Spontaneous reporting of adverse drug reactions: who reports and what?

Life‐threatening drug‐associated hyperkalemia: a retrospective study from laboratory signals

“…Indeed, none of the drug‐associated hyperkalemia of the study had been spontaneously reported to the official pharmacovigilance system. This is presumably because hyperkalemia is a well‐known and ‘expected’ ADR of these drugs that health professionals do not feel the need to report . The presence of clinical risk factors could also lead physicians to underestimate the contribution of the drug although the suspected drugs were often stopped.…”

“…This is presumably because hyperkalemia is a well-known and 'expected' ADR of these drugs that health professionals do not feel the need to report. 13,18 The presence of clinical risk factors could also lead physicians to underestimate the contribution of the drug although the suspected drugs were often stopped. Choice of the medical units included in the study (i.e.…”

Cutaneous Adverse Effects of Ketoprofen for Topical Use

“…However, this pharmacovigilance study was certainly affected by a huge under-reporting, especially since these cutaneous reactions are well known adverse effects of topical ketoprofen. [16] Their well described nature may also explain why only a few patients were patch tested or photopatch tested after the episode. Most physicians probably considered it unnecessary to test the patients, as the causal link was obvious in most cases.…”

A potential competition bias in the detection of safety signals from spontaneous reporting databases

“…For these events, signal detection thresholds were evaluated for a hypothetical new drug for which the total number of reports for the period would have been 50, 100, or 200. As the signals thresholds of detection were estimated for a hypothetical new drug and consideration of an entire pharmacovigilance database may introduce a bias for drugs with a long market history,20 the study was restricted to reports that occurred between 1 October 2005 and 30 September 2006.…”

“…Because the numbers of events that are being studied are the events that are reported, not necessarily those that actually occur, the numerator (number of reports of interest with drug of interest among all reports with this drug) can be influenced by notoriety and other biases15–22 that might spuriously increase the visibility of the drug‐event pair. Actually, most discussions concerning biases that can affect signal detection focus on those related to reporting of reactions to the drug of interest.…”