EDITOR: Maternal hypotension is the most frequent complication of a spinal anaesthetic for Caesarean section. A range of stratagems is currently used to prevent or minimize hypotension, but there is no established ideal technique. Different approaches include prehydration with crystalloids or colloids, lower limb compression, and pharmacological prophylactic protocols using ephedrine, phenylephrine, angiotensin II or metaraminol [1-4]. Loughrey and colleagues described the prophylactic administration of intravenous (i.v.) boluses of ephedrine for elective Caesarean section under spinal anaesthesia [1]. They concluded that a prophylactic bolus of ephedrine 12 mg i.v. given at the time of the intrathecal block - associated, if necessary, with rescue boluses - leads to a lower incidence of hypotension during spinal anaesthesia compared with i.v. rescue boluses alone. We use compound sodium lactate solution (500 mL) given as a preload over 10 min before the intrathecal injection of hyperbaric bupivacaine. The bolus method is chosen over an i.v. infusion to provide a fast and reliable early prophylactic dose of vasopressor as the rapid sympathetic blockade develops. In the recent study of Kee and colleagues, the smallest prophylactic i.v. ephedrine dose was 30 mg. In that study, the parturient was preloaded with lactated Ringer's solution 20 mL kg−1 and the spinal block was performed using hyperbaric bupivacaine 10 mg and fentanyl 15 μg [5]. In another trial by Vercauteren and colleagues, a single i.v. dose of 5 mg limited the severity of hypotension in prehydrated subjects receiving hyperbaric bupivacaine 6.6 mg and sufentanil 3.3 μg [6]. Jackson and colleagues concluded that volume preloading was not essential to prevent spinal-induced hypotension in Caesarean section. They used a prophylactic infusion of ephedrine 60 mg in Hartmann's solution (500 mL) given according to maternal arterial pressure [7]. During 2001, our Medical Centre performed 181 Caesarean section deliveries of which 94% were under spinal anaesthesia. The usual policy we adopted consisted of prehydration with lactated Ringer's solution (300 mL) over 10 min. At the same time, hyperbaric bupivacaine 0.8-1 mL 1%, combined with fentanyl 25 μg, was given intrathecally plus a slow baseline infusion of ephedrine 25 mg in 500 mL lactated Ringer's solution i.v. The aim of the ephedrine infusion was to keep arterial pressure at a baseline level and never below 90 mmHg systolic. This infusion was continued until the extraction of the baby and then it was gradually stopped. We did not observe rebound hyper- or hypotension. The mean total dose of ephedrine used was 12.5 ± 2.3 mg (mean ± SD). Only five patients required rescue boluses of ephedrine 5 mg because of symptoms suggestive of hypotension, e.g. nausea and vomiting, without any recorded fall of arterial pressure. We agree that the prophylactic ephedrine dose is safe and efficacious in reducing the incidence of maternal hypotension, but we also believe that an infusion of the drug, according to the patient's requirements, is much safer than the bolus technique. In our patients, the preload was reduced to 300 mL - similar to Jackson and colleagues' study [7] in which the minimal preload was 200 mL - while the concentration of ephedrine in our protocol was 0.05 versus 0.12 mg mL−1. In conclusion, further randomized control trials would be useful to determine the optimal dose of ephedrine, the method of administration, the volume and the kind of preload. Y. Leykin F. Rubulotta Department of Anesthesia and Intensive Care; Santa Maria degli Angeli Hospital; Pordenone, Italy