Developing a Machine Learning System for Identification of Severe Hand, Foot, and Mouth Disease from Electronic Medical Record Data

Abstract

Children of severe hand, foot, and mouth disease (HFMD) often present with same clinical features as those of mild HFMD during the early stage, yet later deteriorate rapidly with a fulminant disease course. Our goal was to: (1) develop a machine learning system to automatically identify cases with high risk of severe HFMD at the time of admission; (2) compare the effectiveness of the new system with the existing risk scoring system. Data on 2,532 HFMD children admitted between March 2012 and July 2015, were collected retrospectively from a medical center in China. By applying a holdout strategy and a 10-fold cross validation method, we developed four models with the random forest algorithm using different variable sets. The prediction system HFMD-RF based on the model of 16 variables from both the structured and unstructured data, achieved 0.824 sensitivity, 0.931 specificity, 0.916 accuracy, and 0.916 area under the curve in the independent test set. Most remarkably, HFMD-RF offers significant gains with respect to the commonly used pediatric critical illness score in clinical practice. As all the selected risk factors can be easily obtained, HFMD-RF might prove to be useful for reductions in mortality and complications of severe HFMD.

Development of Prediction Models Using Machine Learning Algorithms for Girls with Suspected Central Precocious Puberty: Retrospective Study

“…It should be noted that different combinations of variables may produce models with similar predictive accuracy, relating to the uncertainty analysis of the solutions in any decision-making problem [35-37]. This is not rare in machine learning models in medicine [38-40]. Moreover, the most important features in the clinic such as basal LH, IGF-I, and FSH levels were all sorted out by both models, demonstrating that they are both reliable and effective in predicting response to the GnRHa test.…”

Identifying risk factors for neurological complications and monitoring long-term neurological sequelae: protocol for the Guangzhou prospective cohort study on hand-foot-and-mouth disease

“…This is a particularly important concern considering the fact that most HFMDs develop neurological complications only 1 day after admission. 15 A number of single nucleotide polymorphisms have been identified as risk factors for severe EV-A71 infection, 25–27 but polymorphism testing may take several days. With the detailed clinical and biological information it collects, the GPCS-HFMD study is expected to discover some meaningful and applicable early biomarkers that can contribute to early prediction and prevention of disease deterioration.…”

“…12–14 Based on these factors, several risk scores and algorithms have been developed for early warning purpose. 15–17 However, all these risk factors were identified retrospectively in small studies. Thus their usefulness and accordingly the value of the existing risk scores/algorithms are questionable.…”

A Simple Scoring System for Quick, Accurate, and Reliable Early Diagnosis of Hand, Foot, and Mouth Disease

“…A previous report presented a mortality risk score model comprising 4 laboratory parameters with good discrimination (AUC >0.9) [16]; however, the model can only discriminate children with high mortality risk from the severe HFMD cases. Another study developed a prediction system for identification of the severe HFMD based on 14 variables with an AUC of 0.916 [17], but this system can only discriminate the severe HFMD cases from the mild HFMD cases. Compared to these studies previous [16,17], our study had a different purpose – to discriminate HFMD cases from non-HFMD cases.…”