A randomised controlled trial compared the effectiveness and toxicity in pulmonary tuberculosis of two drug regimens containing rifampicin and isoniazid given daily or twice-weekly for 4 months after a 2-month period of intensive treatment with daily isoniazid, rifampicin, and pyrazinamide. 667 patients with newly diagnosed pulmonary tuberculosis were randomly allocated to continue daily treatment with isoniazid (400 mg) and rifampicin (600 mg) or to twice-weekly treatment with isoniazid (900 mg) and rifampicin (600 mg). 544 of the 667 patients (81%) completed the 6-month course (287 of 337 [85%] treated daily and 257 of 330 [79%] treated twice-weekly). Drug toxicity was not a great problem; the treatment was permanently discontinued in only 2% of patients. There was no significant difference at the end of months 5 and/or 6 of chemotherapy between the groups treated daily and twice-weekly in the proportions with bacteriological failure (at least one positive sputum culture with more than 20 colonies) or who had died from tuberculosis (17 [6%] vs 10 [3%]). Nor was there a significant difference in the relapse rate (17 [7%] treated daily vs 10 [4%] treated twice-weekly) during follow-up of 12 months. Thus, the twice-weekly regimen was at least as effective as the daily regimen for treatment of pulmonary tuberculosis.

In a randomised, placebo-controlled, double-blind study, 822 obese patients of both sexes were given either dexfenfluramine (dF), 15 mg twice daily (404), or placebo (418) in addition to a calorie-restricted diet for 1 year. Patients in both groups lost weight significantly in the first 6 months; after 6 months dF patients had a higher cumulative mean weight loss. Dropout rates were lower in dF patients than in placebo patients, mainly because of dissatisfaction with weight loss in the latter group. More than twice as many dF patients as placebo patients achieved a given weight loss; but more dF patients than placebo patients had transient side-effects (tiredness, diarrhoea, dry mouth, polyuria, and drowsiness).

A randomised, double-blind, placebo-controlled study of L-tyrosine was done in ten subjects with narcolepsy and cataplexy. Of twenty-eight visual analogue scales rating mood and arousal, the subjects' ratings in the tyrosine treatment (9 g daily) and placebo periods differed significantly for only three (less tired, less drowsy, more alert). Ratings of daytime drowsiness, cataplexy, sleep paralysis, night-time sleep, overall clinical response, and measurements of multiple sleep latency and tests of speed and attention did not differ significantly between tyrosine and placebo periods. Dietary supplementation with tyrosine 9 g daily for 4 weeks seems to have a mild stimulant action on the central nervous system but this effect is not clinically significant in the treatment of the narcoleptic syndrome.

A carefully designed set of simulators showing cervical effacement and dilatation was used to assess the error within and between observers in a group of 36 midwives and 24 obstetricians. No observer was correct in every case. There was no significant difference between the obstetricians and the midwives in assessment of effacement or overall assessment of dilatation. However, midwives were significantly more likely than obstetricians to assess dilatation inaccurately by more than 1 cm. Inaccuracy was greatest in the simulators 5-7 cm dilated. These findings have implications for labour management and teaching.

NG monomethyl-L-arginine (L-NMMA), a specific inhibitor of the synthesis of endothelium-derived nitric oxide (NO), was infused into the brachial arteries of healthy volunteers to study the role of NO in the control of forearm blood flow. L-NMMA caused a 50% fall in basal blood flow and attenuated the dilator response to infused acetylcholine but not that to glyceryl trinitrate. These results indicate that the dilator action of endothelium-derived NO contributes to the control of basal and stimulated regional blood flow in man. Impairment of production of NO might account for the abnormalities in vascular reactivity that characterise a wide variety of disease states.

In a study to determine the effect of saturation of fats on their ability to stimulate cholecystokinin (CCK) release six normal volunteers ate five test meals containing different fats with intervals of 1 week. Plasma CCK levels were measured by a specific radioimmunoassay and the gallbladder volume was calculated from ultrasound measurements. The sodium salt of the monounsaturated fatty acid oleic acid (3.5 g) produced a significantly greater integrated CCK response than that of the saturated fatty acid stearic acid (mean [SEM] 103 [41] vs 8[41] pmol.l-1.min). The gallbladder contracted to 42 (3)% of its initial volume after oleate but remained at 89 (8)% of its initial volume after stearate. Integrated CCK responses to dietary triglycerides (30 g) also differed significantly according to the degree of saturation--277 (58) pmol.l-1.min after corn oil (predominantly diunsaturated), 143 (14) pmol.l-1.min after olive oil (predominantly monounsaturated), and 44 (12) pmol.l-1.min after suet (predominantly saturated). The finding that unsaturated fats are stronger stimulants of CCK release than saturated fats may explain the promotion of pancreatic carcinogenesis in rats by unsaturated but not saturated fats and may support the role of CCK in this effect.

FK 506 was given for immunosuppression in 14 liver recipients. The drug was used in the first 10 cases because the recipients under conventional immunosuppression had rejection, nephrotoxicity, or both. This salvage therapy was successful in 7 of the 10 attempts. 2 of the 10 patients in the original salvage group as well as 4 new patients underwent fresh orthotopic liver transplantation under FK 506 plus low-dose steroids from the outset. None of these 6 patients had rejection although 1 with preexisting cor pulmonale and coronary atherosclerosis died of a myocardial infarction. In addition, 2 of the 14 liver recipients were given cadaveric kidneys, either from the same donor or from a different donor, and a third was given a pancreas as well as a kidney from the liver donor. There were no rejections of the kidney and pancreas grafts, and serious side-effects were not encountered.

In a double-blind randomised study, eight healthy men received equimolar amounts of human or porcine insulin by infusion (50 mU/kg per hour). Insulin potencies, in terms of the amount of glucose infused to maintain euglycaemia, were almost identical. Hypoglycaemia (blood glucose concentration 2 mmol/l or below) was then induced and the symptoms and hormonal counter-regulatory responses were recorded. The number of sympathoadrenergic (but not neuroglycopenic) symptoms was significantly greater with porcine than with human insulin, as was the integrated noradrenaline response. Glucagon, growth hormone, cortisol, and adrenaline responses were similar for the two insulins.

In a blind trial, 77 patients were randomised to receive first cadaver kidney allografts that had been perfused either with a pair of CD45 monoclonal antibodies (mAbs), in an attempt to reduce the immunogenicity of passenger leucocytes, or with control human albumin solution. No complications of mAb perfusion were observed. Patient and allograft survival were similar in both groups. Rejection episodes were recorded in 7 (18%) of the patient with mAb perfused allografts compared with 24 (63%) of the controls.

A computer program for diagnosis of acute abdominal pain (CAD-A) was compared with an amended program (DIAG) and with clinical performance. The programs were tested at three Scottish hospitals, representing a range of practice, in 6962 cases. Accuracy of CAD-A lay in the range 48-59%, and of DIAG 56-62%. Clinical accuracy was constant at 65%. These figures suggest that computer systems based on Bayes' formula have no useful role in the diagnosis of acute abdominal pain.

The effect of psychosocial intervention on time of survival of 86 patients with metastatic breast cancer was studied prospectively. The 1 year intervention consisted of weekly supportive group therapy with self-hypnosis for pain. Both the treatment (n = 50) and control groups (n = 36) had routine oncological care. At 10 year follow-up, only 3 of the patients were alive, and death records were obtained for the other 83. Survival from time of randomisation and onset of intervention was a mean 36.6 (SD 37.6) months in the intervention group compared with 18.9 (10.8) months in the control group, a significant difference. Survival plots indicated that divergence in survival began at 20 months after entry, or 8 months after intervention ended.