Department of Statistics, Bahauddin Zakariya University, Multan, Pakistan.
Department of Mathematics and Computer Science, Faculty of Science and Technology Prince of Songkla University, Pattani Campus, Thailand.
BMC Med Inform Decis Mak. 2024 Nov 21;24(1):351. doi: 10.1186/s12911-024-02774-y.
The structural abnormality of the heart and its blood vessels at the time of birth is known as congenital heart disease. Every year in Pakistan, sixty thousand children are born with CHD, and 44 in 1000 die before they are a month old. Various studies used different techniques to estimate the risk factors of congenital heart disease, but these techniques suffer from a deficiency of capacity to present human understanding and a deficiency of adequate data. The current study provided an innovative approach by defining the latent variables to handle this issue and building a reasonable model.
Data used in this study has been collected from mothers and hospital records of the children. The dataset contains information on 3900 children who visited the OPD of the Chaudry Pervaiz Elahi Institute of Cardiology (CPEIC) Multan, Pakistan from October 2021 to September 2022. The latent variables were defined from the data and structural equation modeling was used to model them.
The results show that there are 53.6% of males have acyanotic CHD and 54.5% have cyanotic CHD. There are 46.4% of females have acyanotic CHD and 45.5% have cyanotic CHD. The children who have no diabetes in the family are 64.0% and children who have diabetes in the family are 36.0% in acyanotic CHD, the children who have no diabetes in the family are 59.7% and children have diabetes in the family are 40.3% in cyanotic CHD. The value of standardized root mean residual is 0.087 is less than 0.089 which shows that the model is a good fit. The value of root mean square error of approximation is 0.113 is less than 0.20 which also shows the good fit of the model.
It was concluded that the model is a good fit. Also, the latent variables, socioeconomic factors, and environmental factors of mothers during pregnancy have a significant effect in causing cyanotic while poor general health factor increases the risk of Acyanotic congenital heart disease.
出生时心脏及其血管的结构异常被称为先天性心脏病。在巴基斯坦，每年有 6 万名儿童患有 CHD，其中有 44 名儿童在 1 个月前死亡。各种研究使用不同的技术来估计先天性心脏病的危险因素，但这些技术存在人类理解能力不足和数据不足的缺陷。本研究通过定义潜在变量来处理这个问题并建立合理的模型提供了一种创新方法。
本研究使用的数据来自母亲和儿童的医院记录。该数据集包含了 2021 年 10 月至 2022 年 9 月期间访问巴基斯坦木尔坦乔德里·佩尔瓦兹·埃拉希心脏病研究所（CPEIC）门诊的 3900 名儿童的信息。从数据中定义了潜在变量，并使用结构方程模型对其进行建模。
结果表明，患有非发绀性 CHD 的男性有 53.6%，患有发绀性 CHD 的男性有 54.5%。患有非发绀性 CHD 的女性有 46.4%，患有发绀性 CHD 的女性有 45.5%。家庭中没有糖尿病的儿童占 64.0%，家庭中有糖尿病的儿童占 36.0%；家庭中没有糖尿病的儿童占 59.7%，家庭中有糖尿病的儿童占 40.3%。标准化残差均方根为 0.087，小于 0.089，表明模型拟合良好。近似均方根误差值为 0.113，小于 0.20，也表明模型拟合良好。
研究得出结论，模型拟合良好。此外，母亲在怀孕期间的社会经济因素和环境因素以及一般健康状况较差的潜在变量都会显著导致发绀，而一般健康状况较差则会增加非发绀性先天性心脏病的风险。
