ASN's Mission

To create a world without kidney diseases, the ASN Alliance for Kidney Health elevates care by educating and informing, driving breakthroughs and innovation, and advocating for policies that create transformative changes in kidney medicine throughout the world.

learn more

Contact ASN

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

email@asn-online.org

202-640-4660

The Latest on X

Kidney Week

Please note that you are viewing an archived section from 2022 and some content may be unavailable. To unlock all content for 2022, please visit the archives.

Abstract: SA-PO386

Comparison of Performance of an Artificial Intelligence Risk Prediction Model and Surprise Question: A Prospective Study

Session Information

Category: Dialysis

  • 701 Dialysis: Hemodialysis and Frequent Dialysis

Authors

  • Chen, Huan-Sheng, An Hsin Qing Shui Clinic, Taichung, Taiwan
  • Lei, Chen-Chou, An Hsin Nan Zi Clinic, Kaohsiung, Taiwan
  • Chou, Ming-Hsien, He Tai Clinic, Taipei, Taiwan
  • Lin, Chu-Cheng, Ho Hospital, Changhua, Taiwan
  • Lu, Hsueh-Wen, An Hsin Da Se Clinic, Taoyuan, Taiwan
  • Wu, Wen-Chieh, Shin Loong Clinic, New Taipei, Taiwan
  • Lin, Tsai-Ying, An Hui Clinic, Taoyuan, Taiwan
  • Yu, Tsuan-Shih, He Yang Clinic, Taoyuan, Taiwan
  • Liou, Hung-Hsiang, Hsing-Jen Hospital, New Taipei, Taiwan
  • Cheng, Chen-Ting, Saint Paul's Hospital, Taoyuan, Taiwan
  • Paik Seong, Lim, Tungs' Taichung MetroHarbor Hospital, Taichung, Taiwan
  • Zhang, Hanjie, Renal Research Institute, New York, New York, United States
  • Tan, Weihao, Fresenius Medical Care Asia Pacific Ltd, Hong Kong, Hong Kong
  • Ho, Kakiu, Fresenius Medical Care Asia Pacific Ltd, Hong Kong, Hong Kong
  • Li, Margaret, Fresenius Medical Care Asia Pacific Ltd, Hong Kong, Hong Kong
  • So, Carmen, Fresenius Medical Care Asia Pacific Ltd, Hong Kong, Hong Kong
Background

Hemodialysis (HD) patients encounter a significantly high risk of morbidity and mortality. The surprise question (SQ) (“Would you be surprised if this patient were still alive in 6 or 12 months?”) is used as a mortality prognostication tool in HD patients. We aimed to apply a risk prediction model (RPM) and compare its performance with SQ on 6- and 12-month mortality prediction for HD patients in Taiwan.

Methods

We conducted a 1-year prospective observational study in 422 chronic HD patients from 8 dialysis clinics. Demographic, clinical, laboratory and dialysis treatment indicators (158 features) were used to model 6- and 12-month mortality probability using logistic regression. All patients were assessed by SQ upon enrolment and, subsequently, by RPM every month. The performance of RPM against SQ was evaluated independently using area under the receiver operating characteristics curve (AUC). We compared sensitivities and specificities of RPM and SQ.

Results

A total of 207 high-risk patients were identified. During the 12-month follow-up, 47 patients died. For 6-months mortality prediction, AUC for SQ and RPM were 0.56 (95%CI 0.48, 0.63) and 0.70 (95%CI 0.62, 0.79), respectively (p=0.006). For 12-month mortality prediction, AUC for SQ and RPM were 0.56 (95%CI 0.51, 0.62) and 0.73 (95%CI 0.68, 0.78) respectively (p<0.0001). Sensitivities of the 6- and 12-month RPM were 0.88 (95%CI 0.73, 1) and 0.89 (95%CI 0.81, 0.98), respectively. The sensitivities of the 6- and 12-month SQ were 0.12 (95%CI 0, 0.27) and 0.17 (95%CI 0.063, 0.28), respectively. The specificities of 6- and 12-month RPM were 0.53 (95%CI 0.48, 0.57) and 0.56 (95%CI 0.51, 0.61), respectively. The 6- and 12-month SQ specificities were 0.99 (95%CI 0.98, 1) and 0.95 (95%CI 0.93, 0.97), respectively.

Conclusion

RPM is a more reliable predictor of mortality in HD patients compared to SQ. Sensitivity of SQ was rather low in this HD patient population.

Funding

  • Commercial Support –