. 2011 Mar 8;6(3):e17604. doi: 10.1371/journal.pone.0017604

Table 5. Risk factors of Streptococcus suis infection on univariate analysis.

Exposure	Cases versus Hospital controls				Cases versus Community controls
	OR⁽¹⁾ (95%CI)	p value	OR⁽²⁾ (95%CI)	p value	OR⁽¹⁾ (95%CI)	p value	OR⁽³⁾ (95%CI)	p value
Occupations related to pigs	9.68(4.13–22.67)	<0.001	7.51(2.85–19.82)	<0.001	11.50(4.31–30.65)	<0.001	11.01(4.03–30.12)	<0.001
Medical history
Diabetes mellitus	3.06(0.61–15.41)	0.175	0.82(0.13–5.23)	0.830	2.25(0.50–10.05)	0.288	3.75(0.75–18.73)	0.107
Alcoholism	2.55(1.22–5.33)	0.013	1.31(0.54–3.16)	0.547	2.50(1.15–5.45)	0.021	1.48(0.63–3.31)	0.381
Skin injuries	7.68(4.08–14.46)	<0.001	8.16(3.72–17.92)	<0.001	22.09(7.79–62.64)	<0.001	22.30(7.55–65.84)	<0.001
Breeding pigs at home	2.41(1.34–4.35)	0.003	2.34(1.09–5.00)	0.028	1.95(1.04–3.65)	0.036	1.99(1.04–3.80)	0.036
Any exposure to pigs/pork in the last 2 weeks	5.66(3.38–9.49)	<0.001	4.69(2.43–9.07)	<0.001	4.51(2.55–7.97)	<0.001	4.16(2.30–7.52)	<0.001
With skin injuries	14.72(5.36–40.42)	<0.001	12.16(3.74–39.50)	<0.001	30(7.01–128.35)	<0.001	26.95(6.14–118.23)	<0.001
Without skin injuries	2.74(1.55–4.86)	0.001	2.06(0.99–4.27)	0.052	1.66(0.92–3.00)	0.090	1.57(0.85–2.91)	0.152
Eating any “high risk” dish in the last 2 weeks	3.25(2.02–5.24)	<0.001	2.48(1.35–4.52)	0.003	6.00(3.33–10.81)	<0.001	4.38(2.72–8.08)	<0.001
Ill pigs at home in the last 4 weeks ⁽⁴⁾	24.62(2.85–212.24)	0.004	30.10(2.72–333.64)	0.006	-	-	-	-
Pigs at home with *S. suis* serotype 2 (confirmed by PCR) ⁽⁵⁾	2.31(0.49–10.82)	0.289	7.83(0.68–90.19)	0.099	-	-	-	-

⁽¹⁾

Crude OR based on logistic (hospital controls) or conditional logistic regression (community controls).

⁽²⁾

Adjusted for age, sex and rural/urban residence, using logistic regression.

⁽³⁾

Adjusted for sex (matched for age and residence), using conditional logistic regression.

⁽⁴⁾

Only individuals with pigs at home were analyzed. OR could not be analyzed for community controls because none of them reported ill pigs at home.

⁽⁵⁾

Only individuals who had pig swab samples at their houses were analyzed. OR could not be analyzed for community controls because there was no discordant pairs included in the analysis.