The study protocol received the approval of an independent ethics committee (IRB 2020-A01819-30, CPP Sud Ouest Outre Mer III, ClinicalTrials.gov Identifier: NCT04680728). We conducted this prospective, observational study in four ICUs of university-affiliated and tertiary hospitals from October 2020 to October 2022. The patient or their kin provided written informed consent for study participation. The study was conducted in accordance with the 1964 Declaration of Helsinki. We included adult patients within 24 h of ICU admission. The exclusion criteria were: (1) person not affiliated to national health insurance, (2) age < 18 years, (3) pregnant or breastfeeding women, (4) poor echogenicity confirmed by the ultrasound operator, (5) chronic atrial fibrillation, (6) mechanical cardiac assistance, and (7) uncontrolled blood pressure (MAP < 65 mmHg).

Patients underwent clinical, biological, echocardiographic, and venous ultrasound assessments at several points in time: ICU admission (first 24 h), day 1 (24–48 h), day 2 (48–72 h), and the last day of ICU stay. The patients’ demographic characteristics (age, gender, weight, BMI), comorbidities, the reason for admission, and SAPS II scores were collected at inclusion. The clinical assessment included haemodynamic parameters [blood pressure, heart rate, central venous pressure (CVP)], diuresis, 24-h hydric balance, use of renal replacement therapy, ventilatory parameters (invasive or non-invasive mechanical ventilation use, FiO2, tidal volume, respiratory rate, end expiratory pressure), type of shock (sepsis, postoperative, haemorrhagic, cardiogenic), catecholamines (type, dose, duration), and mortality. The biological (laboratory) parameters included: haemoglobin, haematocrit, sodium, arterial lactate, and creatinine.

The VExUS score was calculated as previously described [13]. We quantified the type C of the VExUS because it demonstrated an association with AKI for grade ≥ 2 [12, 13]. The ultrasound measurements were performed according to the current guidelines [8, 15, 16]. The acquisition of ultrasound images was performed by a board-certified physician using the Philips Envisor ultrasound system (Affinity ultrasound system Philips Medical System, Suresnes, France) and concomitant ECG. The ultrasound parameters were calculated as the average of five measurements (if not specified otherwise), regardless of the respiratory cycle. Data were acquired, stored, and reviewed offline by an experienced operator blinded to the study outcomes.

The diameter of the inferior vena cava (IVC) was measured in the subcostal view at 1.0 cm from its junction with the right atrium, with the patient in the supine position. The maximum and minimum diameters of the IVC were measured, and the percentage of the change in diameter was calculated. The portal vein flow pulsatility index (PI) was assessed by performing a pulsed-wave (PW) Doppler in the liver hilum [17]. The maximum velocity (Vmax) and minimum velocity (Vmin) of the portal vein were also measured, allowing the calculation of portal PI with the following formula: PI = (Vmax − Vmin)/(Vmax). PW Doppler measurements were performed in the interlobar veins of the upper, median and lower segments of each kidney [12]. The intrarenal venous flow was classified according to the waveform pattern [12, 18].

The study primary endpoint was the VExUS score. It was calculated four times during the ICU stay: within 24 h of ICU admission, after day 1 (between 24 and 48 h), after day 2 (between 48 and 72 h), and at ICU discharge. The patients were classified as being congestive (VExUS ≥ 2) or not being congestive (VExUS < 2) [13]. A VExUS grade 2 was defined as moderate congestion, and a VExUS grade 3 was defined as severe congestion [13].

The secondary endpoints were AKI during the first week in ICU and 28-day mortality. AKI was defined according to Kidney Disease Improving Global Outcomes (KDIGO) criteria. AKI stage 1 is defined as increase in serum creatinine (SCr) by ≥ 0.3 mg/dL or increase in SCr to ≥ 1.5 times baseline, or urine output < 0.5 mL/kg/h for 6–12 h. AKI stage 2 is defined as increase in SCr to ≥ 2.0–2.9 times the baseline, or urine output < 0.5 mL/kg/h for ≥ 12 h. AKI stage 3 is defined as a SCR of up to 3.0 times the baseline or SCR increased to ≥ 4.0 mg/dL or initiation of renal replacement therapy or urine output < 0.3 mL/kg/h for ≥ 24 h or anuria for ≥ 12 h [19]. Baseline SCr was determined using the value obtained within a 3-month period before ICU admission (from hospital medical records, family physician records) and considered to be representative of the baseline kidney function according to the treating physician [20]. If no baseline creatinine value was available and patients had no documented history of renal disease, the baseline creatinine level was estimated to be 1 mg/dL. The patients with AKI at ICU admission were considered to have AKI at the moment of ICU admission. All outcomes were collected at the time of hospital discharge, death was noted within 28 days of admission to the ICU.

Based on previous studies assuming the prevalence of venous congestion being around 15%, and the prevalence of AKI being of 45%, it was estimated that a cohort of 145 patients would be sufficient to identify an association between VExUS and AKI with a statistical power of 80% and a significance level of 95% [10, 11, 14]. The data are presented as means (standard deviation), medians (25%;75% interquartile range), or numbers (percentage), as appropriate. The quantitative variables are compared using the Student’s t-test or the Mann–Whitney test, as appropriate. The proportions were compared using the chi-square or Fisher test, as appropriate. The association between the VExUS and outcomes (AKI, 28-day mortality) was evaluated using logistic regression models. All analysed patients for the primary objective had complete ultrasound measurements. For later timepoint analyses, we excluded the patients with incomplete echographic assessments, due to poor echogenicity, ICU discharge or death. The VExUS measured within 24 h of ICU admission (admission VExUS) was used as a categorial variable (0, 1, 2, 3). The higher VExUS scores were compared with VExUS 0 scores. Sensitivity analyses were performed by dichotomizing the VExUS (VExUS ≥ 2 versus VExUS < 2), and using the scores measured at admission, on day 1 and on day 2 after ICU admission. The dichotomization was based on Beaubien-Souligny et al. [13] study that demonstrated an association between AKI and VExUS (type C) grade ≥ 2. Furthermore, several Cox regression sensitivity analyses were performed. Statistical associations were evaluated at time of the events, thus patients who developed events (AKI or death) before the day of evaluation were excluded from these analyses. The sensitivity analyses were completed by a mixed-effects logistic model with AKI as the dependent variable, the patients as random effect, and the time and VExUS as fixed effects. The statistical analyses were performed using R software version 3.4.4 (R Foundation for Statistical Computing, Austria). A bilateral p value < 0.05 was considered statistically significant.