Continuous variables are provided as mean and standard deviation (SD), and categorical variables as absolute and relative frequencies.
All analyses were adjusted for age and gender. Linear regression models were employed for continuous variables (e.g., number of children), multinomial logistic regressions for multi-categorical variables (e.g., current position), and binary logistic regressions for binary variables (e.g., part-time). Independent binary logistic regressions were used for the association of socioeconomic aspects and work conditions with job expectations, their fulfillment, exhaustion, and satisfaction with support systems. Likert scale responses were dichotomized into, e.g., very important/important vs. not so important/not important, or always/mainly fulfilled vs. hardly/not fulfilled. To increase group size and thus statistical reliability, for comparisons between participants at different career levels, the categories senior physicians and leading senior physicians from the questionnaire were combined. For employed and self-employed radiologists in ambulatory care, separate and combined analyses were conducted.
Concerning missing variables, participants who did not indicate their gender (n = 6) were excluded from the analyses as were participants of diverse gender due to small numbers (n = 2). Radiologists working in other countries than Germany were few (n = 84 vs. n = 510), originated from 33 different countries, and were thus excluded to ensure validity and comparability of the analyses.
Because of the explorative study design, p-values were not adjusted for multiplicity and should be considered as descriptive summary measures. All calculations were performed in SAS 9.4 (SAS Institute).
All analyses were adjusted for age and gender. Linear regression models were employed for continuous variables (e.g., number of children), multinomial logistic regressions for multi-categorical variables (e.g., current position), and binary logistic regressions for binary variables (e.g., part-time). Independent binary logistic regressions were used for the association of socioeconomic aspects and work conditions with job expectations, their fulfillment, exhaustion, and satisfaction with support systems. Likert scale responses were dichotomized into, e.g., very important/important vs. not so important/not important, or always/mainly fulfilled vs. hardly/not fulfilled. To increase group size and thus statistical reliability, for comparisons between participants at different career levels, the categories senior physicians and leading senior physicians from the questionnaire were combined. For employed and self-employed radiologists in ambulatory care, separate and combined analyses were conducted.
Concerning missing variables, participants who did not indicate their gender (n = 6) were excluded from the analyses as were participants of diverse gender due to small numbers (n = 2). Radiologists working in other countries than Germany were few (n = 84 vs. n = 510), originated from 33 different countries, and were thus excluded to ensure validity and comparability of the analyses.
Because of the explorative study design, p-values were not adjusted for multiplicity and should be considered as descriptive summary measures. All calculations were performed in SAS 9.4 (SAS Institute).
Free full text: Click here
