Affiliations: Social Audit Ltd., London, UK | DIPEx Project, Department of Primary Health Care, University of Oxford, Oxford, UK
Note: [] Corresponding author. Charles Medawar, Social Audit Ltd., P.O. Box 111, London NW1 8XG, UK. Tel./Fax: +44 20 7586 7771; E‐mail: charles@socialaudit.org.uk.
Abstract: We have previously described the value of patients' reports relating to withdrawal problems and dependence, and violent or suicidal ideation or acts, and their linkage with paroxetine [1]. Here we describe how Yellow Card reporters – health professionals – perceived and reported suspicions of the same kinds of harm. This analysis was made possible only because the organisers of the Yellow Card scheme – the Medicines and Healthcare Products Agency (MHRA) and Committee on Safety of Medicines (CSM) – agreed to release data. National drug regulatory agencies rarely do so, and we believe this to be the first such analysis of the operation of the scheme. This analysis had two main objectives: to compare the value of professionals' reports with patients' reports of the same suspected adverse drug reactions (ADRs), and to learn more about the effects of paroxetine. Our analysis was limited for many reasons, but sufficient to form a robust preliminary view. In this particular case, the overall quality of professional reporting and interpretation of data seemed poor, providing intelligence that was in some ways inferior to that provided in spontaneous reports from patients. We give new evidence to suggest that miscoding and flawed analyses of Yellow Cards have led to under‐estimation of the risk of suicidal behaviour, and have impeded recognition of what appears to be a close relationship between suicidal behaviour and changes in drug concentration. An increased risk of suicidal behaviour during the first few days of treatment with an SSRI has been suspected for some years: we suggest that comparable risks may also exist outside this ‘window’, when drug dosages are either increased or lowered (during withdrawal). The implications for dosing strategies are discussed. Our analysis relates only to one drug and to two suspected ADRs, but suggests that the Yellow Card scheme is, in important respects, both chaotic and misconceived. Further research is essential.
