To the Editor:
We thank Dr. Haroon, et al for their thoughtful letter and the data presented. They state that 11 of 63 patients (17.5%) with IJD were missed in his initial screening. This was largely due to missing crucial clinical information and laboratory data. Our findings were very similar, with 18 cases out of 45 (40%) missed in our initial screening. Clearly, Dr. Haroon, et al are doing better than we are.
The challenge with “screening” is to create a strategy that is sensitive enough to identify most cases of IJD. Ideally a screening strategy for IJD should have a sensitivity of greater than 90%, indicating that it identifies almost all patients with IJD.
Sensitivity is defined as number of true-positives/(number of true-positives + number of false-negatives). Therefore the whole idea is to limit the number of false-negatives. However, in doing so the number of false-positives may rise, which will result in the rheumatologist seeing more patients with noninflammatory joint disease. What is the answer?
We have attempted to educate our general practitioners about information to include in a referral. The reality is that individual general practitioners see very few cases of new-onset IJD (perhaps 1 or 2 a year). In such an environment, educational interventions are limited by their applicability. We have chosen a different strategy of creating an “Arthritis Referral Tool” that we require for all rheumatology referrals. A referral tool has 3 potential benefits: (1) it identifies information rheumatologists feel is important to screen for IJD; (2) it will educate general practitioners (through repetition) about the clinical and laboratory information rheumatologists would like in a referral; and (3) it allows researchers to further refine and identify what information is important to detect IJD.
The real question is not the diagnosis of IJD, but rather if the patient is started on a therapy that will significantly affect their longterm quality of life. Perhaps a better way to refine our search is to ask, “What information best predicts those patients who require treatment with disease modifying antirheumatic drugs or biologic agents?”
