Hart LG; Association for Health Services Research. Meeting.
Abstr Book Assoc Health Serv Res Meet. 1999; 16: 96-7.
WWAMI Center for Health Workforce Studies, Family Medicine, University of Washington, Seattle 98195-4696, USA.
RESEARCH OBJECTIVE: Although national workforce analysis and methods have made significant progress during the last two decades, intrastate workforce analysis and methods can be characterized as being in their infancy. The development of the BHPr's Integrated Requirements Model (IRM) has been a significant step forward but much remains to be done to help identify those areas within states that are underserved by health care providers. The overall objective of this ongoing development project is to carefully build on the adjusted requirements model we have implemented in Washington State to develop tools that will better answer pressing state health workforce policy questions. An illustrative example of the use of the model's results will be presented that shows the relative contributions of generalist physician assistants (PAs) and nurse practitioners (NPs) to rural and urban generalist care in Washington. Improvements and enhancements are currently being made to the model. STUDY DESIGN: The current version of the Washington generalist model (small area FTE and requirements adjusted model) and data (1997 and 1998) were utilized in this study. The model is basically one in which Department of Health (DOH) licensing data and license renewal survey data are carefully applied to FTE adjust the supply of physicians, PAs, and NPs based on reported visits (and alternatively direct patient hours) and to impute data for non responders. The model includes FP/GPs, general internists, and general pediatricians as well as generalist PAs and NPs. A survey was conducted to validate data for places with fewer than 20 providers. Population requirements are based on age/sex specific expected generalist ambulatory visits. The supply and requirements data are aggregated by ZIP codes areas into the State's 125 generalist health service areas (HSAs) for the analyses. Model parameters (e.g., productivity estimates for survey non responders) are estimates from the license renewal survey data. PRINCIPAL FINDINGS: The model's results identified many HSAs where population requirements exceeded supply (i.e., 44% of the HSAs) and the results were used in combination with other HSA level data (e.g., inadequate prenatal care rates) to identify state health care provider training loan repayment placement areas. In the illustrative use of the model's results per the contribution of generalist PAs and NPs to Washington's generalist care, it was found that they provided over 30 percent of the generalist care in rural areas and slightly less in urban areas. FP/GPs provided half (50%) of the total rural generalist care (and a higher percentage in the more remote and sparsely populated rural HSAs) while there was a more even split between provider types in the urban areas. CONCLUSIONS: The Washington State generalist model produced adequate results at the small area level but was extremely sensitive to the data quality and timeless. Adjustments, such as the one to adjust generalist providers for their FTE production, were critical to the model's accuracy. For state-wide policy issues, the aggregation of the model's small area results provided excellent and useful information. However, the usefulness of enhancements to the model were also apparent. For instance, the model is currently being revised to produce estimates 15 years into the future (e.g., introducing population projections and retirement estimates) and versions of the model are being built especially for the production of estimates for prenatal care and for some types of specialty care (e.g., general surgery and cardiology). IMPLICATIONS FOR POLICY, DELIVERY, OR PRACTICE: This ongoing project is of both substantive policy and methodological importance. Substantively it provides improved and detailed small area comparative information on the generalist physician, PA, and NP workforce distribution, adjusted supply, and care requirements for Washington. Among other things, this is helping policy leaders better target program interventions and plan appropriate educational training output. On the methodological front, the project is enhancing a workforce requirements model that performs better for generalist analyses for substate areas. The generalist model is being adapted for use with the available data in Montana and it is planned that versions of the model will be implemented in Wyoming, Alaska, Idaho, and perhaps Oregon. The use of these generalist models coupled with improved data systems will enable state and federal policy makers, program adminstrators, and others to better understand the dynamics of intra state health workforce systems. This information provides input into the policy making arena that can make the results of decisions better targeted and more predictable and effective.
Publication Types:
Keywords:
- Alaska
- Efficiency
- Forecasting
- Health
- Health Services Needs and Demand
- Health Services Research
- Humans
- Idaho
- Montana
- Nurse Practitioners
- Oregon
- Physician Assistants
- Physicians
- Physicians, Family
- Primary Health Care
- Rural Health Services
- Urban Health Services
- Washington
- Wyoming
- manpower
- hsrmtgs
Other ID:
UI: 102194496
From Meeting Abstracts