confidence intervals (CI)

Category*	Crude OR	95% CI	Adjusted OR	95% CI	P (Wald's test)
Precipitation (mm)
[0,22]	Ref.		Ref.
(22,26]	0·57	0·37–0·88	0·72	0·44–1·18	0·19
(26,31]	0·45	0·28–0·73	0·33	0·19–0·57	<0·001
(31,37]	0·16	0·06–0·41	0·19	0·07–0·51	<0·001
Animal movements
0	Ref.		Ref.
[1,3]	1·52	0·96–2·42	1·57	0·97–2·53	0·06
⩾4	2·97	1·87–4·71	3·40	2·01–5·74	<0·001
Wind speed (m/s)
[0,1]	Ref.
(1,1·8]	3·53	1·38–11·56
(1·8,3·3]	19·28	7·77–61·81
Open landscape (%)
[0,0·25]	Ref.
(0·25,0·50]	1·56	0·70–3·85
(0·50,1]	6·45	3·19–14·77
Cattle density (cows/km2)
[0,7]	Ref.
(7,25]	3·08	1·8–5·45
(25,71]	15·11	8·60–27·41
Sheep density (sheep/km2)
[0,2]	Ref.
(2,5]	1·97	1·14–3·44
(5,16]	6·44	3·88–10·97
Temperature (°C)
[0,5]	Ref.
(5,8]	1·49	0·55–5·04
(8,10]	11·49	4·65–36·69
Cluster variable†
1	Ref.		Ref.
2	2·79	1·1–9·09	4·23	1·65–10·85	<0·001
3	27·97	11·15–90·44	42·07	16·30–108·5	<0·001

Adjusted ORs were estimated using the output of the logistic regression model. The P value calculated using Wald's tests corresponded to the adjusted OR. There are no adjusted results for the wind speed, the percentage of open landscape, animal densities (cattle and sheep) and temperature because they were not included in the multivariable model. All the 1537 dairy herds were included in the bivariate analysis, and 1443 dairy herds that had no missing values were included in the multivariable analysis. The study focused on risk factors for Coxiella burnetii infection in Swedish dairy cattle, sampled in 2008–2009.

^*Cut-off values used to categorize the continuous variables.

^†The cluster variable resulted from the aggregation of the following variables: wind speed, percent of open landscape, animal densities (cattle and sheep) and temperature (see Materials and Methods section for details).