A simple real-time model for predicting acute kidney injury in hospitalized patients in the US: A descriptive modeling study
PLoS Medicine Aug 08, 2019
Simonov M, Ugwuowo U, Moreira E, et al. - Through a retrospective analysis of data obtained from 169,859 hospitalized adults admitted to 1 of 3 study hospitals in the United States from December 2012 to February 2016, researchers created an implementable predictive model to correctly prognosticate acute kidney injury (AKI) in hospitalized patients, which could be easily integrated within an existing electronic health record (EHR) system. The training set and the internal validation set cohort included 60,701 and 30,599 people, respectively. External validation data sets included 43,534 and 35,025 people. About 19.1% and 18.9% of people, respectively, in the training set and external validation set, developed AKI. With AUCs of 0.74, 0.77, 0.79, and 0.69, the full model, including all covariates, had a good ability to prognosticate imminent AKI for the validation set, sustained AKI, dialysis, and death, respectively. A simple model using only readily available, time-updated laboratory values had a very comparable auspicious performance to the complete model. Hence, to prognosticate imminent AKI with good differentiation, a simple model using readily available laboratory data could be created. Moreover, this model may lend itself well to integration into the EHR without compromising the performance observed in more complex models.
-
Exclusive Write-ups & Webinars by KOLs
-
Daily Quiz by specialty
-
Paid Market Research Surveys
-
Case discussions, News & Journals' summaries