Skip to main content

Table 1 Variables considered for inclusion into vigiRank

From: Improved Statistical Signal Detection in Pharmacovigilance by Combining Multiple Strength-of-Evidence Aspects in vigiRank

Variable Concept Motivation VigiBasea implementation
Informative reports (INF) Reports on the drug and the ADR with sufficient information to allow a causality assessment of the individual case Reporters may be more likely to provide detailed information when they have a strong suspicion that the adverse event was drug related Reports with vigiGrade completeness score ≥0.9 (for details, see Sect. 2.1.1)
Narrative (NAR) Number of reports with free text information available Free text information may strengthen the causality assessment of a case. In addition to this, reporters may be more likely to provide free text information when they have a strong suspicion that the adverse event was drug related. Reports with narrative information, excluding purely numerical narratives and standard phrases, e.g. ‘none provided’
Dechallenge (DCH) Reports indicating that the adverse event subsided upon withdrawal of the drug Resolution of the adverse event upon withdrawal of the drug strengthens the causality assessment for the individual case Reports with positive dechallenge (by definition including positive rechallenge, see below)
Rechallenge (RCH) Reports indicating that the adverse event recurred upon re-exposure to the same drug Repeated occurrence of the adverse event upon exposure to the drug strengthens the causality assessment for the individual case Reports with positive rechallenge
Causality assessment (CAU and CAU+) Reports indicating a positive result of causality assessment of the individual case Strengthens the causality assessment of the individual case Implemented as two separate variables: number of reports with causality probable/certain (CAU) and number of reports with causality certain (CAU+)
Time-to-onset (TTO) Reports with a plausible time between the intake of the drug and the adverse event Time-to-onset information may strengthen the causality assessment of the individual case Reports with reported time-to-onset less than 90 daysb
Solely reported (SOL) Reports with no concomitant or co-suspected drugs To capture reports with low likelihood of other drugs having contributed to the reaction Reports with no concomitant or co-suspected drugs
Multiple reporting elements (MUL) Reports fulfilling multiple defined criteria, strengthening the causality of the case Several criteria that speak in favour of a causal relationship naturally strengthens the overall causality assessment of the individual case Reports fulfilling at least two of the following: solely reported, dechallenge, narrative, causality probable/certain
Recent reporting (REC) Reports entered during the last 3 years To capture emerging safety issues; lack of recent reports may speak against a causal relationship Reports entered during the last 3 years
Disproportional reporting (DIS) Information on whether the drug–ADR pair is reported more often than expected An unexpectedly large number of reports on the drug–ADR may strengthen the likelihood of a causal relationship Disproportionate reporting as measured by a lower limit of the credibility interval for the IC either on the full dataset or on a subset of the data (for details, see Sect. 2.1.2)
Geographic spread (GEO) Number of geographic regions contributing reports on the drug–ADR pair of interest True ADRs might be expected to occur not just in a single geographic region Countries with IC > 0 for the drug–ADR pair of interestc
Time trend (TRE) Increase in the reporting frequency of the drug–ADR pair Emerging safety issues may be expected to exhibit an increase in reporting with time Growing IC values over the three 6-month periods up to the dataset end date
  1. ADR adverse drug reaction, IC information component, WHO World Health Organization
  2. aVigiBase®, the WHO global individual case safety report database, is the particular collection of individual case reports considered in this study. As of March 2014 it contained more than 8.5 million reports from 118 countries
  3. bAdmittedly a crude attempt to capture plausible time-to-onset, which may rather filter out implausible temporal relations, with the exception of long-latency reactions
  4. cFor rare drugs and ADRs this is likely to equal the number of countries with at least one report of the drug–ADR pair