Abstract: FR-PO841

Identifying Patients at Risk for Intradialytic Hypotension

Session Information

Category: Dialysis

  • 606 Dialysis: Epidemiology, Outcomes, Clinical Trials - Cardiovascular

Authors

  • Neri, Luca, Fresenius Medical Care, Bad Homburg, Germany
  • Chermisi, Milena, Fresenius Medical Care, Bad Homburg, Germany
  • Barbieri, Carlo, Fresenius Medical Care, Bad Homburg, Germany
  • Mari, Flavio, Fresenius Medical Care, Bad Homburg, Germany
  • Stuard, Stefano, Fresenius Medical Care, Bad Homburg, Germany
Background

Despite Intradialytic Hypotension is a prominent clinical problem for patients on dialysis, there currently is no valid risk prediction tool

Methods

All FME patients registered in Spain (2013 - 2015) have been enrolled in a historical cohort. We extracted medical data from de-identified electronic clinical charts (EuClid database). We partitioned the initial dataset in a training (90%) and validation (10%) sample. We implemented a two-part regression to evaluate correlates of Intradialytic Hypotension (IDH) and derived a risk score. Stage 1 identified patients at high risk of experiencing at least 1 IDH. Stage 2 predicted IDH rate in patients identified in stage 1 as the IDH cluster. AIC minimization was used for model selection

Results

Among 7582 patients (Validation set: 758), 4346 had at least one IDH during follow up time (mean=1.58 ±1.03 years). On average there were 5.1±9.5 events per patient during the follow up (incidence: 0.023±0.036 IDH/person-treatment). Factors associated with increased IDH risk: ESRD vintage, previous IDH, variability of intradialytic body weight drop, intradialytic blood pressure drop, ultrafiltration volume, pre-dialysis blood pressure, serum potassium, extracellular water volume, calcium, ferritin, body fat, female sex, use of diuretics, diabetes, CRP, history of stroke, dialysate bicarbonate and sodium, dementia, over-hydration. Factors associated with reduced IDH risk: intradialytic blood pressure variability, dry body weight, lower intracellular water volume, intradialytic heart rate variability, urea distribution volume, hematocrit, serum potassium. The risk score had excellent calibration and accurately discriminated across different risk classes (fig1)

Conclusion

We identified potentially modifiable risk factors for IDH. Additionally, our prediction algorithm provides a reliable tool for risk stratification