Objective: To characterise the haemodynamic profile after epidural injection of high-dose clonidine for postoperative pain management and to establish recommendations for the therapy of haemodynamic instabilities.
Design: 20 patients with major surgery on pancreas, stomach or infrarenal aorta took part in the study. Anaesthesia was a combined epidural/inhalational regimen with bupivacaine 0.25%, enflurane, oxygen/nitrous oxide, fentanyl 0.1 mg and pancuronium. Postoperative analgesia consisted of morphine 50 micrograms/kg in 10 ml NaCl 0.9% for the first 12 postoperative hours; if pain > = 5 points on the VAS occurred after > 12 h postoperatively clonidine 8 micrograms/kg in 10 ml NaCl 0.9% was injected epidurally and the pain intensity (self-assessment by the patient using the visual analog scale) and circulation (invasive pressure monitoring, pulmonary artery catheter) was monitored for 60 minutes in ten minutes intervals.
Results: The reduction of the initial VAS score of 6 was 50% after 20 minutes and 100% after 60 minutes. We observed a significant decline in heart rate (87 +/- 11 (t0), 74 +/- 10 min-1 (t60)), mean arterial pressure (97 +/- 17 (t0), 72 +/- 15 mmHg (t60)) and cardiac output (8.7 +/- 1.3 (t0), 7.0 +/- 1.3 l.min-1 (t60)) (all p < 0.001) and no change of systemic vascular resistance. Filling pressures (CVP and PCWP) remained stable. In 9 patients the mean arterial pressure fell below 60 mmHg (always within the first 40 min); 6 of these patients responded to infusion of a colloid (500 ml of hydroxyethyl starch at > = 2 ml/kg.min) whereas the other 3 patients needed a bolus injection of a betamimetic catecholamine (theodrenaline/cafedrine, Akrinor).
Conclusion: Epidural clonidine 8 micrograms/kg causes rapid and intense analgesia. Haemodynamic instability is a consequence of a drop in heart rate and has to be treated accordingly. The application of a pure vasopressor does not seem to be indicated taking in account the fact that the total peripheral resistance remains unchanged and in the normal range.