Background: Reports of clinical trials usually emphasize benefits and give less attention to harms.
Purpose: To compare the reporting of harm in trials of pharmacologic and nonpharmacologic treatment.
Data sources: MEDLINE and the Cochrane Central Register of Controlled Trials.
Study selection: Reports of randomized, controlled trials assessing treatment of rheumatic disease that were published between January 1999 and January 2005.
Data extraction: A standardized abstraction form was used to extract data.
Data synthesis: 193 articles were analyzed. After adjustment for medical area, sample size, funding source, and multicenter trials, data on harm were more often described in pharmacologic treatment reports than in nonpharmacologic treatment reports in reporting adverse events (odds ratio, 5.2 [95% CI, 2.1 to 12.9]), reporting withdrawals due to adverse events (odds ratio, 4.6 [CI, 2.0 to 10.9]), reporting severity (odds ratio, 3.7 [CI, 1.5 to 9.1]), and allocating space for describing harm (odds ratio, 1.6 [CI, 1.2 to 2.3]).
Limitations: Extrapolating results to trials in areas other than rheumatic disease is questionable.
Conclusions: The lack of reporting harm in trials assessing nonpharmacologic treatment in rheumatic disease is an important barrier to evaluating the benefit-harm balance of nonpharmacologic treatments.