Study Develops Model for Predicting Acute Kidney Injury Risk

July 27, 2019

CMS is considering hosptial-acquired acute kidney injury as a quality measure, and the study offers a model using electronic health records that could help predict imminent risk of this condition using readily available laboratory values.

Acute kidney injury (AKI) is an adverse effect resulting in increased costs and mortality; however, AKI may be predicted using a simple model implemented through an electronic health record, according to research published by PLoS Medicine.

Researchers analyzed data from 169,850 hospitalized adults who were admitted to 1 of 3 study hospitals in the United States in order to develop a model to predict AKI within the first 24 hours of observation. Mortality, AKI severity, and requirement for renal replacement therapy were the outcomes being predicted.

“Hospital-acquired AKI is being evaluated as a potential quality measure by the Centers for Medicare and Medicaid Services. AKI is diagnosed in relation to a rise in creatinine, but this marker rises late in the course of the syndrome,” noted the authors. “Real-time prediction of AKI prior to a creatinine increase holds promise to preempt such events through medication adjustment, avoiding nephrotoxins, optimizing hemodynamics, or engaging in other diagnostic or therapeutic procedures, including biomarker measurement.”

A training set cohort of 60,701 patients for the model was internally validated with a set of 30,599 patients, then externally validated against data sets of 43,534 and 35,025 patients. Of the training set, 19.1% of patients developed AKI, while 18.9% of the external validation set did.

According to the authors, the full developed model was able to successfully predict imminent AKI for the validation set, sustained AKI, dialysis, and death. The simple model, using readily available laboratory values, had a similar performance.

These results demonstrate that the predictive model could be incorporated with an electronic health record to alert healthcare providers of a patient’s risk for developing AKI. Using this information, healthcare providers will be better informed to make guided decisions on how to best support their patient and what preventative measures to take.

“Modern electronic health record systems can provide readily accessible data (eg, demographics, laboratory studies) to fuel scientific study and prediction modeling,” the researchers explained. “There have been several attempts to utilize large medical data sets for predicting which patients will develop AKI; however, none are widely implemented in clinical settings.”

The researchers concluded that future studies should assess different interventions randomized to individual patients at high risk for developing AKI in order to develop novel therapies for prevention treatment.

Reference

Simonov M, Ugwuowo U, Moreira E, et al. A simple real-time model for predicting acute kidney injury in hospitalized patients in the US: a descriptive modeling study [published online July 15, 2019]. PLoS Medicine. doi.org/10.1371/journal.pmed.1002861