MAGGIC-Predicted Mortality of >20% at One Year is Associated with Poor Prognosis in Ambulatory Patients with Advanced Heart Failure
2020
Purpose The MAGGIC score might help select patients who would gain prognostic benefit from heart transplantation or mechanical circulatory support (MCS). We examined the association between MAGGIC predicted prognosis and actual outcomes, in order to assess predictive performance of the score and determine a threshold for identifying patients at risk. Methods We included 702 consecutive outpatients who were assessed for heart transplantation over ten years. The MAGGIC score was calculated for each patient according to the scoring matrix and patients were stratified into quartiles. The co-primary endpoints were death, urgent heart transplantation or need for MCS. Follow up was censored at routine heart transplantation. Kaplan-Meier curves were constructed and calibration testing was performed. Results 320 (45.6%) patients reached a primary endpoint during a median follow-up of 547 days. MAGGIC-predicted mortality was associated with event-free survival (figure, log rank P 20% and these patients had the worst outcomes. At this threshold, the MAGGIC score had high specificity (93.5%) but low sensitivity (20.9%), giving a positive likelihood ratio of 3.22 for death, urgent heart transplantation or the need for MCS. Conclusion MAGGIC-predicted mortality is associated with real world outcomes in ambulatory patients with advanced heart failure. Clinicians should not be reassured by low MAGGIC-predicted mortality because sensitivity is poor and event-free survival at three years is lower than predicted. However, high MAGGIC-predicted mortality (>20% at one year) is a specific marker of very poor outcomes and may be used to select patients who might derive prognostic benefit from surgical treatment of advanced heart failure, such as implantation of a left ventricular assist device.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
0
Citations
NaN
KQI