Multivariate epidemiological approach to coccidiosis in broilers.

A retrospective, case-control study into risk factors of coccidiosis was undertaken using data from 189 broiler flocks. A case flock was defined as a flock in which at least one bird had intestinal lesions on 1 of 6 wk in a 42-day cycle. Flocks wherein such birds could not be detected were defined as controls. There were 187 variables, measured or derived. These were assigned to subsets of data, each subset being a group of variables representing related information. Uni- and bivariate analyses were performed in each subset. Variables and interactions that were significant in these analyses were entered into a multivariate model across subsets. In the final model, seven variables appeared to be significantly associated with detecting lesions in birds of a flock. Differences among breeds covered a range of about an 80-fold change in risk of being a case. At intermittent lighting, the risk of being a case increased about sevenfold compared with continuous lighting. A higher initial (Week 1) environmental temperature decreased the risk of finding lesions in a flock (about .8-fold per degree Celsius). This risk was also lower at a lower average aerial ammonia content (below versus above 14 ppm) and higher maximum carbon dioxide content (above versus below .4 vol%) changing the risk about .3- and .4-fold, respectively. The risk of being among cases increased with more litter (about twofold per kilogram of litter per square meter). Flocks in houses of 600 to 800 m2 were about 9.8 times more at risk of being scored as lesion-positive than those in smaller houses.(ABSTRACT TRUNCATED AT 250 WORDS)