. 2009 Feb 3;89(1):8–15. doi: 10.1016/j.prevetmed.2009.01.001

Table 3.

Final logistic-regression models in a clinical trial of the effect of single cow calving pens (treatment) vs. multiple cow calving pens (control) on the risk of morbidity, diarrhea, and pneumonia in Holstein heifer calves born on three Minnesota dairy farms in 2005.

Effect of intervention	Level	Odds ratio	95% Confidence interval of odds ratio		P value
			Lower	Upper
Diarrhea^a
Single cow calving pen	Yes	0.93	0.57	1.47	0.75
	No	Baseline	–	–	–

Birth season	Winter	0.59	0.37	0.94	0.03
	Summer	Baseline	–	–	–

Herd	1	Baseline	–	–	–
	2	45.5	16.1	128	<0.001
	3	17.6	5.7	54.4	<0.001

Pneumonia^b
Single cow calving pen	Yes	1.23	0.52	2.91	0.64
	No	Baseline	–	–	–

Lactation	1	0.29	0.09	0.91	0.03
	≥2	Baseline	–	–	–

Morbidity due to any cause^c
Single cow calving pen	Yes	0.93	0.63	1.40	0.74
	No	Baseline	–	–	–

Herd	1	Baseline	–	–	–
	2	4.18	2.61	6.67	<0.001
	3	1.82	0.94	3.51	0.07

Lactation, calving difficulty, and two-way interactions were not significant and were consequently excluded from this model. The overall significance of herd variables based on likelihood ratio test (LR test) was P < 0.0001. Birth of a calf in winter was significantly protective against diarrhea after adjusting for the treatment and herd effects, respectively.

Birth season, calving difficulty, and two-way interactions were not significant and were consequently excluded from this model. Note that records from two farms were excluded from this model because pneumonia outcomes were recorded in a single herd (Herd 1). In this herd (Herd 1) calves born to first lactation cows were significantly less likely to suffer an event of pneumonia after adjusting for the treatment and herd effects, respectively.

Lactation, calving difficulty, season of birth, and two-way interactions were not significant and were consequently excluded from this model. The overall significance of herd based on LR test was P < 0.0001.