Q.01

How should I interpret MHRA iDAP reports?

Treat an iDAP as a list of suspected reactions, not as evidence of risk. A report doesn't show that the medicine caused the reaction. You can't work out how likely a reaction is from it. And MHRA says you can't compare the risks of different medicines by comparing the numbers. MHRA asks you to read each iDAP with the "essential guidance" printed at the bottom of the report (Yellow Card, What is being reported).

Approach: I searched the MHRA Yellow Card iDAP data we hold and read MHRA's official pages on iDAPs, the iDAP release notes and MHRA's published Drug Analysis Print interpretation guide. I couldn't extract the footer guidance text from a live iDAP, so the caveats below come from the Drug Analysis Print guide.

What an iDAP contains

  • iDAPs replaced the older Drug Analysis Prints. Each one lists every suspected adverse reaction reported to MHRA through the Yellow Card scheme for one drug substance. That includes reports from healthcare professionals, the public and pharmaceutical companies. A report takes about a month to appear (Yellow Card iDAPs).

  • Medicines are listed by active ingredient, not brand name. The data covers the drug given alone ("single constituent") and in combination products ("multi-constituent"). Reactions are grouped by MedDRA (the standard medical terminology for adverse event coding), from System Organ Class down to specific terms, with a fatal-outcome count per reaction (MHRA interpretation guide).

  • Reports and reactions are different counts. Each report covers one patient but can list several reactions, so the reaction total can be higher than the number of reports (same guide).

  • Each reaction record carries these fields:

    • the MedDRA levels: preferred term, high-level term, high-level group term and System Organ Class
    • a fatal flag and case seriousness
    • single or multi-constituent
    • the year received, sex and age group
    • whether a non-health professional reported it
    • route of administration

    One example is a sotorasib record from 2024 coded "Death", which is flagged fatal, single-constituent and reported by a health professional [1].

MHRA's interpretation caveats (all from the interpretation guide)

  • You can't estimate how likely a reaction is, because MHRA has limited information on how many people took the medicine without one.
  • Reporters are asked to report on suspicion alone. A report "does not necessarily mean that the medicine has caused the reaction." GOV.UK guidance also says causality doesn't need to be proven to report (GOV.UK Yellow Card guidance).
  • A reaction may have happened anyway, or may be part of the condition being treated.
  • Deciding whether a medicine caused a reaction means weighing many factors. MHRA staff do that analysis as part of safety monitoring.
  • "It is not possible to compare the risks of different medicines by comparing the numbers presented."
  • The prints don't give a complete overview of a medicine's risks, and you can't draw safety conclusions from them alone. MHRA assesses signals alongside other evidence, such as the worldwide literature.

Changes in the current iDAP version (release notes v1.0.3, 25 August 2026, iDAP release notes)

  • The data is shown in three views:

    • a Yellow Card Overview (only reports sent in directly by the public and healthcare professionals)
    • a Total Report Profile
    • a Total Reaction Profile

    Make sure you know which view a figure comes from.

  • MHRA now counts each report under one "Primary Reporter", so counts per reporter category may be lower than in earlier publications. Don't read that drop as a trend.

  • Seriousness now uses the current list of "serious" terms, so a report's classification can change when terms are added or removed.

  • A breakdown by ethnicity is published using NHS categories. It only covers reports from 2010 onwards, when MHRA started collecting it.

  • The yearly breakdown is based on the most recent revision date, not the original logic, so year totals may differ from older figures.

  • Known issues: CSV downloads are temporarily disabled, and some products are wrongly linked to more than one product type.

If you cite or publish iDAP data MHRA asks that you include its guidance whenever you circulate the data. It encourages research use as long as the limits on interpreting the data are made clear. It also offers advice, and feedback on manuscripts before publication, through its Vigilance and Risk Management of Medicines Division (same guide).

If you tell me which substance you're looking at, I can pull its iDAP reaction data and break it down by MedDRA class, seriousness and fatal outcome.

Q.02

Can I compare side effect rates between medicines?

No. iDAP figures can't be used to compare side-effect rates between medicines. MHRA states: "It is not possible to compare the risks of different medicines by comparing the numbers presented in Drug Analysis Prints" (MHRA interpretation guide).

Approach: I searched the MHRA Yellow Card iDAP reaction data, which holds individual reaction records and none of MHRA's interpretation guidance. I then reopened MHRA's published Drug Analysis Print interpretation guide.

Why the numbers aren't rates (all from the same guide)

  • No denominator. You can't estimate the likelihood of a reaction from the prints, because MHRA has "limited information about how many people have taken the medicine without experiencing a reaction." A count of 500 reports for drug A and 50 for drug B says nothing about which drug is riskier per patient.
  • Suspicion, not causation. Yellow Cards are filed on suspicion alone. A report "does not necessarily mean that the medicine has caused the reaction."
  • Confounding by the condition treated. It can be hard to tell a reaction apart from something that happened naturally, and some reactions are part of the condition being treated.
  • Reports and reactions are different counts. One report covers one patient but can list several reactions, so the reaction total can be higher than the report count. If you line up two drugs, use the same unit for both.
  • Not a full risk picture. The prints "do not present a complete overview of the risks," and safety conclusions "cannot be made on the information in Drug Analysis Prints alone."

What the record-level data can show Each iDAP record shows how a single reported case is coded:

  • MedDRA hierarchy (the standard terminology for coding adverse events)
  • fatal flag and case seriousness
  • year received, sex, age group and route
  • whether a health professional or a member of the public reported it

For example, one naproxen record from 2018 is coded "Micturition frequency decreased" and marked non-serious [2]. On a trastuzumab record from 2012, a serious case is coded "Metastases to skin" [3]. That second case shows how a report can reflect the underlying disease rather than the drug. Within one medicine, you can describe how the reported cases break down. Across medicines, those breakdowns still aren't incidence.

Where comparative risk is actually assessed

  • MHRA evaluates Yellow Card reports together with other evidence, such as the worldwide literature. When it finds a new side effect, it considers it "in context of the overall side effect profile for the medicine, and how it compares with other medicines used to treat the same condition" (guide, p. 2).
  • For a medicine's known side effects, the guide points you to the patient information leaflet or to a healthcare professional.

If you plan to publish a comparison MHRA asks that you make the limits of interpreting the data clear and include its guidance whenever you circulate the data. It also offers advice, and feedback on manuscripts before publication, through the Director of its Vigilance and Risk Management of Medicines Division (same guide, p. 2).

If you tell me which medicines you want to look at, I can pull their iDAP records and set out how the reported cases break down side by side. It would be a description of the reports, not a rate comparison.

Q.03