This BSIF was used in all analyses except for analyses based on an IDIF.
Analyses were performed using Statistical Package for Social Science SPSSS, Chicago, Ill., USA) version 12.0 and RGui version 2.7.0 except for analyses of the activity difference between mutant and wild type constructs which was calculated using a t-test for paired samples.
All efficacy analyses, except for analyses of duration of response (which was assessed only in responders), were performed on the ITT population.
* Adjusted for age, sex, smoking status, alcohol use, education level, history of cardiovascular disease or stroke and CRP (except for analyses on the High CRP).
Participants with missing data (eg, sex, self-reported alcohol consumption or participant estimation of result) were included except for analyses involving the missing component itself.
In all analyses, the global P values were obtained from models using the generalized estimation equation (GEE) to account for familial correlation, except for analyses for which sample sizes were too small to permit the GEE.
All analyses were conducted using SAS, version 9.3 (SAS Institute Inc ., except for analyses incorporating penalized splines for the continuous lead term, which were completed using R, version 3.0.1 (R Core Team 2013).
Observations in which respondents selected a "prefer not to answer" option were excluded from analyses except for analyses of income and sharing where missing income data were substantial and responses separately categorized.
All statistical analyses, except for analyses of missing data and sensitivity, were restricted to complete cases (i.e. GPs who completed the questionnaire and who were also involved in the diagnostic process).
Since several measures were not normally distributed, we used the Mann-Whitney U-test for comparisons of ratings for stress and burnout, stratifying by sex (except for analyses of stage of education).
