Use of Machine Learning to Differentiate Children With Kawasaki Disease From Other Febrile Children in a Pediatric Emergency Department

Abstract

ImportanceEarly awareness of Kawasaki disease (KD) helps physicians administer appropriate therapy to prevent acquired heart disease in children. However, diagnosing KD is challenging and relies largely on subjective diagnosis criteria.ObjectiveTo develop a prediction model using machine learning with objective parameters to differentiate children with KD from other febrile children.Design, Setting, and ParticipantsThis diagnostic study included 74 641 febrile children younger than 5 years who were recruited from 4 hospitals, including 2 medical centers and 2 regional hospitals, between January 1, 2010, and December 31, 2019. Statistical analysis was performed from October 2021 to February 2023.Main Outcomes and MeasuresDemographic data and laboratory values from electronic medical records, including complete blood cell count with differential, urinalysis, and biochemistry, were collected as possible parameters. The primary outcome was whether the febrile children fulfilled the diagnostic criteria of KD. The supervised eXtreme Gradient Boosting (XGBoost) machine learning method was applied to establish a prediction model. The confusion matrix and likelihood ratio were used to evaluate the performance of the prediction model.ResultsThis study included a total of 1142 patients with KD (mean [SD] age, 1.1 [0.8] years; 687 male patients [60.2%]) and 73 499 febrile children (mean [SD] age, 1.6 [1.4] years; 41 465 male patients [56.4%]) comprising the control group. The KD group was predominantly male (odds ratio, 1.79; 95% CI, 1.55-2.06) with younger age (mean difference, −0.6 years [95% CI, −0.6 to −0.5 years]) compared with the control group. The prediction model’s best performance in the testing set was able to achieve 92.5% sensitivity, 97.3% specificity, 34.5% positive predictive value, 99.9% negative predictive value, and a positive likelihood ratio of 34.0, which indicates outstanding performance. The area under the receiver operating characteristic curve of the prediction model was 0.980 (95% CI, 0.974-0.987).Conclusions and RelevanceThis diagnostic study suggests that results of objective laboratory tests had the potential to be predictors of KD. Furthermore, these findings suggested that machine learning with XGBoost can help physicians differentiate children with KD from other febrile children in pediatric emergency departments with excellent sensitivity, specificity, and accuracy.

Hemoglobin and Its Z Score Reference Intervals in Febrile Children: A Cohort Study of 98,572 Febrile Children

“…We have also observed lower Hb and higher CRP levels in children with bacteremia [7], urinary tract infection [29], or Kawasaki disease in our previous studies [24,28]. The combination of Hb and CRP can be an effective diagnostic approach for these cases [24,30].…”

“…The World Health Organization (WHO) classifies anemia in children under 5 years of age as a condition where the Hb concentration is below 11 g/dL without arbitrary references [ 3 ]. Anemia or decreased Hb levels can be observed in febrile children and may indicate various underlying causes, including infectious or inflammatory conditions [ 23 , 24 ]. These low Hb levels are a non-specific indicator that warrants further investigation to identify the precise cause, enabling appropriate diagnosis and treatment.…”

“…As a result, anemia associated with inflammation can occur due to a combination of hepcidin-induced hypoferremia and cytokine-mediated inhibition of erythropoiesis, leading to a shortened lifespan of erythrocytes [ 23 , 26 , 27 ]. Monitoring Hb levels in febrile children can provide valuable insights into the severity of the illness and guide appropriate management strategies [ 7 , 24 , 28 ]. We have also observed lower Hb and higher CRP levels in children with bacteremia [ 7 ], urinary tract infection [ 29 ], or Kawasaki disease in our previous studies [ 24 , 28 ].…”

Utilizing Multimodal Data for Diagnosis of Kawasaki Disease: An AI Approach

“…Due to the generation of large volumes of diverse types of data in clinical practice, multimodal deep learning models have been widely applied and have seen vigorous development in the medical eld (8-11). In the eld of assisting the diagnosis of Kawasaki disease, existing research has mainly focused on developing single-modal models using either laboratory examination indices or clinical symptom images alone for identifying and aiding in the diagnosis of Kawasaki disease patients (12)(13)(14)(15)(16)(17). However, these models exhibit poor generalization, as relying solely on one clinical data type cannot fully diagnose Kawasaki disease; comprehensive assessments involving multiple types of data are necessary to make informed judgments.…”

Contextualizing Depression in Pacific Islander Sexual and Gender Minority Youth—Location, History, and Culture

“…We applaud the authors for investigating a vital intersection of pediatric, sexual and gender minority, and mental health. They implemented the χ 2 automatic interaction detection technique—a decisional tree approach developed in the 1980s that was foundational to modern-day gradient boosting machine learning techniques used in health care research . Using surveys from California and Minnesota, they found depressive symptoms among the aggregate grouping of Filipinx, Korean, and Japanese youth.…”