Using a multidimensional prognostic index (MPI) based on comprehensive geriatric assessment (CGA) to predict mortality in elderly undergoing transcatheter aortic valve implantation

Background Selection of appropriate elderly who can benefit from transcatheter aortic valve implantation (TAVI) is challenging. We evaluated the prognosis of this procedure according to the comprehensive geriatric assessment (CGA) based on the multidimensional prognostic index (MPI). Methods Prospective observational monocentric study from January 2013 to December 2015. Consecutive patients aged ≥ 75 who underwent TAVI and a complete CGA were included. Baseline demographic, geriatric and cardiologic data were collected. CGA was used to calculate the MPI score that is divided in three groups according to the mortality risk. Follow up was performed until December 2016 and mortality rate was assessed at one, six and 12 months. Results 116 patients were included. Mean age was 86.2 ± 4.2 years, mean European system for cardiac operative risk evaluation (EuroSCORE) was 19.2 ± 11.3%, mean MPI score was 0.39 ± 0.13. Forty-five (38.8%) patients belonged to MPI-1 group, 68 (58.6%) to MPI-2 group and three to MPI-3 group. MPI score and Euroscore were moderately correlated (Spearman correlation coefficient r_s = 0.27, p = 0.0035). Mortality rate was significantly different between MPI groups at six and 12 months (p = 0.040 and p = 0.022). Kaplan Meier survival estimates at one year stratified by MPI groups was significantly different (hazard ratio HR = 2.83, 95%confidence interval (CI) 1.38–5.82, p = 0.004). Among variables retained to perform logistic regression analysis, the score of instrumental activities of daily living appeared the most relevant (p < 0.001). Conclusion This study indicates that CGA based on MPI tool is accurate to predict prognosis in elderly patients undergoing TAVI procedure.