Class of 2012 / Specific Aims: To evaluate the presence of statistical information from clinical studies in official product labeling specific for disease-modifying anti-rheumatic drugs (DMARDs) used in the treatment of rheumatoid arthritis. Methods: Data were abstracted from official product labeling DMARDs with FDA approval for treatment of rheumatoid arthritis. Each document was examined for the presence of statement regarding a priori type 1 error rate, p-values, and measures of variance. Medications were classified as either biologic or non-biologic. Main Results: A total of 14 DMARDs, 7 biologics (50%) and 7 non-biologics (50%), were found to be FDA approved for the treatment of rheumatoid arthritis. Primary outcomes consisted of American College of Rheumatology (ACR) response rates, radiographic changes, and health assessment questionnaire score (HAQ). Any measure of variance and the presence of a p-value were both found in six (43%) of the drug labels. Inclusion of p- values was found to be significantly greater in biologics compared to non-biologics for both ACR and radiographic results. Inclusion of variance was found to be significantly greater in biologics compared to non-biologics for radiographic changes only. No package inserts contained statements regarding the a priori type I error rate. Conclusions: Measures of variance are not frequently included in product labeling for either biologic or non-biologic DMARDs. However, inclusion of variance and p-values for ACR response rates and radiographic changes were more likely to be reported for biologics therapies as compared to non-biologics. A statement regarding Type 1 error rates were absent from labels regardless of outcome assessed.

Class of 2012 Abstract / Specific Aims: To evaluate the presence of statistical information from clinical studies in official product labeling specific for disease-modifying anti-rheumatic drugs (DMARDs) used in the treatment of rheumatoid arthritis. Methods: Data were abstracted from official product labeling DMARDs with FDA approval for treatment of rheumatoid arthritis. Each document was examined for the presence of statement regarding a priori type 1 error rate, p-values, and measures of variance. Medications were classified as either biologic or non-biologic. Main Results: A total of 14 DMARDs, 7 biologics (50%) and 7 non-biologics (50%), were found to be FDA approved for the treatment of rheumatoid arthritis. Primary outcomes consisted of American College of Rheumatology (ACR) response rates, radiographic changes, and health assessment questionnaire score (HAQ). Any measure of variance and the presence of a p-value were both found in six (43%) of the drug labels. Inclusion of p-values was found to be significantly greater in biologics compared to non-biologics for both ACR and radiographic results. Inclusion of variance was found to be significantly greater in biologics compared to non-biologics for radiographic changes only. No package inserts contained statements regarding the a priori type I error rate. Conclusions: Measures of variance are not frequently included in product labeling for either biologic or non-biologic DMARDs. However, inclusion of variance and p-values for ACR response rates and radiographic changes were more likely to be reported for biologics therapies as compared to non-biologics. A statement regarding Type 1 error rates were absent from labels regardless of outcome assessed.

No description available.