Regional variation in intervention rates: what are the implications for patient selection?

Quantitative evaluation of prostatectomy for benign prostatic hypertrophy under a national health insurance law: a multi-centre study

Assessing regional variation between various medical centres in diagnostic and surgical processes is an approach aimed at evaluating the quality of care. This study analyses the differences between eight medical centres in Israel, where all citizens are covered by medical insurance, through the National Health Insurance Law (NHIL). The analysis refers to the diagnostic process, type of surgery and immediate post-surgical complications associated with prostatectomy for benign prostatic hypertrophy (BPH), which is the most frequent surgical procedure performed on men aged 50+. The study sample was comprised of 261 consecutive prostatectomy patients operated on in eight Israeli medical centres (MC), located in various parts of the country, between November 1996 and April 1997. Co-operation with participating directors of surgical wards was obtained after confidentiality of information had been assured. Surgeons in selected departments abstracted data routinely recorded in the patient's file and filled-out a standard one-page questionnaire. The following items were included: age, the presence of accompanying chronic diseases, preoperative tests, type of operation, and post-surgical complications. In the various MCs 32.6% of the patients underwent more than five preoperative tests ranging from 8.9% to 88.9% (<0.01). Assessment of kidney and bladder normality ranged from 75% to 100% (P < 0.01). The rate of patients whose prostatic symptoms (I-PSS) were assessed ranged from 0% to 79% (P < 0.01). There were also differences in severity of prostatism between the MCs, with severe symptoms ranging from 54.0% to 89.3% (P < 0.05), for type of operation performed (for 'open' prostatectomies, 35.4% to 68.0%, P < 0.01) and post-operative complications (19.0% to 41.6%, P = 0.07). After controlling for case-mix, type of operation was the most important predictor for post-surgical complications. MCs with low volume of surgeries had a higher rate of postoperative complications. We conclude that diagnostic and type of operation and post-surgical complications differed between various MCs. Participating surgeons were willing to fill out a one-page standard questionnaire from data routinely recorded in patients' files.