Predictive value of the new Hemodynamic Index
During the ramp test, the final step at 6400 RPM was obtained in 3 (10%) patients in the RVF - group. No patient in the second group tolerated a pump speed over 6100 RPM. In 30 patients (78%) the test was stopped because frequent suction events occurred. In the other 8 patients the test was interrupted when a significant leftward shift of the SIV were diagnosed and an LVEDD less than 3 cm was echocardiographically assessed. Once speed optimization was obtained, the final increase in PS resulted significant for both groups (Table 2). Only 2 patients in the RVF + group and 5 patients in the RVF group – showed residual pulsatility. Given the small number of subjects, the level of pulsatility was not considered in our inferential analysis.
Hemodynamic variables, before and after a ramp test for patients with and patients without RVF after LVAD implantation were compared (Table 2). As shown in the table, a significant increase in pump flow (PF) and pump speed (PS) was observed in both the RVF+ group and the RVF- group. A significant reduction in the PCWP was also detected. In the RVF – group, we observed an increase in the cardiac output (CO) and a significant reduction in CVP. There were no statistical differences for the two groups between values estimated before and after the ramp test in terms of PI, MAP, SPAP, DPAP, MPAP, PVRI, SVRI, RVSWI.
Using the parameters obtained from catheterization, the hemodynamic index was calculated, as already mentioned above, before and after ramp tests with the following results:
Group RVF-: The Hemodynamic Index calculated before and after the ramp resulted respectively 96,10 ± 26,24 and 80,10 ± 13,45 (p=0,010).
Group RVF+: HI calculated pre ramp test (Fig 1) resulted 87,9 ± 19,5. In this group, the HI decreased significantly (p=<0,001) to 51,8 ± 5,3 after speed optimization.
The values of the different hemodynamic variables obtained after the ramp test, including the HI, were then analyzed in univariate analysis to evaluate the predictive effect on the right failure. Factors with a statistically significant difference between patients with and patients without RVF in the univariate analysis were considered for multivariate analysis. Multivariate analysis revealed that a low value of HI was independent risk factors for late RVF after LVAD implantation (Table 3). Other factors failed to show significant discriminatory capacity. In the resulting ROC curves, an HI of 55 provided a sensitivity of 83,2% and a specificity of 85% (Fig 2) for predicting late RVF. To examine the association between the Hemodynamic Index and outcomes, we divided patients into tertiles of the Hemodynamic Index (lowest tertile = HI < 50, middle tertile = HI between 50-80, highest tertile = HI > 80). The patients in the first group showed a worse outcome in terms of freedom from right ventricular failure when compared to the other two groups (Fig 3).