Purpose: We evaluate the agreement between self-reported physician diagnosed prostatitis and pain questions from the National Institutes of Health Chronic Prostatitis Symptom Index (CPSI).
Materials and methods: A randomly selected cohort of white men 47 to 90 years old from Olmsted County, Minnesota completed a study questionnaire on a history of physician diagnosed prostatitis in the preceding 2 years, including the CPSI questions. The medical records were also reviewed for physician diagnosis of chronic prostatitis during the preceding 10 years.
Results: Of 1,543 men 27 (1.7%) reported a physician diagnosis of prostatitis in the preceding 2 years. There were strong associations between self-reported prostatitis and pain at the tip of the penis (OR 6.3, 95% CI 1.4, 28.5), ejaculatory (5.9, CI 1.3, 26.6) and testicular (3.6, 1.2, 10.8) pain. The chance corrected agreement between self-reported prostatitis and pain symptoms was low at 0.01 (pubic pain) to 0.07 (pain at the tip of the penis, ejaculatory pain and testicular pain). Agreement in positive responses was also low at 3.7% (perineal pain) to 8% (pain at the tip of the penis, ejaculatory pain and testicular pain). The predictive value of the modified total CPSI score for prostatitis was 0.67, which was better than chance (0.5). Associations based on medical record ascertainment of prostatitis during followup were weak, and the area under the curve was 0.57, which was no better than chance.
Conclusions: These findings demonstrate low agreement between CPSI-like pain measures and self-reported physician diagnosed prostatitis. The stronger associations between pain symptoms and self-reported diagnosed prostatitis, and the moderate predictive ability of the modified total CPSI score measures for self-reported prostatitis compared to medical record ascertainment demonstrate the sensitivity of the CPSI to prevalent symptoms. Thus, the tool may be best used to evaluate the severity of current symptoms rather than to assess the presence or absence of prostatitis.