A systematic review of health manpower forecasting models.

Martins-Coelho, G., Greuningen, M. van, Barros, H., Batenburg, R. A systematic review of health manpower forecasting models.: , 2011. 108 p. Abstract. In: Abstract Book. EHMA Annual Conference 2011: 'Integration in Health and Healthcare', Porto, 22-24 juni 2011.
Read online
Context: Health manpower planning (HMP) aims at matching health manpower (HM) supply to the population’s health requirements. To achieve this, HMP needs information on future HM supply and requirement (S&R). This is estimated by several different forecasting models (FMs). In this paper, we review FMs for physician manpower planning (PMP) and their variation. Methods: We performed a systematic literature review (SLR) of scientific papers published worldwide after 1970, describing FMs for PMP. Studies were included if: they described at least one FM; the FM was used for macro level PMP; the FM forecast S&R; and the FM was dynamic. From each study, we extracted information on country, year and type of physicians studied, data sources searched, and parameters used to estimate current and future S&R. From the data gathered, we analysed the variation between FMs. Results: From the 994 papers identified, 35 were included in the review. They described 69 FMs, covering mostly specialists practicing in North America. Publication trends suggest the topic gained interest in the late 1970s and early 1980s, and again in the late 1990s. FMs overlapped regarding the parameters used to estimate current and future S&R. All but two FMs used workforce size to measure current supply. Future supply was estimated mostly from expected inflow of new graduates and outflow of retiring physicians. Current requirement was most frequently estimated by population size and age, age- and diseasespecific utilisation, care standards or duration of care. Changes in population size and age were the most used to estimate future requirement. Discussion: Our SLR provides the first knowledge base of international experience in physician S&R forecasting. This allows the development of a classification framework of FMs for PMP. Such a framework, unlike the traditional need/demand division, can be derived from the variation found by this SLR. Planners can benefit from previous experience, e.g. regarding data collection requirements, assumptions made, and outputs of FMs. Our knowledge base is also useful internationally to inform HMP and improve health labour market efficiency.