Work and health-related determinants of midwives' intention to leave: Nationwide cross-sectional and prospective cohort studies in sweden.

Background: Midwives are essential for achieving the Sustainable Development Goal on sexual and reproductive health. However, midwifery shortage challenge healthcare organisations globally. The aim was to analyse cross-sectional and prospective associations between independent variables: 'possibilities of development'; 'quality of work'; 'role conflict'; 'recognition'; 'self-rated health' and 'burnout' and dependent variables: 'intention to leave current position' and 'intention to quit midwifery'.

Methods: In 2020 a nation-wide web-survey was distributed to midwives in Sweden (n = 5076) listed as members in unions that organise midwives. This was repeated in 2023 (n = 4398 midwives). Cross-sectional (n = 3696) and prospective (n = 531/406) uni- and multivariable logistic regressions were performed.

Results: We found a lower probability of intention to leave current position in respondents who reported possibilities for development, quality of work and recognition while those reporting role conflicts and burnout had a higher probability of intention to leave current position. In the first prospective analysis (n = 531) we found that reporting good quality of work and recognition at baseline were associated with a lower probability of reporting intention to leave current position at follow-up while burnout at baseline was associated with a higher probability of reporting intention to leave current position at follow-up. In the second prospective analysis (n = 406) we excluded respondents who reported intention to leave current position at baseline and found that only quality of work at baseline predicted a low probability of intention to leave at follow-up. For the second dependent variable, intention to quit midwifery, we found that possibilities for development and recognition was associated with a lower probability to quit while burnout was associated with a higher probability. In the first prospective analysis, we found that recognition at baseline was associated with a lower probability to quit while burnout at baseline was associated with a higher probability. When excluding those who reported intention to quit midwifery at baseline, we found that recognition and self-rated health was associated with a lower probability to report intention to quit at follow-up.

Conclusion: Quality of work, recognition and health appeared to be important for retention among midwives. Recognition stands out as a humble but relevant wish from a professional group of great importance for health care.