Preanesthetic ultrasonography assessment of inferior vena cava diameter in the supine position, left lateral tilt position, and with the left uterine displacement maneuver in full‐term pregnant women: A randomized cross‐over design study

Preanesthetic ultrasonography assessment of inferior vena cava diameter (IVCD) in the supine position (SP), left lateral tilt position (LLT), and with the left uterine displacement maneuver (LUD) in full‐term pregnant women: a randomized cross‐over design study.


Introduction
Aortocaval compression by the gravid uterus can result in hemodynamic disturbances in full-term pregnant women. Particularly, a decrease in venous return secondary to inferior vena cava (IVC) compression is recognized as a major cause of decrease in blood pressure. 1,2 In addition, the decrease in blood pressure usually becomes more serious during spinal anesthesia in patients undergoing cesarean section. 3 Left lateral tilt position (LLT) and the manual left uterine displacement maneuver (LUD) are commonly applied in pregnant women to attenuate hypotension secondary to IVC compression. 4,5 In fact, it has been demonstrated that cardiac output and stroke volume are greater in parturients placed on an operating table tilted 15 to the left as LLT, when compared with those in the supine position (SP), 6,7 LUD has also been recommended in practice guidelines for obstetric anesthesia, 5 since it is effective in reducing both the incidence of hypotension and administration of vasopressors during spinal anesthesia. 8,9 Therefore, the patients undergoing cesarean section has routinely been placed on LLT or LUD during spinal anesthesia in clinical settings. However, only few studies have directly and quantitatively evaluated the effect of posture on alleviation of IVC compression by using IVC diameter (IVCD) as an index, one of which has suggested that the effect of LLT on enlargement of the IVCD is inconsistent, with differences between individual patients. 10 In this cross-over design study, we pre-anesthetically measured IVCD using ultrasonography in each posture, including the SP, LLT, and with LUD, in full-term pregnant women, to quantitatively evaluate the effectiveness of LLT and LUD on alleviation of IVC compression, and to evaluate the presence of interindividual differences among patients in the effect of posture on IVCD. We hypothesized that the IVCD with both LLT and LUD would be larger than that in SP. In addition, we hypothesized that due to interindividual differences in the IVCD with different postures, preanesthetic assessment of IVCD in different postures might help us to estimate which is the best posture to maintain blood pressure during spinal anesthesia in individual patients.

Methods
This randomized, single-blinded, cross-over design study was approved by the Hospital Ethics Committee on Human Rights in Clinical Trials and Research of Nihon University Hospital (Tokyo, Japan), and was registered in the "UMIN Clinical Trials Registry" (ID no.: 000024344). Our study adheres to CONSORT guidelines.

Study participants
We obtained written informed consent from all patients. Twenty-two full-term pregnant patients, aged 20 years old or older, American Society of Anesthesiologists (ASA) physical status II, who were scheduled for nonemergency cesarean section under spinal anesthesia were included in this study. Exclusion criteria included multi-fetus pregnancy, preeclampsia, pregnancy-induced hypertension, diabetes mellitus, and obesity (body mass index > 30 kg/m 2 ).
All patients fasted overnight and were given an infusion of 500 ml of crystalloid solution in the ward within an hour before their arrival at the operating room. After arriving at the operating room, patients were connected to standard monitors using ECG electrodes, an arm cuff on the right arm for noninvasive blood pressure measurement, and a pulse oximeter (SpO 2 ) probe on the fingertip (Life Scope ® , Nihon Kohden Corp., Tokyo, Japan). IVCD indices were measured by ultrasonography (C60X, 2-5 MHz convex probe; MicroMaxx ® , SonoSite, Washington, USA) in each posture in a predetermined order with the numbered container method (Figure 1). In the SP, patients lay face up on an operating table fixed perfectly horizontal; LLT was achieved by tilting the operating table 15 to the left; LUD was accomplished by the method described by Kundra et al., 11 in which an attending anesthetist stands on the left side of the patient, and retracts the right border of the patient's uterus upward and to the left, resulting in the uterus being displaced approximately 3-5 cm to the left of the midline. Patients were maintained in each posture for 5 min, following which IVCD and hemodynamic measurements, including heart rate (HR) and blood pressure (systolic blood pressure; sBP, diastolic blood pressure; dBP) were measured and recorded. A crystalloid solution was infused at the rate of 1 ml/kg/h throughout the study. After the experiment, patients received standard combined spinal-epidural anesthesia for cesarean section, and were maintained LLT position after anesthesia induction.

Ultrasonography
Two attending anesthetists who measured IVCD had experienced measurement of IVCD using ultrasonography, according to American Society of Echocardiography guidelines, 12 more than 50 times. One of the assessors was deployed to visualize the IVC, while the other assessor measured IVCD, without getting involved with visualization. They exclusively deported to visualize the IVC and measure the IVCD, respectively. The IVC was visualized in the subxiphoid window under ultrasonography using a convex probe in the long-axis view. Pulse wave Doppler was used to differentiate the IVC from the aorta. The IVCD was measured using M-mode imaging at the place that was 20-30 mm distal to the junction between right atrium and IVC. We directly measured and recorded the largest and smallest values of the IVCD by ultrasonography, that is, IVCDmax and IVCDmin, and estimated the value of respiratory variation in IVCD to indicate the collapsibility index (CI). IVCDmax and IVCDmin were respectively defined as delineated maximum diameters of the IVC in the expiratory and inspiratory phase, respectively. While the IVCD was measured by ultrasonography in each posture, patients held their steady breathing. CI was calculated using the following formula: CI (%) = [(IVCDmax -IVCDmin) / IVCDmax] Â 100. These corrected data were blinded by using the prospective randomized open blinded end-point (PROBE) method, and were analyzed by the triers other than measurers. The values of these data were compared between the three postures in each patient. Additionally, the distribution of the measured values of IVCDmax in each posture was described in each patient, since IVCDmax might be a reliable indicator of intravascular volume status. 13-15

Statistical analysis
The sample size in this study was based on the fact that one of the primary outcomes in this study was the difference in the mean value of IVCDmax between the three postures.

Results
The outcomes of 20 patients were analyzed and reported, since the data of two patients whose IVC image on ultrasonography was obscure were excluded from the study (Figure 1). The patients' characteristics and clinical data are shown in Table 1. None of the cases experienced critical changes in hemodynamic parameters or subjective symptoms. IVC parameters and hemodynamics in the three postures are presented in Tables 2 and 3. Mean values of IVCDmax observed with both LLT and LUD were significantly higher than that in SP (SP vs. LLT: p = 0.0353; 95% CI (0.17-5.91), SP vs. LUD: p = 0.008; 95% CI (0.84-6.58): Tukey-Kramer). The mean value of IVCDmax observed with LUD was larger than that  in LLT, although the difference was not statistically significant. There was also no statistically significant difference between the postures in terms of IVCDmin, CI, sBP, dBP, and HR. The values of IVCDmax in each posture in all patients are shown in Figure 2. IVCDmax was highest in the LUD in 11 patients (55%), in the LLT in 7 patients (35%) and in the SP in 2 patients (10%).

Discussion
The present study showed that the mean values of IVCDmax observed with both LLT and LUD were significantly larger than that in SP. Based on reports that IVCDmax measured by ultrasonography is a reliable indicator of intravascular volume, 13-15 the present results suggest that both LLT and LUD can effectively relieve compression of the IVC by the gravid uterus that occurs in SP in full-term pregnant patients. Additionally, the fact that a few patients (10%) demonstrated the largest IVCDmax in SP in the present study cautions us that both LLT and LUD are not necessarily appropriate treatments for relieving IVC compression in some pregnant women. Based on previous reports that showed that LLT or LUD suppressed both maternal BP decrease and vasopressor requirements observed in SP, 6,7,11 it is believed that both LLT and LUD lead to an increase in venous return to the right atrium via relief of the SP-induced IVC compression. A previous report showed that IVCDmax measured by ultrasonography might be a reliable indicator of intravascular volume status, since it is almost directly proportional to central venous pressure. 12,16 A previous study that quantitatively evaluated IVCD using ultrasonography, as in our study, demonstrated that the IVCDmax observed with both LLT and right lateral tilt positioning were significantly larger than that observed in SP. 10 Conversely, several studies showed that the angle of LLT by 15 and 20 had no effect on IVCD enlargement. 17,18 Another study also showed that the IVCD measured by magnetic resonance imaging significantly increased with LLT by 30 and 45 compared with that in SP, although a 15 LLT had no significant effect. 19 In addition, it has been demonstrated that LLT might lead to narrowing of the IVCD due to compression of the IVC by the liver. 20 Furthermore, while there is no report evaluating the effect of LUD on enlargement of the IVCD using quantitative measurement methods, some studies demonstrated that LUD was more effective in maintaining hemodynamics during spinal anesthesia than LLT. 11 Considering these above facts, the present results showing that IVCDmax was significantly larger with LLT and LUD than that in SP suggested that both LLT and LUD are effective in increasing blood flow volume in the IVC due to enlargement of the IVCD. Also, in the present study, more patients had the largest IVCDmax with LUD as compared to LLT although the mean values of IVCDmax with LLT and LUD were statistically equivalent, suggesting that LUD might be a more reliable method to relieve IVC compression than LLT.
There were no statistically significant differences in measured IVCD parameters other than IVCDmax in the three postures in this study. In ultrasonography for the IVC, the combined evaluation of IVCDmax and CI value, calculated as the difference between IVCDmax and IVCDmin, which denotes respiratory variation in IVCD, can estimate the status of circulating blood volume and right atrial pressure (RAP), as below. 12 It is estimated that RAP might decrease to <5 mmHg and hypovolemia may develop when IVCDmax ≤ 21 mm and CI > 50%, and RAP may  increase to >15 mmHg and hypervolemia may develop when IVCDmax > 21 mm and CI < 50%. Therefore, since the IVCDmax and CI in each posture of all patients observed in the present study did not fulfill the above condition, circulating blood volume in all the patients might have been within the normal range and their RAP might have been suitably maintained in each posture. However, the validity of the above assessment is controversial since there has not been enough evidence whether the general evaluation criteria of CI can be applied in full-term pregnant women. In addition, a previous study showed that the mean value of CI before spinal anesthesia in the SP was higher than that in the LLT. 21 This result was contrary to ours that the mean value of CI tended to be larger in LUD, followed by LLT and the least in SP. It is hard to compare between the two since the previous study did not show the values of IVCDmax and IVCDmin. However, the present results showing that both the IVCDmax in LLT and LUD were larger than that in SP while the IVCDmin values were equally among postures suggested that LLT and LUD increased distensibility of IVC due to reduction of IVC compression, and resulting in that the calculated CI values in LLT and LUD were larger than that in SP. Also considering the fact that there were no significant differences in values of sBP, dBP and HR among the three postures, the effect of both LLT and LUD on increase in IVC blood flow secondary to enlargement of IVCDmax might be slight, which would explain the lack of significant improvement in the hemodynamic decline seen in the SP by adoption of LLT or LUD. However, the measured values in this study were not obtained under spinal anesthesia and were evaluated in healthy parturients. After induction of spinal anesthesia, blood pressure often suddenly decreases due to decrease in central blood volume, 22,23 and both LLT and LUD suppress this decrease in blood pressure in clinical settings. 6,7,11 Additionally, a previous study showed that LLT attenuated decrease in cardiac output and blood pressure during spinal anesthesia in full-term pregnant women, even though it did not affect their hemodynamic parameters before anesthesia induction. 24 Furthermore, there exists a phenomenon whereby parturients with pre-existing and pregnancy-induced hypertension have a greater tendency to develop hypovolemia. 25 Considering the above facts, it is easy to assume that the effectiveness of LLT and LUD on the enlargement of IVCD observed in this study would be more remarkable in anesthetized and/or parturients with complications. Therefore, further studies are needed in order to define the real effect of LLT and the LUD maneuver on IVCD measured using ultrasonography in actual clinical settings. The present study also showed that more patients had the largest IVCDmax with LUD as compared to LLT and SP, as seen in the individual evaluations of patients. This order of incidences of the largest observed values of IVCDmax were not inconsistent with the results of the mean values of IVCDmax in the respective postures. Meanwhile, it should be noted that the effectiveness of the three postures on IVCDmax enlargement was not uniform between patients, as indicated by the fact that there were two patients (10%) in whom the largest IVCDmax was found in SP. This suggests that both LLT and LUD are not necessarily appropriate treatments to relieve compression of the IVC in some cases, and the optimal posture to increase IVCD should be individually evaluated. Although the reason for the individual differences in the effects of each posture on IVCD is still unclear, this could be due to variations in certain factors, such as placental site, size of the uterus and uterine adnexa, and collateral circulation via the vertebral venous system. 9,26 Ultrasonography for evaluation of the IVC can be performed noninvasively and comparatively easily, and economically. 13 In fact, we were able to measure IVCDmax and IVCDmin in all postures within a total of about 20 min in almost all our patients, although we were unable to visualize the IVC clearly in two patients. Therefore, ultrasonography for assessment of the IVC before induction of spinal anesthesia can be easily performed not only in the operating room, but also at the patient's bedside in the ward as a point-of-care method. 27,28 Preanesthetic ultrasonography assessment of the IVC might help us to preoperatively estimate which is the best posture to maintain IVCD during spinal anesthesia for cesarean section in individual patients.
There are several limitations to this study. First, the present data were obtained from patients who were not under spinal anesthesia. Since it is possible that the effect of each posture on IVCD, as described in the study, might be different in patients under spinal anesthesia, the part of our discussion related to the effects of these postures on IVCD under spinal anesthesia is only speculative. Second, it is desirable that ultrasonography for IVC assessment should be performed under standardized respiratory conditions in order to obtain accurate values of IVCD. 29,30 The measured values of IVCD parameters in this study might have been affected by variations in respiratory conditions in the three postures in each patient, even though we urged all patients to hold their steady breathing during ultrasonography in all three postures.

Conclusions
We used ultrasonography to evaluate the effects of three postures, SP, LLT, and LUD, on IVCD in healthy full-term pregnant patients, using a cross-over design. LLT and LUD might be equally effective in preventing compression of the IVCD in SP. However, both LLT and LUD might not necessarily be appropriate treatment to relieve IVC compression in some cases. Preanesthetic ultrasonography assessment of IVCD might help us to estimate which is the best posture to maintain blood pressure during spinal anesthesia for cesarean section in individual patients.

Author Contributions
Tomonori Furuya: conception and design of the study, data analysis, drafting of the manuscript. Noriya Hirose: conception and design of the study, data analysis, statistical analysis of data, drafting of the manuscript, supervision. Hanae Sato: patient recruitment, acquisition of data, data analysis. Risa Niikura: patient recruitment, acquisition of data, data analysis. Miho Kijima: data analysis, statistical analysis of data. Takahiro Suzuki: data analysis, statistical analysis of data, supervision.