Mortality after transurethral and open prostatectomy in Scotland.

Abstract

To use the linked medical and death records in Scotland to investigate the possible increased mortality that has been reported after transurethral prostatectomy (TURP) compared with open prostatectomy. Scotland has maintained linkable hospital, cancer and death records for more than 20 years, representing one of the largest such databases in the world. From these computerized records, data on various cohorts of men aged 55-84 years selected from 81,997 men who underwent prostatectomy in Scotland between 1968 and 1989 were analysed. The risk of late mortality was calculated for each type of operation, whether there was prior comorbidity and for a range of specific causes (cancer, respiratory and circulatory conditions) after prostatectomy. Among the largest cohort, consisting of 65,519 men who underwent prostatectomy between 1968 and June 1989, the relative risk of late mortality after TURP compared with open prostatectomy was 1.13 (95% CI, 1.10-1.16), after controlling for age and the presence of a diagnosis of cancer. A more restricted cohort of 18,732 men who underwent prostatectomy between 1974 and 1979 allowed adjustment for prior hospitalization with, or concurrent diagnosis of, circulatory and respiratory conditions. In this cohort, the relative risk of late mortality after TURP as compared with open prostatectomy was 1.15 (95% CI, 1.11-1.19) after adjusting for prior and comorbidity and age. Finally, a cohort of 'healthy patients' restricted to the 6932 men who underwent prostatectomy from 1974 to 1979 and with no evidence of hospitalization in the previous 5 years or any current diagnosis other than benign hypertrophy of the prostate, showed a relative risk of 1.14 (95% CI, 1.07-1.21). There was no evidence of an increased risk of dying from circulatory disease in general, ischaemic heart disease or acute myocardial infarction after TURP as opposed to open prostatectomy. However, there was an increased risk of dying from respiratory conditions and from cancer, especially of the prostate and bladder. The analysis suggested the possibility that open prostatectomy may have cured some patients with early prostatic cancer, because the late death rate from prostatic cancer was greater in patients who underwent TURP than open prostatectomy. The present analysis confirmed the increased risk of late mortality after TURP compared with open prostatectomy, as shown in previous studies based on administrative records. However, limitations in the coding of comorbidities and the absence of coding of more subtle aspects of the condition of the patient which may influence the choice between the forms of prostatectomy mean that the differential mortality after the two procedures could still be a reflection of the pre-operative selection of patients rather than the effects of the surgical procedure.