基于机器学习算法构建慢性阻塞性肺疾病吸入剂治疗患者不良吸入风险预警模型

目的构建并验证接受吸入剂治疗的慢性阻塞性肺疾病(COPD)患者不良吸入的风险预测模型,为不良吸入的个体化预防提供决策支持工具。方法采用横断面研究,收集COPD吸入剂治疗患者相关数据,形成数据集。按4:1将数据集随机分为训练集和测试集,通过采用4种不同的缺失值填补方法、3种变量特征筛选方法以及18种机器学习算法,在训练集上构建模型。在测试集中使用蒙特卡罗模拟法进行重采样,验证模型,以曲线下面积(AUC)、准确率、精准率、召回率和F1值评估模型性能。选择最优模型用于构建不良吸入预测平台。结果共纳入COPD患者308例,135例(43.8%)存在不良吸入风险。根据33个特征变量构建了216个风险预警模型,其中,集成学习算法的平均AUC最大,为0.844±0.058[95%CI=(0.843,0.845)]。216个模型在预测性能方面差异有统计学意义(P<0.01)。在集成学习算法下,吸入剂使用依从性(38.0874%)、吸入剂满意度(25.6801%)、教育水平(24.0313%)、吸入剂数量(5.4823%)、年龄(4.2045%)和过去一年急性加重频次(2.1847%)对预测模型贡献最大。该模型展现出良好的预测性能,其AUC 0.8693、准确率83.87%、精准率86.96%、召回率74.07%、F1值0.8。结论该研究构建的COPD吸入剂治疗患者不良吸入风险的预测模型预测能力良好,具有一定潜在临床应用价值。

品管圈在小儿哮喘糖皮质激素吸入治疗中的应用

目的探讨品管圈在小儿哮喘患者糖皮质激素吸入治疗中的应用价值,为临床治疗提供参考。方法选取广州中医药大学第一附属医院2016年1月—2018年2月收治的160例小儿哮喘作为研究对象,所有患儿均采取糖皮质激素吸入治疗,随机分为观察组和对照组,每组80例。对照组患儿配合常规护理,观察组在对照组基础上接受品管圈干预。观察并比较两组患儿雾化吸入不良次数、储药罐内残余药量、临床疗效;干预后,评估并比较两组医疗操作恐惧、医疗环境恐惧、自我恐惧、服药依从性及护理满意度。结果观察组雾化吸入缺陷率为3.75%,明显低于对照组20.00%,差异有统计学意义(P<0.05)。观察组的雾化吸入结束后平均残余药量为(0.33±0.05) ml,明显低于对照组的(0.60±0.18) ml,差异有统计学意义(P<0.01)。观察组医疗操作恐惧、医疗环境恐惧和自我恐惧评分均低于对照组评分,差异有统计学意义(P<0.01)。观察组服药依从性、抵触情绪和沟通意向评分均高于对照组评分,差异有统计学意义(P<0.01)。观察组护理满意率为97.50%,明显高于对照组的62.50%,差异有统计学意义(P<0.05)。观察组总有效率为96.25%,明显高于对照组的85.00%,差异有统计学意(P<0.05)。结论小儿哮喘患者糖皮质激素吸入治疗实施品管圈活动后能够有效降低氧气雾化吸入患者缺陷发生率,减轻儿童医疗恐惧及抵触情绪,提高护理满意度及临床治疗效果。

Influence of moxa smoke on mitochondrial transmembrane potential and Bax/Bcl-2 in alveolar type Ⅱ epithelial A549 cells

Objective： To investigate the influence of moxibustion products on mitochondrial transmembrane potential （MTP） and mRNA expression of Bax/Bcl-2 in alveolar type Ⅱ epithelial A549 cells, and to further explore influence of moxibustion products on the oxidative damage of A549 cells. Methods： Smoke and particles generated by moxibustion were collected using the filter box for gas sampling. The moxa smoke extract （MSE） was diluted sequentially to the final concentrations of 0.05 mg/mL, 0.2 mg/mL, 0.2 mg/mL, 0.3 mg/mL and 0.4 mg/mL using the cell culture medium, and A549 cells were then intervened by the above MSE solution. Cell MTP was detected by JC-1 staining. Fluorescence quantitative polymerase chain reaction （PCR） was used to detect Bax/Bcl-2 mRNA expression of A549 cells. Results： Compared with cells in the normal control group, MTP was significantly decreased in cells of 0.3 mg/mL and 0.4 mg/mL MSE intervention groups （P〈0.01）; while MTP showed no significant changes in cells of 0.05 mg/mL, 0.1 mg/mL and 0.2 mg/mL MSE intervention groups （P〉0.05）; compared with cells in 0.05 mg/mL MSE intervention group, MTP was decreased significantly in cells of 0.1 mg/mL, 0.2 mg/mL, 0.3 mg/mL and 0.4 mg/mL MSE intervention groups （P〈0.05）; compared with cells in 0.1 mg/mL MSE intervention group, MTP was decreased significantly in cells of 0.4 mg/mL MSE intervention group （P〈0.01）. Bax mRNA expression of cells in each concentration of MSE intervention group all showed no significant difference compared to that in the normal control group; Bcl-2 mRNA expression of cells was reduced with the increase of MSE intervention concentration. Wherein, Bcl-2 mRNA expressions of cells in 0.4 mg/mL and 0.3 mg/mL MSE intervention groups were significantly reduced compared with that of cells in the normal control group （P〈0.05）; Bcl-2 mRNA expression of cells in 0.4 mg/mL MSE intervention group was significantly reduced compared to that in 0.05 mg/mL MSE intervention group （P〈0.05）. Conclusion： Certain higher concentration of moxa smoke could reduce MTP and mRNA expression of the anti-apoptosis gene Bcl-2 in alveolar type Ⅱ epithelial A549 cells. Oxidative damage may be the important mechanism of apoptosis caused by the high concentration of moxa smoke solution, and further studies are necessary on the specific mechanisms.