Improving Current Glycated Hemoglobin Prediction in Adults: Use of Machine Learning Algorithms With Electronic Health Records

Abstract

Background Predicting the risk of glycated hemoglobin (HbA1c) elevation can help identify patients with the potential for developing serious chronic health problems, such as diabetes. Early preventive interventions based upon advanced predictive models using electronic health records data for identifying such patients can ultimately help provide better health outcomes. Objective Our study investigated the performance of predictive models to forecast HbA1c elevation levels by employing several machine learning models. We also examined the use of patient electronic health record longitudinal data in the performance of the predictive models. Explainable methods were employed to interpret the decisions made by the black box models. Methods This study employed multiple logistic regression, random forest, support vector machine, and logistic regression models, as well as a deep learning model (multilayer perceptron) to classify patients with normal (<5.7%) and elevated (≥5.7%) levels of HbA1c. We also integrated current visit data with historical (longitudinal) data from previous visits. Explainable machine learning methods were used to interrogate the models and provide an understanding of the reasons behind the decisions made by the models. All models were trained and tested using a large data set from Saudi Arabia with 18,844 unique patient records. Results The machine learning models achieved promising results for predicting current HbA1c elevation risk. When coupled with longitudinal data, the machine learning models outperformed the multiple logistic regression model used in the comparative study. The multilayer perceptron model achieved an accuracy of 83.22% for the area under receiver operating characteristic curve when used with historical data. All models showed a close level of agreement on the contribution of random blood sugar and age variables with and without longitudinal data. Conclusions This study shows that machine learning models can provide promising results for the task of predicting current HbA1c levels (≥5.7% or less). Using patients’ longitudinal data improved the performance and affected the relative importance for the predictors used. The models showed results that are consistent with comparable studies.

Using model explanations to guide deep learning models towards consistent explanations for EHR data

“…The second EHR dataset we use is KAIMRC: a private EHR dataset collected from King Abdulaziz Medical City located in the central and western regions of Saudi Arabia 37 . The dataset spans 2016 to 2018, and includes patient demographics (e.g.…”

“…cholesterol levels) and vital signs during this period. For a detailed description of the features included in the dataset, and their clinical relevance, we refer the reader to 37 . The dataset was collected to aid the development of ML models for diabetes prediction.…”

“…Similarly, the KAIMRC regression task is used to verify our proposed methods work on regression as well as classification. We follow the methods presented in 37 to create our MLP models. The KAIMRC classification MLP uses 3 hidden layers of sizes 48, 48, and 24 respectively, using ReLU activation functions after each hidden layer and Sigmoid on the output.…”

Personalising monitoring for chemotherapy patients through predicting deterioration in renal and hepatic function

“… 9 We chose to develop a MLP model rather than a prognostic multivariable logistic regression model 10 due to the multiple levels within our dataset, limiting the performance of statistical models. 11 MLPs have been shown to outperform multivariable regression models in similar applications 12 and so were chosen in order to maximise the performance of the model. Additionally, recent developments in the field of explainable machine learning would allow for clinical interpretation.…”