135 Context analysis of interventions among lower socioeconomic position employees Data Collection Semi-structured interviews were conducted by telephone or videoconference between May and November 2020. A topic guide was used to examine the perceptions of stakeholders on preventive health interventions that consider multiple life domains and to explore related contextual factors. The following topics were discussed: (1) addressing problems on multiple life domains in occupational health practice (2) how problems on multiple life domains are dealt with and which stakeholders play a role in dealing with these problems; (3) the extent to which prevention is important in occupational health practice, (4) the implementation of preventive services in organizations; (5) collaboration between organizations, OHSs and OHPs in occupational health practice; (6) the organization of occupational healthcare in the Netherlands in relation to addressing problems on multiple life domains; and (7) the collaboration between occupational healthcare and curative healthcare. Within these topics, questions were based on contextual factors that could affect implementation, which were identified by Fleuren et al. (12). Furthermore, in case employees with a lower SEP were discussed within these topics, this group of employees were conceptualized as workers with blue-collar occupations or a lower educational level, who more often have health problems on multiple life domains and an increased risk to drop out of the labor market, as compared to workers with white-collar occupations or a higher educational level. For each stakeholder the interview topics were the same, but questions were stakeholder-specific to align the questions to the profession and background of the stakeholder. Interviews lasted around 30–60 min and were conducted in Dutch by the primary researcher (RS). Data Analysis The interviews were audiotaped and transcribed verbatim. The data was coded using Atlast.ti. Thematic analysis was used to analyze the data (22). The analysis started with re-reading the transcripts, listening to audiotapes, and making summaries of each transcript to become familiar with the data. Thereafter, open coding of the transcripts was performed using an inductive approach. During this process an initial list of codes was produced by the first coder (RS). Another coder (FS) read several transcripts and checked the codes. Next, the data was searched for similarities and discrepancies to combine and group codes. There were several meetings to discuss and categorize the codes into sub-themes (RS, FS, MH). This ultimately resulted into broader themes, which were depicted in code matrices. After 12 interviews, the themes were discussed with the whole research team (RS, FS, MH, CB, JA), wherein we came to the conclusion that we did not yet achieve data saturation. Some underlying factors influencing implementation were still unclear and it became clear that trade associations were an important stakeholder that were not yet interviewed. Therefore, four extra interviews, of which three with representatives of trade associations, were 5
RkJQdWJsaXNoZXIy MTk4NDMw