Description of the Evidence
Repetitive hematuria testing
Two studies using Ames Hemastix, a chemical reagent strip for hemoglobin that correlates with microscopic urinalysis in detecting hematuria, were conducted in geographically defined (Madison, Wisconsin and Leeds, England) populations of middle-aged and elderly men using repetitive home reagent strip testing. In each program, patients were solicited from patient care registries. Men with histories of previous urologic malignancies, or known causes of hematuria, or who were noncompliant were eliminated. In the four studies performed (one pilot study and one larger study at each site), 45% to 55% of solicited individuals took part. In these studies, 1.2% to 1.3% of all participants were found to have bladder cancer (all TCC). Only 1 of the 21 patients in the first study [49,58,63] and none of the 26 detected in the second study had stage T2 or greater malignancy. As a limitation of repetitive hematuria screening in a general population of men aged 50 years and older, more than 90% of individuals with positive tests upon initial work up were found not to have bladder cancer. In the Wisconsin hematuria screening studies, all patients who were hematuria positive with negative work ups or who were found to have no serious disease were followed for at least 24 months, with no findings of developing bladder cancer. Similarly, at least 18 months after their last testing, no screening participant (with or without hematuria) had died of bladder cancer. It is possible that longer follow-up is necessary to prove that these participants did not have bladder cancer; however, such studies are not available. The relatively low positive predictive value of repetitive hematuria testing (7.6% for bladder cancer and 11.6% for all malignancies) [49,63,64] raises questions about the practicality of this mode of screening.
Other possible screening modalities
The accuracy of voided urine cytology in detecting bladder cancer has been evaluated primarily in patients with histories of bladder cancer who are undergoing cystoscopic surveillance or as a routine test performed in all patients attending a large urology office in a multispecialty clinic. In the studies of patients with histories of bladder cancer, voided urinary cytology was effective in diagnosing 20% to 40% of grade I TCCs, 20% to 50% of grade II malignancies, and 60% to 80% of grade III/Tis cancers.[65,66] Although such studies were not performed in patients without either hematuria or histories of recurrent bladder tumors, a major concern for screening purposes is the lack of sensitivity for well-differentiated and moderately differentiated TCCs and the large proportion of specimens in which an insufficient number of cells were present for any cytologic diagnosis to be offered. Although false-positive results were exceedingly rare, the lack of sensitivity even in this highly suspect population make voided urine cytology an inappropriate test for screening the general population. No studies have looked at outcome of cytologic screening on disease-related mortality in a non-industrially exposed population. Outcomes of patients screened at the urology clinic are also not available.