5. Data analysis

Descriptive analysis using IBM SPSS ver. 23.0 (IBM Corp., Armonk, USA) was applied to the data of gender, type of residency program, marital status, phase of study, the number of working hour/night shifts, and the number of death and difficult cases. Burnout score and WHOQOL-BREF scores were also calculated. The burnout scores were distributed into the three subscales (domains) within the MBI-HSS, which were EE, DP, and PA. Each item in the subscale was measured using a 7-point Likert scale and the scores of the subscales were grouped based on the cut-off score for each category of low, moderate, and severe/high according to Maslach et al. [7]. A multivariate analysis using linear regression test (for QoL variable) and logistic regression test (for burnout variable) was also carried out to examine the relationship between the demographic variable with burnout and QoL.

Verbatim transcriptions were made based on the interview recordings. A thematic analysis was conducted manually by all authors. The first author read the first two transcripts and identified relevant themes and subthemes. A coding system was developed and agreed by all authors. Subsequent analysis of the remaining transcripts followed the coding system and any new themes and subthemes were added. Data display and conclusions drawing and verification, including member checking, followed the process [13].