Secondary Logo

Journal Logo

Original papers

Ketaminemidazolam total intravenous anaesthesia for prolonged abdominal surgery

Atallah, M. M.; El-Mohayman, H. A.; El-Metwally, R. E.

Author Information
European Journal of Anaesthesiology: January 2001 - Volume 18 - Issue 1 - p 29-35



Total intravenous anaesthesia (TIVA) for prolonged surgery involves the simultaneous infusion and administration of drugs to achieve the components of anaesthesia, in patients ventilated with oxygen enriched air, namely: hypnosis, amnesia, analgesia and muscle relaxation. Various modalities of TIVA have been available since the introduction of barbiturates. TIVA is always achieved by a combination of drugs with similar pharmacokinetic profiles and complementary additive or synergistic pharmacodynamic effects. It cannot be obtained with a single drug without undesirable effects on system function.

Ketamine is a good analgesic and anaesthetic agent [1–2] but its clinical usefulness has been limited because of its cardiovascular stimulating properties and high incidence of psychomimetic emergence reactions. Benzodiazepines provide the most complete protection against both the circulatory and psychomimetic side-effects of ketamine. Midazolam, a benzodiazepine derivative, has a sedative, hypnotic, and amnesic properties associated with a hypotensive effect [3]. Total intravenous anaesthesia can be effectively conducted, in prolonged surgery, by the simultaneous infusion of ketamine and midazolam supplemented by muscle relaxants. The complementary effect of both agents may result in steady-state anaesthesia. However, the popularity of total intravenous anaesthesia was hampered by the cumulative effects of these drugs, inevitably resulting in long recovery times.

Our hypothesis was that ketaminemidazolam TIVA – administered by an infusion regimen using a multistepped decreasing scheme – may provide steady-state anaesthesia, stable haemodynamic variables, and may minimize the cumulative effects of both drugs and consequently shorten recovery time. We aimed to develop an approximation regimen for the infusion of ketamine and midazolam, derived from their pharmacokinetic variables, to obtain steady-state anaesthesia for prolonged abdominal surgery followed by acceptable recovery. This has been tested in patients undergoing radical cystectomy.


Patients undergoing radical cystectomy with urinary diversion were the subject of this prospective cohort open study which comprised a pilot study and two further parts. Exclusion criteria were: patients with mental disorders or a history of family troubles, patients with cardiac (chronic congestive heart failure), hepatic (repeated serum bilirubin > 1.3 mg dL−1) or renal (repeated serum creatinine > 1.5 mg dL−1) insufficiency, and patients with endocrine malfunction except type II diabetes mellitus (Table 1). The study protocol was approved by the Hospital Research Committee and written informed consent was obtained for Part II patients.

Table 1
Table 1:
The initial regimen doses of ketamine (K) and midazolam (M) used in the five subclasses (each comprised three patients) of the pilot study

For the purpose of assessing the recovery of mental activities after anaesthesia, memory span was selected as a suitable measure. This was carried out, the day before surgery, by the digit forward and digit backward subtest of the Wechsler–Bellevue adult intelligence scale [4]. Patients were asked to repeat an increasing array of randomly arranged digits in a forward and backward manner. The maximum number of digits the patient could recall was scored. Patients were premedicated with 5 mg diazepam orally 2 h before transfer to the operating suite. Anaesthesia was induced with i.v. midazolam followed by ketamine. Tracheal intubation was facilitated with succinylcholine 1 mg kg−1 and the patients lungs were ventilated with oxygen enriched air (FiO2 = 0.35). Anaesthesia was maintained by the infusion of ketamine and midazolam; each drug was given from a separate infusion pump. Further skeletal muscle paralysis was obtained with vecuronium.

Pilot study

This was aimed to determine the appropriate ketaminemidazolam regimen compatible with stable haemodynamic variables and early recovery. The initial regimen doses were roughly calculated from the anaesthetic doses and pharmacokinetic variables [5–10]. As drug elimination depends on distribution and clearance and proceeds at constant fractional rate, we reckoned that the elimination half-life can be used to forecast the required dose needed to maintain drug action. This was used, on an hourly basis, to replace the presumed eliminated amount of drugs calculated from the induction doses, and this necessitated an hourly adjustment of the infused does to obtain – on an empirical (clinical) basis – minimal fluctuations in drug action.

The wide variability of the pharmacokinetic variables made it judicious to start initially with small doses and then manipulate regimen doses to be compatible with minimal fluctuations in drug action, manifested by adequate analgesia and minimal variations in haemodynamic variables (< 20%) together with avoidance of delayed recovery. Consequently, five subclasses, each comprising three patients, were used in the pilot study (Table 1). The obtained approximation regimen doses were used to maintain anaesthesia in the study population (Table 2).

Table 2
Table 2:
The approximation regimen doses used for maintenance of ketaminemidazolam TIVA

Part I

Twenty patients were studied and we aimed to validate the approximation regimen doses of ketamine and midazolam. Anaesthesia was induced with i.v. midazolam 150 µg kg−1 followed by ketamine 2 mg kg−1, and maintained by the approximation regimen doses (Table 2). If the infusion rate proved to be insufficient to maintain adequate anaesthesia, as manifested by increases in heart rate and arterial pressure (> 20%), then a bolus dose was given and the infusion rate was increased. If, on the other hand, the infusion rate was deemed to great, as manifested by a decrease in arterial pressure (> 20%), then the midazolam infusion was stopped for a short period and subsequently resumed at a lower rate.

Part II

Eleven patients were studied in our aim to profile the full haemodynamic and oxygenation variables during the perioperative period. When the patients arrived in the operating suite, a pulmonary artery catheter was inserted via the right internal jugular vein, and the radial artery of the non-dominant forearm was cannulated. Induction and maintenance of anaesthesia were the same as patients in group I.

All patients were monitored by continuous five-lead ECG, pulse oximetry, non-invasive arterial pressure in group I and invasively in group II, end-tidal CO2, central venous pressure and mid-oesophageal temperature. Patients in group II were monitored additionally for haemodynamic and oxygenation variables. The blood loss which occurred during the posterior dissection of the urinary bladder was replaced by two units of homologous blood. Rarely, a third unit was needed. The requirement for crystalloids was calculated as 2 mL kg−1 h−1 during the preoperative fasting period, and 8–10 mL kg−1 h−1 during surgery. When surgery had finished the residual neuromuscular block was antagonized with neostigmine. All patients were extubated in the operating room.

During their stay in the recovery room, patients were observed and assessed for their monitored physiological parameters, sedation score [11], any abnormal behaviour and the need for analgesics. The digit forward and digit backward test was performed every 30 min. On the second morning, the patients were sent to the ward and routinely observed.

Data were summarized as mean and 95% confidence interval of the mean. The difference between two means was considered insignificant when their 95% confidence intervals overlapped.


Demographic data and preoperative medical status are shown in Table 3. The anaesthetic time ranged from 4½ h to 6 ½ h with a mean of 5½ h. During the first hour, the infusion of midazolam was stopped in 12 patients for variable periods ranging from 5 to 12 min as a precaution against further reductions in arterial pressure. In another seven patients, the rate of drug infusion was increased by 20% during the first 30 min with 2–3 i.v. boluses (ketamine 5 mg each, midazolam 250 µg each). Every hour, the rate of infusion of ketamine and midazolam was reduced (Table 2). If anaesthesia was prolonged more than 5 h, the rate of drug infusion during the sixth hour was made the same as the fourth hour, and the seventh hour was the same as the fifth hour. The total amount of drugs used for every patient was about 1150 mg ketamine and 35 mg midazolam.

Table 3
Table 3:
Demographic data and preoperative condition of both patient groups

Good haemodynamic stability with minimal fluctuations in heart rate and arterial pressure was obtained (Table 4). There was no other autonomic hyperactivity in the form of sweating, lacrimation, or papillary dilatation. Requests for muscle relaxants was judged by triggering the ventilator or finger movement. The colon and intestinal loops were not distended, and body temperature remained close to 35°C.

Table 4
Table 4:
Perioperative heart rate, mean arterial pressure and indices of vascular resistance in patients of group II

The haemodynamic and oxygenation variables are shown in Tables 4 and 5, and Figures 1 and 2. Adoption of the head-down tilt for the surgical procedure increased the central venous pressure and this continued during most of the operation. This effect was temporary reflected in the readings of pulmonary artery wedge pressure during the first hour. The oxygen consumption index decreased mostly during surgery. However, the work of both sides of the heart, the dynamics of the pulmonary and systemic circulations and mixed venous oxygen variables were reasonably steady during the perioperative period.

Table 5
Table 5:
Perioperative oxygen variables in patients of group II
Figure 1.
Figure 1.:
Perioperative central venous pressure (CVP), mean pulmonary artery pressure (MPAP), cardiac index (CI), and pulmonary artery wedge pressure (PAWP) in patients of group II.
Figure 2.
Figure 2.:
Perioperative right and left ventricular stroke work indices (RVSW1 and LVSW1 respectively), mixed venous oxygen saturation and tension (SvO2 and PvO2 respectively) and shunt fraction in patients of group II.

When muscle tone had been regained the patients remained sedated. After 30 min, patients were assessed for sedation: 16 patients scored III (drowsy, but responsive to command), 15 patients scored V (asleep, with sluggish response to auditory stimulus). However, all but five patients scored III by the second hour, and II (co-operative, oriented and tranquil) by the third hour. Spontaneous verbalization was observed in five patients and lasted for 3–5 h: they were given boluses of midazolam.

The digit forward and digit backward tests returned to preoperative values after a mean of 3 h (average 2.30 h−4 h). During their stay in the recovery room, 22 patients did not complain of any pain; three patients complained of slight pain not requiring an analgesic, while six patients needed an analgesic within 3–4 h after surgery and were given demerol 50 mg i.m. There was not a single incidence of intraoperative awareness or recall of events. In the ward, patients were subjected to routine management and pain was relieved with demerol 25 mg i.m.


For prolonged surgery, TIVA is now a technique used routinely and a valid option to inhalational anaesthesia. The induction doses are necessarily loading doses to achieve therapeutic blood concentrations as rapidly as possible, while the continuous infusion minimizes the fluctuations in concentrations of the drugs in the blood, and hence in the brain. The induction drugs can be administrated as either boluses or rapid infusions over 10 min: the later is the better associated with minimal haemodynamic fluctuations.

Patients displayed moderate variations in dosing needed for a defined response, hence individual variability is expected. This is dealt with by changing the rate of drug infusion in an analogous way to inhalational anaesthesia where the inspired concentration is adjusted to obtain a targeted response. Arterial pressure responses to surgical stimulation and the need for muscle relaxants are useful guides to judge the depth of anaesthesia, while autonomic hyperactivity is not concomitant with these variables.

Ketamine produces a significant increase in heart rate and mean arterial pressure [12] caused by central sympathetic stimulation [13], while midazolam causes venodilatation, a decrease in systemic vascular resistance and a reduction in arterial pressure [14]. The net effect of ketaminemidazolam TIVA provides haemodynamic stability with minimal fluctuations in heart rate and arterial pressure. This has been reported in cardiac surgical patients [15]. Pulmonary vascular resistance is increased by ketamine [16], but – in this study – the concomitant ketaminemidazolam infusion does not significantly change the pulmonary vascular resistance.

Ketaminemidazolam TIVA has been previously reported by Restall and his colleagues during short elective surgery [17]. They used a low dosage of ketamine and midazolam for both induction (midazolam 0.07 mg kg−1, ketamine 1.0 mg kg−1) and maintenance (midazolam 50 µg kg−1 h−1, ketamine 2.0 mg kg−1 h−1) of anaesthesia. This resulted in cardiostimulatory actions in the form of tachycardia and hypertension after tracheal intubation and persisting for the first 30 min of anaesthesia. The very low total drug dosage and the relatively short mean duration of anaesthesia (73.6 min) resulted in rapid awakening (12.5 min). The authors did not recommend this technique to be adopted to the exclusion of other methods, but suggested that it could be used when other techniques are not available or unsuitable.

Ketamine acts as non-competitive antagonism of glutamate action at the NMDA-type glutamate receptor complex. The analgesic and anaesthetic properties of ketamine are primarily attributed to NMDA channel blockade [18–19]. Midazolam acts on the benzodiazepine receptor – part of the GABA receptor complex. The combined ketaminemidazolam TIVA obtunded surgical-induced pain and provided satisfactory analgesia after surgery.

The avoidance of inhalational anaesthetic agents, either gases or vapours, and the use of oxygen-enriched air for lung ventilation resulted in a lack of distension of both colon and ileum. This made surgical access easier. Body temperature remained about 35°C thus obviating the occurrence of prolonged shivering and associated hypoxaemia in the early postoperative period.

The recovery profile proved satisfactory. Generally, patients are quiet with a profound analgesia. A small dose of an analgesic proved quite effective in those patients who did experience pain. Five patients spoke incoherently: they were not deemed to have recovered. The digit forward and digit backward test is part of Wechsler adult intelligence scale and Wechsler memory scale used as a general intelligence test and memory test [4–20]. It is used to measure the digit span of memory, and the Wechsler memory scale was applied to measure recovery/delay of memory after thiopental anaesthesia for electroconvulsive therapy [21].

It is our contention that for a successful completion of prolonged abdominal surgery, anaesthesia has to provide adequate relaxation, haemodynamic stability, and a ‘roomy’ abdominal cavity. This is best provided by ketaminemidazolam TIVA in the approximation regimen doses. This technique is simple, effective, and quite inexpensive. It should be adaptable for use, generally in prolonged surgery and especially for prolonged abdominal surgery. Once established, minimum interference is required and junior anaesthetists can be taught the method easily. Ketaminemidazolam TIVA, with its low cost and safety, is a comprehensive pharmaco-economic approach to anaesthesia. Ketaminemidazolam TIVA provides a receptor-targeted type of prolonged anaesthesia that provided a satisfactory operative field, stable haemodynamic and oxygenation variables albeit a demanding recovery and postoperative profile.


We would like to acknowledge the help offered by M. E. Khater, Professor of Psychiatry, University of Mansoura, in the interpretation of the recovery data.


1 Knox JWD, Bovill JG, Clarke RSJ, Dundee JW. Clinical studies of induction agents. XXXVI. Ketamine. Br J Anaesth 1970; 42: 875–885.
2 Tweed WA, Minuck M, Mymin D. Circulatory responses to ketamine anesthesia. Anesthesiology 1972; 37: 613–619.
3 McNulty SE, Gratch D, Kim JY. Comparative vascular effects of midazolam and lorazepam administered during cardiopulmonary bypass. Anesth Analg 1994; 79: 675–680.
4 Meleika LK, Ismail ME. Wechsler-Bellevue Intelligence Scale for Adults and Adolescents, Arabic version. Cairo: Al-Nahda, 1991: 19–21.
5 Clements JA, Nimmo WS. The pharmacokinetics and analgesic effect of ketamine in man. Br J Anaesth 1981; 53: 27–30.
6 Grant IS, Nimmo WS, McNicol LR, Clements JA. Ketamine disposition in children and adults. Br J Anaesth 1983; 55: 1107–1111.
7 White PF. Clinical use of newer intravenous induction drugs. Anesth Analg 1988; 67 (Suppl.): 106–112.
8 White PF. What's new in intravenous anaesthesia. Anesth Analg 1990; 69 (Suppl.): 105–114.
9 Nimmo WS. Practical pharmacokinetics for the clinician. Anesth Analg 1992; 71 (Suppl.): 25–28.
10 Schüttler J, Zsigmond EK, White PF. Ketamine and its isomers. In: White PF, ed. Textbook of Intravenous Anesthesia. Baltimore: Williams & Wilkins, 1997: 171–188.
11 Ramsay MAE, Savage TM, Simpson ERJ, Goodwin R. Controlled sedation with alphaxalone-alphadolone. Br Med J 1974; 2: 656–659.
12 Stanley TH. Blood pressure and pulse rate responses to ketamine during general anesthesia. Anesthesiology 1973; 39: 648–649.
13 Chodoff P. Evidence for central adrenergic action of ketamine. Anesth Analg 1972; 51: 247–250.
14 Lebowitz PW, Cote ME, Daniels AL et al. Comparative cardiovascular effects of midazolam and thiopental in healthy patients. Anesth Analg 1982; 61: 771–775.
15 Chai M, Thanga D, Morgan E, Viljoen J. Ketamine-midazolam continuous infusion in cardiac surgical patients. Anaesthesia 1989; 44: 364–364.
16 Gooding JM, Dimick AR, Ravakoli M, Corssen G. A physiological analysis of cardiopulmonary responses to ketamine anesthesia in non-cardiac patients. Anesth Analg 1977; 56: 813–816.
17 Restall J, Tully AM, Ward PJ, Kidd AG. Total intravenous anaesthesia for military surgery. A technique using ketamine, midazolam and vecuronium. Anaesthesia 1988; 43: 46–49.
18 Irifune M, Shimizu T, Nomoto M, Fukuda T. Ketamine-induced anesthesia involves the N-methyl-D-asparatate receptor-channel complex in mice. Brain Res 1992; 569: 1–9.
19 Oye I, Paulsen O, Maurset A. Effects of ketamine on sensory perception: evidence for a role of N-methyl-D-asparate receptors. J Pharmacol Exp Ther 1992; 260: 1209–1213.
20 Wechsler D, Stone C. Manual of Wechsler Memory Scale. Cleveland: The Psychological Corporation, 1955.

ANAESTHETIC TECHNIQUES; infusion; ANAESTHETICS; ketamine; midazolam; SURGERY; abdominal.

© 2001 European Academy of Anaesthesiology