Intra-Session recurrence in intradialytic hypotension prediction: evaluation implications and recurrence-aware modeling - Scorecard - MDSpire

Intra-Session recurrence in intradialytic hypotension prediction: evaluation implications and recurrence-aware modeling

  • By

  • Siun Kim

  • Jiwon Ryu

  • Sejoong Kim

  • Su Hwan Kim

  • Myeongju Kim

  • Hyung-Jin Yoon

  • July 7, 2026

  • 0 min

Share

Clinical Scorecard: Predicting Intra-Session Recurrence of Intradialytic Hypotension: Implications for Evaluation and Modeling with Recurrence Awareness

At a Glance

CategoryDetail
ConditionIntradialytic Hypotension (IDH)
Key MechanismsRecurrence patterns of IDH events influence predictive model performance.
Target PopulationPatients with end-stage renal disease undergoing hemodialysis.
Care SettingHemodialysis monitoring program.

Key Highlights

  • Study analyzed 12,767 hemodialysis sessions from 66 patients.
  • Recurrent IDH defined as events occurring ≥30 min after the initial IDH.
  • Incorporating recurrence information improved AUROC by up to 8.0 percentage points.
  • Conventional evaluation methods overestimated model performance for IDH predictions.
  • Standardized evaluation protocols for IDH prediction models are recommended.

Guideline-Based Recommendations

Diagnosis

  • IDH can be defined using various criteria including nadir blood pressure levels and changes from baseline.

Management

  • Real-time prediction models should incorporate recurrence patterns for better accuracy.

Monitoring & Follow-up

  • BP measurements should be taken more frequently during dialysis to improve IDH detection.

Risks

  • Frequent IDH occurrence can impair quality of life and increase mortality risk.

Patient & Prescribing Data

Outpatients aged 18 years and above with end-stage renal disease.

Patients receiving maintenance hemodialysis for at least 3 months.

Clinical Best Practices

  • Utilize deep learning models that account for IDH recurrence in predictions.
  • Implement standardized definitions of IDH to facilitate research comparisons.
  • Adopt adversarial training to enhance model robustness across patient subgroups.

Related Resources & Content

Original Source(s)

Related Content