关键词: 结核,肺, 发病率, 模型,统计学, 预测

Abstract:

Objective: To analyze the epidemiological trends of pulmonary tuberculosis (PTB) in Daxing District, Beijing, from 2015 to 2024, and to predict the disease incidence by Seasonal Autoregressive Integrated Moving Average (SARIMA). Methods: PTB cases data from 2015 to 2024 were collected from the “Infectious Disease Monitoring System”, a subsystem of the “China Disease Prevention and Control Information System”. Descriptive epidemiological methods were used to analyze the distribution of PTB, while Joinpoint regression was employed to analyze annual incidence trends. The SARIMA model was constructed using RStudio software with data from 2015—2023 as the training set to predict the number and incidence rate in 2024, which were then compared with actual observations. Results: From 2015 to 2024, a total of 5548 PTB cases were notified in Daxing District, with an average annual incidence of 32.1/100000(5548/17303714). Among these, 2299 cases were etiologically confirmed PTB, 1829 cases were etiologically negative, and 1420 cases had no etiological results. The overall PTB incidence decreased by 49.3% from 2015 to 2024 was, with an average annual decrease rate of 7.3%. The proportion of etiologically positive cases increased from 30.7% (189/615) to 69.0% (256/371)(Z=22.387,P<0.001), while the proportion of etiologically negative cases decreased from 38.9% (239/615) to 20.2% (75/371)(Z=-12.400,P<0.001), and the cases without etiologically results decreased from 30.4% (187/615) to 10.8% (40/371)(Z=-11.914,P<0.001). The highest PTB incidence rates occurred in the ≥80 age group (126.4/100000, 324/256359), followed by the 70-79 (69.1/100000, 457/661535), and 60-69 age groups (40.4/100000, 588/1454785). The top five occupations among PTB cases were homemakers and unemployed individuals (38.1%, 2116/5548), farmers (14.5%, 805/5548), workers (11.6%, 646/5548), retirees (10.6%, 587/5548), and cadre staff members (8.2%, 453/5548). The top five town/subdistricts with the highest number of TB cases were Huangcun Town (12.4%, 689/5548), Jiugong Town (10.9%, 606/5548), Xihongmen Town (10.0%, 556/5548), Qingyuan Subdistrict (6.9%, 381/5548), and Yizhuang Town (6.7%, 372/5548). The ARIMA(0,1,1)(1,0,0)_[12] model predicted the 2024 PTB cases and incidence rate with Mean Absolute Percentage Errors (MAPE) of 24.008% and 32.493%, respectively, and actual observed values fell within the 95% confidence intervals of prediction. Conclusion: PTB incidence in Daxing District has shown a annually decline, with an increasing proportion of etiologically confirmed cases from 2015 to 2024. The elderly (≥60 years) exhibited higher incidence rates, and homemakers/unemployed individuals and farmers were the most affected occupational groups. High-incidence areas included densely populated regions such as Huangcun and Jiugong. The ARIMA (0,1,1)(1,0,0)_[12] model demonstrated reasonable accuracy for PTB incidence prediction in the district.

Key words: Tuberculosis, pulmonary, Incidence, Models, statistical, prediction