Background Recent data suggest an increased risk of melanoma and of cutaneous carcinoma in patients receiving TNF blockers. Thus, there is a need that rheumatologists improve their abilities to recognize skin cancers.
Objectives To demonstrate that an on-line training dedicated to skin tumors increase the abilities of rheumatologists to recognize skin cancers.
Methods This is a nationwide randomized web-based survey involving French rheumatologists. All performed a baseline evaluation (Test 1), including 5 multiple choice questions (MCQ) to assess their knowledge on risk factors and prevention of skin cancers, and 20 short cases with skin lesion picture. For each case, they had to indicate if the skin lesion was benign or premalignant/malignant (Score 1; range: 0-20), their level of confidence in this diagnosis measured on a 10-points Likert scale, and to give the precise diagnosis among 5 propositions (Score 2; range: 0-20). Then, they were randomized in 2 arms. In Arm A, they had access to an on-line formation on skin tumors consisting in 4 e-learning modules, of 15 minutes each. In Arm B, they did not receive any formation. Participants were reevaluated (Test 2, similar to Test 1) 3 weeks after the end of the course (Arm A) and 3 weeks after the initial evaluation (Arm B). The primary end-point was the number of adequate diagnoses for benign vs premalignant/malignant lesions (Score 1) at Test 2. The secondary end-points the number of (i) adequate precise diagnoses of the skin lesions (Score 2), (ii) correct responses to MCQ (Score 3; range: 0-25), (iii) adequate precise diagnoses of the skin lesions and correct responses to MCQ (Score 4; range: 0-45), and (iv) the level of rheumatologists' self-confidence in discriminating between benignity and malignancy (Score 5). Analysis was performed according to the ITT principle.
Results Altogether, 141 rheumatologists were randomized: 71 in Arm A, 70 in Arm B. No statistical difference for all the 5 scores was observed between the 2 arms at Test 1. In Arm A, 61 participants (86%) completed the course and 57 (93%) performed Test 2. In Arm B, 67 participants (96%) performed Test 2. The mean number of adequate diagnosis for benign vs malignant/premalignant lesions (Score 1) at Test 2 was higher in Arm A than in Arm B (13.4 vs 11.2 points; p value <0.0001). Scores 2, 3 and 4 were also significantly higher in Arm A. No statistical difference was observed on the level of self-confidence between the 2 arms.
Conclusions The on-line formation was effective to improve skin cancer diagnosis in French rheumatologists.
Disclosure of Interest M. Viguier Grant/research support: Abbvie, S. Rist Grant/research support: Abbvie, F. Aubin Grant/research support: Abbvie, M.-T. Leccia Grant/research support: Abbvie, M.-A. Richard Grant/research support: Abbvie, P. Gaudin Grant/research support: Abbvie, T. Pham Grant/research support: Abbvie, P. Richette Grant/research support: Abbvie, D. Wendling Grant/research support: Abbvie, F. Tubach Grant/research support: Abbvie