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: PO0536

Development and Internal Validation of a Mortality Risk Prediction Model in Older Adults with Advanced Non-Dialysis-Dependent (NDD) CKD

Session Information

Category: CKD (Non-Dialysis)

  • 2102 CKD (Non-Dialysis): Clinical, Outcomes, and Trials

Authors

  • Li, David Yujie, Vanderbilt University Medical Center, Nashville, Tennessee, United States
  • Prigmore, Heather Leanne, Vanderbilt University, Nashville, Tennessee, United States
  • Stewart, Thomas G., Vanderbilt University, Nashville, Tennessee, United States
  • Abdel-Kader, Khaled, Vanderbilt University Medical Center, Nashville, Tennessee, United States
Background

Older adults with CKD expect practitioners to share prognostic estimates to inform decision-making regarding future care. The availability of useful mortality prediction models in NDD-CKD could reduce prognostic uncertainty and aid in identifying patients who would benefit from advance care planning (independent of dialysis initiation).

Methods

699 patients with NDD-CKD stages 4-5 and age >60 were enrolled and followed between 2014 and 2019. Cox proportional hazards regression was used to model the risk of 1-year mortality. Candidate predictor variables included age, gender, race, Charlson Comorbidity Index (CCI), common labs and the provider’s response to the Surprise Question (“Would you be surprised if this patient died in the next 12 months?”, SQ, recorded using binary and 5-point Likert response scales). Optimism-corrected measures of model performance were calculated with bootstrap resampling. Model calibration was assessed visually.

Results

In the derivation cohort, age, CCI, hemoglobin values and the provider’s Likert scale response to the SQ were predictive of 1-year mortality (Table 1). The C-statistic in the derivation sample was 0.76 and the optimism corrected C-statistic obtained by bootstrap resampling was 0.73. Visual examination of model calibration demonstrated good calibration.

Conclusion

A 1-year mortality risk prediction model in older adults with advanced NDD-CKD performed reasonably well and was well calibrated. Studies are needed to understand how to best leverage information on mortality risk to enhance patient-provider communication and ensure that future care delivered to patients is aligned with their priorities.

1-year Mortality Hazard Ratios
 Hazard Ratio (95% CI)
Age (per 12 yr increase)1.57 (1.18 - 2.09)
CCI (per 2-point increase)1.37 (1.12 - 1.67)
Hemoglobin (per 2.3 g/dl increase)0.80 (0.65 - 0.99)
SQ - not at all surprised: neutral2.06 (1.27 - 3.33)
Not surprised: neutral1.12 (0.71 - 1.75)
Surprised: neutral0.55 (0.33 - 0.92)
Very surprised: neutral0.48 (0.28 - 0.82)