Secondary Logo

Journal Logo

Original Article

The analgesic and sedative effects of intrathecal midazolam in perianal surgery

Yegin, A.; Sanli, S.; Dosemeci, L.; Kayacan, N.; Akbas, M.; Karsli, B.

Author Information
European Journal of Anaesthesiology: August 2004 - Volume 21 - Issue 8 - p 658-662


Epidural or spinal anaesthesia are common techniques for perianal surgery, and local anaesthetic agents are generally used alone or in combination with opioids for these procedures. When local anaesthetics are combined with opioids, the analgesic effect begins in a shorter time and the duration of analgesia is prolonged. However, the use of opioids is limited in spinal anaesthesia because of side-effects such as respiratory depression, nausea, vomiting, urinary retention and pruritus [1].

The administration of midazolam, a short acting benzodiazepine, has been shown to produce segmental antinociception by a central neuraxial route [2]. It was also shown that epidural administration of midazolam has either an analgesic or a sedative effect in post-gastrectomy patients [3], and its intrathecal administration produces tranquillity in patients undergoing Caesarean delivery [4].

After perianal surgery, parenteral or oral non-steroidal anti-inflammatory drugs or opioids are generally used for analgesia. In the literature, there are relatively few reports regarding the use of intrathecal midazolam for perianal surgery [5]. The purpose of this study was to evaluate the analgesic and sedative effects of intrathecal midazolam as an adjunct to intrathecal bupivacaine in patients undergoing perianal surgery under spinal anaesthesia.


The study protocol was approved by the medical Ethics Committee of our hospital and written informed consent was obtained from each patient. Forty-four patients, who were classified as physical status I or II according to the ASA and undergoing perianal surgery (anal fissure, perianal fistula, perianal abscess or haemorrhoidectomy) under spinal anaesthesia, were included in this study. Patients presenting with one or more of the following criteria were excluded: known allergy or contraindication to any of the test drugs; contraindication to spinal anaesthesia (e.g. coagulation defects, infection at puncture site and pre-existing neurological deficits in the lower extremities).

This study was conducted in a randomized double-blind controlled fashion and all drug solutions were prepared by an anaesthesiology nurse who was not involved in the administration of spinal anaesthesia. Thus, both the observers and the patients were blinded to the drug solutions and patient groups. Our patients were randomly allocated into two groups of 22: Group B received 2 mL of 10 mg hyperbaric bupivacaine (Marcaine heavy 0.5%, in dextrose, 8%; Astra Zeneca, Turkey) + saline 0.9% 1 mL, in a total volume of 3 mL intrathecally; Group BM received 2 mL of 10 mg hyperbaric bupivacaine + 1 mL of 2 mg mL−1 preservative-free midazolam (Dormicum; Hoffman-La Roche, Basle, Switzerland) in a total volume of 3 mL intrathecally.

No pre-medication was given and the spinal puncture was performed at the L3-4 interspace with a 25-G Quincke needle with the patient in the sitting position. After injection of the spinal anaesthetic solution over a 30 s period, the patients were maintained in the sitting position for 5 min. They were then placed into the lithotomy position and remained in the same position throughout the surgery. Patients were given 10 mL kg−1 Ringer's lactate solution as a circulatory preload followed by an infusion of 6-10 mL kg−1 h−1, and they were monitored with electrocardiography, pulse oximetry, heart rate (HR) and non-invasive blood pressure (BP) measurements.

In all patients, the level of sensory block was tested by loss of pinprick sensation bilaterally at the mid-clavicular line, and the interval from the time of drug administration to the onset of T10 sensory block was recorded. Anaesthesia was defined as loss of pinprick sensation using a 27-G short-bevel dental needle. The degree of motor block was measured using a modified Bromage scale (0: full movement; 1: inability to raise extended leg but can bend knee; 2: inability to bend knee but can flex ankle and 3: no movement), and the degree of sedation was recorded on four-point scale (0: wide awake and alert; 1: drowsy on occasion but easily aroused; 2: somnolent but easily aroused and 3: somnolent and difficult to arouse).

After the operation, the patients were observed in the recovery room until the complete resolution of the motor block. They were then admitted for 24 h, and analgesia continued using tramadol 50 mg every 6 h if the visual analogue pain score was 3 or greater. The severity of postoperative pain was measured using a 10 cm visual analogue scale (VAS; 0 = no pain to 10 = the worst possible pain) at 2, 4, 8, 12 and 24 h. The period between the time of spinal injection and the first awareness of pain was considered to be the period of complete analgesia, and the time of administering the first dose of tramadol was recorded in each patient. The regression of the sensory block to the L5-S1 dermatome was regarded as the time to full recovery of sensory block. The duration of motor block was considered the time from drug administration to the time when the patient was able to lift his or her legs above the stretcher. Any adverse effects such as pruritus, nausea, vomiting, hypotension, bradycardia and respiratory depression were recorded for 24 h after drug administration. After the operation, patients were followed up for 6 weeks to monitor neurological sequelae, including persistent paraesthesia, sensory or motor deficits, and bowel or bladder dysfunction.

Statistical analysis was performed using the computer program SPSS (version 10; SPSS Inc., Chicago, IL, USA). Data were analysed using paired or unpaired t-test, U-test or χ2-test, as appropriate. No a priori power analysis was performed. A value of P < 0.05 was regarded as a statistically significant difference. Data was expressed as mean ± standard deviation (SD).


All patients had successful spinal anaesthesia, and no one was excluded from the study due to a technical failure. In the two groups, there were no statistically significant differences with regard to patient characteristics data. Similarly, the average times of operation as well as BP, HR and oxygen saturation values did not differ significantly between the two groups (Table 1).

Table 1
Table 1:
Patient characteristics and operation times (expressed as mean ± SD, no significant differences).

No statistically significant differences were observed between the two groups with respect to the degree and duration of motor block and the beginning and duration of sensory block (Table 2). However, a more profound and longer analgesic effect was seen after the operation in Group BM compared to Group B; in Group BM, the time to the first analgesic administration was significantly longer (199.3 ± 51.1 min vs. 167.5 ± 41.5 min), and the VAS scores at the first 4 h were significantly lower compared to Group B (Tables 2 and 3, respectively). In Group BM, an increased sedative effect was also noted (P < 0.001; Table 4).

Table 2
Table 2:
Median times in minutes for sensory and motor blocks, and time to the first requirement of analgesia (mean ± SD).
Table 3
Table 3:
VAS scores during postoperative 24 h period (mean ± SD).
Table 4
Table 4:
Degree of sedation.

In Group B, two patients had urinary retention and one patient had pruritus, while in the Group BM, three patients had urinary retention. Bradycardia developed in one patient in each group, and a decrease in BP was seen in one patient in Group BM. Nausea was seen in two patients in Group B, and in one patient in Group BM. However, the frequency of these side-effects was not statistically significant in the two groups. Pruritus and respiratory depression were not detected in any of the patients. None of the patients developed short-term neurological impairment or deficit during the 6-week follow-up periods.


Gamma-aminobutyric acid (GABA) is the inhibitory amino acid neurotransmitter which is found in neurones of laminae I, II and III of the dorsal horn [6]. Three types of GABA receptors have been identified, and the GABA-A receptors are modulated by benzodiazepines, barbiturates and alcohol [7]. Intrathecal midazolam, a short acting GABA-A agonist, has been found effective in treating postoperative pain and chronic mechanical low back pain [5,8,9]. Long-term midazolam administration has also been used successfully for the relief of chronic non-malignant musculoskeletal and neurogenic pain without any neurotoxicity [10].

In the literature, a number of animal studies revealed conflicting results on the use of intrathecal midazolam. In several studies it was reported that midazolam depresses spinal nociceptive neurotransmission and has a good analgesic effect [11,12]. In addition, intrathecal midazolam does not produce any neurotoxic or inflammatory effects in the spinal cord, nerve roots or meninges [13,14]. In the study by Nishiyama [13], no acute histopatological changes or neurotoxicity were found after spinal administration of midazolam in cats, even when used in high doses (10 mg). In contrast, several other studies reported potential hazards with chronic administration of midazolam or preservative-free midazolam [15,16]. Erdine and colleagues [15] demonstrated significant spinal lesions with light and fluorescence microscopy in rabbits that received midazolam or preservative-free midazolam and concluded that chronic intrathecal administration of midazolam should be avoided in human beings. Similarly, Svensson and colleagues [16] showed that chronic subarachnoid administration of midazolam may result in neurotoxicity in the rat spinal cord, and it may be unsafe to use it in human beings.

Despite these concerns, the use of intrathecal midazolam appears safe in human beings, even with long-term administration; Borg and Krijnen [10] used continuous intrathecal midazolam infusion in four patients with neurogenic and musculoskeletal pain without any significant complications. Similarly, Serrao and colleagues [9] reported that a single dose of 2 mg of midazolam used intrathecally did not cause any neurological deficits and achieved pain relief and analgesia for 2 months in patients with chronic low back pain. In our study, a single dose of intrathecal preservative-free midazolam was used, and no neurologic deficit was observed in any patients during the 6-week follow-up periods. Although there is little data regarding the neurotoxicity of such a midazolam dose in the literature, our results support the view that a single 2 mg intrathecal dose of midazolam can be safely used in human beings.

In the literature, there are relatively few reports on the safety and efficacy of intrathecal midazolam. Kim and Lee [5] demonstrated that the addition of 0.2 mL (1 mg) or 0.4 mL (2 mg) of 0.5% preservative-free midazolam to intrathecal bupivacaine significantly increased the time to first analgesic requirement in haemorrhoidectomy operations. In another study, it was reported that intrathecal administration of midazolam caused a decrease in postoperative analgesic requirements [4]. Our results were in agreement with those reported by these authors. We found the duration of complete analgesia and the time to the first dose of analgesic administration to be significantly longer in the intrathecal midazolam group, compared with the bupivacaine group.

Batra and colleagues [2] compared the postoperative analgesic effect of intrathecal midazolam-bupivacaine mixture vs. bupivacaine alone in patients undergoing knee arthroscopy, and showed that the addition of midazolam to intrathecal bupivacaine provided better postoperative analgesia without any adverse effects. In their study, time to regression of sensory analgesia to the L5-S1 level was found to be longer in the group that received midazolam. In our study, the addition of 2 mg of midazolam to bupivacaine for spinal anaesthesia did not affect the degree and recovery time of motor block or the onset time of sensory block, while the recovery time of the sensory block was prolonged, as in the study of Batra and colleagues. However, this difference was not statistically significant in our study.

In the literature there seems to be very little data on the sedative effect of intrathecal midazolam. In a prospective and randomized study, Nishiyama and colleagues [3] investigated the analgesic, sedative and amnesic effects of continuous epidural midazolam with different doses of bupivacaine in post-gastrectomy patients and found that the addition of midazolam increased either the analgesic or sedative effect in the postoperative period. In another study, Sen and colleagues [4] showed that intrathecal administration of lignocaine 5% 1.5 mL + 2 mg preservative-free midazolam produced significant postoperative pain relief together with an anti-emetic and tranquillizing effect in patients undergoing Caesarean section. In our study, we observed that intrathecal midazolam and bupivacaine combination also produced a mild but statistically significant sedative effect probably due to the supraspinal and non-segmental action of midazolam. To our knowledge, the sedative effect of midazolam combined with bupivacaine has not previously been reported.

In conclusion, the results of our study suggest that intrathecal midazolam combined with intrathecal bupivacaine produces a longer and more effective analgesia compared to bupivacaine alone in patients undergoing perianal surgery. The use of intrathecal midazolam in a single dose appears to be safe and may increase sedation.


This study was supported by the Akdeniz University Scientific Research Projects Unit, Antalya, Turkey.


1. Chaney MA. Side effects of intrathecal and epidural opioids. Can J Anaesth 1995; 42: 891-903.
2. Batra YK, Jain K, Chari P, Dhillon MS, Shaheen B, Reddy GM. Addition of intrathecal midazolam to bupivacaine produces better postoperative analgesia without prolonging recovery. Int J Clin Pharmacol Ther 1999; 37: 519-523.
3. Nishiyama T, Matsukawa T, Hanaoka K. Effects of adding midazolam on the postoperative epidural analgesia with two different doses of bupivacaine. J Clin Anesth 2002; 14: 92-97.
4. Sen A, Rudra A, Sarkar SK, Biswas B. Intrathecal midazolam for postoperative pain relief in caesarean section delivery. J Ind Med Assoc 2001; 99: 683-684.
5. Kim MH, Lee YM. Intrathecal midazolam increases the analgesic effects of spinal blockade with bupivacaine in patients undergoing haemorrhoidectomy. Br J Anaesth 2001; 86: 77-79.
6. Raj PP. Practical Management of Pain, 3rd edn. Missouri: Mosby Inc., 2000: 689-709.
7. Johnston GAR. GABA-A receptor pharmacology. Pharmacol Ther 1996; 69: 173-198.
8. Valentine JM, Lyons G, Bellamy MC. The effect of intrathecal midazolam on postoperative pain. Eur J Anaesthesiol 1996; 13: 589-593.
9. Serrao JM, Marks RL, Morley SJ, Goodchild CS. Intrathecal midazolam for the treatment of chronic mechanical low back pain: a controlled comparison with epidural steroid in a pilot study. Pain 1992; 48: 5-12.
10. Borg PA, Krijnen HJ. Long term intrathecal administration of midazolam and clonidine. Clin J Pain 1996; 12: 63-68.
11. Kohno T, Kumamoto E, Baba H. Actions of midazolam on GABAergic transmission in substantia gelatinosa neurones of adult rat spinal cord slices. Anesthesiology 2000; 92: 507-515.
12. Goodchild CS, Serrao JM. Intrathecal midazolam in the rat: evidence for spinally-mediated analgesia. Br J Anaesth 1987; 59: 1563-1570.
13. Nishiyama T. Acute phase histopatological study of spinally administered midazolam in cats. Anesth Analg 1999; 89: 717-720.
14. Serrao JM, MacKenzie JM, Goodchild CS, Gent JP. Intrathecal midazolam in the rat: an investigation of possible neurotoxic effects. Eur J Anaesthesiol 1990; 7: 115-122.
15. Erdine S, Yucel A, Ozyalcin S, et al. Neurotoxicity of midazolam in the rabbit. Pain 1999; 80: 419-423.
16. Svensson BA, Welin M, Gordh T Jr, Westman J. Chronic subarachnoid midazolam (Dormicum) in the rat. Morphologic evidence of spinal cord neurotoxicity. Reg Anesth 1995; 20: 426-434.

ANAESTHESIA, spinal; MIDAZOLAM, drug effects; PAIN, postoperative

© 2004 European Academy of Anaesthesiology