The ultrasound assessment of SVR is not validated. However, based on physiological models proposing arterial compliance and elastance theory as major SVR determinants, different approaches based on the RI have been proposed. The RI can be obtained using pulsed Doppler examination on various arterial structures based on the application of the Pourcelot index [14].

The RI is an objective parameter indicating the amount and velocity of flow towards a tissue. Using pulsed Doppler technique on vascular structures, the maximum systolic and diastolic velocities can be determined (Vmax - Vmin), and the RI formula can be solved as (Vmax - Vmin) / Vmax [15]. This examination can be performed on any arterial vascular bed with an acceptable correlation with SVR measurement [13, 16]. Early experimental models concluded that the RI could be associated with SVR [14, 17]. In 2005, Ban et al. compared the behavior of the RI and SVRI correlation in experimental models and in patients in cardiac surgery POP, showing that the correlation varies in in vivo models (r = 0.98 vs. r = 0.58). Additionally, they showed how the insonation angle strongly interferes with the strength of correlation [13].

In 2019, Lee et al. compared the SVRI measured using PAC with the RRI measured in the anatomical snuffbox in septic patients, and establishing a strong correlation between an RRI ≤ 0,97 with an SVR < 1700 din-seg-m2/cm5, as well as a RRI < 1,1 with a SVR < 2400 din-seg-m2/cm5 [12]. In 2021, the RRI was used in a sepsis resuscitation protocol, where patients with an RRI measurement < 0,9 were assumed as patients with a low SVR. Vasopressor agents were administered early, achieving a lower volume of crystalloids and a shorter hospital stay [18].

Unlike what has been reported in the literature [12, 13, 17, 18], in this study the correlation between the SVRI and the RRI is low, with wide data scattering, limiting its use in the clinical practice. However, SVRIs in this study were higher than those previously presented [12, 13]. Ban et al. reported SRVIs no higher than 1528 din-seg-m2/cm5 [13], while the studies by Lee et al. and Devia-Menendez were conducted on patients with distributive shock for whom a low SVRI is assumed [12, 18].

In patients with vasoplegia due to septic shock, distal arteries have been described to exhibit altered elastic properties associated with changes in pulsatility and resistance indexes [19]. Likewise, early SVR and RRI correlation studies using experimental models are consistent with this premise, since it has been shown that the more compliant vessels are, the higher the correlation between the SVRI and the RRI is [17]. This suggests that patients with marked vasoplegia may experience greater alterations in peripheral RIs (including RRI), accounting for the difference in this study compared to previous research.

In this study, the use of vasoactive agents and temperature did not alter the association between SVRI and RRI. Based on these findings and previous evidence [12, 13, 17, 18], it appears that a low RRI (≤ 1.1) is a good predictor of vasoplegia in the clinical context of septic shock. However, if the predominant shock mechanism is not vasoplegia but involves multiple shock mechanisms or if there is no shock, the behavior of the elastic properties of distal vessels is unpredictable, and the RRI does not show a strong correlation with SVRI.

Ultrasound measurement is known to be dependent on the skill of the observer and their ability to reproduce the measurement method with the lowest number of variations; this becomes even more important when the change in estimation magnitude varies by a few decimal places. Flow estimation using pulsed Doppler requires the flow direction to be parallel to the insonation angle. Any insonation angle in ultrasound assessments must be under 30° to obtain a reliable measurement [20]. In this study, all the measurements were made with no need to correct the insonation angle and at the same depth, which is why it is considered a standardized measure. The intraclass coherence coefficient and the Bland-Altman analysis among different evaluators showed a good aggrement for the application of this measurement.

The study being conducted at a single site in patients whose hemodynamic profiles had SVRIs close to normal is considered a study limitation. Nonetheless, the sample size is appropriate to determine the correlation magnitude and to conclude that the RRI is not a useful measure to identify SVR changes. A selection bias occurred, so its use is not recommended to approach shock patients.