Infiltration with local anaesthetics with addition of adrenaline or noradrenaline has been used since 1910 prior to skin incision for craniotomy. In a controlled series of patients scheduled for intracranial operations, systemic blood pressure was measured intra-arterially after infiltration with 15-20 ml 0.5% lignocaine with nonradrenaline 1 mu gr/ml. A marked increase in systemic blood pressure was seen (125% of values prior to infiltration). In a control series, where saline was injected, a slight fall in blood pressure was registered (p smaller than 0.01). As hypertensive episodes increase intracranial pressure in patients with impaired cerebral autoregulation, or promote the risk of haemorrhage in aneurysm surgery, this technique is considered a hazard to patients undergoing intracranial operations.

Patients wit h severe traum ati c br ain injuries still present an impo rtant prob lem in neuro sur ger y. The aim o f rrearmenr is to prevenr or diminish seconda ry brain da mage caused by cerebraI oedema. Treat ment consists of ba rbiru rares, sreroids and osmotic agents combined with hyper ventil ati on. Clinical evidence has been p resenred to show the favour able infl uence of hyperventil ation on int ra cerebral acidosis and hypoxia (6, 7). Barbiturat es a re considered to protect da maged brain tissue against hypo xia du ring br ain swelling after acute severe head injuri es, and ma y ea use a rapid reducrion of inereased intracra nial pressure (16 , 17). In 1976 srudies were presented sho wing th at high-dose dexa meth ason e rreatm enr co uld reduee morrali ty and impr ove rhe final ourco me th rou gh a beneficial effeet on br ain oed ema (4, 5). A compa rison of the result s in series o f pa tienrs with head injuri es has been diffieult, eithe r beea use of inadcq uare pa rienr definiti on or differences in clinical classificat ion. Th e recent inrroducrion of rhe Glasgo w co ma scale and th e Glasgow o utcome seale has met rhis difficulty (9, 10, 11,18). In ou r neuro surg ica l int ensive ca re unit pati ents with tra umatic brain inju ry ar e given a combined thera py consisring of highdose dexa methaso ne, ba rbiturate s, hyperventil ation and fluid restriction. At the sta rt o f th e rreatment the patients a re evaluated by means of the Glasgow co ma seale (18), a nd th e final ourco me deserib ed aceording to the ou rcome score system described by [ennet et al. (9).T his paper reports rhe ourcome in 45 patien rs wirh severe brai n in juries, rrearmenr is discussed, and rhe results are compared wirh rhose from orh er cenrres p reviously pu blished.CIinieaI materia l From J anu ar y 1977 to No vember 1978 67 conseeutive patients wirh severe head injuri es were admitt ed to our intensive ea re unit. All pati ents fulfilled the crireria for severe head injury described by th e Glasgow group (18 ), a nd scored 6 points or less acco rding ro rhe Glasgow coma scale. In all cases injur ies had been caused b y tr affic accidenrs. T went y-t wo patients had to be excluded fro m th e srudy. Ten were dead on a rrival or died within 12 hours, from in juries not amenab le to rrear ment . T welve pati ent s admitred mor e th an 24 hours afrer injur y, were not included in the follo w-up study. SummaryThe outcome in 45 consecutive pa tients followi ng severe head injuries is pr esented . All patient s we re evaluated acco rding to the Glasgow Coma Sca le and the Glasgow O utco me Scale. All pa tients reeeived a eombined treatment consisting of dexam ethason e, ba rbit urate and hyperventil ation. A primary operative interve ntion wa s perfo rmed in 14 parients with intraeran ial haernat om as. T here wa s a good recovery in 53 %, 20% were mod erat ely disabled, 11% we re severely disabl ed . 16% died or remained in a vegeta tive srare. Th e pati ents were aged berween 5 a nd 83 years , 4 9% were under 20 years, a...