Methods: Electronic databases including Medline, Embase, Google Scholar and Iran Medex were searched to identify and summaries relevant studies.

Results: Different methods have been used to identify adverse events and their nature, causes and consequences. Record review seems to be the most common and the best method for measuring the rate and consequences of adverse events. However, this method is relatively expensive and time consuming and still underestimates the rate of adverse events. It is also not suitable for identifying the underlying causes of adverse events. The other method commonly used is reporting system including critical incident reporting system. This method is relatively cheap and more suitable for exploring the underlying causes of adverse events, but it is not suitable for identifying the rate of adverse events as it may underestimate many adverse events. Other methods include critical incident technique, interview, questionnaire, direct observation and review of claims and none of them are suitable for identifying the rate of adverse events but they might be useful for other purposes including assessment of the causes of adverse events.

Conclusion: Several methods can be used for study of adverse events. No single method can serve all purposes. Each method has some strengths and weaknesses. Using a combination of methods is more suitable, but this approach is more expensive and time consuming. Which combination of methods are more appropriate, depends on several factors including the aim of the study and resources available.