254 provide improved immunosuppression or merely more immunosuppression? The higher incidence of opportunistic infections in the mycophenolate mofetil-treated patients suggests the latter. There are established methods of testing whether immunosuppressant drugs act synergistically or additively when used in combination but the design of this study precludes their use.’ I C G Winearls Oxford Regional Renal Unit, Oxford Radcliffe Hospital Trust, Headington, Oxford OX3 7LI, UK 1 Berenbaum MC. Synergy, additivism and antagonism in immunosuppression: a critical review. Clin Exp Immunol 1977; 28: 1-18. High-dose frusemide for cardiac failure SiR-Many patients, especially elderly people, need further diuretic treatment in addition to angiotensin-converting- enzyme (ACE) inhibitors. Among elderly people with advanced disease, a lack of response to low-dose frusemide is common. The options are to use a larger dose of frusemide or to add another diuretic agent. High-dose frusemide therapy (250-400 mg per day), given by mouth, by intravenous bolus, or by continuous infusion is effective in refractory congestive cardiac failure.’ Many physicians are, however, reluctant to use high-dose frusemide, mainly because of concerns about renal function. The alternative approach is to use a combination of diuretics, usually frusemide with a thiazide or potassium- sparing agent. Although this approach can be effective, severe electrolyte disturbances can occur. 2 ’ We reviewed 15 geriatric inpatients (eight male, six female; mean age 83 [65-93] years) admitted to a rehabilitation unit with left, congestive, or biventricular cardiac failure, who were treated with moderate to high doses of frusemide. All 15 patients were on clinically optimum doses of ACE inhibitors at discharge and 12 were being so treated on admission. 13 had atrial fibrillation and were receiving digoxin. The average daily frusemide dose was 193 mg (SD 136, maximum 500) on admission and 297 mg (181, 750) at discharge (p<0006). Mean serum creatinine was 161 jjLmol/L (56-7, 325) on admission and 160 jjLmol/L (77-7, 392) at discharge (not significant). There was an average weight loss of 5-6 kg and an average rise in the Barthel score of 4 points. 11 patients returned to their previous home (six) or hostal (five), three were relocated to a hostel (one) or nursing home (two), and one patient had an acute myocardial infarction and died in hospital. There were no significant electrolyte disturbances apart from hypokalaemia that responded to potassium supplementation. In the setting of cardiac failure resistant to an optimum dose of an ACE inhibitor and low-dose frusemide, although the addition of a thiazide diuretic may be a useful adjunct, we conclude that the use of high-dose frusemide was an effective and safe approach to management of severe cardiac failure. *G Waterer, M Donaldson Departments of *Neurology and Geriatrics, Royal Perth Hospital, Perth, Western Australia 6001 1 Gerlag PG, van Meijel JJM. High-dose frusemide in the treatment of refractory congestive heart failure. Arch Intern Med 1988; 148: 286-91. 2 Oster JR, Epstein M, Smoler S. Combined therapy with thiazide-type and loop diuretic agents for resistant sodium retention. Ann Intern Med 1983; 99: 405-06. Ibuprofen versus sumatriptan for high-altitude headache SiR-Ibuprofen, a non-steroidal anti-inflammatory drug, has been shown to be superior for treatment of high-altitude headache than placebo.’ Bartsch and colleagues2 recently proposed that sumatriptan a 5-HTB receptor agonist, might be even more effective than ibuprofen.- To test this hypothesis, we compared oral ibuprofen and sumatriptan in a randomised, double-blind, within-patient crossover trial. 33 volunteers (18 men, 15 females, mean age 29, range 19-52) were transported from 200 m to an altitude of 3480 m (by bus and cable car) and stayed there for 18 h. Medical history and clinical examination of the subjects revealed no evidence of existing acute or chronic diseases. Nutrition and physical activity were standardised 12 h before and during the altitude sojourn. 5-15 h after arrival at altitude, 13 (9 males, 4 females, mean age 36, range 23-52) developed headache. Treatment was started when a headache score of 2 (0=none, l=mild, 2=moderate, 3=severe) and an acute mountain sickness score3 greater than 2 were reached. The patients were randomised to ibuprofen (600 mg) and sumatriptan (100 mg), respectively. Scoring was repeated 2 h after drug intake. If patients needed further medication, they were given the drug not initially administered. Heart rate, blood pressure, and oxygen saturation were measured in all subjects before and during the altitude sojourn. Score changes within groups were analysed with Wilcoxon’s signed rank tests. To compare differences for continuous variables we used t tests. All statistical tests were 2-tailed. After intake of ibuprofen, nearly complete relief of the headache occurred in these patients (n=7) within 2 h. In contrast, no decrease of the headache score was found in patients (n=6) treated with sumatriptan. 5 of these patients needed further medication and achieved complete relief within 2 h of subsequent ibuprofen intake (figure). A significant reduction of the acute mountain sickness score was observed within the ibuprofen group (mean 4-3 [SD 049] vs 0-57 [0-98], p=0018). No such reduction occurred within the sumatriptan group. No assessments except headache score were made for the sumatriptan group after ibuprofen intake. The mean values of heart rate, blood pressure, and oxygen saturation did not differ between the group with headache and the group without headache throughout the investigation. However, heart rates tended to decrease 2 h after treatment with ibuprofen (88-7 [16.7] vs 75-0 [15-3], p=0.1) probably due to the decrease in pain- related sympathetic adrenergic stimulation. No heart rate changes occurred in the group treated with sumatriptan. No adverse effects were observed. Figure: Changes in mean (SD) headache scores after treatment with sumatriptan or ibuprofen, or ibuprofen after previous sumatriptan p values calculated with Wilcoxon’s signed rank tests.