Reporting of harms in randomized controlled trials of psychological interventions for mental and behavioral disorders: a review of current practice.

Background: Data suggest that certain psychological interventions can induce harm in a significant number of patients. While the need for adequate reporting of harms in clinical trials has repeatedly been emphasized, it is uncertain whether such information is routinely collected and reported in trials within this research field.
Method: We used the two major databases in clinical psychology and medicine (PsycINFO and PubMed) to identify original publications from 2010 reporting randomized controlled trials of psychological interventions for patients with mental and behavioral disorders. Two reviewers searched the full-text reports for information about monitoring of adverse events, side effects, and deterioration.
Results: A total of 132 eligible trials were identified. Only 28 trials (21%) included information that indicated any monitoring of harms on patient level. Four (3%) of these trials provided a description of adverse events as well as the methods used for collecting these data. Five of the trials (4%) reported adverse events but did not give complete information about the method. An additional four reports (3%) briefly stated that no adverse events occurred, whereas 15 trials (11%) only provided information on deterioration or indicated monitoring of deterioration. The probability of including harm-related information was related to the journal impact factor.
Conclusion: Important information about harms is not reported systematically within this research field, suggesting that the risk of reporting bias is nontrivial in conclusions about the risk-benefit ratio of psychological treatments. Guidelines on how to define, detect, and report harms related to psychological interventions could facilitate better reporting.
(Copyright © 2014 Elsevier Inc. All rights reserved.)