The present study is part of an ongoing prospective survey of a cohort of graduates of the three medical schools in German-speaking Switzerland, beginning in 2001 (T1). Of the 1,004 registered final-year students, 719 (72%) participated in the first assessment (T1) . Subjects were re-evaluated after two years in 2003 (T2) [20,21]. The present paper refers to results of the third assessment (T3), conducted in the fourth year of residency (in 2005). The postal questionnaire consisted of multiple-choice and free-response items. The free-response items were embedded in the multiple-choice questions addressing the issues of what arguments still speak in favour of or against a career in medicine. A total of 577 subjects (306 females, 53%; 271 males, 47%) participated in the third assessment (T3).
The study was approved by the research and ethical committee of the Zurich University. The deans of the involved three medical faculties Basel, Bern and Zurich supported the survey by a letter of recommendation which was enclosed to the first inquiry. As the survey addressed healthy people on a voluntarily and anonymous basis without a planned intervention no further special ethical considerations had to be followed. To ensure participants' anonymity, the returned questionnaires were only identified by a code. The respondents sent their addresses to an independent address-administration office, allowing for follow-up.
The free-response items of the questionnaire read as follows: "In your opinion, what arguments are there still in favour of or against a career in medicine? ("Welche Argumente sprechen Ihrer Meinung nach heute noch für bzw. gegen den Arztberuf?") Please mention three arguments for and three against."
In addition, participants answered the question of whether or not they would again choose to go into medicine ("yes" or "no") ("Würden Sie sich heute nochmals für den Arztberuf entscheiden?").
Not all of the (n = 577) participants of the third assessment (T3) were included in the qualitative analysis. Ten of the respondents did not answer this question. The study sample therefore consisted of 567 residents (n = 303 females, 53.4%; n = 264 males, 46.6%). The mean age was 31.3 years (SD 2.4 y, range 27–46 y).
Clinical fields and distribution of the residents
Residents were grouped according to the speciality qualification they aspired to. "Internal Medicine fields" comprised all sub-specialities of Internal Medicine and fields related to Internal Medicine, as well as Primary Care; Surgery, Obstetrics & Gynaecology, Urology, Orthopaedics, Ophthalmology and ENT were categorized as "Surgical fields"; "High-technology medicine" comprised Anaesthesiology, Emergency Medicine, Intensive Care, Nuclear Medicine, and Radiology. Further speciality groups were identified: "Paediatrics", "Psychiatry", "Other Specialities" such as Preventive Medicine, Pathology, and "Speciality not yet decided". Distribution of the aspired-to specialities of the 567 residents (100%) was as follows: Internal Medicine, n = 197 (34.7%), Surgical fields, n = 106 (18.7%), High-technology medicine, n = 51 (9.0%), Paediatrics, n = 48 (8.5%), Psychiatry, n = 27 (4.8%), Other specialities, n = 53 (9.3%), and Not yet decided, n = 85 (15.0%).
Content and statistical analysis
The qualitatively assessed data (reasons for and against entering the medical profession) were evaluated according to Mayring's content analysis  by a researcher not involved in the quantitative data analysis, as follows: First, the respondents' handwritten answers (headwords or whole sentences) were transcribed into an Excel file. In a second step, content categories were inductively formulated, and their descriptions written down in a code manual (definition, examples, and rules for coding). In a further step, the passages of text were encoded according to the code manual and assigned to the content categories. Frequency distributions were given for categories and tested with Chi2-tests for differences in gender, speciality fields, and whether or not the respondent would again choose a career in medicine. Inter-rater reliability: A random sample of 20% of the analyzed statements was submitted to two additional raters (staff from the department experienced in qualitative analyses). The index of concordance (ratio of identically rated answers to all rated answers) was 0.79 and Cohen's Kappa 0.76.