The outcome in bleeding gastric and duodenal ulcer has been compared in two hospitals in Nottingham which take their patients randomly from a common source according to a fixed daily admission rota. Bleeding seemed equally severe in the two groups but 66 (32%) of 206 patients were operated upon at hospital A compared with 44 (46%) of 96 at hospital B (P=0.03), where operations also tended to be done earlier. Postoperative mortality-rates were virtually identical at the two hospitals, but at hospital B the overall death-rate from bleeding ulcer was higher, partly because of its higher operation-rate. The value of an aggressive surgical policy in bleeding peptic ulcer is questionable.

Severe systemic hypertension developed in a patient who took indomethacin shortly after ingesting an appetite suppressant ('Trimolets') containing phenylpropanolamine. The hypertension was attributed to a drug interaction whereby the inhibition of prostaglandin synthesis by indomethacin exacerbated the sympathomimetic effects of phenylpropanolamine. It is suggested that withdrawal of preparations containing phenylpropranolamine from general use should be considered in view of their potential for adverse reactions with other commonly used drugs and their doubtful therapeutic value.

106 patients with anxiety, phobic, and depressive neuroses referred to the outpatient clinics of 6 psychiatrists were randomly assigned to outpatient care or two forms of psychiatric day care. Ratings of symptoms and social adjustment were recorded before treatment and after 4 and 8 months in 89 patients. There were no important differences in the outcome of day care and outpatient treatment although patients were more satisfied with the outpatient service. Because outpatient care is more economical of time and personnel it should be preferred to psychiatric day care for many neurotic disorders.

Apomorphine in combination with a peripheral dopamine receptor blocker (domeperidone) was administered to four parkinsonian patients in a double-blind placebo-controlled study. The therapeutic efficacy of apomorphine was not reduced by domperidone, while nausea, drowsiness, sedation, and arterial hypotension were prevented. Combination of domperidone with dopamine agonists may result in more effective treatment of Parkinson's disease.

20 randomly selected infants of 33 weeks gestation and under at birth were allowed to go home from the neonatal unit provided they were clinically well and had passed the nadir of their postnatal weight-loss and provided home conditions were satisfactory. Weight-gain at home was satisfactory and there was no increased rate of hospital readmission compared with 20 randomly selected pre-term infants who were discharged at a more traditional weight of 2200 g. There is therefore little justification for the widespread practice in Britain of delaying discharge of preterm infants until they reach a predetermined weight (usually 2000--2500 g).

In a prospective randomised trial in which 93 patients undergoing elective colorectal operations were given a short prophylactic course of metronidazole and kanamycin orally or systemically, postoperative sepsis occurred in only 3 (6.5%) of those given antimicrobials systemically, compared with 17 (36%) of those given oral prophylaxis (P less than 0.01). 15 of the 17 infections in patients who received antimicrobials orally were due to kanamycin-resistant bacteria present in the colon at operation. Bacterial overgrowth of Staphylococcus aureus was recorded in 6 of the patients who received oral therapy. Antibiotic-associated pseudomembranous colitis occurred in 7 patients, 6 of whom had received prophylaxis orally. These results indicate that oral administration of prophylactic antimicrobials in colon surgery should be avoided because of the risks of bacterial resistance, superinfection, and antibiotic-associated pseudomembranous colitis. Systemic per-operative antimicrobial prophylaxis is safer and more effective.