This study was conducted at the Multipurpose Medical intensive care unit of La Rabta University Hospital. Our unit consists of 8 ICU beds and receives medical pathologies of all types (hemodynamic, respiratory, neurological and metabolic). The traumatic-surgical pathologies do not enter in our activity.

This was designed as an observational prospective cohort study, over a period of four months (from January 2017 to April 2017). The study was approved by the Ethics Committee of the institutional review board of La Rabta hospital. As the protocol was considered part of routine practice, informed consent was waived following the recommendations of the Ethical Committee. Patients or their legal relatives were informed by the participation in this study.

Patients were enrolled from the medical ICU and any patient requiring VE as determined by the ICU attending physician was eligible for enrolment. No specific criteria for circulatory insufficiency were required for study entry. The decision of the ICU attending to administer fluid was based on clinical signs of inadequate global perfusion: mean arterial pressure (MAP) below 60mmHg, oliguria (urine output less than 0.5ml/kg per hour for more than 2h), cold extremities, mottled skin, and tachycardia [heart rate (HR) higher than 100/min]. Exclusion criteria were known aortic valve disease, known ascending aortic aneurysm, contraindication to PLR for any reason, and poor cardiac echogenicity.

The echocardiographic examination was performed by the same operator (AT) following training on the device and using a transthoracic ultrasound device (Hitachi Aloka – ARIETTA V60), equipped with the tissue Doppler imaging program and a phased array transducer C251 Convex 5-1Mhz. Conventional echocardiography including M-mode, two-dimensional, and Doppler measurements was performed. Echocardiographic images were recorded together with the electrocardiogram. The device used directly measures the blood flow through the aortic valve.

For each patient studied, echocardiographic measurements were taken in four stages. In stage one the patient was placed in a semi-recumbent position with the head elevated at 45 degrees. In stage two, the patient was positioned supine with the legs straight and elevated at 45 degrees for 90s. The highest plateau value of velocity-time integral of aortic blood flow (VTIAo) recorded during PLR was obtained within the first 90s following leg elevation. Stage three readings were taken 2min after the patient was returned to the baseline position, and stage four immediately following expansion of 500ml of crystalloids over 15min. The different stages of passive leg raising as well as the volume expansion were in accordance with the method described by Monnet et al.7 (Fig. 1). Vasopressors and ventilator settings were not changed at any time while a patient was being studied.

Figure 1.

Used method for passive leg raising as described by Monnet et al.7

(0.18MB).

The stroke volume was calculated as the product of the VTIAo by the aortic valve area. Using the apical five-chamber view the VTIAo was computed from the area under the envelope of the continued-wave doppler signal obtained at the level of the aortic annulus. The VTIAo value was averaged over five consecutive measurements in sinus rhythm patients and over ten measurements in patients with atrial fibrillation. Using the parasternal long axis view the diameter of the aortic orifice was measured at the insertion of the aortic cusp and the aortic valve area was calculated [π×(Diameter/2)2]. As the diameter of the aortic orifice is assumed to remain constant in a given patient, the diameter was measured once at baseline. The cardiac output was calculated as the product of stroke volume by the heart rate. Thereafter, the studied patients were classified according to their responsiveness to the endogenous (by PLR) and/or exogenous (by VE). Likewise, the left ventricular ejection fraction (LVEF) was measured by the technique of Teicholz in the large-axis parasternal view or simpson on the 4-cavity view (in the presence of segmental kinetics disorders). Also, the inferior vena cava (IVC) diameter was examined within 0.5–3cm from the caval–right atrial junction in the subcostal view and the distensibility index (dIVC) was calculated as the ratio of (Dmax−Dmin)/Dmin.

Patients were classified according to their hemodynamic response to VE. Responders had a SV increment of at least 15% in response to VE (an increase in SV from stage one to stage four), while non-responders had a SV increase of less than 15%. Cut-off values of 15% have been previously used as a significant difference between two measures of CO in similar studies.10–13 Indeed, this value is considered as the cut-off that predict volume responsiveness in the recent revision on echocardiography in shock management.8

Continuous data are expressed as mean±standard deviation or median [interquartile range 25th–75th percentile (p25–p75)] as appropriate. Hemodynamic variables were compared between responders and non-responders using the nonparametric Mann–Whitney test. The Kruskal–Wallis test was used to test differences between 3 subgroups. The quantitative variables before/after PLR and before/after EV were compared using the non parametric Wilcoxon test for paired series. The area under receiver operating characteristic (ROC) curve was used to measure the predictive values of ultrasound variation of SV after PLR and VE and expressed as the area±standard error. The Spearman test was used to assess the correlation between the SV variations (before/after PLR and before/after VE). Differences with a p value less than 0.05 were considered statistically significant.

Statistical analyses were performed using the SPSS© version 20 software package.

Twenty-eight patients were examined and 26 were studied. Two patients had to be excluded because of satisfactory doppler signals could not be obtained. Thus, a total of 30 volume challenges in 26 consecutive patients were evaluated. Three patients had more than one study. Repeat studies performed on the same patient were separated in time by at least 24h. The indications of a volume expansion were: septic shock in 21cases, sepsis without shock: 4 cases, metabolic complications of diabetes (ketoacidosis: 1 case and hyperosmolar state occurred during hospitalization: 2 cases), dehydration with deglobulisation following accidental arterial puncture: 1 case (the time waiting blood transfusion) and pancreatitis with a 3rd sector: 1 case. The clinical characteristics are reported in Table 1.

The initial hemodynamic measurements are summarized in Table 2. The responders had a significantly lower initial VTIAo and SV compared with the non-responders (23±5.5 vs 29±4cm, p=0.012 and 66±18 vs 89±35ml, p=0.025). Likewise, the distensibility index of inferior vena cava (dIVC) was higher in responders group (42% vs 18.8%, p=0.009).

Although the MAP (68±13 vs 75.05±14mmHg, p=0.09), the CO (6.3±2.3 vs 7.32±2.39L/min, p=0.35) and the cardiac index (CI) (3.45±1.2 vs 3.96±1.46L/min/m2, p=0.3) were not different between the groups.

The hemodynamic readings taken throughout the four stages of measurements are summarized in Table 3.

During the first step (PLR): there were 21 responders and 9 non-responders. For the responders (n=21), PLR induced a significant increase in VTI Ao (23±5.5cm vs 29±6.1, p=0.004), in SV (66±18ml vs 86±17ml, p=0.001), in CO (6.3±2.3L/min vs 7.6±2.9L/min, p=0.04) and in CI (3.45±1.2L/min/m2 vs 4.32±1.57L/min/m2, p<0.0001). Although, the heart rate and the MAP were unchanged after PLR in this same group (95.3±24beats/min vs 95.7±24beats/min, p=0.74) and (75±16mmHg vs 84±25mmHg, p=0.17) respectively. In the non-responders (n=9), PLR did not induce a significant change in any of the hemodynamic values measured.

During the second step (VE): there were 19 responders and 11 non-responders. For the responders group (n=19), VE induced a significant increase in VTI Ao (26±5.6cm vs 31±7, p=0.019) and in CI (3.86±1.53L/min/m2 vs 4.72±1.95L/min/m2, p=0.001). No significant increase in SV (78±27ml vs 92±21ml, p=0.06) nor in CO (7.1±2.7L/min vs 8.6±3.4L/min, p=0.07) was revealed. Any changes were showed with HR and MAP. For the non-responders, all the parameters were unchanged before or after VE (Table 3).

The proportional changes in SV induced by VE were correlated positively with the proportional changes in SV induced by PLR (Rho score=0.77, p<0.0001), with a linear correlation equation calculated at y=0.82 x+1.68 (Fig. 2).

Figure 2.

Relationship between proportional changes in SV induced by VE and proportional changes in SV induced by PLR. SV, stroke volume; PLR, passive leg raising; VE, volume expansion.

(0.1MB).

A correlation study between the SV variations induced by PLR and SV variations induced by VE was performed separately: in ventilated patients (n=23) versus in SBP (n=7). Correlation was found to be better in ventilated patients (Spearman coefficient: 0.836 and p<0.0001 versus 0.75 and p=0.052).

The area under the ROC curve for the percent change in SV during PLR predicting a response to VE was 0.84±0.04. A SV increase induced by PLR of 15% or more predicted volume response with a sensitivity of 94.7%, specificity of 72.7%, positive predictive value of 85.7%, and a negative predictive value of 88% (Fig. 3).

Figure 3.

ROC curve showing the ability of PLR induced changes in SV of at least 15% to predict fluid responsiveness. ROC, receiver operating characteristics; PLR, passive leg raising; SV, stroke volume; AUC, area under the curve; PPV, positive predictive value; NPV, negative predictive value.

(0.09MB).