Adverse Reaction Signaling and Disproportionality Analysis: An Update
- PDF / 230,686 Bytes
- 7 Pages / 602.986 x 782.986 pts Page_size
- 109 Downloads / 183 Views
Adverse Reaction Signaling and Disproportionality Analysis: An Update
Drug Information Journal 46(6) 708-714 ª The Author(s) 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0092861512453041 http://dij.sagepub.com
Gerald Faich, MD, MPH1, and Jonathan Morris, MD1
Abstract There is an increased focus on signal detection, signal management, and the use of use of large collections of adverse event databases to meet the evolving regulatory mandates. The basic elements of signaling require an understanding of the different signaling methods, data mining approaches, and the strengths and limitations of disproportionality analysis. The application of signaling thresholds and the resulting statistical interpretation require an understanding of the methods themselves, as well as the strengths, limitations, and biases inherent in the adverse event data sources. This paper will highlight recent developments in these areas as well as provide cautions related to biases that affect reporting and analyses of adverse event databases. Keywords data mining, signal detection, disproportionality, AERS, pharmacovigilance
Introduction Adverse event reporting and analysis is a crucial postmarketing surveillance activity that is key to assuring pharmaceutical safety. In recent years, global expansion in reporting has caused an increased use of signaling methods, data mining, and disproportionality analysis. This paper will highlight some developments in these areas as well as provide cautions related to biases that affect reporting and analyses of adverse event databases. Adverse event (AE) databases consist of electronically stored reports of AEs related to pharmaceuticals that are submitted by physicians, study investigators, other health care providers, company field representatives, patients (consumers), lawyers, and others. Report submissions are made directly to regulatory authorities or pharmaceutical manufacturers. AE databases have grown substantially since the mid-1980s, in part due to the 1985 Food and Drug Administration’s (FDA) reporting regulations and subsequent European regulations specifying how manufacturers are required to submit AE reports to regulators. In the United States, most reporting is done verbally or by email, surface mail, or telephone to the pharmaceutical manufacturers or their representatives. Typically, reports are screened by manufacturers at the time of receipt and decisions are made as to whether or not they need to obtain more information about the specific AE or the clinical scenario described in the report. Qualitative and semi-quantitative (counts, trending) review of individual reports and groups of reports is then done mainly to detect signals of new, important toxic effects
and to assess relative changes in the frequency of case reporting over specified periods of time. Because thousands of reports accumulate in AE databases, a variety of quantitative algorithmic methods have evolved to determine if unusual (relative to other drugs or relative to expected rates
Data Loading...
