Applicability of the universal thermal climate index for predicting the outbreaks of respiratory tract infections: a mathematical modeling approach

Respiratory tract infections (RTI) are one of the most frequent reasons for medical consultations. As air temperature decreases, but also in connection with other meteorological parameters, evident seasonal fluctuations in the number of consultations for RTI can be observed. The Universal Thermal Climate Index (UTCI) is a complex meteorological index derived from an analysis of human thermal balance that depends on air temperature, air humidity, and wind speed. Our aims were to check if this index, although never used before for that purpose, is an adequate tool for forecasting seasonal increases in RTI prevalence. This study is a retrospective analysis of patients’ consultations with general practitioners in the period of 2012–2015 (453,674 records) recorded in the city of Olsztyn (Poland), which is characterized by a cold climate type (Dfb). The values of air temperature, atmospheric pressure, relative air humidity, wind speed, and UTCI were used for a statistical analysis and a mathematical analysis of curve fitting in order to determine correlations between analyzed meteorological parameters and a number of medical consultations for RTI. Analysis of the number of medical consultations for RTI revealed an evident seasonal pattern in a 4-year observation period, with a strong inverse correlation between the number of patients with RTI and the UTCI. A statistically significant increase in the number of patients with RTI appeared when the UTCI decreased, especially when it reached the classes of strong cold stress and very strong cold stress. In conclusion, the UTCI is a valuable predictive parameter for forecasting seasonal increases in RTI cases. Its decrease may initiate a seasonal increased prevalence. This effect is strongest about the 10th day following a change in the thermal climate conditions and is not continuous. A larger number of consultations for RTI after weekends and holidays (the Monday effect) may blur the results of statistical analyses.