RESUMO
High-quality information on the potential benefit and harm of a drug is required for patients and clinicians to make informed treatment decisions and to enable cost-effectiveness modeling to be undertaken. This systematic review describes the collection and reporting of adverse event data as presented in published clinical trials of neuropathic pain for the evaluation of antidepressant or antiepileptic drugs. A total of 74 studies in 16,323 patients published between 1965 and 2012 were identified, of which 43 were published from 2004 onwards. The review found that methods used to collect adverse event data, the frequency of collection, and the selection criteria used by authors for reporting adverse events vary substantially, and these events are often inadequately reported. Consequently, a potential synthesis of valuable harm information across trials is hampered. We make recommendations regarding the reporting of methods used to collect, assess, select, and present adverse event data in publications. Through the Core Outcome Measures in Effectiveness Trials (COMET) initiative, core outcome sets (which include effectiveness and harm) are developed by disease condition. To facilitate data synthesis for adverse events of drug therapies, we suggest that core outcome sets for harms could be developed by therapeutic class (ie, individualized for each class of drug). To improve comparability of information across trials collection methods need to be standardized for patient reports (spontaneous or prompted) and active surveillance (clinical examinations and laboratory tests). Uniform methods for presenting summary information regarding recurrent events, duration and timing of events requires further research.