Abstract
Purpose: :
To evaluate the utility of various diagnostic tests in patients with uveitis.
Methods: :
New patient referrals to the uveitis clinic at Duke University were reviewed. Final diagnosis, initial visit date, final follow–up date, diagnostic testing results and review of systems were recorded. The sensitivity, specificity and predictive values of diagnostic tests were calculated.
Results: :
A total of 1202 patients were reviewed. Mean follow–up was approximately 18 months. An anti–nuclear antibody (ANA) test was positive in 81 out of 410 tested patients, but only assisted in making a diagnosis in 1 patient. An angiotensin converting enzyme was positive in 55/444 tested patients. A positive chest x–ray was obtained in 52/702 patients tested. A fluorescent treponemal antibody (FTA) was positive in 23/475 tested patients and a rapid plasmin regain (RPR) test was positive in 16/396 patients. However, only 4 patients were established to have syphilis. A positive Lyme antibody test was obtained in 3/190 patients; all were determined to be false positive. A positive PPD was found in 36/410 patients; only 4 patients were determined to have tuberculosis. Other diagnostic tests such as elevated erythrocyte sedimentation rate, positive rheumatoid factor or positive ANCA did not help to establish a diagnosis in the vast majority of patients tested. The practice of obtaining multiple laboratory tests also did not increase the chances of obtaining a diagnosis. 198 patients had at least 5 laboratory tests with 16 of them obtaining a diagnosis. An additional 13 of these patients had a diagnosis established after an evaluation by a uveitis specialist. The positive predictive values (PPV) were calculated for each of these tests. Assuming 10% prevalence of sarcoidosis among uveitis patients, the PPV of a chest x–ray and an angiotensin converting enzyme were 87% and 68% respectively. Assuming a 1% prevalence of syphilis and tuberculosis, the PPV of a FTA and a PPD was 25% and 11% respectively.
Conclusions: :
Broad testing in patients with uveitis does not assist in establishing a firm diagnosis in the majority of patients. Screening uveitis patients with laboratory tests such as ANA, ESR, rheumatoid factor, ANCA does not assist in establishing a systemic diagnosis. The predictive values of a positive PPD and FTA are very low given the low prevalence of syphilis and tuberculosis in this study population.
Keywords: uveitis-clinical/animal model • clinical laboratory testing