Yang Liping, Wang Chunxia, Zhou Pan, Xie Na, Tian Maozai, Wang Kai
College of Public Health, Xinjiang Medical University, Urumqi, 830017, China.
College of Medical Engineering and Technology, Xinjiang Medical University, Urumqi, 830017, China.
Sci Rep. 2025 Jan 30;15(1):3830. doi: 10.1038/s41598-025-88508-0.
Brucellosis is a significant global challenge, but there has been a lack of epidemiological studies on brucellosis in Xinjiang from a change point perspective. This study aims to bridge this gap by employing sequence decomposition and identifying significant change points, with datasets sourced from the Xinjiang Disease Prevention and Control Information System. This study employed the BEAST algorithm to decompose the brucellosis time series in Xinjiang from 2010 to 2023, while simultaneously identifying change points in the decomposed seasonal and trend components. The probability of four change points occurring within the seasonal component is 0.8950. And the locations where these four change points occur and the probabilities associated with each change point are August 2013 ([Formula: see text]), August 2017 ([Formula: see text]), February 2022 ([Formula: see text]), and May 2023 ([Formula: see text]), respectively. The probability of the existence of five change points in the trend factors of brucellosis in Xinjiang is highest ([Formula: see text]). The times at which these five change points occur, along with the probabilities of change at those moments, are as follows: March 2013 ([Formula: see text]), August 2015 ([Formula: see text]), July 2017 ([Formula: see text]), February 2020 ([Formula: see text]), and May 2023 ([Formula: see text]). Change point analysis holds significant utility within the field of epidemiology. These discoveries furnish pivotal insights for epidemiological investigations and the development of early warning systems tailored to brucellosis.
布鲁氏菌病是一项重大的全球性挑战，但从变化点角度来看，新疆缺乏关于布鲁氏菌病的流行病学研究。本研究旨在通过采用序列分解并识别显著变化点来弥补这一差距，数据集来自新疆疾病预防控制信息系统。本研究采用贝叶斯进化分析采样树（BEAST）算法对2010年至2023年新疆布鲁氏菌病时间序列进行分解，同时识别分解后的季节性和趋势成分中的变化点。季节性成分中出现四个变化点的概率为0.8950。这四个变化点出现的位置以及每个变化点的相关概率分别为2013年8月（[公式：见原文]）、2017年8月（[公式：见原文]）、2022年2月（[公式：见原文]）和2023年5月（[公式：见原文]）。新疆布鲁氏菌病趋势因素中存在五个变化点的概率最高（[公式：见原文]）。这五个变化点出现的时间以及当时变化的概率如下：2013年3月（[公式：见原文]）、2015年8月（[公式：见原文]）、2017年7月（[公式：见原文]）、2020年2月（[公式：见原文]）和2023年5月（[公式：见原文]）。变化点分析在流行病学领域具有重要作用。这些发现为布鲁氏菌病的流行病学调查和预警系统的开发提供了关键见解。
