Adverse event recording failed to reflect potential harms: A review of trial protocols of behavioural, lifestyle and psychological therapy interventions

Objective : To explore how potential harms are assessed in trials of behavioural, lifestyle and psychological therapy interventions.

Study Design and Setting : This study was a review of protocols from the National Institute of Health Research Health Technology Assessment and Public Health Research programmes. Protocols were included if the study was a randomised controlled trial and the intervention intended to change lifestyle or behaviour to improve health or improve psychological outcomes.

Results : 95 of 151 protocols planned to record adverse events (AEs). Definitions of AEs were often not given and varied widely. Serious AEs were mostly defined using standards originally devised for pharmacological trials. Twenty-two protocols listed expected AEs. Few protocols described assessment of causation between AEs and intervention. Examples of useful AE recording practice were identified.

Conclusion : Monitoring and recording AEs in behavioural intervention trials was variable and frequently based on reporting guidelines for pharmacological trials. This may mean potential harms are being missed. Future trials should consider: 1) Potential harms posed by the intervention 2) How to define serious AEs 3) What are expected AEs. Further research to achieve consensus on AE recording is required, including identification of core adverse outcomes in clinical areas or caused by interventions.

• Harms or unintended consequences are known to be possible from trials of behavioural, lifestyle and psychological therapy interventions, yet are not always appropriately or consistently evaluated in trials.

• Reliance on definitions for harms (typically termed adverse events) originally devised for pharmacological trials, may result in omission of important harms.

• The level of risk or likelihood of harm from an intervention should always be assessed in trials of these interventions.

• Better details on adverse event recording in protocols is needed to increase transparency and consistency and improve practice, but consensus is needed on the appropriate approach.

• Examples of useful AE recording practices identified from trials included in this review are discussed, for example: alternative definitions of what constitutes a serious adverse event or methods of deriving lists of expected adverse events.