Diagnostic value of a 3-day course of prednisolone in patients with possible rheumatoid arthritis – the TryCort study

Uta Kiltz, Christine von Zabern, Xenofon Baraliakos, Frank Heldmann, Bernd Mintrop, Michael Sarholz, Dietmar Krause, Friedrich Dybowski, Ludwig Kalthoff, Jürgen Braun

In patients with tender and swollen finger joints, the differential diagnosis between rheumatoid arthritis (RA) and osteoarthritis (OA) of the hands can be initially difficult. This prospective study (the TryCort study) was performed to study the diagnostic value of prednisolone in differentiating between RA and hand OA. We present the results of this potentially diagnostic test in patients with possible RA in daily clinical practice by demonstrating the results of a pilot and a validation part of this 'prednisolone test' (pred-test). We investigated the response to a 3-day course of 20 mg of prednisolone in patients with suspicion of RA. All patients received 1 g of paracetamol per day for 5 days for pain relief. On days 3-5, a morning dose of 20 mg of prednisolone was added. Hand pain was quantified on a 0-10 Numerical Rating Scale, and the subjective percentage of improvement (0-100%) was recorded. Thresholds for response to prednisolone were investigated in a pilot phase with differentiation in response between patients with RA and patients with OA of the hands, both with pain in the hands ≥4. In a validation phase, the best differentiating cut-off of the pilot phase was applied to discriminate responders from non-responders in consecutive patients with hand pain ≥4 referred because of suspected RA. Final diagnoses were made by the expert upon re-examination at week 12. Primary outcomes were the sensitivity and specificity of a positive test in relation to the diagnosis. A percentage of 40% subjective improvement of pain in the hands on day 3 discriminated best between RA and OA in the pilot phase. Among 95 patients with complete data in the validation phase, RA was diagnosed in about 50%. Patients with RA had more swollen joints, higher C-reactive protein levels and slightly higher Health Assessment Questionnaire scores. The pred-test was positive in 42.1% of all patients (40 of 95). The median percentage of improvement on day 5 was higher in RA than in non-RA: 50% (IQR 30-60%) vs. 20% (IQR 10-30%) (p < 0.001). The sensitivity and specificity of the pred-test were 0.6 (95% CI 0.5-0.8) and 0.8 (95% CI 0.7-0.9), respectively, and the positive and negative predictive values were 0.77 and 0.70, respectively. To our knowledge, this is the first evaluation of the widely used pred-test that has ever been performed. The pred-test had a moderate sensitivity and a good specificity. We conclude that rheumatologists may use this test in unclear clinical situations to better differentiate between inflammatory and other conditions. ClinicalTrials.gov identifier: NCT01395251 . Registered on 14 Jul 2011. EudraCT number: 2011-002633-19. Registered on 21 Dec 2011.