Spontaneous adverse event signaling methods: classification and use with health care treatment products.

AE signal detection and its techniques are part of the continuum of public health surveillance, borrowing from both its theory and application (171). Like public health surveillance networks, whose major goals are to identify early signs of new outbreaks, pinpoint new organisms, and monitor designated illnesses, AE signaling and surveillance systems attempt to provide early warnings of previously unsuspected product-AE pairs, hypothesize potential drug-event relations, identify populations "at risk," and facilitate case ascertainment and definition. In both examples, definitive research is often subsequently undertaken to quantify the strength of relations that may be proposed. As with any public health surveillance effort, AE surveillance provides an infrastructure for the ongoing collection of health data and its direct integration into the health regulatory policy (172), including its keystone role in risk assessment and management. However, unlike many surveillance systems, postmarketing AE systems collect case information that is often relatively incomplete and imperfect, estimate exposure based on surrogate values (e.g., sales data), and are used by both governmental and the private sector for preventive planning. These factors make AE signaling and surveillance more ambiguous, regulatory oriented, and complex than its disease counterparts (173). Despite such issues, AE signaling methods continue to evolve in extent, complexity, and acceptance (4, 131, 174). Undoubtedly, this is largely due to the widespread practical experience that has been gained with spontaneous reporting systems over the past 4 decades and the cumulative usefulness that has been demonstrated.