| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Department of Anesthetics, Royal Free Hospital, London, and Department of Anesthesia and Intensive Medicine, South Manchester University Hospital, Wythenshawe, United Kingdom
Address correspondence and reprint requests to Roshan Fernando, FRCA, Department of Anesthetics, Royal Free Hospital, Pond Street, London NW3 2QG, United Kingdom. Address e-mail to r.fernando{at}btinternet.com.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Hyperbaric bupivacaine is popular in non-obstetric practice, attaining higher sensory levels of intrathecal anesthesia than equal doses of plain (glucose-free) bupivacaine when anesthesia is induced with the patient in the lateral position (2–5). Plain bupivacaine, however, is unpredictable in its behavior, often spreading to cervical dermatomal levels (6). In the obstetric population, however, neither posture during induction of spinal anesthesia nor the density of the intrathecal drugs used has been shown to have any effect on their subsequent spread within the CSF (7–10)
Plain, or glucose-free, bupivacaine has been frequently referred to as "isobaric" in the literature (11,12), even after Blomqvist and Nilsson (13) demonstrated its hypobaricity. More recently, several studies using high precision equipment to accurately measure the density of commonly used intrathecal drugs and human CSF at 37°C have confirmed that plain bupivacaine is indeed hypobaric in comparison with human CSF (14–16). Therefore the term "plain solution" should be used for these local anesthetic solutions (17). Although both plain and hyperbaric bupivacaine have been evaluated in clinical studies, the effect of a true isobaric solution has not been evaluated. Using data from these studies, the precise dose of glucose required to produce a true isobaric solution of bupivacaine at 37°C was established in a laboratory study (18).
The aim of this study was to compare the spread of a true isobaric bupivacaine solution with the commercially available preparations of hypobaric and hyperbaric bupivacaine in parturients undergoing elective cesarean delivery. To identify any influence of patient posture during the induction of spinal anesthesia, the right lateral and sitting positions were also compared.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Using sealed envelopes, patients were randomly allocated to receive hypobaric, isobaric, or hyperbaric bupivacaine solutions (vide infra) with CSE induced in either the sitting or right lateral positions. This created six distinct groups.
The densities of the 3 solutions were determined at 37°C (± 0.01°C) using a density meter (DMA 450; Paar Scientific Ltd, London, UK) accurate to ± 0.00001 g/mL, as validated in a previous study (18). The baricity of these solutions (spinal drug density relative to CSF density) were based on the results of a study by Richardson and Wissler (15) who found the mean (sd) density of term pregnant CSF to be 1.00030 (± 0.00004) g/mL. Three stock solutions were freshly prepared at the start of each day of the study. The hyperbaric solution was prepared by adding 8 mL 0.5% w/v hyperbaric bupivacaine containing glucose 80 mg/mL (Marcain Heavy®; AstraZeneca, King’s Langley, Herts, UK)) to diamorphine (Evans, Leatherhead, Surrey, UK) 1 mg in 2 mL 5% w/v glucose (Macoflex®; Macopharma, Twickenham, Middlesex, UK). The isobaric solution was prepared by adding 8 mL plain bupivacaine 0.5% w/v (Antigen, Hillside, Southport, UK) to diamorphine 1 mg in 1.6 mL 0.9% w/v normal saline plus 0.4 mL of 5% w/v glucose. The hypobaric solution was prepared by adding 8 mL 0.5% w/v plain bupivacaine to diamorphine 1 mg in 2 mL 0.9% normal saline. Each patient received an intrathecal injection of 2.5 mL of the allocated stock solution, which therefore contained 10 mg bupivacaine and 250 µg diamorphine.
Before starting the study, the mean (sd) density of each stock solution was estimated using 5 separate density measurements. The mean (sd) density of the hyperbaric, isobaric and hypobaric solutions were 1.01930 (0.00002), 1.00036 (0.00002), and 0.99969 (0.00003) g/mL, respectively.
To ensure that the procedure could be correctly blinded, one anesthesiologist was responsible for patient randomization and performing the CSE, and a second anesthesiologist, unaware of patient group allocation, was responsible for preoperative and intraoperative data collection. All patients were unaware of group allocation.
After initiation of routine monitoring, each patient was administered 1 L of 0.9% w/v saline via a 16-gauge cannula. In all patients, a CSE was performed at the L3-4 interspace in either the sitting or right lateral positions, according to randomization. A 16-gauge Tuohy needle (Portex, Hythe, Kent, UK) was first placed in the epidural space using loss of resistance to saline. A 27-gauge, 119-mm Whitacre (Becton Dickinson, Franklin Lakes, NJ) spinal needle was then introduced through the epidural needle into the subarachnoid space and observed for CSF flow, after which 2.5 mL of the appropriate spinal solution was injected. Finally, the spinal needle was removed and an epidural catheter threaded through the Tuohy needle such that 4 cm remained in the epidural space. Immediately after the epidural catheter was secured, the patient was placed in a supine position with a 15° left lateral tilt. Oxygen at 4 L/min was administered via a Hudson mask if maternal hemoglobin saturation decreased to less than 94%.
Anesthetic assessment of the regional block before surgery included maximum upper sensory level using loss of cold sensation to ethyl chloride spray. Lower limb motor block was assessed using a modified Bromage score (1 = able to raise legs above table, 2 = able to flex knees, 3 = able to move feet only, 4 = no movement in legs or feet). All assessments were made at 5-min intervals for 15 min. Failure of block was defined as a maximal sensory level using loss of cold sensation below T4 or a Bromage score less than grade 3 at 15 min after the spinal block. In such cases, the epidural catheter was topped-up with incremental 5-mL boluses of 0.5% w/v bupivacaine. The first injection was given at 15 min postspinal injection and then at 10-min intervals to achieve a sensory level (ethyl chloride) to T4 and a Bromage score of 3/4. Additional data collected were the times of spinal injection, to lie supine, to surgical incision, uterine incision to delivery, and to the end of surgery. The use of intraoperative supplementation for discomfort was noted and included treatment with incremental 5-mL boluses of bupivacaine 0.5% w/v via the epidural catheter, incremental 20-µg boluses of IV fentanyl, or general anesthesia. Hypotension (a systolic blood pressure less than 90 mm Hg or a 20% decrease to less than baseline values) or nausea and vomiting (not related to surgical stimulation) were treated with IV boluses of ephedrine 6 mg. All patients received diclofenac sodium 100 mg rectally on completion of surgery. Neonatal condition was assessed using Apgar scores and umbilical cord blood gases.
Data were analyzed using two-way analysis of variance with baricity and posture as independent factors and analysis of covariance. Post hoc tests included Tukey-Kramer and Cuzick’s trend (adjusted for baricity) as appropriate. Variations in median maximal upper sensory level and ephedrine use in groups were analyzed using two-way analysis of variance, with baricity and posture as independent factors. For clarity in the graphs, the maximal sensory block heights are presented as median, interquartile range (IQR), and range or simply as means. Ephedrine use was also analyzed using two-way analysis of variance. The incidences of hypotension, high block (cervical analgesia) and failure of block were analyzed using 
2 and expanded Fisher’s exact tests. Analyses were performed using the following software: Excel 2000 (Microsoft, Redmond VA) and Number Crunching Statistical System (NCSS) 2000 (NCSS Inc., Kaysville UT). The study had an 80% power to detect a 2-segment difference in sensory level in the groups. Statistical significance was defined for an overall 
error at the 0.05 level.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
|
There were no significant differences in patient characteristics, time of spinal injection to supine, surgical times, or neonatal outcome among the groups (Table 1). No postdural puncture headaches or other neurological sequelae were noted. Oxygen was given intraoperatively to 6 patients whose baseline oxygen saturations were <94% after CSE.
Using two-way analysis of variance, with baricity and posture as independent factors, baricity, but not posture, was associated with a statistically significant difference in maximal upper sensory level (P = 0.015 and 0.64, respectively). The interaction of baricity and posture (P = 0.059, analysis of variance) is shown in Figure 1, with data representing mean maximum sensory levels. Post hoc analysis demonstrated a significant trend of decreasing sensory block height with increasing baricity in the sitting groups (P = 0.002, Cuzick’s trend) but not the lateral groups (P = 0.297). For clarity, the median (IQR, range) maximum sensory block levels for the six groups are presented in Figure 2.
|
|
There was significant variability (P = 0.016, expanded Fisher’s exact test) in the frequencies of sensory levels involving the cervical dermatomes: six patients in the hypobaric sitting group, two in the isobaric sitting group, three in the hyperbaric lateral group, and one in the hypobaric lateral group (Table 2).
Baricity, but not posture, was associated with a statistically significant difference in motor block (P = 0.023 and P = 0.109, respectively, analysis of variance). Post hoc analysis demonstrated an overall significant trend of decreasing motor block with increasing baricity (P = 0.029, Cuzick’s trend), which was significant (P = 0.033) only for the lateral position.
There was a statistically significant difference among the groups in terms of hypotension incidence (P = 0.036, expanded Fisher’s exact test), with an increasing incidence of hypotension with decreasing baricity (P = 0.001, 2 for trend).
Baricity, but not posture, demonstrated a statistically significant difference in total ephedrine requirements (P = 0.004 and P = 0.580, respectively, analysis of variance). Decreasing baricity was associated with increasing ephedrine requirements (P = 0.004, Cuzick’s trend test).
There was no significant difference in the incidence of nausea and vomiting among the groups (P = 0.709, 2).
The incidences for failure of block whereby epidural supplementation was required before surgery could commence are shown in Table 2. There were no significant differences in groups with respect to failure of block (P = 0.356, 2). All patients developed adequate levels of anesthesia for surgery after 10 mL 0.5% w/v bupivacaine epidurally. Six patients experienced intraoperative discomfort or pain. All of these occurred within 75 min of the spinal injection and in all 6 cases surgery was allowed to continue after 10 mL 0.5% w/v bupivacaine, with none requiring IV supplementation or general anesthesia. There were no significant differences among the groups.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Studies in the non-obstetric population have found the spread of hyperbaric bupivacaine to be higher (often considerably higher) than hypobaric bupivacaine (2,5,12,19), in contrast to our study. A proposed mechanism for the increased cephalad spread of hyperbaric bupivacaine compared with hypobaric bupivacaine in the non-obstetric patients has been explained by the curvature of the spine (2). When a nonpregnant individual adopts the supine position, the highest part of the spinal column is at the level of L3, whereas the lowest part lies at T5-6 (thoracic depression). Hyperbaric solutions injected around the L3 level are more likely to spread into this thoracic depression under the influence of gravity, compared with hypobaric solutions (19).
During pregnancy, the fetus compresses the lumbar vertebrae such that they are displaced posteriorly (20). This causes a general flattening of the spinal column with loss of the thoracic depression, with the result that postural effects seen in the non-obstetric population may not be seen in the pregnant patient. This may explain why previous studies in the obstetric population have found no differences in spread of bupivacaine with either posture or baricity (7–10)
Overall, our study demonstrates that changes in baricity of the spinal injection are more important than the posture the patient adopts during the induction of spinal anesthesia, with decreasing baricity causing higher sensory block and increasing hypotension and concomitant ephedrine usage. Further analysis of the data shows that the positional influences on baricity occur mainly in the sitting position for sensory level, incidences of cervical blocks, and hypotension.
Although our study failed to show any differences in spread for sensory level in the lateral position it did show that in the sitting position the hypobaric solution spreads more cephalad than the hyperbaric solution. Stienstra et al. (21) used a spinal canal model where hypobaric and hyperbaric solutions were injected into a vertical tube. Three minutes later, samples were taken. It was found that solutions that were hyperbaric moved downward under the influence of gravity, whereas those that were hypobaric moved upwards. The same may have happened to the patients in our study, with gravity having no effect on the spread of the isobaric solution. The mean time from spinal injection to supine wedged position was approximately 2 minutes. It is quite possible that the extra time required to secure an epidural catheter after the spinal component of a CSE may have affected local anesthetic spread. Because there were no significant differences among the six groups, we do not believe that variations in the time to lie supine would have influenced the results. All CSE procedures were performed by a single senior anesthesiologist experienced in these techniques.
There was significant variability in the incidence of cervical analgesia among the groups. A post hoc contrast analysis adjusted for multiple comparisons suggests that the variability was attributable to the patients in the hypobaric sitting group (P = 0.032). Of the 12 cases of cervical analgesia, the hypobaric sitting group accounted for six, compared to none in the hyperbaric sitting group.
Other potential confounding variables in our study included the manner in which our stock solutions were prepared and the determination of the sensory block at 15 minutes. Naturally, the preparation of fresh stock spinal solutions at the start of each day of the study could have caused changes in the baricity of each individual solution. The ideal may therefore have been to request the hospital pharmacy to prepare all three stock solutions at the beginning of the trial to minimize variations in baricity. On the other hand, all stock solutions were prepared in exactly the same way for each baricity used by the same operator throughout the study period using standard drugs and syringes. A previous laboratory study by Hallworth et al. (18) found that the variations in the density of the bupivacaine/opioid solutions during multiple measurements (inter/intra-batch) were extremely small and therefore unlikely to be of clinical significance. In addition our methodology has been similar to other workers in this context.
It is possible that waiting longer than 15 minutes may have altered the final level of block height, but we felt that this was a clinically relevant amount of time before an epidural top-up was administered in the context of preparing a patient for surgery within a busy obstetric unit. All six groups were also treated in exactly the same way in terms of block assessment and thresholds for epidural top-ups.
The isobaric solution used in our study was injected at L3-4 but achieved a median maximum sensory level of T2. Unlike the hypobaric and hyperbaric solutions, the spread of the isobaric solution was not affected by posture. Because it has a similar density to that of the CSF in pregnant patients, gravity effects alone cannot explain such cephalad movement. It is much more likely that bulk movement of CSF initially carries any drug injected intrathecally, regardless of baricity, to the thoracic level, perhaps after placement of a patient in the supine wedged position. Adopting the supine position in pregnancy leads to inferior vena caval compression, which in turn causes engorgement of the epidural venous plexus (22). The consequent dural sac compression would probably aid any bulk movement of drugs injected into the CSF. As demonstrated in our study, the differences in maximal cephalad sensory spread among the three baricities used was small (only one segment), so it is tempting to hypothesize that, at least in the obstetric population, simple bulk flow of CSF is more important for achieving spread of drug, with baricity performing a more minor role.
Two non-obstetric studies have examined the effect of the spread of isobaric solutions in men and women undergoing general and gynecological surgery. Brown et al. (19) claimed to use an isobaric solution with a baricity of 1.0069 (density 1.00025 g/mL). This isobaric density was based on CSF measurements by Levin et al. in 1981 (23). However, Richardson and Wissler (15) used more sophisticated equipment to measure CSF density in 1996 and found that mean CSF density for premenopausal women was 1.00049 ± 0.00004 g/mL, for postmenopausal women 1.00070 ± 0.00018 g/mL, and for men 1.00064 g/mL. Thus the isobaric solution used in the study by Brown et al. would have been hypobaric in many of their patients. Bannister et al. (4) compared bupivacaine solutions with final glucose concentrations of 0.33, 0.83, and 8% w/v, with baricities (at 23°C) of 1.0018, 1.0045, and 1.02303, respectively. Again, the 0.33% w/v solution is likely to have been hypobaric and the 0.83 and 8% w/v solutions hyperbaric with respect to the mean CSF density in their study groups. The density of our isobaric solution was measured using the same sophisticated equipment as that used by Richardson and Wissler (14). The density of this solution was 1.00036 g/mL compared with mean CSF density of pregnant individuals of 1.00030 ± 0.00004 g/mL. Thus our isobaric solution was most likely to have been truly isobaric, unlike the hypobaric solutions used in previous studies.
Hypotension is a common problem after spinal anesthesia. Decreasing baricity of the solution significantly increased the incidence of hypotension (P = 0.001, 2 test for trend). The incidence of hypotension was mirrored by the amount of ephedrine administered. The mean ephedrine requirements for the hyperbaric, isobaric and hypobaric groups were 7.6, 10.1, and 13.6 mg, respectively.
In summary, our study has shown that in the sitting position there was a statistically significant difference in spread with the hypobaric solution resulting in higher levels of (often cervical) analgesia than the hyperbaric one. A higher level of analgesia was reflected in a more frequent incidence of hypotension. Because a CSE technique was used, the difference in behavior between hyperbaric and hypobaric bupivacaine in the sitting position could be explained by the time delay from the spinal injection to getting the patient in the supine wedged position. The isobaric solution had the least variation in that it had the least incidence of both cervical analgesia and failure although it was associated with a moderate incidence of hypotension.
In the lateral position, baricity had no effect on the spread of sensory levels for bupivacaine. This is in contrast to all non-obstetric studies cited, in which hyperbaric bupivacaine has consistently been shown to spread to a higher sensory level than plain bupivacaine in the lateral position. It may be that a generalized flattening of the spinal canal during late pregnancy helps to account for these differences, explaining why previous studies in the obstetric population have also failed to find any differences with baricity or posture.
DMA 450 density machine loaned by Paar Scientific Ltd, London, UK.
|
Footnotes |
|---|
Accepted for publication October 20, 2004.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
= hyperbaric; 
= isobaric; 
= hypobaric. *P = 0.002, Cuzick’s trend.
