SUBENDOCARDIAL VIABILITY RATIO PREDICTS MORTALITY IN OLDER NURSING HOME RESIDENTS

Objective: The subendocardial viability ratio (SEVR), estimate by arterial tonometry, can quantify the degree of myocardial perfusion relative to left-ventricular workload. This study aims to evaluate the ability of SEVR to predict long-term mortality in the older population, in the PARTAGE (Predictive values of blood pressure (BP) and ARTerial stiffness in institutionalized very AGEd population) study. Additionally, we aimed to identify the SEVR cutoff value best predicting mortality. Design and method: Patients were enrolled from the PARTAGE, a multicenter, longitudinal study aimed at determining the predictive value of blood pressure and arterial mechanical parameters in a large population of individuals aged 80 years and over living in nursing homes. Data from central blood pressure derived from arterial tonometry (PulsePen, DiaTecne, Milan) were used to elaborate the SEVR. The mortality over a 10 years follow-up was analyzed according to SEVR, using Kaplan–Meier curves and Cox regression multivariate models. Results: A total of 828 people were enrolled (mean age: 87.7±4.7 years, 78% female). 735 patients died within 10 years and 24 were lost to follow-up. SEVR was inversely associated with mortality at univariate Cox-regression model (risk ratio, 0.683 per unit increase in SEVR; 95% confidence interval (CI) [0.502–0.930], p = 0.015) and in a model including age, sex, body mass index, Activity of Daily Living index and Mini-Mental State Examination score (risk ratio, 0.647; 95% CI [0.472–0.930]). The lowest tertile of SEVR was associated with higher 10-years total mortality than the middle (p < 0.001) and the highest (p < 0.004) tertile. A SEVR cutoff value of 83% was identified as the best predictor of total mortality. Conclusions: SEVR is significantly and independently associated with mortality among nursing home residents aged 80 years and over, with a proposed cutoff value of 83%. An accurate non-invasive estimation of SEVR could be a useful and independent parameter to assess survival probability in very old adults.