RESUMEN
Wrong dose, a common prescription error, can cause serious patient harm, especially in the case of high-risk drugs like oral corticosteroids. This study aims to build a machine learning model to predict dose-related prescription modifications for oral prednisolone tablets (i.e., highly imbalanced data with very few positive cases). Prescription data were obtained from the electronic medical records at a single institute. Cluster analysis classified the clinical departments into six clusters with similar patterns of prednisolone prescription. Two patterns of training datasets were created with/without preprocessing by the SMOTE method. Five ML models (SVM, KNN, GB, RF, and BRF) and logistic regression (LR) models were constructed by Python. The model was internally validated by five-fold stratified cross-validation and was validated with a 30% holdout test dataset. Eighty-two thousand five hundred fifty-three prescribing data for prednisolone tablets containing 135 dose-corrected positive cases were obtained. In the original dataset (without SMOTE), only the BRF model showed a good performance (in test dataset, ROC-AUC:0.917, recall: 0.951). In the training dataset preprocessed by SMOTE, performance was improved on all models. The highest performance models with SMOTE were SVM (in test dataset, ROC-AUC: 0.820, recall: 0.659) and BRF (ROC-AUC: 0.814, recall: 0.634). Although the prescribing data for dose-related collection are highly imbalanced, various techniques such as the following have allowed us to build high-performance prediction models: data preprocessing by SMOTE, stratified cross-validation, and BRF classifier corresponding to imbalanced data. ML is useful in complicated dose audits such as oral prednisolone. Supplementary Information: The online version contains supplementary material available at 10.1007/s41666-023-00128-3.
RESUMEN
Objectives The influential factors for anti-severe acute respiratory syndrome coronavirus 2 spike protein antibody (S-ab) levels were assessed after the administration of BNT162b2 mRNA coronavirus disease-2019 (COVID-19) vaccine at short and medium terms. Methods A total of 470 healthcare workers (118 males, mean age 41.0±11.9 years) underwent serum S-ab level measurement at 3 and 8 months after two inoculations of BNT162b2 vaccine given 3 weeks apart, who had no history of COVID-19 were enrolled in this study. The changes and differences after vaccination due to gender and adverse reactions of S-ab were analyzed. Results Systemic adverse reactions incidence (48%) was significantly higher after the second dose than after the first dose (8%). S-ab levels decreased as the age increased (from the 20s to 60s) in both measurements. S-ab level 8 months after the second inoculation [median 476.3 (interquartile range (IQR) 322.4-750.6) U/mL] was significantly lower than that after 3 months [977.5 (637.2-1,409.0) U/mL; p<0.001]. The median decrease rate of S-ab levels in 5 months was 50.3% (IQR 40.3-62.6) and those differences were not observed among all generations. Gender-associated differences in S-ab levels were not observed; however, a significant relationship between higher S-ab levels and the systemic adverse reactions was observed at both measurements. Conclusions The systemic adverse reaction is an independent factor for higher S-ab levels at short and medium terms after BNT162b2 vaccination as demonstrated in our data.
