Deep Learning Model for Predicting Intradialytic Hypotension Without Privacy Infringement: A Retrospective Two-Center Study.

Objective: Previously developed Intradialytic hypotension (IDH) prediction models utilize clinical variables with potential privacy protection issues. We developed an IDH prediction model using minimal variables, without the risk of privacy infringement. Methods: Unidentifiable data from 63,640 hemodialysis sessions (26,746 of 79 patients for internal validation, 36,894 of 255 patients for external validation) from two Korean hospital hemodialysis databases were finally analyzed, using three IDH definitions: (1) systolic blood pressure (SBP) nadir <90 mmHg (Nadir90); (2) SBP decrease ≥20 mmHg from baseline (Fall20); and (3) SBP decrease ≥20 mmHg and/or mean arterial pressure decrease ≥10 mmHg (Fall20/MAP10). The developed models use 30 min information to predict an IDH event in the following 10 min window. Area under the receiver operating characteristic curves (AUROCs) and precision-recall curves were used to compare machine learning and deep learning models by logistic regression, XGBoost, and convolutional neural networks. Results: Among 344,714 segments, 9,154 (2.7%), 134,988 (39.2%), and 149,674 (43.4%) IDH events occurred according to three different IDH definitions (Nadir90, Fall20, and Fall20/MAP10, respectively). Compared with models including logistic regression, random forest, and XGBoost, the deep learning model achieved the best performance in predicting IDH (AUROCs: Nadir90, 0.905; Fall20, 0.864; Fall20/MAP10, 0.863) only using measurements from hemodialysis machine during dialysis session. Conclusions: The deep learning model performed well only using monitoring measurement of hemodialysis machine in predicting IDH without any personal information that could risk privacy infringement.

Comparison of early and late Pneumocystis jirovecii Pneumonia in kidney transplant patients: the Korean Organ Transplantation Registry (KOTRY) Study.

Late Pneumocystis jirovecii pneumonia (PJP) is not rare in the era of universal prophylaxis after kidney transplantation. We aimed to determine the nationwide status of PJP prophylaxis in Korea and compare the incidence, risk factors, and outcomes of early and late PJP using data from the Korean Organ Transplantation Registry (KOTRY), a nationwide Korean transplant cohort. We conducted a retrospective analysis using data of 4,839 kidney transplant patients from KOTRY between 2014 and 2018, excluding patients who received multi-organ transplantation or were under 18 years old. Cox regression analysis was performed to determine risk factors for early and late PJP. A total of 50 patients developed PJP. The number of patients who developed PJP was same between onset before 6 months and onsets after 6 months. There were no differences in the rate, duration, or dose of PJP prophylaxis between early and late PJP. Desensitization, higher tacrolimus dose at discharge, and acute rejection were associated with early PJP. In late PJP, old age as well as acute rejection were significant risk factors. In conclusion late PJP is as common and risky as early PJP and requires individualized risk-based prophylaxis, such as prolonged prophylaxis for old patients with a history of rejection.

Nationwide Implementation of Nonpharmaceutical Interventions During the Coronavirus Disease 2019 Pandemic Is Associated With Decreased Incidence of Pneumocystis jirovecii Pneumonia in Kidney Transplant Recipients.

This study aimed to investigate the impact of nationwide nonpharmaceutical interventions against coronavirus disease 2019 (COVID-19) on the incidence of Pneumocystis jirovecii pneumonia (PCP) in kidney transplant recipients. The monthly incidence of PCP during the COVID-19 period decreased significantly compared to that of the pre-COVID-19 period in kidney transplant recipients.

Association between Body Mass Index, Waist Circumference and Prevalence of Microalbuminuria in Korean Adults of Age 30 Years and Older without Diabetes, Hypertension, Renal Failure, or Overt Proteinuria: The 2013 Korean National Health and Nutrition Examination Survey.

BACKGROUND: Microalbuminuria and obesity markers are known risk factors for cardiovascular or renal disease. This study aimed to evaluate the prevalence of microalbuminuria according to body mass index (BMI) and abdominal obesity criteria. METHODS: The study subjects included 3,979 individuals aged 30 years or older who did not have diabetes, hypertension, renal failure, or overt proteinuria, from among those who participated in The Korean National Health and Nutrition Examination Survey in 2013, a cross-sectional, nationally representative, stratified survey. Microalbuminuria was defined as a urinary albumin to creatinine ratio of 30 to 300 mg/g. BMI and waist circumference were classified according to the Asia-Pacific criteria. RESULTS: The prevalence of microalbuminuria was found to be 5.1%. In the normoalbuminuria group, 3.4%, 41.7%, 24%, 27.6%, and 3.2% of participants were included in the underweight, normal, overweight, obesity 1, and obesity 2 groups, respectively. These percentages in the microalbuminuria group were 7.1%, 34.5%, 19.2%, 28.6%, and 10.6%, respectively (P<0.001). The waist circumference in men was 21.4% in the normoalbuminuria group and 36.5% in the microalbuminuria group (P=0.004). Logistic regression analyses were performed to evaluate the relationship between the presence of microalbuminuria and BMI or waist circumference groups. The risk of microalbuminuria was significant only in the underweight group (odds ratio, 13.22; 95% confidence interval, 2.55-68.63; P=0.002) after adjusting for confounding factors, abdominal obesity was not significantly associated with microalbuminuria. CONCLUSION: The prevalence of microalbuminuria in a general population in Korea was associated with underweight in men and was not associated with waist circumference in either men or women.