Diagnostic Performance and Clinical Utility of Referral Rules to Identify Primary Care Patients at Risk of an Inflammatory Rheumatic Disease.

Abstract

To determine the diagnostic performance and clinical utility of the Rotterdam Early Arthritis Cohort (REACH) and the Clinical Arthritis Rule (CARE) referral rules in an independent population of unselected patients from primary care. This study consisted of adults who were suspected of the need for referral to a rheumatologist by their general practitioner. Diagnostic accuracy measures and a net benefit approach were used to compare both rules to usual care for recognizing inflammatory arthritis and inflammatory rheumatic diseases (IRDs). Using the least absolute shrinkage and selection operator method and cross-validation we created an optimal prediction rule for IRD. This study consisted of 250 patients, of whom 42 (17%) were diagnosed with inflammatory arthritis and 55 (22%) with an IRD 3 months after referral. Considering inflammatory arthritis, the area under the receiver operating characteristic curve (AUC) was 0.72 (95% confidence interval [95% CI] 0.64-0.80) for REACH and 0.82 (95% CI 0.75-0.88) for CARE. Considering IRD, the AUC was 0.66 (95% CI 0.58-0.74) for REACH and 0.76 (95% CI 0.69-0.83) for CARE. CARE was of highest clinical value when compared to usual care. The composite referral rule for IRD of 10 parameters included sex, age, joint features, acute onset of symptoms, physical limitations, and duration of symptoms (AUC 0.82 [95% CI 0.75-0.88]). Both validated rules have a net benefit in recognizing inflammatory arthritis as well as IRD compared to usual care, but CARE shows superiority over REACH. Although the composite referral rule indicates a greater diagnostic performance, external validation is needed.