Me Or Your Own Eyes: RNA-Seq and the Kidney
| العنوان: | Me Or Your Own Eyes: RNA-Seq and the Kidney |
|---|---|
| المؤلفون: | R. J. Kossmann, Adrian Guinsburg, Jeffrey Hymes, M. I. D. Bessone, J. W. Larkin, Len A. Usvyat, J. C. Berbessi, Caitlin Monaghan, Franklin W. Maddux, Yue Jiao, M. A. Kraus, Peter Kotanko |
| المصدر: | J Am Soc Nephrol |
| بيانات النشر: | Ovid Technologies (Wolters Kluwer Health), 2021. |
| سنة النشر: | 2021 |
| مصطلحات موضوعية: | Coronavirus disease 2019 (COVID-19), Demographics, business.industry, medicine.medical_treatment, General Medicine, medicine.disease, Dialysis patients, Blood pressure, Nephrology, Up Front Matters, Diabetes mellitus, medicine, Hemodialysis, Presentation (obstetrics), medicine.symptom, business, Weight gain, Demography |
| الوصف: | Background: We aimed to build machine learning (ML) models to understand the predictors of short-and longer-term mortality among hemodialysis (HD) patients affected by COVID-19 in four countries in the Americas. Methods: We used data from adult HD patients treated at regional institutions of a global provider in Latin (LATAM) & North America (NA) who had COVID-19. We used data on 96 variables from Jul-2019 through Dec-2020 to develop XGBoost models (60%:20%:20% random split for training, validation, & testing) to predict the likelihood of death in 0-14, 15-30, >30 days after COVID-19 presentation, and identify importance of predictors. Models were developed in a side-by-side manner and used same programing for datasets in LATAM (Argentina, Columbia, Ecuador) & NA (United States) countries. Results: Among HD patients with COVID-19 in LATAM (n=12,121) and NA (n=21,624), 15.8% and 7.3% died within 0-14 days, 8.2% and 4.6% died within 15-30 days, and 4.8% and 8.6% died >30 days after presentation, respectively. Models in LATAM & NA had area under cure (AUC) in testing datasets of 0.64 & 0.70 for death within 0-14 days;top predictors at presentation were diabetes, lower interdialytic weight gain (IDWG) in LATAM, and higher age and longer vintage in NA. AUCs were similar across models. Top predictors of death 15-30 days were higher pre-HD weight and post-HD systolic blood pressure (SBP) in LATAM, and same as 0-14 days in NA. Top predictors after >30 days were diabetes and higher pre-HD SBP in LATAM, and higher age and lower dry weight in NA. Conclusions: Profiles of mortality in HD patients after COVID-19 were distinct in LATAM & NA. Mortality more often occurred within 0-14 or 15-30 days after COVID-19 in LATAM versus NA. About 5% to 9% of COVID-19 patients died >30 days after presentation. Top two predictors of mortality differed in LATAM & NA during earlier and later periods after COVID-19, albeit when considering top 15 predictors similarities exist. Comorbidities, demographics, weight, and BP appear risk factors for death after presentation. Use of underexplored follow-up timeframes along with ML modeling techniques that account for collinearity and missingness provide novel insights related to mortality in COVID-19. |
| تدمد: | 1533-3450 1046-6673 |
| URL الوصول: | https://explore.openaire.eu/search/publication?articleId=doi_dedup___::01d34ea962c95c510925b2c20b3f1052 https://doi.org/10.1681/asn.2021010017 |
| حقوق: | OPEN |
| رقم الأكسشن: | edsair.doi.dedup.....01d34ea962c95c510925b2c20b3f1052 |
| قاعدة البيانات: | OpenAIRE |
| تدمد: | 15333450 10466673 |
|---|