The multiple question survey was distributed to all residents at Oregon Health & Science University (OHSU) in December 2002  and December 2004, before and after the implementation of the ACGME duty hour requirements, respectively. Specific questions compared included responses to levels of career satisfaction (5-point Likert scale), 31 questions (Y/N) about positive (14) and negative (17) aspects of training, and 15 responses to questions about positive or negative emotional states (5-point Likert scale). Both surveys took place at mid-year, a time that typically represents the peak period of negative feelings and attitudes among residents about their career choices. The 2004 survey included duty hours reporting averaged over four weeks but did not include the other ACGME requirements. Duty hours averaged over a four-week period was chosen because it reflects most closely total working hours and it is the first question in the ACGME residents' survey.
The sets of survey data for the two cohorts were explored by using descriptive statistics. They were compared for pre and post intervention analysis. For further evaluation of the effect of the work hours' requirements, the subgroups of programs where there were reported violations were compared. For the pairwise comparisons between the two groups, the Fisher's exact test was used for both 2 × 2 and 2 × 5 contingency tables.