Improved Prediction of Fluid Responsiveness in Ventilated Patients With Low Tidal Volume: The Role of Preload Variation

Abstract

OBJECTIVES:

To analyze whether two levels of preload, one reduced by the application of tourniquets with sphygmomanometer cuffs and the other increased by passive leg elevation, improve the predictive capacity of pulse pressure variation (PPV) and stroke volume variation (SVV) of fluid responsiveness in patients ventilated with low tidal volume (Vt).

DESIGN:

Prospective cohort study.

SETTING:

ICU at the University Hospital of Cádiz (Spain).

PATIENTS:

Patients diagnosed with septic shock, on controlled invasive mechanical ventilation without spontaneous breathing, with a Vt of 6 mL/kg predicted body weight and considered for an intravascular volume load due to hemodynamic instability.

INTERVENTIONS:

Patient position changes: supine position and passive leg raise. Placement of pressure cuff compression at 60 mm Hg in one upper limb and the two lower limbs. Administration of 10 mL/kg of saline solution in 10 minutes.

MEASUREMENTS AND RESULTS:

Twenty-eight tests were obtained. The baseline characteristics of the responders and nonresponders were similar. The baseline variables PPV and SVV had a limited ability to predict the response to fluids, with areas under the curve of 0.71 and 0.66, respectively. However, its predictive capacity increases significantly with different maneuvers, with the best prediction of the difference between the PPV value during the application of tourniquets and the PPV value in the supine position, with an area under the receiver operating characteristic curve of 0.97.

CONCLUSIONS:

Lowering preload using tourniquets improves the predictive capacity of PPV and SVV for fluid responsiveness in patients ventilated with low Vt.

KEY POINTS

Question: Could the predictive ability of pulse pressure variation (PPV) and stroke volume variation for fluid response in ventilated patients with low tidal volume (Vt) be increased by using two levels of preload (passive leg raise [PLR] and tourniquets)?

Findings: The selection criteria for this prospective cohort study include patients in septic shock who require volume administration and are ventilated with low Vt. The use of two different levels of preload (PLR and tourniquets) improved the predictive capacity of PPV in response to volume, increasing the area under the receiver operating characteristic curve from 0.71 to 0.97.

Meaning: We can predict fluid responsiveness in patients ventilated with Vt with the difference in PPV value between PLR and the use of tourniquets.

The administration of IV fluids remains the first-line treatment for many patients with shock, especially in septic shock, although many associated problems are known. IV fluid administration aims to increase venous return to the heart and therefore cardiac output and thus improve organic perfusion, a situation known as preload dependency or volume responder. However, with the exception of cases of hemorrhage, significant dehydration, or the very early phase of septic shock, only half of the cases will demonstrate a clear increase in cardiac output (1) after fluid loading, as this response is temporary and unpredictable, depending on the dynamic relationship between cardiac output and venous return at a given time. In addition, we have long been aware of the numerous harmful effects of fluid accumulation in the body; thus, positive fluid balance is an independent risk factor of mortality in critically ill patients (2), particularly in those with septic shock (3) and acute respiratory distress syndrome (ARDS) (4).

To predict the response of cardiac output to fluid administration, different dynamic variables of hemodynamics have been used in recent years, which are derived from the influence that mechanical ventilation has on the arterial waveform; among them, the most studied and used variables are pulse pressure variation (PPV) and stroke volume variation (SVV). The dynamic behavior of these variables is based on heart-lung interactions. Mechanical ventilation with positive pressure induces cyclic changes in intrathoracic pressure (5–7), which peak during inspiration, leading to an increase in left ventricular (LV) stroke volume. Conversely, during expiration, these pressures are minimal, resulting in a decrease in LV stroke volume. The magnitude of the change in the LV stroke volume, or its surrogates such as pulse pressure, will be greater when the patient depends on the preload, but this respiratory oscillation is of greater intensity the greater the tidal volume (Vt) used in ventilation (8). Thus, a high PPV value should be associated with fluid responsiveness and a low PPV value with a lack of fluid responsiveness (9), as long as there is adequate transmission of pressure from the airway to the pericardium, which occurs with high Vt or in cases of low respiratory compliance (8).

These dynamic variables have several drawbacks in the prediction of preload dependence since they require the absence of spontaneous ventilation, the absence of arrhythmias and a Vt greater than or equal to 8 mL/kg of predicted body weight (9). Different maneuvers have been used to overcome these limitations in the prediction of the response to liquids, such as passive leg raising (PLR), the end-expiratory respiratory occlusion test, and changes after mini volume loads (10). Even so, they remain less reliable when a low Vt (< 8 mL/kg) is used (11, 12), a strategy that we now know improves outcomes and reduces pulmonary complications, both in patients with ARDS (13, 14) and in those without lung damage (15, 16). To circumvent this limitation, Myatra et al (17) proposed a momentary increase in Vt from 6 to 8 mL/kg and thus assessed the change in PPV to predict patients who are volume responders, a maneuver that has been used in other studies with different results (18).

We propose to keep ventilation constant with a Vt of 6 mL/kg and use two momentary preload levels that induce variability in the PPV and SVV and thus determine how they behave at these extremes. We will use a rapid, transient increase in preload with PLR and a transient decrease in preload by decreasing venous return from the extremities with pressure cuffs at 60 mm Hg for one minute. With this method, we increase the preload by approximately 300 mL with the PLR (19, 20), and we decrease it by approximately 900 mL with the application of pressure cuffs on the extremities (21, 22). This technique has been used in the past for the treatment of cardiogenic pulmonary edema (21–23), and from different studies, the volume of venous return that is decreased by applying this pressure in three extremities (one arm and both legs in the area closest to the trunk) is known.

We hypothesize that the predictive capacity of PPV and SVV for fluid responsiveness in patients ventilated with low Vt can be improved by using two levels of preload (PLR and tourniquets).

PATIENTS AND METHODS

A prospective cohort study was carried out in a university hospital with a 27-bed ICU. The study was approved by the ethics committee of Cádiz (Spain) as “Usefulness of two preload levels in predicting volume response in patients ventilated with low tidal volume” on May 23, 2012 (Institutional Review Board number 132.12), and written informed consent was requested from the patients’ legal representatives. All procedures in the current study were conducted in accordance with the ethical standards of the responsible institutional committee on human experimentation and with the Helsinki Declaration of 1975.

Patients

Patients were eligible for inclusion if they met the following criteria: age 18 years old or older, diagnosed with septic shock based on The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) (24), receiving controlled invasive mechanical ventilation with no spontaneous breathing, ventilated with a Vt of 6 mL/kg of predicted body weight, undergoing hemodynamic monitoring with the pulse index contour continuous cardiac output (PICCO) system and considered for intravascular volume loading due to hemodynamic instability. Patients were excluded if they presented with any of the following: cardiac arrhythmia, spontaneous breathing, hemorrhagic shock, or any condition that made PLR or pressure cuff placement difficult or contraindicated.

Methods

All patients were on invasive mechanical ventilation, and a Vt of 6 mL/kg was checked before the test, with adjustment of the respiratory rate, Fio₂, positive end-expiratory pressure (PEEP), and flow or inspiratory time to achieve normocapnia and acceptable oxygenation in control arterial blood gas analysis, with a plateau pressure (Pplat) less than 30 cm H₂O. At baseline, Pplat, driving pressure (Pplat–PEEP), and respiratory system compliance (Crs) were recorded using the formula: tidal volume/(Pplat–PEEP). The respiratory parameters did not change during the test.

Patients were monitored with a transpulmonary thermodilution device (PICCO) with a central venous catheter and an iliac-tipped arterial catheter inserted through the femoral artery. The transpulmonary thermodilution measurements were performed by injecting 15 mL of cold saline solution (< 8°C) through the central venous catheter, performing initial calibration with the average of three measurements, from which variables derived from these measurements, such as the cardiac index (CI), global end-diastolic volume index, intrathoracic blood volume index, extravascular lung water index, cardiac function index, and global ejection fraction, were obtained. Subsequent measurements of the CI and other study variables were obtained continuously by pulse contour analysis. The following variables were recorded at the different study positions: heart rate, systolic blood pressure, diastolic blood pressure, mean arterial pressure, systemic vascular resistance index, maximum left ventricular contractility, central venous pressure (CVP), arterial PPV, and SVV. These variables were obtained after 1 minute in the following positions (Fig. 1): baseline situation 1 (bed position at 40°), supine, baseline 2, PLR, baseline 3, pressure cuff compression at 60 mm Hg (in one upper limb and the two lower limbs below the arterial cannula), and finally, after a saline solution bolus (10 mL/kg) for 10 minutes. The pressure cuffs were made with sphygmomanometers for manual blood pressure measurement, using a standard-sized inflatable velcro cuff for upper limb occlusion and two larger-sized cuffs for the lower limbs, all of which were inflated with air using the corresponding manometers until reaching the pressure of 60 mm Hg.

Figure 1.

Positions of the study chronologically and the study variables. Cuff = tourniquets with pressure cuff compression at 60 mm Hg in one upper limb and the two lower limbs, PLR = passive leg raising.

Patients were classified as volume responders or nonresponders if the change in CI (ΔCI) was greater than or equal to 15% after serum loading with respect to baseline or if the ΔCI was less than 15%, respectively (10, 18). No more than two tests were performed on any patient, and an interval of at least 24 hours was required between tests when two tests were performed on the same patient. During the test, the doses of vasoactive agents, sedatives, or mechanical ventilation parameters were not modified. The change in PPV and SVV was calculated between the basal position and PLR (ΔPPV basal-PLR), basal and cuffs (ΔPPV basal-cuff), between the PLR and cuffs (ΔPPV PLR-cuff), and between the supine position and cuffs (ΔPPV sup-cuff).

Statistical Analysis

Quantitative variables are shown as the means ± sds when they are normally distributed or as medians and interquartile ranges when they are nonnormally distributed. Qualitative variables are shown as numbers and percentages. Variables before and after fluid administration were compared using a paired Student t test in cases of normal distribution or a Wilcoxon test in cases of nonnormal distribution. Variables between volume responders and nonresponders were compared with a two-sample Student t test when they had a normal distribution or via the Mann-Whitney U test when they had a nonnormal distribution.

Receiver operating characteristic (ROC) curves were generated to quantify the ability of the variables to detect volume responders, and both the area under the curve (AUC) and the best diagnostic threshold were determined with the value that provided the best Youden index. A p value of less than 0.05 was considered statistically significant. The statistical analysis was performed with SPSS (IBM Corp, Armonk, NY) for Mac software.

RESULTS

Twenty-one patients were studied, resulting in 28 tests, with two tests performed on seven patients. No more than two tests were ever performed on the same patient, and in the seven patients in whom two tests were performed, they were separated by at least 24 hours as indicated in the study protocol. Thirteen of the cases tests (46.4%) indicated fluid responsiveness. All enrolled patients had septic shock, with pneumonia, either with or without ARDS, as the predominant diagnosis. All patients received noradrenaline at varying doses, with no significant differences noted between volume responders and nonresponders. The baseline characteristics of the volume responders and nonresponders are shown in Table 1 and were similar in the two groups, with only a slight difference in age. The study variables (PPV and SVV) were slightly greater in responders, but the difference was not statistically significant.

TABLE 1.

Baseline Clinical, Hemodynamic, and Respiratory Characteristics of Fluid Responders and Nonresponders

Characteristics	Responders (n = 13)	No Responders (n = 15)	p
Age	50.9 ± 13.4	60.7 ± 15.2	0.08
Diagnosis
Pneumonia	10	13
Acute respiratory distress syndrome	5	5
Urinary sepsis	0	2
Peritonitis	3	0
Hemodynamics
Heart rate	104.1 ± 21.4	96.4 ± 27.5	0.42
Mean arterial pressure	85 ± 13.2	83.4 ± 12.9	0.76
Cardiac index	3.74 ± 1.16	3.53 ± 1.47	0.68
Peripheral resistance index	1767 ± 638	1904 ± 995	0.67
Maximum left ventricular contractility	1447 (931)	1490 (400)	0.96
Cardiac function index	5.52 ± 1.26	4.57 ± 2.76	0.32
Global ejection fraction	22.9 ± 7.1	22.5 ± 8.6	0.89
Global end-diastolic index	689 ± 154	744 ± 187	0.41
Intrathoracic blood volume index	852 ± 208	834 ± 244	0.86
Extravascular lung water index	13.54 ± 5.88	14.80 ± 5.93	0.56
Pulse pressure variation	12 (13)	6 (5)	0.10
Stroke volume variation	12 (12)	8 (7)	0.16
Central venous pressure	6.9 ± 2.9	9.1 ± 4.2	0.15
Respiratory
Positive end-expiratory pressure	6 (3)	10 (3)	0.12
Driving pressure	12.62 ± 2.36	13.46 ± 6.19	0.63
Respiratory system compliance	35 (10.6)	33.9 (21.4)	0.82
Others
Norepinephrine	0.67 (0.57)	0.48 (0.68)	0.56
Lactate	1.77 ± 1.24	1.78 ± 3.46	0.81

The evolution of different hemodynamic variables according to the different positions is shown in Table 2, without observing a significant change in most of them except for the two study variables (PPV and SVV) and, logically, the CI, which differentiated the responders after the administration of fluid load. For both PPV and SVV, in the responders, there was a decrease in the values in the supine position, which was more marked in the PLR position as an effect of increased venous return, and a significantly greater increase during the application of the pressure cuffs, which resulted in a decrease in venous return.

TABLE 2.

Evolution of Hemodynamic Variables in Responders and Nonresponders

Variables	Base 1	Supine	Base 2	Passive Leg Raising	Base 3	Cuffs	Volume
Heart rate
Responders	104.1 ± 21.4	103.7 ± 20.9	106.3 ± 20.7	103.9 ± 20.1	104.8 ± 21	105.2 ± 21.3	101.5 ± 20.2
Nonresponders	96.4 ± 27.5	96.6 ± 28.1	96.9 ± 28	96.9 ± 27.8	96.6 ± 27.7	96.7± 27.7	95.9 ± 20.2
Mean arterial pressure
Responders	85 ± 13.2	81.1 ± 12	80.1 ± 27	81.1 ±10.3	85.3 ± 10.3	80.8 ± 10.7	87.5 ± 9.2
Nonresponders	83.4 ± 12.9	79.1 ± 9.1	84.8 ± 10.7	81.4 ± 11.1	85.5 ± 11.8	83.5 ± 10.7	87.7 ± 13.4
Cardiac index
Responders	3.75 ± 1.18	4.03 ± 1.47	3.80 ± 1.49	4.31 ± 1.56	3.72 ± 1.37	3.51 ± 1.28	4.64 ± 1.39
Nonresponders	3.53 ± 1.47	3.83 ± 1.65	3.66 ± 1.68	3.92 ± 1.63	3.55 ± 1.49	3.24 ± 1.18	3.65 ± 1.53
Central venous pressure
Responders	6.9 ± 2.9	9.08 ± 2.39	7.09 ± 3.02	11.75 ± 2.53	6.58 ± 2.78	5.5 ± 2.47	11.17 ± 3.19
Nonresponders	9.1 ± 4.2	11.93 ± 4.42	9.27 ± 4.08	13.53 ± 4.7	8.73 ± 4.4	7.4 ± 4.03	13.13 ± 5.08
Pulse pressure variation		a		b	b	b
Responders	12 (13)	11 (9)	9.5 (11)	8 (11)	13 (9)	16 (8)	6 (5)
Nonresponders	6 (5)	4 (7)	6 (5)	4 (6)	6 (4)	6 (2)	5 (4)
Stroke volume variation					b	a
Responders	12 (12)	11 (8)	11 (14)	9 (10)	15.5 (12)	20 (13)	8 (7)
Nonresponders	8 (7)	7 (7)	6 (6)	7 (7)	8 (8)	9 (8)	7 (7)

^ap < 0.01.

^bp < 0.05.

The capacity to predict the fluids responsiveness of different variables and the changes in response to fluids at different positions are evaluated and described in Table 3 and Figure 2. In this table, the ROC AUC and its CI are analyzed, with the level of significance and the ideal cutoff value for each variable calculated with the Youden index together with the sensitivity and specificity for that cutoff value. Note how the baseline variables of the PPV and SVV have a limited capacity to predict the response to fluids, with AUC of 0.71 and 0.66, respectively. However, this area and therefore its predictive capacity increases significantly with different maneuvers and changes in position, with the best prediction of the difference between the value of the PPV during the application of pressure cuffs and the value of the PPV in the supine position, with a ROC AUC of 0.97 and a cutoff value of 3.5 with a sensitivity of 92.3 and specificity of 93.3. The SVV generally has a lower predictive capacity than the PPV. This is the case for ventilated patients with a Vt 6 mL/kg without having to vary it during the examination and assessment of the response to fluids. CVP, as expected from previous studies, has very little predictive capacity for the response to fluids, even at different positions and their variation in these.

TABLE 3.

Diagnostic Capacity of Variables in Predicting Response to Fluids

Variables	Area Under the Receiver Operating Characteristic Curve	p	Cutoff Value	Sensitivity	Specificity
PPV	0.71 (0.51–0.91)	0.06	10.5	53.8	86.7
PPV in PRL	0.73 (0.54–0.92)	0.04	3.5	92.3	46.7
PPV with cuffs	0.86 (0.74–1)	0.00	8	1	80
ΔPPV cuffs-base	0.90 (0.77–1)	0.00	3.5	84.6	1
ΔPPV cuffs-PRL	0.91 (0.79–1)	0.00	4	92.3	86.7
ΔPPV cuffs-supine	0.97 (0.91–1)	0.00	3.5	92.3	93.3
SVV	0.66 (0.45–0.87)	0.15	15	46.2	86.7
SVV in PRL	0.61 (0.39–0.83)	0.32	12.5	46.2	86.7
SVV with cuffs	0.80 (0.63–0.96)	0.01	9.5	92.3	60
ΔSVV cuffs-base	0.73 (0.45–0.87)	0.04	2.5	61.5	80
ΔSVV cuffs-PRL	0.80 (0.62–0.98)	0.08	5.5	69.2	73.3
ΔSVV cuffs–supine	0.71 (0.5–0.91)	0.06	4.5	69.2	73.3
CVP	0.34 (0.14–0.55)	0.17	—	—	—
ΔCVP PRL-base	0.49 (0.26–0.71)	0.9	—	—	—
ΔCVP base-cuffs	0.49 (0.24–0.71)	0.91	—	—	-
ΔCVP PRL-cuffs	0.47 (0.23–0.7)	0.77	—	—	—
ΔCVP supine-cuffs	0.37 (0.16–0.59)	0.26	—	—	—

cuff = value of the variable during the application of the tourniquets, cuffs-base = value difference between cuff and baseline position, cuffs-PLR = value difference between cuff and passive leg raising, cuffs-supine = value difference between cuff and supine position, CVP = central venous pressure, PLR = passive leg raising, PPV = pulse pressure variation, SVV = stroke volume variation, ΔCVP = change in central venous pressure, ΔPPV = change in pulse pressure variation, ΔSVV = change in stroke volume variation.

Dashes represent data not indicated.

Figure 2.

Receiver operating characteristic curves of the pulse pressure variation (PPV) on the left and the stroke volume variation (SVV) on the right at the different study positions. AUC = area under the receiver operating characteristic curve, cuff = value of the variable during the application of the tourniquets, cuffs-PLR = value difference between cuff and passive leg raising, cuffs-supine = value difference between cuff and supine position.

The best prediction is always produced when the cuffs are used and compared with other positions, ordering them from highest to lowest predictive power: the comparison of the ΔPPV cuff-supine, ΔPPV cuff-PLR, and ΔPPV cuff-base. Even considering only the application of cuffs without comparison with other methods (PPV-cuff), the prediction was greater than the PPV in the basal position or other positions, with an ROC AUC curve of 0.86 (Fig. 2).

We analyzed the behavior of the PPV as the study variable with the best prediction and divided the sample based on whether they had normal or low compliance (Fig. 3). We had ten cases with Crs less than or equal to 30 mL/H₂O, of which four were responders and six were nonresponders, and 18 cases with Crs greater than 30 mL/H₂O, of which nine were responders and nine were nonresponders. We observed better prediction of PPV changes in cases of low compliance, although the PPV variations between cuffs and the PLR and supine positions prediction remained acceptable in cases of low compliance.

Figure 3.

Receiver operating characteristic curves of pulse pressure variation (PPV) in study positions based on normal (respiratory system compliance [Crs] > 30 mL/H₂O) (left) or low compliance (Crs ≤ 30 mL/H₂O) (right). AUC = area under the receiver operating characteristic curve, cuff = value of the variable during the application of the tourniquets, cuffs-PLR = value difference between cuff and passive leg raising, cuffs-supine = value difference between cuff and supine position.

DISCUSSION

In this study, we demonstrated that it is possible to improve the sensitivity of PPV with low Vt using two levels of preload. The predictive capacity was similar to the one obtained in patients with Vt greater than 8 mL/kg (9). Notably, the three comparisons involving cuff application demonstrated a high predictive capacity for fluid responsiveness, with an AUC greater than or equal to 0.9. Among these, the comparison between the cuff application and the supine position showed the strongest predictive value, achieving an AUC of 0.97. Interestingly, this outperformed the comparison with the PLR maneuver, which would typically be expected to provide a greater increase in preload. On the other hand, the prediction of the response to volume with SVV is lower despite the changes in the different positions, which matches the findings of some studies that have compared the two variables (17).

The predictive capacity of the PPV and SVV in the baseline situation with a Vt of 6 mL/kg was low (AUCs of 0.69 and 0.66, respectively). This is very similar to the results reported in previous studies and reflects the low sensitivity of these variables to predict the response to fluids when ventilating with low Vt (17, 18). However, we achieved an increased predictive capacity of PPV and SVV when ventilating with low Vt by changing positions: we compared low preload by applying cuffs and high preload with the supine and PLR positions. The result was a significant improvement of the predictive capacity, which was reflected with an increase of the AUC of PPV and SVV (0.97 and 0.80, respectively). Similarly, the cutoff point of the ΔPPV cuff in the supine position was 3.5, similar to the cutoff point published in other studies (18).

Furthermore, when only the cuffs are used, the prediction of fluid responsiveness has an AUC of 0.86 for the PPV and 0.80 for the SVV, so its isolated use could be sufficient to determine the response to volume without the need to combine with the other maneuvers.

To date, the maneuvers carried out to improve the prediction of these variables were aimed at increasing cardiac preload, such as PLR and prolonged end-expiratory occlusion. In this study, we used PLR, which has been validated as a technique to increase venous return and therefore preload in different studies and is estimated to increase venous return to the heart equivalent to the administration of 300 mL of fluid by autotransfusion of blood from the legs. To the best of our knowledge, this study is the first in which a maneuver has been used to decrease preload and assess the fluid responsiveness of patients. With the venous occlusion maneuver using pressure cuffs on three limbs, venous return and therefore cardiac preload are reduced. This technique was previously widely used to treat pulmonary edema with the use of cuffs on three limbs that were rotated (rotating cuffs). Its use was gradually abandoned during the 1970s, with the appearance of increasingly effective drugs for the treatment of this pathology, although it was still mentioned as part of the nonpharmacological treatment in the 1992 edition of a classic cardiology text such as Braunwald’s (25). Ebert and Stead (21) in 1940 demonstrated that with its use, an average of 750 mL of blood could accumulate, ranging from 580 to 980 mL. Judson et al (26) reported a decrease of 5 mm Hg in pulmonary artery pressure after venous occlusion. In 1993, a study carried out with isotopes confirmed the accumulation of high amounts of blood (approximately 800–1000 mL) in the extremities with the use of cuffs, although there was no improvement in LV function, which in some cases even worsened, probably due to the increase in afterload. Its greatest effect was achieved at 60 mm Hg of pressure 45–60 seconds after its application, with the blood volume of the extremity rapidly returning to its baseline value after decompression (27).

We also observed that in cases of low compliance and low Vt ventilation, the study variables still maintain a good predictive ability, somewhat lower than in cases with normal compliance, but they even have a ROC AUC of 0.96 for the cuff-sup PPV, which is the most reliable measurement. All of these findings contradict what would be expected due to the low transmission of airway pressures to the juxtapericardial region and pulmonary vessels (28) in cases of low pulmonary compliance; however, several studies have also shown these contradictory values (8, 12, 29, 30). This is probably because the changes in pleural pressure, which are those that influence transmission to the vascular area, are greater in cases of high respiratory elastance where the predominant component is the increase in thoracic elastance (31, 32). Although Monnet et al (33) solved this problem of low Crs with the PLR maneuver or with expiratory occlusion, in these cases, monitoring of the CI is needed to determine the response. However, in our study, monitoring cardiac output was not necessary, and it would be feasible to use only the PPV and its change with the placement of cuffs.

As in previous studies, our series also revealed the lack of predictive value of CVP in response to fluids, even though it is still used for this purpose, as demonstrated by some surveys (34). This does not mean that the CVP should not be monitored since it can inform us of the preload-dependent state at extreme values; furthermore, CVP remains key for assessing global cardiac function and right-sided function in particular, and it affects the perfusion pressure gradient of several organs (35–37).

As we have already noted, the greater predictive capacity of the PPV comparison between cuffs and the supine position is striking, as it would be expected that the comparison of cuffs with the PLR would be better, where the increase in preload would theoretically be greater. We do not know the explanation for this apparently contradictory finding; although it could simply be explained by a statistical origin with a loss of predictive power due to the low sample number, it could also be explained by physiologic reasoning. By increasing venous return from the blood stored in the legs, a volume is mobilized from the peripheral muscular compartment, which is characterized by lower compliance and a longer time constant, as described by Krogh’s venous compartment theory (38, 39). This blood is therefore obtained quickly, and the supine position in our study preceded the PLR position, so it is likely that there was already a recovery of peripheral muscular blood in that supine position that increased the venous return and made the subsequent PLR maneuver less effective. In favor of this, in addition to physiology, Table 2 shows that at the basal position 3, there is a significant difference between the PPVs and SVVs of responders and nonresponders, which theoretically should not exist since this position is equal to the basal positions 1 and 2, where this difference was not observed. This could be explained by a persistent increase in the time of autotransfusion caused by the supine position and the PLR, which has not yet been distributed to the area of blood volume without stress, both in the peripheral compartment and in the splanchnic compartment, with greater compliance and a lower time constant.

We acknowledge that reducing venous return through the temporary application of cuffs on the extremities has limited clinical utility in routine practice, unlike adjusting Vt from 6 to 8 mL/kg. However, this study was conducted before Vt variation became standard practice. Despite this, our findings offer valuable insights into the physiologic mechanisms behind dynamic variables in predicting fluid responsiveness. Specifically, our results help explain why these variables are less effective at low Vt and how their predictive accuracy improves with the use of two preload maneuvers.

This study has some limitations. This was a single-center study with a limited number of patients. Only patients with septic shock were studied, and the findings are not applicable to other critical patient populations. Esophageal pressure was not monitored as a measure similar to pleural pressure.

CONCLUSIONS

The application of tourniquets significantly enhances the predictive capacity of PPV and SVV for fluid responsiveness in patients ventilated with low Vt. This improvement is statistically significant with the use of pressure cuffs alone and is further amplified when comparing PPV and SVV changes across different positions, such as the supine position and PLR. These findings highlight the potential value of incorporating tourniquet-induced preload modulation as a practical and effective method to improve fluid responsiveness assessment in clinical practice.