ASN's Mission

ASN leads the fight to prevent, treat, and cure kidney diseases throughout the world by educating health professionals and scientists, advancing research and innovation, communicating new knowledge, and advocating for the highest quality care for patients.

learn more

Contact ASN

1401 H St, NW, Ste 900, Washington, DC 20005

email@asn-online.org

202-640-4660

The Latest on Twitter

Kidney Week

Abstract: PO0058

A Clinical Score to Predict Recovery in ESKD due to AKI

Session Information

Category: Acute Kidney Injury

  • 102 AKI: Clinical, Outcomes, and Trials

Authors

  • Shah, Silvi, University of Cincinnati, Cincinnati, Ohio, United States
  • Leonard, Anthony C., University of Cincinnati, Cincinnati, Ohio, United States
  • Harrison, Kathleen, University of Cincinnati, Cincinnati, Ohio, United States
  • Meganathan, Karthikeyan, University of Cincinnati, Cincinnati, Ohio, United States
  • Christianson, Annette, University of Cincinnati, Cincinnati, Ohio, United States
  • Kramer, Samantha M., University of Cincinnati, Cincinnati, Ohio, United States
  • Thakar, Charuhas V., University of Cincinnati, Cincinnati, Ohio, United States
Background

Acute kidney injury (AKI) is a major contributor to end-stage kidney disease (ESKD). About one-third of patients with ESKD due to AKI recover kidney function. However, there is lack of clinical models to predict kidney recovery in ESKD due to AKI.

Methods

Using data from the United States Renal Data System (2005-2014), we developed a clinical score to predict kidney recovery by 90-days post-dialysis initiation in patients with ESKD due to AKI (N=22,922). We used multivariable logistic regressions to model the effects of patient demographics, comorbidities, and laboratory measures on kidney recovery. The resulting logistic parameter estimates were transformed into integer point totals by doubling and rounding the estimates. The predictive accuracy of the score models was compared to the underlying logistic models by comparing areas under the receiver operating characteristic curves (AUROC) and internal validation was performed.

Results

In ESKD due to AKI, kidney recovery within 90-days occurred in 24% of patients. Median recovery time for patients who recovered was 2 months; 72% recovered within 90-days. In the logistic models of recovery at 90-days, older age, lower body mass index, hemoglobin < 12 gm/dl, Black and Native American race, Hispanic ethnicity, congestive heart failure, amputation, poor functional status, and pre-dialysis nephrology care were associated with a lower likelihood of recovery. Eight patient characteristics were included in the final clinical score- age, body mass index, race, congestive heart failure, amputation, functional status, and prior nephrology care. Recovery scores ranged from zero to 11, with corresponding recovery rates ranging from 6% to 86%. Three risk categories (score range of 0-5, 6-7, and 8-11) exhibited 90-day recovery rates of 11%, 23%, and 45%. The internal validation assessment showed no overfitting of the models. The AUROC of the score was 0.70, similar to the original AUROC of 0.71.

Conclusion

A simple clinical risk score derived from information available at incident dialysis can accurately predict kidney recovery at 90 days in ESKD due to AKI. This predictive tool can be utilized by dialysis providers and policymakers to individualize care, and to improve the quality and processes of care.

Funding

  • Clinical Revenue Support