Time series analysis of correlativity between pulmonary tuberculosis and seasonal meteorological factors based on theory of Human-Environmental Inter Relation

Objective: This paper aims to study the correlativity between the number of pulmonary tuberculosis (PTB) cases and seasonal meteorological factors in Beijing. Methods: Based on theory of Human–Environmental Inter Relation in Huangdi's Internal Classics, we adopted monthly cases of PTB in Beijing from 2004 to 2011, and established a Seasonal Autoregressive Integrated Moving Average (SARIMA) model. Using the cross-correlation function (CCF), we then analyzed the correlation between meteorological factors and number of infected patients. The related meteorological factors were subsequently integrated, to establish a Seasonal Autoregressive Integrated Moving Average with explanatory variables (SARIMAX) model, which was used to estimate and verify the number of PTB cases in 2012. Results: In this study, a SARIMA(0,1,1) (0,1,1)12 model was established; CCF analysis was used to reveal the correlativity between PTB and precipitation with 1 lag, relative humidity with 1 lag. Then, integrated with relative humidity with 1 lag (β = 2.405, 95% confidence interval: 0.433–4.377), the SARIMAX prediction model was proved to be an accurate approach for predicting local situations of PTB occurrence. Conclusions: The occurrence of PTB is correlated with seasonal meteorological factors. Combining these factors, an exact prediction model can be established, to estimate of the number of PTB infected patients. Keywords: Human–Environmental Inter Relation, Pulmonary tuberculosis, Time series analysis, Seasonal Autoregressive Integrated Moving Averag