Machine Learning-Based Prediction of Hospital Admission Among Children in an Emergency Care Center.

OBJECTIVES: Machine learning-based prediction of hospital admissions may have the potential to optimize patient disposition and improve clinical outcomes by minimizing both undertriage and overtriage in crowded emergency care. We developed and validated the predictive abilities of machine learning-based predictions of hospital admissions in a pediatric emergency care center. METHODS: A prognostic study was performed using retrospectively collected data of children younger than 16 years who visited a single pediatric emergency care center in Osaka, Japan, between August 1, 2016, and October 15, 2019. Generally, the center treated walk-in children and did not treat trauma injuries. The main outcome was hospital admission as determined by the physician. The 83 potential predictors available at presentation were selected from the following categories: demographic characteristics, triage level, physiological parameters, and symptoms. To identify predictive abilities for hospital admission, maximize the area under the precision-recall curve, and address imbalanced outcome classes, we developed the following models for the preperiod training cohort (67% of the samples) and also used them in the 1-year postperiod validation cohort (33% of the samples): (1) logistic regression, (2) support vector machine, (3) random forest, and (4) extreme gradient boosting. RESULTS: Among 88,283 children who were enrolled, the median age was 3.9 years, with 47,931 (54.3%) boys and 1985 (2.2%) requiring hospital admission. Among the models, extreme gradient boosting achieved the highest predictive abilities (eg, area under the precision-recall curve, 0.26; 95% confidence interval, 0.25-0.27; area under the receiver operating characteristic curve, 0.86; 95% confidence interval, 0.84-0.88; sensitivity, 0.77; and specificity, 0.82). With an optimal threshold, the positive and negative likelihood ratios were 4.22, and 0.28, respectively. CONCLUSIONS: Machine learning-based prediction of hospital admissions may support physicians' decision-making for hospital admissions. However, further improvements are required before implementing these models in real clinical settings.

Preventive effect of TAK-751S on complications of hemorrhagic colitis (results of clinical study of TAK-751S).

BACKGROUND: The effective therapy for hemolytic uremic syndrome and encephalopathy caused by enterohemorrhagic Escherichia coli have not been established. Great attention has been drawn to the results of the clinical study of TAK-751S, performed in Canada. In Japan, a nationwide clinical study of TAK-751S had been performed since 1997 to investigate the preventive effect on the onset of HUS and the safety. METHODS: TAK-751S was administered in daily doses of 500 mg/kg for one week to 128 pediatric patients with colitis who were suspected of enterohemorrhagic Escherichia coli (EHEC) infection. RESULTS: 1. TAK-751S was confirmed to absorb Shiga toxin (Stx) existing inside the human intestine and to excrete Stx out of the body. 2. The incidence of HUS is 5.9% (4/68) and a tendency to inhibit the onset of HUS was observed as compared with the historical control. The complications of central neuropathy such as encephalopathy were observed in 3 of these patients with HUS. 3. Mild "sweating" and "nausea" were observed. There were 13 mild non-specific abnormalities of laboratory test values in 8 patients. CONCLUSIONS: From these results, it was clarified that TAK-751S absorbed and removed free Stx in the intestinal tract of pediatric patients with EHEC infection. The test drug could not inhibit the onset of HUS completely, but since HUS occurred within 48 hours after the start of administration in 3 of the 4 patients with onset of HUS, TAK-751S is a safe drug for pediatric patients with EHEC infection in which the preventive effect on HUS and encephalopathy are expected when it can be given from an early stage of the diseases. Furthermore, these results suggest that importance of rapid diagnosis of HUS.