Rural health workforce retention: strengthening the evidence base

Context/relevance: The shortage of health care professionals, managers and support workers remains a critical issue in the quest to provide good access to appropriate, sustainable health care in rural and remote areas. To date, surprisingly little use has been made of human resource data routinely collected by rural and remote health services to progress our understanding of patterns of workforce turnover and retention.

Objectives: This paper demonstrates a methodology and analytical framework showing how existing health service human resource records can be used to monitor health workforce turnover and retention, and thereby inform rural health workforce planning strategies.

Method: Individual level de‐identified employment data on health care workers from Australian regional, rural and remote health services were collected in two distinct studies, and separately analysed using simple arithmetic calculations of turnover and stability, Kaplan Meier survival analysis and regression modelling. The first study surveyed 100 primary health services from across Australia collecting length of stay data on all nurses, doctors, allied health professionals, Indigenous health workers and managers employed during the period 1/1/2003 to 31/7/2009. The second study surveyed 16 Victorian rural health services, collecting length of stay data on all allied health professionals employed between 1/1/2004 to 31/12/2009.

Results: Thirty three health services provided length of stay data on 1,479 health professionals in the national study, whilst eleven health services provided length of stay data on 901 allied health professionals in the Victorian study. A suite of five workforce measures (annual turnover rates, stability rates, median length of stay in current position, median survival and survival probabilities) collectively provided comprehensive turnover and retention metrics, whilst regression analysis identified significant predictors of turnover. Nationally, allied health professionals were 1.74 (95%CI 1.27, 2.40) times more likely to leave compared with nurses. Age upon commencement of employment was a significant factor affecting the likelihood of leaving in both studies.

Conclusions: The routine recording of employee commencement and separation dates in human resources records, together with basic demographic information and employment data required for payroll function(such as discipline and grade of employment), provides a valuable source of data for calculating both simple and complex measures of workforce turnover and retention in rural Australia. To maximise use of this resource, rural and remote health services, together with regional health authorities, must have sufficient human resource capacity to support the collection, maintenance and analysis of their human resource data as the foundation for ensuring a solid evidence‐base to underpin health workforce retention support and incentive strategies.