Department of Obstetrics and Gynecology, Chung Shan Medical University Hospital, Taichung, Taiwan.
Institute of Oral Sciences, Chung Shan Medical University, Taichung, Taiwan.
Taiwan J Obstet Gynecol. 2022 May;61(3):479-484. doi: 10.1016/j.tjog.2022.03.014.
In this 3-year longitudinal cohort study, we aimed to evaluate the evolution of overactive bladder in female community residents aged 40 years and above in central Taiwan and identify its risk factors.
Female community residents aged 40 years and above were invited to participate in this study and fill out a yearly Overactive Bladder Symptom Score (OABSS) questionnaire over a 3-year period. A woman was defined to have OAB if the total OABSS was ≧4 and urgency score was ≧2. At the end of the third year, the incidence, remission, persistence, and relapse of OAB in these community residents were calculated. A novel statistical analysis technique, machine learning with data mining, was applied to examine its use in this field. Five machine learning models were used to predict the risk factors associated with persistent OAB and the results were compared with the conventional logistic regression model.
In total, 1469 female residents were included in the first year and 1290 (87.8%) women completed the questionnaires for all 3 years. The prevalence of OAB was 20.2% (n = 260). The second- and third-year incidence rates of OAB were 13.5% and 7.1%. The remission rates were 39.6% and 44.3%. Twenty-two percent of the women reported relapse of OAB in the third year. The two-year OAB persistence rate was 43.8%. For the prediction of risk factors for persistent OAB, the multivariable logistic regression model had better predictive accuracy (AUC = 0.664) than the five machine learning models. Age ≧ 60 was associated with persistent OAB (OR 2.8; 95% CI: 1.34-5.89, P = 0.002).
The yearly incidence, remission, and persistence rates of OAB were high in female community residents aged 40 years and above in central Taiwan. Older women had a higher risk of persistent OAB symptoms in this 3-year longitudinal cohort study.
在这项为期 3 年的纵向队列研究中，我们旨在评估台湾中部 40 岁及以上女性社区居民中膀胱过度活动症的演变，并确定其危险因素。
邀请 40 岁及以上的女性社区居民参加本研究，并在 3 年内每年填写一份膀胱过度活动症症状评分（OABSS）问卷。如果总 OABSS≧4 且急迫评分≧2，则定义女性患有 OAB。在第 3 年末，计算这些社区居民中 OAB 的发生率、缓解率、持续性和复发率。应用一种新的统计分析技术，即数据挖掘的机器学习，来检查其在该领域的应用。使用了 5 种机器学习模型来预测与持续性 OAB 相关的危险因素，并将结果与传统的逻辑回归模型进行比较。
共有 1469 名女性居民参加了第一年的研究，其中 1290 名（87.8%）女性完成了所有 3 年的问卷调查。OAB 的患病率为 20.2%（n=260）。OAB 的第二年和第三年的发生率分别为 13.5%和 7.1%。缓解率分别为 39.6%和 44.3%。22%的女性在第三年报告了 OAB 的复发。第三年 OAB 的两年持续性率为 43.8%。对于持续性 OAB 危险因素的预测，多变量逻辑回归模型的预测准确性（AUC=0.664）优于 5 种机器学习模型。年龄≧60 岁与持续性 OAB 相关（OR 2.8；95%CI：1.34-5.89，P=0.002）。
在台湾中部 40 岁及以上的女性社区居民中，每年 OAB 的发生率、缓解率和持续性率均较高。在这项为期 3 年的纵向队列研究中，年龄较大的女性患持续性 OAB 症状的风险更高。
