Joint inflammation, predominantly of the lower limbs, occurring some weeks after urogenital or gastrointestinal infection is classified as reactive arthritis (ReA) but there is no general agreement on diagnostic criteria, especially if the preceding infections are asymptomatic. The same is true for
Lyme disease (LD) which is caused by Borrelia burgdorferi (BB). Determination of antibody titre or culture of urethral swabs and stools are often used as diagnostic tools. We examined 4 groups of patients: one with undifferentiated arthritis (Group I, n = 55), one with well-defined rheumatic diseases other than ReA (n = 43, Group II), one group without joint
disease (n = 50, Group III) and one with ReA or LD (n = 7). Specific antibacterial antibody titres in serum were measured in all patients; stool and urethral cultures were performed in all groups except the last. A calculation of positive predictive value (PPV) was done for each test. Evidence of present or previous infection with the microbes Chlamydia trachomatis (CT), Mycoplasma urethritidis (MU), Yersinia enterocolitica (YE) and BB were found in all groups. In Group I, Group II and Group III respectively, positive serological results were found for CT IgA (20%, 31%, 16%) and IgG (49%, 51%, 34%), YE (7%, 6%, 0%) and BB (17%, 2%, 10%). Positive cultures were found in Group I and Group II respectively for CT (28%, 29%) and MU (14%, 17%). Therefore no test had a significant positive predictive value for ReA in the general population and even in the rheumatology clinic the PPV for most tests was low. We conclude that these methods are of little value in the diagnosis of reactive arthritis when the preceding infection is asymptomatic.