Abstract: FR-PO908
External Validation of a Short-Term Prognostic Model for Patients Who Are on Maintenance Hemodialysis
Session Information
- Dialysis: Palliative and End-of-Life Care
November 03, 2017 | Location: Hall H, Morial Convention Center
Abstract Time: 10:00 AM - 10:00 AM
Category: Dialysis
- 609 Dialysis: Palliative and End-of-Life Care
Authors
- Kittanamongkolchai, Wonngarm, Mayo Clinic, Rochester, Minnesota, United States
- Gonzalez Suarez, M. Lourdes, Mayo Clinic, Rochester, Minnesota, United States
- Gregoire, James Robert, Mayo Clinic, Rochester, Minnesota, United States
Background
Simple and accurate instruments for prognostication are needed to enable nephrologists to identify dialysis patients who have a poor prognosis and benefit from palliative care and advance care planning. Prognostic models have been developed but external validation studies are scarce. We aimed to externally validate a short-term prognostic model among our dialysis patients.
Methods
Between October 1 to October 31 2015, all of the adult HD patients (n =89) at a single dialysis facility in Rochester, MN were screened. Patient charts were reviewed for actuarial predictors (age, serum albumin, dementia and peripheral vascular disease) and nephrologists answered the “surprise” question, “Would I be surprised if this patient died within the next 6 months?”. One-year survival was calculated using the prognostic calculator tool derived from a study by Cohen et al1( H:\QI palliative\Prospective study\Prognosis calculator.htm ) with each patient’s individual covariates. Survival was monitored for 12 months. We assessed the predictive accuracy of the short-term prognostic model by calculating c-index based on the receiver operating characteristic curve.
Results
The mean predicted 1-year mortality was 19% compared with the observed mortality of 12%. The model had a C-index of 0.82 (p = 0.0001). Thirty eight percent of patients with 1-y predicted survival rate in the lowest quartile (equal or less than 66%) died compared to 3% of those with 1-Y predicted survival rate more than 66%. As seen in Figure 1, the model successfully predicted which patients had worse and better survival over time.
Conclusion
The short-term prognostic tool for dialysis patients may be of value for clinicians to improve end-of-life care by providing more accurate prognostic information.
Reference:
1. Cohen LM, et al. Clin J Am Soc Nephrol. 2010 Jan;5(1):72-9
Figure 1: Survival of patients with predicted 1-Y survival in the lowest quartile (blue line) and higher quartiles (red line)