Scrotal versus prescrotal castration: Data assessment

Data were entered into a spreadsheet (Excel) and analyzed for surgical efficiency and the presence or absence of complications by using statistical software (SAS for Windows-SAS Institute). Multivariable logistic regression, using (PROC LOGISTIC; SAS for Windows), was used to measure the strength of association between the occurrence of each of the complications and the explanatory variables of method, state, and weight of dog. Initially, for each outcome, the model included method, state, and weight as well as interaction terms of method*state and method*weight. If the interaction terms were found to be insignificant, they were dropped from the model. The main effects were retained in the model regardless of significance because odds ratios adjusted for all main effects were desired.

Differences in mean length of surgery by method were assessed by analysis of variance by using (PROC GLM; SAS for Windows). A P value < 0.05 was considered statistically significant for analysis.