Predicting the 14-Day Hospital Readmission of Patients with Pneumonia Using Artificial Neural Networks (ANN).

Unplanned patient readmission (UPRA) is frequent and costly in healthcare settings. No indicators during hospitalization have been suggested to clinicians as useful for identifying patients at high risk of UPRA. This study aimed to create a prediction model for the early detection of 14-day UPRA of patients with pneumonia. We downloaded the data of patients with pneumonia as the primary disease (e.g., ICD-10:J12*-J18*) at three hospitals in Taiwan from 2016 to 2018. A total of 21,892 cases (1208 (6%) for UPRA) were collected. Two models, namely, artificial neural network (ANN) and convolutional neural network (CNN), were compared using the training (n = 15,324; ≅70%) and test (n = 6568; ≅30%) sets to verify the model accuracy. An app was developed for the prediction and classification of UPRA. We observed that (i) the 17 feature variables extracted in this study yielded a high area under the receiver operating characteristic curve of 0.75 using the ANN model and that (ii) the ANN exhibited better AUC (0.73) than the CNN (0.50), and (iii) a ready and available app for predicting UHA was developed. The app could help clinicians predict UPRA of patients with pneumonia at an early stage and enable them to formulate preparedness plans near or after patient discharge from hospitalization.

Overused laboratory test of haemoglobin A1c at one medical centre in southern Taiwan.

AIMS: To investigate overused laboratory test of haemoglobin Ac1 (HbA1c) in one medicine centre in southern Taiwan. METHODS: Data were extracted from the database of the Medical Center from March 2013 to March 2015. These patients were classified into five groups, including group A (diabetic patients with HbA1c value ≥7), group B (healthy people with HbA1c value ≥7), group C (diabetic patients with HbA1c test value <7), group D (healthy people with HbA1c value <6.5) and group E (prediabetic people with HbA1c value 6.5-7). The divisions requested for HbA1c test were divided into four categories, including endocrinology, internal medicine, surgery and the others. Repeat testing at the time of the second test was investigated using survival analysis. RESULTS: The percentage of overall inappropriate repeat testing was as high as 34%. The percentages among the five patient groups were relatively different. Group C had the largest percentage of inappropriate repeat testing (48%) and group A had the second largest (30%), followed by groups D (25%), E (13%) and B (10%). The percentages of inappropriate repeat testing of the five patient groups were also relatively different among the four categories of division, with Kaplan-Meier curves showing significant differences. The time to repeat testing was the shortest for group A and was the second shortest for group C, followed by groups B, E and D. CONCLUSIONS: The results provided detailed information about the percentages of inappropriate repeat testing of HbA1c of the five patient groups among the four categories of division.