This prospective, randomised controlled trial compared a spinal block developing in the left lateral position (LL), after siting a combined spinal-epidural for Caesarean section with the woman sitting – using intrathecal hyperbaric bupivacaine and fentanyl – with a block developing in the supine, wedged position (SW). The median (IQR) block onset times for loss of cold sensation to T4 bilaterally for LL and SW women were 15 (13–19) and 12 (10.8–14) min, respectively (p = 0.004). Unilateral blocks did not occur in the left lateral group. There were no other significant differences between groups, except the median [range] 1-min Apgar scores, which were 9 [8–10] and 9 [6–9] in the LL and SW groups, respectively (p = 0.04). In summary, the LL position is associated with a relatively slower block onset, but produces a spinal block with similar characteristics to that obtained in the SW position.
Compression of the aorta and inferior vena cava by the gravid uterus is a common problem in late pregnancy. The left lateral position has been shown to minimise the degree of aortocaval compression [1, 2] thereby maximising maternal cardiac output and uteroplacental blood flow. A recent publication  compares hypotension, following combined spinal-epidural (CSE) anaesthesia for Caesarean section, in the left lateral position with the supine, wedged position. However, in that study, women were first moved from the sitting position to the right lateral position for 2 min, and then finally to the left lateral position. We wished to determine whether the intermediate step of placing the woman in the right lateral position could be omitted.
The aim of our study was to compare onset times of a spinal block developing with the woman in the full left lateral position (LL) to onset times in the supine, wedged position (SW), following insertion of a CSE in the sitting position. We were also interested in the distribution of block that would develop.
This study was approved by the Hillingdon Local Research Ethics Committee. Women booked for elective Caesarean section were introduced to the study and given a patient information leaflet during routine assessment at the Anaesthetic Antenatal Clinic. Subsequently, written informed consent was obtained on the day of admission. Women were excluded from the study if they were of less than 36 weeks' gestation or had a multiple pregnancy, pre-eclampsia or any known fetal or maternal compromise.
All women were given standard antacid prophylaxis. Baseline blood pressure (BP) was recorded as the lower of two BP readings taken at a 5-min interval, before intravenous cannulation. After securing intravenous access and establishing routine monitoring (ECG, non-invasive blood pressure and pulse oximetry), each woman received a 10 ml.kg−1 preload of crystalloid solution. The CSE was then inserted with the woman sitting on the operating table, at the L3/4 or L4/5 interspace. A 16- or 18-gauge Tuohy needle was used to identify the epidural space using loss of resistance to saline. A 26-gauge pencil-point spinal needle was then inserted with the bevel directed cephalad, and a standard mixture of 2.2 ml 0.5% hyperbaric bupivacaine plus fentanyl 15 μg (total volume 2.5 ml) was injected over 20 s. Aspiration of cerebrospinal fluid was performed before, during and at the end of the injection to confirm correct placement of the needle throughout, although barbotage was avoided. The epidural catheter was then inserted and secured, and the woman placed into the randomly assigned study position: either the full left lateral position, or the supine position with a 10 cm high Crawford wedge under the right hip. The table was kept horizontal, with a single pillow under the woman's head.
Blood pressure was recorded every 3 min and ephedrine and/or fluid boluses were given as required to maintain systolic blood pressure to within 20% of baseline. The height of the block was assessed every 2 min, using ethyl chloride spray, and the time taken for loss of cold sensation to reach a dermatomal level of T4 (time to T4 block) was noted for each side. The lower limit of the block was then ascertained and a three-point motor score (1 − ability to straight leg raise against gravity; 2 − impaired ability to raise leg against gravity; 3 − no leg movement against gravity) recorded for each leg. The block continued to be assessed every 5 min to establish maximum block height. Patients in the left lateral group were turned into the supine, wedged position when surgery was about to commence. The time between attaining bilateral T4 block and the start of surgery was noted.
If the block failed to reach a level of T4 bilaterally after at least 20 min, then epidural supplementation was permitted at the discretion of the anaesthetist. In such cases the time elapsed at the point of epidural supplementation was used as the nominal ‘time to T4 block’ for the primary outcome statistical analysis.
The lowest recorded systolic blood pressure after initiation of the spinal block, ephedrine requirements, fluids administered and the presence of nausea and/or vomiting were recorded. Any epidural or other analgesic supplementation given before or during the Caesarean section was also recorded.
Postoperatively, within 30 min of arrival in the recovery room, a midwife – who was blind to study group – asked the woman to complete two, 10-cm visual analogue scales (VAS) to separately assess discomfort and pain during the procedure (0 cm – no discomfort/pain; 10 cm – worst discomfort/pain imaginable). Fetal well being is routinely assessed in our unit using 1- and 5-min Apgar scores, assigned by a midwife, and umbilical arterial and venous pH.
Data for a power analysis were derived from a preliminary audit (n = 37) of block onset time and pattern of development in the left lateral position. To detect a 20% difference in block onset time between LL and SW groups, with 80% power and at a significance level of α = 0.05, we needed 25 women in each group. Data were analysed using the Student t-test, the Mann–Whitney U-test, Fisher's exact test and the Wilcoxon Signed Ranks test as appropriate.
A total of 60 women were recruited into the study. Seven were excluded due to technical difficulties with the CSE (four in LL group and three in SW group), five because of failure to maintain free aspiration of CSF during and/or at the end of the spinal injection, one because of an inadvertent dural tap with the epidural catheter, and one because of failure to identify the epidural space (single-shot spinal given instead). This left 25 women in the LL group and 28 in the SW group.
Patient characteristics are summarised in Table 1. The median (IQR) [range] time to bilateral T4 block was 15 (13–19) [8–25] min for the LL group and 12 (10.8–14) [7–25] min for the SW group (Fig. 1). This difference was statistically significant (p = 0.004).
Table 1. Maternal characteristics. Data are number of women or mean (SD).
Left lateral (n = 25)
Supine, wedged (n = 28)
All women had loss of cold sensation on perineal testing. There were no statistically significant differences between the two groups in terms of maternal secondary outcomes (Table 2 and Fig. 2).
Table 2. Secondary outcomes. Data are number of women or mean (SD), or median (IQR) [range].
Left lateral (n = 25)
Supine, wedged (n = 28)
Total motor score
6 (4–6) [3–6]
6 (5.5–6.0) [4–6]
Time from T4 block to start of surgery; min
12 (11–15) [5–28]
10 (8–14) [0–38]
% systolic BP decrease from baseline
≥20% decrease in systolic BP
Ephedrine used; mg
10.5 (0–24.0) [0–39]
12.0 (5.3–15.8) [0–36]
Total intravenous fluids; ml
Nausea ± vomiting
Blocks requiring supplementation
Pain VAS score
1 (0.3–1.7) [0–5.1]
0.4 (0.1–0.8) [0–6.4]
Discomfort VAS score
1.3 (0.5–2.9) [0.1–7.6]
0.7 (0.1–2.0) [0–5.4]
Table 3 compares the time to T4 block, motor score and maximum height of block between the left (lower) and right (upper) sides of the women in the LL group. There were no statistically significant differences in any of these variables. Within the LL group, the block reached T4 at the same time on both sides in 17 women (68%), had a faster onset on the lower side in six women (24%) and a faster onset on the upper side in two women (8%).
Table 3. Spinal block development in women in the left lateral position. Data are median (IQR) [range].
Left side (lower)
Right side (upper)
Time to T4 block; min
15 (12–17) [6–24]
15 (13–19) [6–25]
3 (2–3) [1–3]
3 (2–3) [2–3]
Maximum height of block
T3 (T2–T4) [T1–T7]
T3 (T2–T4) [T1–9]
Eight women in the LL group and six in the SW group required supplementary analgesia for the spinal block, but none had to be converted to a general anaesthetic (Table 2). Of these, four required epidural supplementation before starting the Caesarean section, three in the LL group and one in the SW group. Of the three in the LL group, two had supplementation given at 21 min because of maximum block heights at T7 and T8, respectively, and one woman had a time to bilateral T4 block of 24 min, but was then given supplementation after a further 12 min because she could feel pain on testing with surgical forceps. The case of pre-operative supplementation in the SW group was a woman with a time to bilateral T4 block of 12 min, who similarly required supplementation 8 min later when she felt pain on testing with surgical forceps.
The 1-min Apgar scores showed a statistically significant difference between neonates in the two groups (p = 0.04) (Table 4). This difference was no longer present at 5 min and there was no significant difference in umbilical arterial or venous pH.
Table 4. Neonatal outcomes. Data are mean (SD) or median (IQR) [range].
This study confirms recent findings [3, 4] that the onset times for blocks developing in the lateral position are slower than those developing in the supine, wedged position. This difference was statistically significant (p = 0.004), but it may not be clinically significant for elective surgery, particularly when weighed against the further delays that are typical before surgery actually commences (median delay of 12 min in the LL group and 10 min in the SW group).
Traditional anaesthetic teaching is that the intrathecal spread of hyperbaric local anaesthetic solutions follows the effect of gravity . Therefore, it might be anticipated that a significant unilateral block would develop in the LL group, which is presumably why Mendoca et al.  placed their women in the right lateral position for 2 min before assuming the left lateral position. However, unilateral block was not observed in our study. Hence the intermediate use of the right lateral position, which carries manpower implications and possible risks to staff and mother, appears unnecessary. The reason for this finding is unclear. A recent study has found that the addition of fentanyl reduces the density of hyperbaric bupivacaine to a statistically significant degree , but the actual change in density is so small that it is not expected to be clinically significant. We therefore cannot explain this aspect of our results but it is consistent with the experience of one of the authors (J.D.) who routinely positions patients in this way.
There was no statistically significant difference between the two groups in terms of the upper limit of the spinal block; however, it is interesting to note that three blocks spread to the cervical dermatomes in the SW group and none in the LL group (Fig. 2).
As with other studies [1, 3, 4] peri-operative cardiovascular stability, as indicated by incidence and degree of hypotension and overall ephedrine requirements, was not significantly different between the two groups. However, it may be noteworthy that the SW group showed a tendency towards a higher incidence of hypotension, and also of nausea with or without vomiting. These secondary outcome measures did not constitute part of our original power analysis and there is thus a significant possibility of a type II error in these results.
Mendoca et al.  reported that three women in the left lateral group required epidural supplementation of the block to achieve adequate anaesthesia. In our study, two women in the left lateral group whose blocks only reached T7 and T8 (Fig. 2) required epidural supplementation to achieve adequate anaesthesia. This incidence was not statistically significant but is noteworthy as a trend and it could suggest that this method of positioning is more appropriate to the CSE technique than the single shot spinal approach.
One woman in each group had a bilateral loss of cold sensation to T4 but felt pain on testing with surgical forceps and so received an epidural supplement prior to the start of surgery. A further five women in each group received supplementation intra-operatively; although this number is relatively high, none of the regional techniques had to be converted to a general anaesthetic and the median pain and discomfort VAS scores were low in both groups. When this study was designed, a debate was underway over which sensory modality to use in order to assess the block for Caesarean section [7–10]. At the time we routinely used loss of cold sensation to test our blocks. We chose this modality for the study because of our familiarity with the technique, its widespread use and support for it in the contemporary literature [8, 9]. However, if we were to perform the study again, we would assess the block using fine touch sensation and it is assumed that this would have allowed earlier detection of the two cases that had inadequate blocks despite loss of cold sensation to T4. Our overall supplementation rate may be higher than that of Mendoca et al.  because we used a smaller dose of bupivacaine and opioids, which may also explain our lower incidence of hypotension.
There was a statistically significant difference between groups in fetal outcome, in terms of the 1-min Apgar scores being slightly higher in the LL group (p = 0.04). This difference was no longer present at 5 min and there was no significant difference in umbilical arterial and venous pH. This finding was unexpected and was not detected by Rees et al. . It is likely to be due to the reduction in utero placental blood flow which occurs during block development in the supine, wedged position [1, 2]. We cannot be sure whether this finding would have a long-term effect in an already compromised fetus, as we have not studied this population. In addition, Apgar scores are considered poor predictors of outcome, which may explain why other authors have not reported them [3, 4], whereas standard base excess has more recently been identified as a better surrogate marker than pH for predicting significant neonatal morbidity .
In conclusion, a CSE technique incorporating intrathecal injection of hyperbaric bupivacaine and fentanyl in the sitting position, and subsequent block development in the left lateral position, does not produce unilateral blockade and is suitable for Caesarean section. The onset time is slower for this technique than development of the block in the supine, wedged position. However, this consideration may be outweighed by the known improvement in maternal cardiac output obtained in the left lateral position and the possible benefits this confers to the fetus.
We would like to thank Dr Nuala Lucas for help with study design and statistical analysis and all the Hillingdon Hospital anaesthetists, operating department practitioners and midwives who kindly assisted the authors with this study.