External validation of Gout-calculator performance on a cohort of acute arthritis (SYNOLACTATE) sparing distal joints such as hallux and midfoot. A cross-sectional study of 170 patients.

F Robin, O Berthoud, J D Albert, S Cadiou, A Gougeon-Jolivet, C Bendavid, P Guggenbuhl, Guillaume Coiffier

Clinical rheumatology 2021 May

filter terms:

To evaluate the performance of the Gout-calculator in a cohort of consecutive acute arthritis affecting large and intermediate joints (without an attack on hallux or midfoot joints). A retrospective study. Gout-calculator data were collected in medical records of patients included in the prospective consecutive cohort of acute arthritis called SYNOLACTATE. The diagnosis of gout was defined by the presence of sodium urate crystals in synovial fluid. The diagnostic performance of the Gout-calculator was studied by performing an ROC curve with the calculation of its AUC (95% CI) as well as the calculation of Sensitivity (Se), Specificity (Sp), and positive likelihood ratio (LR+). 170 patients with acute arthritis were included. Variables associated with the diagnosis of gout were as follows: serum uric acid > 350 μmol/L (OR 5.52 (2.52-12.1), p < 0.001), joint redness (OR 5.08 (1.85-14.0), p = 0.001), previous patient-reported arthritis attack (OR 4.04 (1.92-8.49), p < 0.001), male (OR 3.00 (1.17-7.69), p = 0.02), hypertension or cardiovascular disease (OR 2.33 (1.07-5.06), p = 0.03). The median (IQR) of Gout-calculator was significantly higher in gouty arthritis (7.0 [5.5-8.1]) than in associated-CPP acute arthritis (4.0 [2.0-5.8]), septic arthritis (3.0 [2.0-5.1]), or others arthritis (3.5 [2.0-5.5]). The AUC was 0.833 (0.768-0.897) with for the threshold ≥ 8, a Se at 27.5% (0.161-0.428), Sp 97.7% (0.934-0.992), and LR+ 11.9 (3.5-40). Despite diagnostic performances close to those published, the use of the Gout-calculator is not sufficient for the diagnosis of gout or to exclude the differential diagnosis of septic arthritis in the SYNOLACTATE cohort. • For a Gout-calculator threshold of ≤ 4, Sensitivity is 92.5%, Specificity 50.8% and LR- 0.15 to the gout diagnosis. • For a Gout-calculator threshold of > = 8, Sensitivity is 27.5%, Specificity 97.7% and LR+ 11.9 to the gout diagnosis. • In a population of acute arthritis affecting large joints, Gout-calculator is not sufficient to discriminate between gouty arthritis and septic arthritis.

Citation

F Robin, O Berthoud, J D Albert, S Cadiou, A Gougeon-Jolivet, C Bendavid, P Guggenbuhl, Guillaume Coiffier. External validation of Gout-calculator performance on a cohort of acute arthritis (SYNOLACTATE) sparing distal joints such as hallux and midfoot. A cross-sectional study of 170 patients. Clinical rheumatology. 2021 May;40(5):1983-1988