THE AGREEMENT AMONG UROLOGICAL EXPERTS ON THE DIAGNOSTIC MANAGEMENT OF PATIENTS WITH COMMON UROLOGICAL PROBLEMS

Objective To assess the level of agreement among randomly selected int ernational urologists on the diagnostic management of patients with pr ostate cancer, bladder cancer, urinary stones or lower urinary tract s ymptoms (LUTS) arising from benign prostatic hyperplasia (BPH). Method s A computer program was used to provide an unbiased format of 53 simu lated patients, comprising 13 with prostate cancer, 10 with bladder ca ncer, 10 with stones in the upper urinary tract and 20 with LUTS from BPH. For each case, the history was provided to the user while informa tion from 60 diagnostic tests could be chosen interactively. Thirty-th ree university-based urologists participated in the study. The probabi lity that a certain test was used by them in a certain patient [P(test )] and the related costs (Swedish 1995 prices) were recorded. The prob ability that two urologists would agree (relative measure of agreement , RMA) on whether or not to use one particular test in a certain case was RMA(test) = P(test)(2) + [1 - P(test)](2) and the mean of this RMA (test) for a certain patient [RMA(case)] was used as a measure of the inter-individual agreement among the urologists on the diagnostic mana gement. The significance levels of the generalized kappa statistic, KG , were also calculated. The correlation between the RMA(case) and the diagnostic groups was analysed. Results The K-G was statistically sign ificant for all cases; the RMA(case) was significantly correlated with the diagnostic groups (rs = 0.86). The agreement in the diagnostic ma nagement was the strongest for stones, then for bladder cancer and pro state cancer, and the weakest for BPH. The mean cost for the diagnosti c evaluation for one case varied from $455 to $1771 (mean 898) and var ied in the diagnostic groups, i.e. $1718 for prostate cancer, $947 for bladder cancer, $400 for stones and $594 for BPH. Conclusion The diag nostic management of urological patients varies greatly among urologic al experts from the industrial world. As a consequence, the related di agnostic costs might vary by about 400% if prices were similar everywh ere. The agreement on the diagnostic management of cases is strongly c orrelated to the diagnosis. LUTS from BPH seems to be managed with the poorest agreement.