Table 5 presents the CCCs, the bias and the standard deviation (SD) for each quantitative answer provided by residents and the expert as a reference.
To avoid any heterogeneity in the answers, qualitative variables were defined if necessary, and units of each quantitative variable were given in the questionnaire.
The assumption of linearity in the logit scale (log-odds) between each quantitative covariate and the outcome in binary logistic regression analysis was assessed by examination of smoothed scatter plots.
Each quantitative real-time PCR assay was performed in triplicate.
DNA standards were run in triplicate on each quantitative plate.
Each quantitative experiment was performed at least in triplicate at least three independent times.
Normality of each quantitative measurement was confirmed prior to analysis and those not conforming to normality underwent data transformations.
Each quantitative real-time RT-PCR was performed using the Chromo4 real time detection system (BIO-RAD).
In each centre, each quantitative variable was transformed into a z-score (equal to (observed value minus the sample mean), divided by the sample standard deviation).
The top associations consistently showed the same direction for each quantitative trait tested, in agreement with our observations of multivariate linkage analysis [24].
To ensure elimination of carry-over contamination, we relied on the UQPCR method replacing dTTP by dUTP and using uracil-N-glycosylase (UNG) treatment prior to each quantitative real-time PCR amplification [12].
