Scientists from Perm Technical University have created the first model in Russia capable of predicting life expectancy with an accuracy of up to 99.7%.
Experts from the National Institute for Public Policy Research at the university explained that the new model would improve the accuracy of healthcare planning and the pension system in Russia.
According to them, the mathematical models currently used to estimate mortality rates date back to the last century and do not take into account modern demographic patterns, which may lead to a decrease in the accuracy of spending estimates and budget allocations.
Most current models rely on the assumption that the risk of death increases with age, which is true for adults, but it doesn't reflect the reality for infants. The risk of death is highest during the first year of life, then drops sharply until age five. Furthermore, traditional models don't account for the annual improvements in healthcare resulting from modern medicines and vaccines, which can lead to an underestimation of life expectancy. The researchers add that Western models are not suitable for application in Russia due to the climatic, social, and economic differences between regions.
The researchers based their model on the principle that the risk of death increases with age, adding two key factors: the annual decline in mortality rates due to medical advancements, and the dynamic changes in mortality risk during early childhood. The model was refined using mortality statistics for all age groups between 2004 and 2009. When compared to actual data from 2010, the model showed a 99.7% match.
"What distinguishes this model is that it takes into account the specificities of the Russian population structure, where infant mortality rates and risk distribution vary according to age," said German Pushkarev, associate professor in the Department of Applied Mathematics at the university. "Also, adjusting it based on local data ensures higher accuracy in forecasts and gives stakeholders more reliable information to support administrative decisions."
According to the researchers, the model does not require massive databases, making it useful even with limited statistics. It can be used to plan healthcare services, pension systems, social security, and to estimate future needs for hospital beds and medications, as well as to prepare pension budgets at both the federal and state levels.
