Evaluation of extrapolation methods for actual state expenditures on health care in Russian Federation

Forecasting methods, extrapolation ones in particular, are used in health care for medical, biological and clinical research. The author, using accessible internet space, has not met a single publication devoted to extrapolation of financial parameters of health care activities. This determined the relevance of the material presented in the article: based on health care financing dynamics in Russia in 2000–2010 the author examined possibility of application of basic perspective extrapolation methods - moving average, exponential smoothing and least squares. It is hypothesized that all three methods can equally forecast actual public expenditures on health care in medium term in Russia's current financial and economic conditions. The study result was evaluated in two time periods: within the studied interval and a five-year period. It was found that within the study period all methods have an average relative extrapolation error of 3–5%, which means high precision of the forecast. The study shown a specific feature of the least squares method which were gradually accumulating results so their economic interpretation became possible only in the end of the studied period. That is why the extrapolating results obtained by least squares method are not applicable in an entire study period and rather have a theoretical value. Beyond the study period, however, this feature was found to be the most corresponding to the real situation. It was the least squares method that proved to be the most appropriate for economic interpretation of the forecast results of actual public expenditures on health care. The hypothesis was not confirmed, the author received three differently directed results, while each method had independent significance and its application depended on evaluation study objectives and real social, economic and financial situation in Russian health care system.