Prostatectomy is effective in reducing symptoms in most men. Men who experience a substantial reduction in symptoms were more likely to report a favourable outcome. The study confirmed that approximately one-third of men reported an unfavourable result 3 months after their operation.

The clinical management of prostatectomy has been defined in a large and representative UK sample. In some circumstances consistent variations have been identified. It is not yet clear whether these variations influence outcome. These data can be used by surgeons wishing to compare their own patient management with that described here.

SUMMARYThree cases of leiomyosarcoma of the duodenum are reported. T h e literature is reviewed, and the clinical, diagnostic, and pathological features of this tumour are described. Wide surgical excision has proved to be the most effective treatment.LEIOMYOSARCOMA of the duodenum is an uncommon malignant tumour and is seldom encountered more than once or twice in the working life of any surgeon.I t receives brief mention in a few standard surgicalWe recently encountered a case in clinical practice and were able to find 2 others in the records of St. Peter's Hospital, Chertsey, from 1952 to 1970 CASE REPORTS Case 1.4vh-s. M. M., a housewife aged 53 years, was admitted to hospital in February, 1969. For a month she had felt faint and giddy and 3 days prior to admission had begun to pass melaena stools. She suffered from rheumatoid arthritis and had been treated with aspirin and prednisone for 9 years.On admission she was found to be anaemic (haemog!obin 5.2 g. per IOO ml.), but had no other abnormal signs.A blood transfusion was given but the melaena continued. A barium meal (Fig. I) showed an unusual 'diverticulum' of the third part of the duodenum, with an irregularly narrowed and possibly neoplastic neck. Atonic double-contrast duodenography (Fig. 2) reinforced the suspicion of malignancy and suggested a mass in the duodenal wall in relation to the ' diverticulum'.As the bleeding continued, laparotomy was performed. A large mobile tumour was found in the third part of the duodenum. No other abnormality was found. The third part of the duodenum containing the tumour was carefully mobilized from the pancreas and adjacent vessels, was resected, and an end-to-end anastomosis was performed. The abdomen was closed with drainage. The patient recovered uneventfully and went home on the tenth postoperative day. textbooks, and the majority of the recorded cases Macroscopic examination of the specimen (Fig. 3) in the literature were neither diagnosed preoperarevealed a tumour, 9 X 5 x 2 cm., in the wall of the tively nor recognized at laparotomy. T h e first case duodenum, the lumen. The was repofled by salis in 1920, and only 28 serosa was not involved, but the mucosa was ulcerated over an area 5 cm. in diameter (Fig. 4). MicroscopicOthers had been recorded before I9.53. Since then examination (Figs. 5, 6) showed a cellular, spindle-celled the lesion has been more frequently described. tumour arising in the smooth muscle of the duodenal wall; occasional mitoses were seen. The diagnosis was leiomyosarcoma (Dr. A. C. Cornsell).The patient is alive and in good health I year after her operation.

A 69-year-old man presented with several months' history of progressive bladder outlet obstruction culminating in acute urinary retention. A DRE revealed a uniformly enlarged prostate gland. He had mild renal impairment with a serum urea level of 9.5 mmol/L and creatinine of 131 mmol/L. His PSA level was 19.3 ng/mL and so, after catheterization, a transrectal biopsy was taken, which revealed a small focus of carcinoma in situ. Subsequently, TURP was performed and 12 g of prostate resected; the GC patient made an uneventful recovery; his PSA level returned to normal and has remained so for 18 months.

(4226) completed and returned the postal questionnaire. The response rate was higher in South West Thames (86.7%) than in the other regions (80.60/6-80/8%).The audit was well received: 91% of patients found the questionnaire easy to complete and only 2-3% of them disapproved. Completeness of data was high with both the hospital and patient questionnaires. Missing data occurred in less than 5% of responses to most questions. The attributable cost was £34.50 per patient identified or £44 for patients in whom either the treatment outcome or vital status was known three months after their prostatectomy. Conclusions-This multicentre audit of process and outcome of prostatectomy proved feasible in terms of surgeon participation, patient identification, and the quantity and quality of data collection. Whether the cost was warranted will depend on how surgeons use the audit data to modify their practice.