Usefulness and limitations of rapid urine dipstick testing for joint-fluid analysis. Prospective single-center study of 98 specimens
Résumé
OBJECTIVE: To evaluate the diagnostic performance of rapid urine reagent strip testing of joint fluid in separating mechanical from inflammatory disease. METHODS: In a prospective single-center 12-month study of joint fluid specimens, leukocyte esterase reagent strip testing (LERST) was compared to leukocyte counts used as the reference standard. Leukocyte counts greater than 2000/mm(3) were taken to indicate inflammation. Reproducibility of LERST was evaluated by testing 73 specimens twice and computing Cohen's kappa coefficient. RESULTS: Ninety-eight joint fluid specimens (26 with mechanical and 72 with inflammatory characteristics) were evaluated. LERST had 79.2% sensitivity, 92.3% specificity, 96.6% positive predictive value, 61.5% negative predictive value, a positive likelihood ratio of 10.3, and a negative likelihood ratio of 0.23. The kappa coefficient was 0.70 (0.53-0.87). Two negative LERSTs a few minutes apart had 80% negative predictive value and a negative likelihood ratio of 0.08. CONCLUSION: LERST of joint fluid is a rapid means of satisfactorily separating mechanical from inflammatory joint fluids.