Application of machine learning methods for the prediction of true fasting status in patients performing blood tests.

The fasting blood glucose (FBG) values extracted from electronic medical records (EMR) are assumed valid in existing research, which may cause diagnostic bias due to misclassification of fasting status. We proposed a machine learning (ML) algorithm to predict the fasting status of blood samples. This cross-sectional study was conducted using the EMR of a medical center from 2003 to 2018 and a total of 2,196,833 ontological FBGs from the outpatient service were enrolled. The theoretical true fasting status are identified by comparing the values of ontological FBG with average glucose levels derived from concomitant tested HbA1c based on multi-criteria. In addition to multiple logistic regression, we extracted 67 features to predict the fasting status by eXtreme Gradient Boosting (XGBoost). The discrimination and calibration of the prediction models were also assessed. Real-world performance was gauged by the prevalence of ineffective glucose measurement (IGM). Of the 784,340 ontologically labeled fasting samples, 77.1% were considered theoretical FBGs. The median (IQR) glucose and HbA1c level of ontological and theoretical fasting samples in patients without diabetes mellitus (DM) were 94.0 (87.0, 102.0) mg/dL and 5.6 (5.4, 5.9)%, and 92.0 (86.0, 99.0) mg/dL and 5.6 (5.4, 5.9)%, respectively. The XGBoost showed comparable calibration and AUROC of 0.887 than that of 0.868 in multiple logistic regression in the parsimonious approach and identified important predictors of glucose level, home-to-hospital distance, age, and concomitantly serum creatinine and lipid testing. The prevalence of IGM dropped from 27.8% based on ontological FBGs to 0.48% by using algorithm-verified FBGs. The proposed ML algorithm or multiple logistic regression model aids in verification of the fasting status.

Prediction of non-responsiveness to pre-dialysis care program in patients with chronic kidney disease: a retrospective cohort analysis.

The responsiveness of patients with chronic kidney disease (CKD) to nephrologists' care is unpredictable. We defined the longitudinal stages (LSs) 1-5 of estimated glomerular filtration rate (eGFR) by group-based trajectory modeling for repeated eGFR measurements of 7135 patients with CKD aged 20-90 years from a 13-year pre-end-stage renal disease (ESRD) care registry. Patients were considered nonresponsive to the pre-dialysis care if they had a more advanced eGFR LS compared with the baseline. Conversely, those with improved or stable eGFR LS were considered responsive. The proportion of patients with CKD stage progression increased with the increase in the baseline CKD stage (stages 1-2: 29.2%; stage 4: 45.8%). The adjusted times to ESRD and all-cause mortality in patients with eGFR LS-5 were 92% (95% confidence interval [CI] 86-96%) and 57% (95% CI 48-65%) shorter, respectively, than in patients with eGFR LS-3A. Among patients with baseline CKD stages 3 and 4, the adjusted times to ESRD and all-cause death in the nonresponsive patients were 39% (95% CI 33-44%) and 20% (95% CI 14-26%) shorter, respectively, than in the responsive patients. Our proposed Renal Care Responsiveness Prediction (RCRP) model performed significantly better than the conventional Kidney Failure Risk Equation in discrimination, calibration, and net benefit according to decision curve analysis. Non-responsiveness to nephrologists' care is associated with rapid progression to ESRD and all-cause mortality. The RCRP model improves early identification of responsiveness based on variables collected during enrollment in a pre-ESRD program. Urgent attention should be given to characterize the underlying heterogeneous responsiveness to pre-dialysis care.