A total of 2101 consecutive patients electively underwent standard comprehensive TTE during the study period. Among them, 775 (37%) were excluded according to the predefined criterion. Thus, our study comprised 1326 patients (age: 73 ± 13 years, 756 men), none of whom were undergoing hemodialysis. The background characteristics of the patients are presented in Table 1. IRVF abnormalities were detected in 13 (1.0%) patients: five had discontinuous waves, three had biphasic waves, and five had monophasic waves. We divided the study cohort into two groups based on the presence or absence of abnormal IRVF. Among the background factors, the frequency of dyslipidemia, and atrial fibrillation during TTE between these groups. Additionally, the frequencies of valvular heart disease (e.g., aortic, mitral, or tricuspid regurgitation) differed significantly between these groups.

As presented in Supplemental Tables 1, 2, the patients excluded from our study were older than those included and had lower body weights and more referrals from non-cardiovascular physicians. Additionally, the excluded cohort contained more men, in-patients, technically difficult TTE cases, and missing TTE values (e.g., left ventricular ejection fraction [LVEF], E/e, and RAP). However, the frequency of missing TTE value in tricuspid regurgitation pressure gradient was not significantly different, nor were the frequencies of severe obesity (body mass index > 30 kg/m2), depressed LVEF (< 50%), or elevated RAP (≥ 15 mmHg).

The echocardiographic features of the study cohort are listed in Table 2. We excluded parameters with > 20% missing values, such as the tricuspid regurgitation pressure gradient, when comparing patients with and without IRVF abnormalities. We found that left atrial volume, E wave velocity, E/e’, and percentage of patients with elevated RAP were larger in the abnormal vs. normal group (53 ± 57 vs. 26 ± 27 mL/m2, p < 0.0001; 98.1 ± 37.3 vs. 65.0 ± 21.0 cm/s, p < 0.0001; and 17.5 ± 14.0 vs. 11.3 ± 4.4, p < 0.0001; 31% vs. 0.5%, p < 0.0001, respectively). No other factors differed significantly between these groups.

The correlations between IRVF abnormalities and RAP are presented in Table 3. Nine of the 13 (69%) patients with abnormal IRVF had non-elevated RAP (< 15 mmHg), and seven of 11 (64%) patients with elevated RAP (≥ 15 mmHg) had normal IRVF. Thus, in total, 16 patients had discrepant IRVF and RAP results. The kappa statistics between IRVF and RAP results was 0.99 ([1326–16]/1326). Comparison of patients with abnormal IRVF and non-elevated RAP with patients without this combination revealed no significant differences in the background factors, including serum creatinine level, eGFR, and frequency of eGFR < 45 mL/min/1.73 m2 (1.07 ± 0.46 vs. 0.95 ± 0.43 mg/dL, p = 0.435; 52.3 ± 17.0 vs. 60.2 ± 18.6 mL/min/1.73 m2p = 0.230; and 22.2% vs. 15.6, p = 0.584, respectively). No patients with IRVF abnormalities and non-elevated RAP had an eGFR < 15 mL/min/1.73 m2.

Twenty-one cardiac events were observed (1.6%, 21/1326): one myocardial infarction and 20 heart failures, either de novo or relapsed. Cardiac-related death occurred in four patients. Table 4 lists the echocardiographic parameters associated with the cardiac events, as determined via linear regression analysis. In model 1, where elevated RAP was one of the covariates, more than mild tricuspid regurgitation, left ventricular mass index, E/e’, and elevated RAP predicted cardiac events (odds ratio [OR]: 5.248, 95% confidence interval CI 1.830–14.971, p = 0.004; OR: 1.022, 95% CI 1.009–1.036, p = 0.002; OR: 1.137, 95% CI 1.069–1.209, p < 0.0001; OR: 18.930, 95% CI 4.104–87.319, p < 0.001, respectively). In model 2, where IRVF abnormality was one of the covariates, more than mild tricuspid regurgitation, left ventricular mass index, E/e’, and IRVF abnormality predicted cardiac events (OR: 5.245, 95% CI 1.732–1.5.797, p = 0.006; OR: 1.023, 95% CI 1.008–1.037, p = 0.001; OR: 1.131, 95% CI 1.053–1.216, p < 0.001; OR: 74.233, 95% CI 14.692–374.959, p < 0.0001, respectively).

The Kaplan–Meier plots to predict the occurrence of cardiovascular events are shown in Fig. 2. They reveal the cumulative probabilities of cardiac-free survival within 6 months after the TTE index day. Differences between groups were assessed using the log-rank test. The plots show significantly better outcomes in patients with non-elevated RAP (vs. elevated RAP, p < 0.001) and with normal IRVF (vs. abnormal IRVF, p < 0.0001).

Cumulative survival without cardiac events. A, right atrial pressure; B, intrarenal vein flow. IRVF intrarenal vein flow, RAP right atrial pressure, TTE transthoracic echocardiography

As determined via ROC analysis (Fig. 3A), E/e’ had moderate predictive potential (AUC: 0.795, cutoff value: 14.7, p < 0.0001). The Delong test revealed that the combination of E/e’ and IRVF abnormality had better predictive potential than did E/e’ alone (p = 0.043), whereas the combination of E/e’ and elevated RAP did not have better predictive value than did E/e’ alone (p = 0.277) (Fig. 3B and C).

Receiver operating curve analysis to predict the occurrence of cardiovascular events. A, E/e’; B, E/e’ vs. E/e’ plus right atrial pressure; C, E/e’ vs. E/e’ plus intrarenal vein flow. The dashed line indicates the curve for E/e’ in panels B and C. The solid line indicates the curve for E/e’ plus RAP in panel B and the curve for E/e’ plus IRVF in panel C. AUC, area under the curve; IRVF intrarenal vein flow, RAP right atrial pressure