Prognostic implication of a novel right ventricular injury score in septic patients

Abstract

Aims: We aim to investigate the prognostic value of a right ventricular (RV) injury score based on the concept of RV dilation, RV systolic dysfunction, and RV-pulmonary arterial (PA) decoupling in septic patients and to explore whether the RV injury (RVI) score can be used to grade the severity of RV dysfunction in these patients.

Methods and results: Septic patients admitted to the ICU were prospectively included. We collected haemodynamic and echocardiographic parameters as well as prognostic information. RV dilation was defined as right and left ventricular end-diastolic area ratio (R/LVEDA) > 2/3. RVSD was defined as tricuspid annular plane systolic excursion (TAPSE) < 17 mm, right ventricular fractional area change (FAC) < 35%, or peak velocity of tricuspid annulus via tissue Doppler (S') < 10 cm/s. RV-PA decoupling was represented by the TAPSE/pulmonary arterial systolic pressure (PASP) ratio. RVI score were determined by the presence of the following findings: RVSD, RV dilation, and RV-PA decoupling, that is, one point for each finding. A total of 327 patients were enrolled in this study, among whom 276 survived and 51 died at 30 days after admission. Overall, 18.0% had RV dilation, 35.8% had RVSD, and 21.4% had RV-PA decoupling, with an appreciable overlap present. A multivariate Cox regression analysis showed that RV dilation (HR: 2.19, 95% CI: 1.19-4.01, P = 0.011), RVSD (HR: 2.25, 95% CI: 1.23-4.13, P = 0.009) and RV-PA decoupling (HR: 2.08, 95% CI: 1.19-3.65, P = 0.011) were independently associated with a 30 day mortality. Furthermore, RVI score was also an independent predictor, displayed additive effect with respect to 30 day mortality (RVI score 1 vs. RVI score 0, HR: 2.94, 95% CI: 1.20-7.20, P = 0.018; RVI score 2 vs. RVI score 0, HR: 3.20, 95% CI: 1.28-7.98, P = 0.013; RVI score 3 vs. RVI score 0, HR: 7.17, 95% CI: 2.65-19.38, P < 0.001), and had the best performance in model goodness of fit, discrimination and variance explained than the other RV indices.

Conclusions: The RVI score was independently related to 30 day mortality and had the potential to grade the severity of RV dysfunction in septic patients.

Keywords: Echocardiography; Prognosis; Right ventricular dysfunction; Sepsis.

Figure 1

Flow chart of this study. LVOT, left ventricular outflow tract; TR, tricuspid regurgitation.

Figure 2

Coexistence of RV dilation, RVSD and RV‐PA decoupling and their additive prognostic relevance. (A) The distribution of RV dilation, RVSD and RV‐PA decoupling. In all, 49% of patients displayed RV abnormality, RV dilation 10% (32), RVSD16% (51), RV‐PA decoupling 3% (10), RV dilation & SD 2% (7), RV dilation & decoupling 1% (2), RVSD & decoupling 12% (40), RV dilation & SD & decoupling 6% (18). (B) Event rates (per 100 person‐years) of death on the basis of RV parameters. RVIs displayed additive effect with respect to 30 day mortality, RVIs 1 vs. RVIs 0, HR: 2.94, 95% CI: 1.20–7.20, P = 0.018; RVIs 2 vs. RVIs 0, HR: 3.20, 95% CI: 1.28–7.98, P = 0.013; RVIs 3 vs. RVIs 0, HR: 7.17, 95% CI: 2.65–19.38, P < 0.001. RV, right ventricular; SD, systolic dysfunction; RV‐PA decoupling, right ventricular‐pulmonary arterial decoupling.

Figure 3

Kaplan–Meier curves for 30 day survival. The RVIs three patients had highest mortality (RVIs 3 vs. RVIs 2, log‐rank: 9.198, P = 0.002; RVIs 3 vs. RVIs 1, log‐rank: 19.155, P < 0.001; RVIs 3 vs. RVIs 0, log‐rank: 90.938, P < 0.001). The RVIs 2 patients had similar mortality as RVIs 1 and had higher mortality than patients with RVIs 0 (RVIs 2 vs. RVIs 1, log‐rank: 1.472, P = 0.225; RVIs 2 vs. RVIs 0, log‐rank: 32.590, P < 0.001). The RVIs 1 patients had higher mortality than patients with RVIs 0 (RVIs 1vs. RVIs 0, log‐rank: 20.108, P < 0.001). RVIs, right ventricular injury score.