Cohort-Derived Machine Learning Models for Individual Prediction of Chronic Kidney Disease in People Living With Human Immunodeficiency Virus: A Prospective Multicenter Cohort Study. (9th May 2020)

• Record Type: Journal Article
• Authors: Roth, Jan A; Radevski, Gorjan; Marzolini, Catia; Rauch, Andri; Günthard, Huldrych F; Kouyos, Roger D; Fux, Christoph A; Scherrer, Alexandra U; Calmy, Alexandra; Cavassini, Matthias; Kahlert, Christian R; Bernasconi, Enos; Bogojeska, Jasmina; Battegay, Manuel
• Is Part Of: Journal of infectious diseases. Volume 224:Number 7(2021)
• Page Start: 1198
• Page End: 1208
• Publication Date: 2020-05-09
• Subjects: chronic kidney disease -- digital epidemiology -- HIV -- machine learning -- prediction; Communicable diseases -- Periodicals; Diseases -- Causes and theories of causation -- Periodicals; Medicine -- Periodicals; Communicable Diseases -- Periodicals; Electronic journals; 616.9
• DOI: 10.1093/infdis/jiaa236
• Languages: English
• ISSNs: 0022-1899

Abstract

Abstract: Background: It is unclear whether data-driven machine learning models, which are trained on large epidemiological cohorts, may improve prediction of comorbidities in people living with human immunodeficiency virus (HIV). Methods: In this proof-of-concept study, we included people living with HIV in the prospective Swiss HIV Cohort Study with a first estimated glomerular filtration rate (eGFR) >60 mL/minute/1.73 m 2 after 1 January 2002. Our primary outcome was chronic kidney disease (CKD)—defined as confirmed decrease in eGFR ≤60 mL/minute/1.73 m 2 over 3 months apart. We split the cohort data into a training set (80%), validation set (10%), and test set (10%), stratified for CKD status and follow-up length. Results: Of 12 761 eligible individuals (median baseline eGFR, 103 mL/minute/1.73 m 2 ), 1192 (9%) developed a CKD after a median of 8 years. We used 64 static and 502 time-changing variables: Across prediction horizons and algorithms and in contrast to expert-based standard models, most machine learning models achieved state-of-the-art predictive performances with areas under the receiver operating characteristic curve and precision recall curve ranging from 0.926 to 0.996 and from 0.631 to 0.956, respectively. Conclusions: In people living with HIV, we observed state-of-the-art performances in forecasting individual CKD onsets with different machine learning algorithms. Abstract : In people living with HIV who participate in the Swiss HIV Cohort Study, we observed state-of-the-art performances in forecasting individual onsets of chronic kidney disease with different machine learning algorithms.