Abstract：
BACKGROUND: The rise in prevalence of atopic dermatitis has been correlated with numerous elements of the exposome, modern-day lifestyle, and familial history. The combined analysis of familial history and other risk elements may allow us to understand the driving factors behind the development of atopic dermatitis.
OBJECTIVE: We aimed to develop prediction models to assess the risk of developing atopic dermatitis using a large and diverse cohort (N=77,525) and easily-assessed risk factors.
METHODS: We analyzed electronic medical record data from Leumit Health System. Documented predictive factors include sex, season of birth, environment (urban/rural), socio-economic status, household smoking, diagnosed skin conditions, number of siblings, a paternal, maternal or sibling history of an atopic condition, and antibiotic prescriptions during pregnancy or following birth. Predictive models were trained and validated on the dataset.
RESULTS: Medium (OR 2.04, CI 1.92-2.17, p<0.001) and high (OR 2.13, CI 1.95-2.34, p<0.001) socioeconomic status, a previous diagnosis of contact dermatitis (OR 2.57, CI 2.37-2.78, p<0.001), presence of siblings with an AD diagnosis (OR 2.21, CI 2.04-2.40, p<0.001) and the percentage of siblings with any atopic condition (OR 2.58, CI 2.09-3.17, p<0.001) drove risk for AD in a logistic regression model. A random forest prediction model with a sensitivity of 61% and a specificity of 84% was developed. Generalized mixed models accounting for the random effect of familial relationships boasted an area under the curve of 0.98.
CONCLUSIONS: Predictive modeling using non-invasive and accessible inputs is a powerful tool to stratify risk for developing atopic dermatitis.
OBJECTIVE: We aimed to develop prediction models to assess the risk of developing atopic dermatitis using a large and diverse cohort (N=77,525) and easily-assessed risk factors.
METHODS: We analyzed electronic medical record data from Leumit Health System. Documented predictive factors include sex, season of birth, environment (urban/rural), socio-economic status, household smoking, diagnosed skin conditions, number of siblings, a paternal, maternal or sibling history of an atopic condition, and antibiotic prescriptions during pregnancy or following birth. Predictive models were trained and validated on the dataset.
RESULTS: Medium (OR 2.04, CI 1.92-2.17, p<0.001) and high (OR 2.13, CI 1.95-2.34, p<0.001) socioeconomic status, a previous diagnosis of contact dermatitis (OR 2.57, CI 2.37-2.78, p<0.001), presence of siblings with an AD diagnosis (OR 2.21, CI 2.04-2.40, p<0.001) and the percentage of siblings with any atopic condition (OR 2.58, CI 2.09-3.17, p<0.001) drove risk for AD in a logistic regression model. A random forest prediction model with a sensitivity of 61% and a specificity of 84% was developed. Generalized mixed models accounting for the random effect of familial relationships boasted an area under the curve of 0.98.
CONCLUSIONS: Predictive modeling using non-invasive and accessible inputs is a powerful tool to stratify risk for developing atopic dermatitis.
摘要：
背景：特应性皮炎患病率的上升与许多因素有关，现代生活方式,和家族史。家族病史和其他危险因素的综合分析可能使我们了解特应性皮炎发展背后的驱动因素。
目的：我们旨在开发预测模型来评估特应性皮炎的风险，使用一个庞大且多样化的队列（N=77,525），并且易于评估的风险因素。
方法：我们分析了LeumitHealthSystem的电子病历数据。记录在案的预测因素包括性别，出生季节,环境（城市/农村），社会经济地位,家庭吸烟,诊断皮肤状况，兄弟姐妹的数量，一个父亲,母亲或兄弟姐妹有特应性疾病史，和抗生素处方在怀孕期间或出生后。预测模型在数据集上进行训练和验证。
结果：中等（OR2.04，CI1.92-2.17，p<0.001）和高（OR2.13，CI1.95-2.34，p<0.001）社会经济地位，先前诊断为接触性皮炎（OR2.57，CI2.37-2.78，p<0.001），在逻辑回归模型中,患有AD诊断的兄弟姐妹(OR2.21,CI2.04-2.40,p<0.001)和患有任何特应性疾病的兄弟姐妹百分比(OR2.58,CI2.09-3.17,p<0.001)导致AD风险.建立了灵敏度为61%、特异性为84%的随机森林预测模型。考虑到家庭关系的随机效应的广义混合模型在曲线下的面积为0.98。
结论：使用非侵入性和可访问输入的预测模型是对特应性皮炎发展风险进行分层的强大工具。
目的：我们旨在开发预测模型来评估特应性皮炎的风险，使用一个庞大且多样化的队列（N=77,525），并且易于评估的风险因素。
方法：我们分析了LeumitHealthSystem的电子病历数据。记录在案的预测因素包括性别，出生季节,环境（城市/农村），社会经济地位,家庭吸烟,诊断皮肤状况，兄弟姐妹的数量，一个父亲,母亲或兄弟姐妹有特应性疾病史，和抗生素处方在怀孕期间或出生后。预测模型在数据集上进行训练和验证。
结果：中等（OR2.04，CI1.92-2.17，p<0.001）和高（OR2.13，CI1.95-2.34，p<0.001）社会经济地位，先前诊断为接触性皮炎（OR2.57，CI2.37-2.78，p<0.001），在逻辑回归模型中,患有AD诊断的兄弟姐妹(OR2.21,CI2.04-2.40,p<0.001)和患有任何特应性疾病的兄弟姐妹百分比(OR2.58,CI2.09-3.17,p<0.001)导致AD风险.建立了灵敏度为61%、特异性为84%的随机森林预测模型。考虑到家庭关系的随机效应的广义混合模型在曲线下的面积为0.98。
结论：使用非侵入性和可访问输入的预测模型是对特应性皮炎发展风险进行分层的强大工具。
