SEARCH · PubMed Health
Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
A large health system is using a two-step forecasting approach to project future demand and capacity. Comparing its 10-year demand and capacity forecasts that were completed in 2001 with current data, the system was able to develop a facility master plan that incorporated projected capacity and physician requirements.
Effective forecasting of demand for healthcare services requires nine steps: 1. Assemble historical data. 2. Analyze historical trends. 3. Identify key demand drivers. 4. Identify relevant benchmarks. 5. Model existing conditions. 6. Develop core assumptions for population-based demand. 7. Develop core assumptions for provider-level demand. 8. Create a baseline forecast of future demand. 9. Test sensitivity of projections to changes in core assumptions.
RATIONALE: There is well-founded concern about the current and future availability of Health Human Resources (HHR). Demographic trends are magnifying this concern -- an ageing population will require more medical interventions at a time when the HHR workforce itself is ageing. The lengthy and costly training period for most health care workers, especially physicians, poses a real challenge that requires planning these activities well in advance. Hence, there is definite need for a good HHR forecasting model. OBJECTIVES: To present a physician forecasting model that projects the Full-Time Equivalent (FTE) demand for and supply of physicians in Nova Scotia to the year 2020 for three specialties: general practitioners, medical, and surgical. The model enables gap analysis and assessment of alternative policy options designed to close the gaps. METHODOLOGY: The methodology for estimating demand fo physician services involves three steps: (i) Establishing the FT for each physician. To this end we calculate the income of each physician using Physician Billings Data and then identify the 40th and 60th percentile income levels for each of the 40 specialties. The income levels are then used to calculate the FTE using a formula developed at Health Canada; (ii) Calculating the FTE for each service by distributing the FTE of each physician at the service level (i.e., by patient age, sex, most responsible diagnosis, and hospital status group); and (iii) Using Statistics Canada's population projections to project future demand for three broad medical disciplines: general practitioners, medical specialist, and surgical specialists. The supply side of the model employs a stock/flow approach and exploits time-series and other data for variables, such as emigration, international medical graduates (IMGs), medical school entrants, retirements, mortality, and so on, which in turn allow us to access a host of policy parameters. RESULTS: Under the status quo assumption, demand for physician services will outstrip the growth in supply for all three specialties. CONCLUSIONS: The model can simulate supply-side policy changes (e.g. more IMGs, delayed retirements) and can also reflect changes in demand (e.g. a cure for leukemia; different work intensities for physicians). The model is highly parameterized so that it can accommodate shocks that may influence the future requirements for physicians. Once a future requirement is determined, the supply model can identify the policy levers (new entrants, immigration, emigration, retirement) necessary to close the gap between demand and supply. The model is a user-friendly tool made for policy makers to formulate appropriate physician workforce planning.
A shortage of physicians is predicted in the near future, as there is a mismatch between the supply and demand for physicians. Our objective was to examine what effect an aging Louisiana physician population might have on these forecasts using data from the American Medical Association and population projections from the Louisiana Population Data Center. Our results propose a considerable rise in the number of Louisiana physicians who are older, with growth in younger physicians being continually flat. We forecast 44.2 percent of Louisiana physicians will be over 55 years of age in the year 2020. Acknowledging many limitations of a study such as this, we believe an aging physician population might lead to an overall decrease in the effective physician supply of Louisiana in the future.
Spanish Aerobiology Network (REA) includes 22 research groups implied in the biological analysis of the air from 50 observation sites extended over the Iberian Peninsula and Baleares Islands. To offer in-advance information for pollen allergy sufferers, doctors and health authorities is one of REA priorities. One of our objectives is the study of the influence of meteorological variations on the quantity and occurrence of pollen in the atmosphere. Nevertheless, to include only weather parameters is not enough for build accurate forecasts. Field phenology data, vegetation land cover, topography data and previous aerobiological data should be included in the models. Meteorological conditions control the start, duration, intensity and end of the pollen season, but their effect depends on the different types of vegetation. The biological response of tree pollination is highly differentiated from the herbaceous one. Another important point is the adaptation of a given species phetnology to different climatic areas. We showed the different phenoclimatic variables determining tree flowering in the Mediterranean climate areas and in the Atlantic climate ones of the Iberian Peninsula, and the consequences on the development of our forecasting models. Finally we comment our more recent studies that use geostatistic algorithms as a tool for including all the cited parameters and also space variables as altitude, latitude and orientation to sun.
This paper uses Grey Model to divide the malaria incidence of 1956-1986 into interval in Xuzhou City. The Grey forecasting GM (1,1) model is used to calculate the malaria incidence in 1987 and 1988. The accuracy of forecasting is 94.83% and 82.44% respectively. Furthermore, the Grey Verhulst model is used to fit the malaria incidence for the study of the declining tendency of vivax malaria incidence. Based on the fitting models, the decline tendency estimation models of vivax malaria are worked out.
The personal medical care costs of those diagnosed with acquired immunodeficiency syndrome (AIDS) in 1988 are forecast to be $2.2 billion, an amount that will increase to $4.5 billion in 1991. This is the first study to include the cost of purchasing azidothymidine (AZT), also called zidovudine, a palliative treatment for AIDS. The forecasts of this study are lower than those reported by Rice and Scitovsky, and other researchers, because the data are more recent and AIDS patients are receiving more care on an outpatient basis and staying in the hospital fewer days. They are also lower because projections for the number of AIDS cases diagnosed in future years are lower than those made by the Centers for Disease Control (CDC). This study projects that about 38,000 AIDS cases will be diagnosed in 1988 and 73,000 in 1991. The projections in this study are derived using data on the number of AIDS cases reported to CDC from January 1984 to October 1987, while the CDC projections employed by Rice and Scitovsky were derived using data from June 1981 to May 1986. It is also projected that the lifetime cost of treating an AIDS patient will increase from $57,000 in 1988 to $61,800 in 1991 due to the wider use of AZT.
This article analyzes three methods used to forecast the transition of long-term care clients through a variety of possible home and facility placements and levels of care. The test population (N = 1,653) is derived from the larger population of clients admitted in 1978 to British Columbia's newly established Long-Term Care program. The investigators have accumulated 5 years of service-generated data on moves, discharges, and deaths of these clients. Results show that the first-order Markov chain with stationary transition probabilities yields a superior forecast to state-by-state moving average growth and state-by-state regression analyses. The results of these analyses indicate that the Markov method should receive serious consideration as a tool for resource planning and allocation in long-term care.
Population projections and age-stratified cancer rates were used to forecast the trends for cancer occurrence for the United States during the coming decades: A progressive increase in the number of new cases is already predetermined by the high birth rate that occurred during the middle part of the century, and it will lead to nearly a doubling in the number of cases in about 4 decades. The demographic circumstances that determine the per capita costs of cancer are likely to worsen, because expected population trends will produce a deteriorating per capita economic base. Furthermore, the high costs of emerging treatment technologies are likely to magnify the deterioration ever further. Development of methods for cancer trend forecasting can provide a sound and accurate foundation for planning a comprehensive national strategy for optimal partitioning of research resources between the need for development of new treatments and the need for new research directed toward primary preventive measures.
THREE MODELS FOR FORECASTING HOME HEALTH CARE (HHC) NEEDS ARE ANALYZED: HSA/SP model (Health Systems Agency of Southwestern Pennsylvania); Florida model (Florida State Department of Health and Rehabilitative Services); and Rhode Island model (Rhode Island Department of Community Affairs). A utilization approach to forecasting is also presented.In the HSA/SP and Florida models, need for HHC is based on a certain proportion of (a) hospital admissions and (b) patients entering HHC from other sources. The major advantage of these models is that they are relatively easy to use and explain; their major weaknesses are an imprecise definition of need and an incomplete model specification.The Rhode Island approach defines need for HHC in terms of the health status of the population as measured by chronic activity limitations. The major strengths of this approach are its explicit assumptions and its emphasis on consumer needs. The major drawback is that it requires considerable local area data.The utilization approach is based on extrapolation from observed utilization experience of the target population. Its main limitation is that it is based on current market imperfections; its major advantage is that it exposes existing deficiencies in HHC.The author concludes that each approach should be tested empirically in order to refine it, and that need and demand approaches be used jointly in the planning process.
A broad range of economic assumptions are used to project the future income and outgo of the Social Security system. The assumptions adopted by the Board of Trustees of the Old-Age and Survivors Insurance and Disability Insurance (OASDI) Trust Fund were rather consistently on the optimistic side of the actual experience that emerged. This article examines the experience of several key economic indicators during the 1970's. Acknowledging that forecasting such quantities is an inexact science at best, the authors present a formula for making estimates of OASDI fund ratios, given the necessary assumptions. The formula is used to project fund ratios from 1981 to 1986. It shows where the fund would stand if forecasting errors were to continue at the magnitudes experienced in 1970-76.
Concern about controlling health care costs is a driving force in forecasting workload. The availability of computers and spreadsheet software now makes it possible for middle managers to forecast future workload using past workload data.
OBJECTIVE: The Health Resources and Services Administration's Bureau of Health Professions developed a demographic utilization-based model of physician specialty requirements to explore the consequences of a broad range of scenarios pertaining to the nation's health care delivery system on need for physicians. DATA SOURCE/STUDY SETTING: The model uses selected data primarily from the National Center for Health Statistics, the American Medical Association, and the U.S. Bureau of Census. Forecasts are national estimates. STUDY DESIGN: Current (1989) utilization rates for ambulatory and inpatient medical specialty services were obtained for the population according to age, gender, race/ethnicity, and insurance status. These rates are used to estimate specialty-specific total service utilization expressed in patient care minutes for future populations and converted to physician requirements by applying per-physician productivity estimates. DATA COLLECTION/EXTRACTION METHODS: Secondary data were analyzed and put into matrixes for use in the mainframe computer-based model. Several missing data points, e.g., for HMO-enrolled populations, were extrapolated from available data by the project's contractor. PRINCIPAL FINDINGS: The authors contend that the Bureau's demographic utilization model represents improvements over other data-driven methodologies that rely on staffing ratios and similar supply-determined bases for estimating requirements. The model's distinct utility rests in offering national-level physician specialty requirements forecasts.
We argue that over the past 300 years human physiology has been undergoing profound environmentally induced changes made possible by numerous advances in technology. These changes, which we call technophysio evolution, increased body size by over 50%, and greatly improved the robustness and capacity of vital organ systems. Because technophysio evolution is still ongoing, it is relevant to forecasts of longevity and morbidity and, therefore, to forecasts of the size of the elderly population and pension and health care costs.
BACKGROUND: To forecast adult patient outcomes in the management of hyperlipidemia using adult National Health and Examination Survey III (NHANES III) population statistics and National Cholesterol Education Program (NCEP) guidelines for goals of therapy. METHODS: Review of the hyperlipidemia drug therapy English-language medical literature with emphasis on randomized controlled trials of more than 6 weeks' duration published in the last 7 years, product package inserts, US Food and Drug Administration submission information, and NHANES III population statistics. Data were extracted from studies of lipid-lowering therapy to modify low-density lipoprotein (LDL) levels for primary and secondary prevention of coronary heart disease. The data that were evaluated included sample size, study design, therapeutic intervention, length of study, percentage change in LDL levels, and patient demographics. RESULTS: Cumulative frequency curves of the LDL distribution among the US adult population were constructed. The mean efficacy of drug therapy from qualified studies was used to extrapolate the percentage of the population expected to respond to the intervention and to forecast the patient outcome. CONCLUSIONS: A useful tool for clinicians was constructed to approximate the percentage of patients, based on risk stratification, who would reach NCEP target goal after a given pharmacotherapeutic intervention to decrease LDL levels.
Forecasts of the requirements for and supply of ophthalmologists in 1990 have produced conflicting results because of varying assumptions about the future utilization of eye care services, incidence and prevalence of ophthalmic disease, physician productivity, and availability of residency training positions. A typical "utilization-based" models, founded on present consumer behavior, predicts a substantial 1990 surplus of ophthalmologists at current rates of residency training. Two "need-based" models, founded on ideal rather than actual use, reach different conclusions because of varying use of a fragile data base and the need to rely heavily on the subjective judgment of experts with regard to norms of care. The 1980 Graduate Medical Educational National Advisory Committee forecasted a surplus, while the 1978 American Academy of Ophthalmology predicted a deficit. Utilization-based models may slightly underestimate future ophthalmologist requirements. However, analysis of the factors that will influence future use suggests that need-based models are likely to overestimate the requirements. It is risky to accept the need-based model projections because of the high cost of a surplus, which include not only the expenses of training unneeded ophthalmologists but also the cost of their decreased exposure to disease and of declining physician morale, acumen, and thresholds for surgical procedures. Because free market mechanisms are ineffective in governing the supply of health providers, it will be necessary for the profession itself to review the current and projected supply and to set limits on the number of persons in training.
Every tenth person in Central Europe is a pollinosis patient. The time of ripening and release of pollen, as well as pollen flight, all depend on the weather. Because each year is different from every other, mean values from pollen calendars do not provide any practical help for allergy-sufferers. For this reason, in many European countries, measuring networks have been established during the last 10 years as a basis for forecasting the prevalence of airborne pollen for the following 2-3 days, in connection with the weather forecast. Cooperation and communication also exist on a European level, and a European Pollen Database had been established.
Prediction of ambient ozone concentrations in urban areas would allow evaluation of such factors as compliance and noncompliance with EPA requirements. Though ozone prediction models exist, there is still a need for more accurate models. Development of these models is difficult because the meteorological variables and photochemical reactions involved in ozone formation are complex. In this study, we developed a neural network model for forecasting daily maximum ozone levels. We then compared the neural network's performance with those of two traditional statistical models, regression, and Box-Jenkins ARIMA. The neural network model for forecasting daily maximum ozone levels is different from the two statistical models because it employs a pattern recognition approach. Such an approach does not require specification of the structural form of the model. The results show that the neural network model is superior to the regression and Box-Jenkins ARIMA models we tested.