The prevalence of variables was assessed by the chi square test or Yates' test, as appropriate.
For some of the variables e.g. teat skin quality and udder consistency observers did see similar patterns in prevalence of variables, although there were differences in values.
For sample size calculation, the objective we have considered is to estimate the prevalence of variables with proportions as little as 0.02 (P).
As a preliminary assessment of the prevalence of variables reporting the number of events, we considered all randomized, controlled trials (RCTs) published in January and February 2010 in the New England Journal of Medicine, Lancet, Journal of the American Medical Association, and BMJ.
To assess likelihood of reporting bias we compared the prevalence of variables reported in Tables 1 and 2, among the cases only, according to informant's gender, age (< 50 vs. 50+ years) and relationship to deceased (1st degree relative vs. other type).
There are several reasons why adding a new predictive variable may not result in a better accuracy of a predictive model including a low prevalence of the variable, the absence of predictive value of this variable, and the presence of a correlation between the predictive variable and the variable(s) originally included in the predictive model (i.e. APACHE-II).
After calculating the prevalence of each variable for clients with or without recidivism, respectively (Table 1), all variables included were entered into a Cox regression survival analysis (Table 2).
Figure 1 shows, by mailing wave, the absolute difference in the log OR of all variables relative to an analysis using all controls, plotted against the unadjusted prevalence of each variable.
The prevalence of each variable was calculated from each source.
Nevertheless, this adjustment may still have yielded an overestimate of the prevalence of this variable.
