Combined spinal epidural anesthesia (CSE) is a relatively new technique. After Brownridge  published his results with this method in patients undergoing elective cesarean section, CSE became increasingly popular, especially in obstetrics and orthopedic surgery. Although the technique inevitably combines some of the disadvantages of both spinal and epidural anesthesia, those in favor point out the advantages, i.e., the speed of onset and intensity of spinal anesthesia combined with the flexibility of continuous epidural anesthesia and postoperative epidural pain relief. Several modifications of the CSE technique have been described, but the needle-through-needle-single interspace technique as described by Coates  is probably the most popular at present.
When using CSE, the level of sensory blockade after subarachnoid injection can be rapidly extended by injection of a relatively small amount of local anesthetic into the epidural space [3,4]. This increase of sensory blockade is rapid and is difficult to explain in terms of the epidural administration of local anesthetic itself. In a study comparing the effect of an epidural top-up with 10 mL saline or bupivacaine 5 min after subarachnoid injection to patients receiving no epidural top-up, Blumgart et al.  found the maximum level of sensory blockade after 20 min for both saline and bupivacaine to be similar, but significantly higher, than the maximum level of sensory blockade in patients receiving no epidural top-up. Based on their observations, they claimed that the mechanism of action by which an epidural top-up extends spinal anesthesia is largely a volume effect, causing compression of the dural sac. However, the epidural top-up was administered before spinal anesthesia had been fully established and it is not clear whether compression of the dural sac after reaching maximum sensory blockade would yield the same effect.
The purpose of the present study was to answer this question and to elucidate the mechanism by which an epidural top-up extends spinal anesthesia after maximum sensory blockade has been established.
The study design was double-blind and prospective. After local ethics committee approval and informed consent, 30 patients (18-80 yr, ASA physical status I and II) scheduled for orthopedic lower limb surgery under CSE were randomly allocated to three groups of 10 patients each. Randomization was achieved using sealed envelopes numbered 1 to 30, each envelope containing one of three codes: "bupivacaine," "saline," or "nothing." Premedication consisted of midazolam 5-7.5 mg orally 1 h before the institution of regional anesthesia. Electrocardiogram, heart rate, and peripheral oxygen saturation were monitored continuously and blood pressure was measured at 5-min intervals using an automatic cycling device (Datex). All patients received isotonic saline, 500 mL intravenously.
With the patient in the sitting position, a 16-gauge Tuohy needle (Portex) was introduced into the epidural space at the fourth lumbar interspace, using the loss of resistance to saline method and a paramedian approach. Through the epidural needle, a long 27-gauge Whitacre spinal needle (Becton Dickinson) was introduced into the subarachnoid space. After obtaining a free flow of cerebrospinal fluid (CSF), 10 mg of plain bupivacaine 0.5% was administered into the subarachnoid space and the Whitacre needle withdrawn. Before removing the Tuohy needle, an 18-gauge lateral eye epidural catheter (Portex) was inserted 5 cm into the epidural space. The patient was then turned into the supine horizontal position.
After subarachnoid injection, the level of sensory blockade was assessed at 5-min intervals by an independent observer by determining loss of sensation to temperature using an ice cube. Establishment of the maximum level of sensory blockade was defined as 1) no further increase during three consecutive measurements, and 2) > 20 min after subarachnoid injection. The onset time to maximum sensory blockade was defined as the time from subarachnoid injection to the time where the maximum level of sensory blockade was first recorded. When the maximum level of sensory blockade had been achieved, the observer measuring sensory blockade left the room. After careful aspiration of the epidural catheter, patients in Group 1 received an epidural top-up with 10 mL bupivacaine 0.5% without adrenaline; patients in Group 2 received an epidural top-up with 10 mL saline; patients in Group 3 served as an active control and received no epidural injection. The patients in Groups 1 and 2 were unaware of the solution being injected. For patients in Group 3, an epidural injection was simulated by manipulating the epidural catheter and telling the patient that an injection was given. Upon completion of the simulated or real epidural injection, time was designated as t = 0 and the observer measuring sensory blockade returned and continued to assess sensory blockade at 5-min intervals during 30 min. The onset time of maximum sensory blockade during this phase was defined as the time from t = 0 to the time when the maximum level of sensory blockade was first recorded. After this observation period, the study was concluded and the epidural catheters were used for intraoperative top-ups and/or postoperative pain relief when necessary as judged by the attending anesthesiologist.
Data on patient characteristics, the maximum level of sensory blockade, and the time to onset of maximum sensory blockade are expressed as mean +/- SD. Statistical analysis comparing the increase in the maximum level of sensory blockade after epidural top-up within each group was done with the Student's paired t-test, whereas comparisons between the three groups were done with one-way analysis of variance and Tukey-Kramer multiple comparisons test when appropriate. A P < 0.05 was considered statistically significant.
There were no significant differences among the three groups regarding age, height, or weight Table 1.
Before the epidural top-up, the maximum level of sensory blockade and the time to onset of maximum sensory blockade were similar among the three groups (spinal phase, Table 2). After the epidural top-up (epidural phase), the maximum level of sensory blockade increased in all patients in Group 1, the average increase being 4.8 +/- 1.6 segments (P < 0.05 versus spinal phase). In Group 2, the maximum level of sensory blockade increased in seven patients, whereas in three patients there was no change; the average increase in Group 2 (calculations including all patients) was 2.0 +/- 2.0 segments (P < 0.05 versus spinal phase). In Group 3, the maximum level of sensory blockade increased by one segment in three patients 5 min after the simulated epidural injection, whereas in six patients the maximum level showed no change; in one patient, the maximum level decreased six segments during the epidural phase. For comparison, the patient showing a decrease was scored as having no increase, resulting in an average increase of 0.3 +/- 0.5 segments (calculations including all patients; not significant versus spinal phase). Intergroup comparisons showed that the increase in Group 1 was greater than the increase in Groups 2 and 3 (P < 0.05) and that the increase in Group 2 was greater than that of Group 3 (P < 0.05). A graphical representation of the maximum levels of sensory blockade during the spinal and epidural phase of individual patients is shown in Figure 1.
The onset time to maximum sensory blockade after epidural top-up was 17.0 +/- 6.7 min in Group 1 and 9.3 +/- 4.5 min (calculations comprising the seven patients showing an increase) in Group 2 (P < 0.05). Data on the maximum levels of sensory blockade and onset times are summarized in Table 2. Figure 2 shows a graphical representation of the maximum levels of sensory blockade during the epidural phase versus time.
In Group 1, one patient developed hypotension (blood pressure 82/45 mm Hg), which was successfully treated with intravenous ephedrine. (Heart rate, blood pressure, and peripheral oxygen saturation remained stable in all other patients.) Analgesia during surgery was excellent in all patients. None of the patients studied developed postspinal headache.
The mechanism of action of epidural anesthesia has been studied extensively . The principle sites of action of epidurally administered local anesthetics are thought to be the spinal roots as they traverse the epidural space and the spinal cord itself. Relevant amounts of local anesthetics injected into the epidural space have been shown to appear in the cerebrospinal fluid (CSF) with peak concentrations occurring 10-30 minutes after injection. Epidurally administered local anesthetics are thought to enter the subarachnoid space by diffusion through the sleeves of dura mater that cover the spinal roots as they pass through it. Diffusion of local anesthetic from the epidural to the subarachnoid space probably plays a marginal role in establishing initial sensory and motor blockade after a single epidural injection; however, it has been suggested that the presence of local anesthetic in the CSF may explain the augmenting effect on sensory and motor blockade when an epidural top-up containing half the initial dose is administered 35-60 minutes after the initial dose .
Various mechanisms have been proposed to explain the rapid extension of spinal sensory blockade when administering an epidural top-up during CSE. Transfer of local anesthetic through the hole in the dura mater made by the spinal needle  or the existence of "subclinical" analgesia being brought to full analgesic strength by small increments of local anesthetic  have been suggested. Blumgart et al.  offered a different mechanism, postulating that the extension of sensory blockade occurs primarily by a cephalad shift of local anesthetic within the CSF caused by compression of the dural sac by the epidurally injected volume. They based their conclusions on the observation that the epidural injection of 10 mL bupivacaine or saline five minutes after subarachnoid injection of bupivacaine resulted in similar maximum levels of sensory blockade that were significantly higher compared with the maximum levels of sensory blockade seen in their control group receiving only subarachnoid bupivacaine. In support of this volume effect is the observation that during the later stages of pregnancy the dose requirements for spinal and epidural anesthesia are reduced, a phenomenon explained in part by a cephalad shift of CSF due to epidural venous engorgement caused by inferior vena cava compression [7-9].
In our study, the epidural injection of 10 mL bupivacaine or saline resulted in a significant increase in the maximum level of sensory blockade. The epidural injections were made after maximum sensory blockade after subarachnoid injection had been established. The fact that under those conditions the injection of epidural saline causes a significant increase in the maximum level of sensory blockade is most likely explained by a volume effect as described by Blumgart et al. . In three patients in Group 2 (epidural saline) there was no change in the maximum level of sensory blockade after epidural top-up. This may simply indicate that a volume effect cannot be demonstrated in all patients. Alternatively, the lack of increase in these three patients may be explained by an epidural catheter malpositioned outside the epidural space although we consider this unlikely. The increase in the maximum level of sensory blockade in the patients receiving epidural bupivacaine was significantly greater compared to the patients receiving epidural saline, even when the three patients in the saline group showing no increase would have been excluded. Therefore, the extension of subarachnoid block by epidural top-up of a local anesthetic solution cannot be explained solely by a volume effect; apparently the local anesthetic itself exerts an effect as well.
The onset time to maximum sensory blockade after the epidural top-up was significantly shorter after epidural saline when compared to epidural bupivacaine. Compression of the dural sac by a volume effect resulting in a cranial shift of CSF already containing local anesthetic seems a plausible explanation for the rapidity by which the maximum level of sensory blockade starts to increase after administration of the epidural top-up, as can be seen in Figure 2. Based on our findings, it appears that the increase in the maximum level of sensory blockade after an epidural top-up with bupivacaine initially occurs by a volume effect, augmented by a local anesthetic effect that takes more time to develop.
There is a possibility that in the absence of an epidural top-up, the maximum level of sensory blockade increases further beyond the time where maximum subarachnoid spread is thought to be final. For this reason, as well as to compensate for investigator's bias, we included a control group receiving no epidural top-up. In this group of patients (Group 3), the maximum level of sensory blockade was scored one dermatome higher in three patients, five minutes after the simulated epidural top-up. However, in the remaining seven patients in Group 3 the maximum level of sensory blockade showed no increase, resulting in a statistically not significant increase of 0.3 dermatomes in this group.
Due to the design of the study, the clinical relevance of our findings is questionable; when the maximum level of sensory blockade after spinal anesthesia is sufficient for surgery, there is obviously no reason to extend sensory blockade by administering an epidural top-up until sensory blockade starts to regress. However, because we included a control group and because the effect of an epidural top-up with saline was unknown at the time the study was undertaken, we wanted to complete the epidural phase of the study early in order to minimize the possibility of regression of the sensory level below the level necessary for surgery during the observation period.
In conclusion, the extension of sensory blockade induced by an epidural top-up with a local anesthetic in CSE appears to be effectuated by a dual mechanism, the initial rapid increase being caused by a volume effect; the local anesthetic itself accounts for an increase in the maximum level of sensory blockade that takes more time to develop.
1. Brownridge P. Epidural and subarachnoid analgesia for elective Caesarean section. Anaesthesia 1981;36:70.
2. Coates MB. Combined subarachnoid and epidural techniques: a single space technique for surgery of the hip and lower limb. Anaesthesia 1982;37:89-90.
3. Rawal N, Schollin J, Wesstrom G. Epidural versus combined spinal epidural block for cesarean section. Acta Anaesthesiol Scand 1988;32:61-6.
4. Carrie LES. Epidural versus combined spinal epidural block for cesarean section. Acta Anaesthesiol Scand 1988;32:595-6.
5. Blumgart CH, Ryall D, Dennison B, Thompson-Hill LM. Mechanism of extension of spinal anaesthesia by extradural injection of local anaesthetic. Br J Anaesth 1992;69:457-60.
6. Bromage PR. Mechanism of action. In: Bromage PR, ed. Epidural analgesia. Philadelphia: WB Saunders, 1978:119-59.
7. Bromage PR. Mechanism of action of extradural anaesthesia. Br J Anaesth 1975;47:199-212.
8. Barclay DL, Renegar OJ, Nelson EW. The influence of inferior vena cava compression on the level of spinal anesthesia. Am J Obstet Gynecol 1968;101:792-800.
© 1996 International Anesthesia Research Society
9. Tunstall ME. Incremental spinal anaesthesia and Caesarean section--relevance to the test dose for extradural analgesia. Br J Anaesth 1991;67:227-8.