Individual Learning Curve Reduces the Clinical Value of Urinary Cytology

BackgroundTo define the learning curve of a single cytologist as a limitation of urinary cytology and to compare his results with a reference center for urinary cytology.Patients and MethodsCytologic and histologic findings of patients undergoing transurethral resection of the bladder for suspicion of bladder cancer were reviewed. Cytologic evaluations from a single local cytologist during his learning period and of a cytologist of a national reference center were compared. Changes of sensitivity and specificity of cytology from the local cytologist over the time period were assessed. Differences were estimated by using kappa statistics.ResultsThe local cytologist evaluated 1034 cytologic findings. Sensitivity and specificity of cytology estimated by the local cytologist were 86% and 66% at the beginning of the learning period, but 68% and 84% at the end of it (P ≤ .05). In high-grade carcinomas, sensitivity did not decrease over the years (P > .05). In contrast, the sensitivity in the diagnosis of low-grade tumors decreased from 86% to 56% (P ≤ .05). The reference center estimated a constant sensitivity and specificity of 59% and 97%, respectively. Agreement of findings between local and national cytologists increased significantly over the learning period (P < .05).ConclusionThe individual learning curve has a significant impact on the quality of urinary cytology. Both specificity of cytology and sensitivity for low-grade tumors changed significantly when done by a local cytologist at the beginning of learning period. These findings emphasize the impact of the individual learning curve on the clinical value of urinary cytology in diagnosis of bladder cancer.