Non-Race-Corrected eGFR Sharpens AKI Risk Prediction in Black People Non-Race-Corrected eGFR Sharpens AKI Risk Prediction in Black People

The study covered in this summary was published in medRxiv.org as a preprint and has not yet been peer reviewed.

Key Takeaways

  • The authors evaluated the impact of removing race from the estimated glomerular filtration rate (eGFR) calculation for predicting acute kidney injury (AKI) after percutaneous coronary intervention (PCI), leveraging data from American College of Cardiology (ACC) National Cardiovascular Data Registry (NCDR) CathPCI Registry.

  • Current AKI models developed from data from the NCDR that include race correction in eGFR underestimate AKI risk among Black patients while slightly overestimating risk among non-Black patients. Removal of the race correction in the analysis improved the underestimation.

Why This Matters

  • Reduced eGFR is a strong predictor for AKI after PCI, and clinicians routinely use eGFR to predict a patient’s postprocedure outcome.

  • The calculation of eGFR has traditionally depended on the patient’s race. Nephrology experts now recommend the removal of race-correction factors in eGFR calculation, acknowledging that race is a social construct and rarely significant in medicine. However, this removal will affect calculated risk estimates and clinical decisions, so current model performances must be reevaluated.

  • When the race correction is removed, the underestimation of the risk for AKI after PCI for Black patients is improved. Still, Black patients have an elevated risk for AKI, so more detailed data and better modeling techniques are needed to help quantify this risk.

Study Design

  • Using data from the ACC NCDR CathPCI registry, the authors analyzed the effect of removing race correction in eGFR calculations on two current AKI models, one based on logistic regression and the other on gradient-boosting machine learning.

  • Model performance was assessed with and without race correction, and race was included as an independent predictor in a model without race-corrected eGFR.

  • The primary outcome of post-PCI AKI was defined as change in postprocedural creatinine greater than 0.3 mg/dL or 50% from before the procedure.

  • The eGFR was calculated according to the Modification of Diet in Renal Disease (MDRD) equation, the same formula used to estimate eGFR as a predictor in the NCDR model. A race-correction factor of 1.212 was used for Black Americans in this equation, and the race variable from the NCDR data was used for the calculation.

Key Results

  • A total of 947,091 PCI procedures in 915,223 patients were assessed (mean age, 64.8 years; 7.9% Black; 32.8% female); the AKI rate was 7.4%.

  • In the NCDR model, inclusion of race correction in eGFR significantly underestimated AKI risk among Black patients (predicted 7.6% vs observed 10.2%) while slightly overestimating risk among non-Black patients (predicted 7.4% vs observed 7.1%).

  • Removing the race correction partially corrected the underestimation among Black patients (predicted 8.2%). The underestimation was further reduced (predicted 10.1%) by including race as an independent predictor and introducing interaction terms.

  • The area under the ROC curve (AUC) was similar in the two models but was consistently lower in Black patients. The machine-learning model had better calibration in Black patients (8.6% predicted with race correction vs 8.7% without) than the logistic model, but AKI risk was still underestimated. The machine-learning model achieved similarly good calibration in Black patients (predicted 10.1%) with race included as a predictor. AUCs of machine-learning models were better than the logistic models but did not differ based on inclusion of race correction.

Limitations

  • NCDR is not representative of PCI patients in non-US centers, and the performance of risk models overall and the effect of race inclusion in eGFR calculation on AKI risk may differ for these populations.

  • Risk models were based on in-hospital AKI. Several patients did not have postprocedure creatinine assessed, so AKI could have been missed in those subjects and in those who might have developed AKI after discharge. Performance of the risk models could have been compromised by missing these patients and events.

Study Disclosures

  • This study did not receive any funding.

  • The lead author and three of the secondary authors have no competing interests to declare. The remaining three secondary authors served as consultants and/or have received grants/expenses from multiple firms in the public and private sectors.

This is a summary of a preprint research study, Effect of removing Race Correction Factor in glomerular Filtration Rate Estimation on predicting Acute Kidney Injury After Percutaneous Coronary Intervention, written by Chenxi Huang from Center for Outcomes Research and Evaluation, Yale–New Haven Hospital, Connecticut, and colleagues, and provided to you by Medscape. This study has not yet been peer-reviewed. The full text of the study can be found on Medrxiv.org.