Shi Caifeng, Chen Lijun
Department of Ophthalmology, Suzhou Ninth People's Hospital Suzhou 215200, Jiangsu, China.
Am J Transl Res. 2024 Oct 15;16(10):5418-5426. doi: 10.62347/WXHN4015. eCollection 2024.
To identify the influencing factors of dry eyes after cataract surgery and construct a prediction model to provide a reference for ophthalmologists in assessing the risk of postoperative dry eyes.
A retrospective study was conducted from January 2023 to April 2024, involving 219 patients (219 eyes) who underwent phacoemulsification with intraocular lens implantation at the Department of Ophthalmology, Ninth People's Hospital of Suzhou. Patients were divided into two groups based on the presence or absence of dry eyes at 2 weeks postoperatively. Data from both groups were analyzed to determine the influencing factors of dry eyes after cataract surgery. A nomogram prediction model was constructed using R software. The model's discrimination was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC), and model calibration was assessed using the Hosmer-Lemeshow (H-L) goodness-of-fit test and the Bootstrap method (self-sampling technique). Decision curve analysis was employed to evaluate the clinical utility of the model.
Among the 219 cataract patients, 53 (24.20%) developed dry eyes during the 2-week follow-up period. Multivariate logistic regression analysis identified smoking (OR = 1.809, P = 0.037), diabetes mellitus (OR = 3.248, P = 0.002), elevated IL-6 (OR = 3.019, P = 0.016), a high Hospital Anxiety and Depression Scale (HADS) score (OR = 2.147, P = 0.029), and longer surgical incision length (OR = 2.995, P = 0.014) as significant risk factors for postoperative dry eye. The AUC of the nomogram model was 0.857 (95% CI: 0.803-0.913), and the H-L goodness-of-fit test showed no statistical significance (χ = 4.472, P = 0.812), indicating good discrimination and calibration of the model. The average absolute error between predicted and actual probabilities after 1000 Bootstrap iterations was 0.021. Decision curve analysis demonstrated that the net benefit of the model was higher than the two extreme scenarios.
Postoperative dry eyes in cataract patients is associated with smoking, diabetes, elevated IL-6, high HADS scores, and longer incision lengths. The nomogram model demonstrates good predictive capability for assessing the risk of dry eyes after cataract surgery.
识别白内障手术后干眼的影响因素并构建预测模型，为眼科医生评估术后干眼风险提供参考。
进行一项回顾性研究，时间跨度为2023年1月至2024年4月，纳入在苏州市第九人民医院眼科接受白内障超声乳化吸除联合人工晶状体植入术的219例患者（219只眼）。根据术后2周时是否存在干眼将患者分为两组。分析两组数据以确定白内障手术后干眼的影响因素。使用R软件构建列线图预测模型。采用受试者操作特征（ROC）曲线下面积（AUC）评估模型的区分度，使用Hosmer-Lemeshow（H-L）拟合优度检验和Bootstrap方法（自助抽样技术）评估模型校准。采用决策曲线分析评估模型的临床实用性。
在219例白内障患者中，53例（24.20%）在2周随访期内发生干眼。多因素logistic回归分析确定吸烟（OR = 1.809，P = 0.037）、糖尿病（OR = 3.248，P = 0.002）、白细胞介素-6升高（OR = 3.019，P = 0.016）、医院焦虑抑郁量表（HADS）评分高（OR = 2.147，P = 0.029）以及手术切口长度较长（OR = 2.995，P = 0.014）是术后干眼的重要危险因素。列线图模型的AUC为0.857（95%CI：0.803 - 0.913），H-L拟合优度检验无统计学意义（χ = 4.472，P = 0.812），表明模型具有良好的区分度和校准度。1000次Bootstrap迭代后预测概率与实际概率之间的平均绝对误差为0.021。决策曲线分析表明，该模型的净效益高于两种极端情况。
白内障患者术后干眼与吸烟、糖尿病、白细胞介素-6升高、HADS评分高以及切口长度较长有关。列线图模型在评估白内障手术后干眼风险方面具有良好的预测能力。
