Trust

Methodology

Sources

We archive published sources twice daily: DHSC Medicine Supply Notifications as relayed publicly by Community Pharmacy England; the NHSBSA Serious Shortage Protocol register and its protocol documents. NHS dm+d (the Dictionary of Medicines and Devices) product coding is in progress and not yet applied to published records. Every original document is kept verbatim, content-addressed, before any parsing — so our history can always be re-derived and re-checked.

What we publish

Re-expressed facts (which product, which reference, which tier, which dates) and our own derived statistics (durations, extension counts, change feeds). We do not reproduce source editorial or clinical advice — every entry links to the source document instead. Dates the sources state fuzzily (“late August”) stay fuzzy here; we never invent precision.

Coverage & honesty

Coverage is England-first (DHSC notifications as relayed by CPE, plus Serious Shortage Protocols). We now also cover Scotland separately: NHS Scotland issues its own Medicine Supply Alert Notices (MSANs), genuinely Scotland-authored, so they sit on a distinct Scotland page and are never merged into the England register or counts. We deliberately do notadd Northern Ireland as separate coverage: NI's HSCNI/BSO supply alerts re-publish the same DHSC Medicine Supply Notifications (they match England's by title and date), so listing them as NI-specific shortages would double-count one shortage as two nations. Wales is not yet included. UK-wide claims appear only where the source is UK-wide. Source typos — and they exist — are preserved and flagged, not silently corrected. Where our record is incomplete, the gap is shown, not smoothed over. Our archive began in July 2026; Serious Shortage Protocol documents extend back to 2019.

For each Serious Shortage Protocol we also read the signed protocol document for the few facts a pharmacist needs first: which UK nations it covers, its legal category, and the medicine that may be supplied instead. These sit next to a link to the full signed protocol — never our own advice. The substitution is published only when it appears in the source PDF as one contiguous run of text and survives completeness checks; anything we cannot prove word-for-word is withheld rather than paraphrased, because a half-quoted clinical instruction is worse than none. That is a deliberately high bar and it costs coverage: we show the substitution for 25 of 89 protocols (at last check, 20 Oct 2026). Quantity is not reproduced at all — those cells run to several paragraphs and their boundaries are not reliably derivable, so the protocol itself states them. The per-protocol scope is why UK-wide vs England-and-Wales is now stated from the document itself, not assumed.

A note on counting backwards. Our own capture began in July 2026, and CPE takes a notice down once the shortage resolves — so at first the historical record here was only the notices still on public display, skewed toward longer-running shortages. We have since backfilled the back-catalogue from the Internet Archive: at last count (23 July 2026) we held 277 of about 364 Medicine Supply Notification posts it had ever recorded — roughly three quarters, reaching back to March 2021. First-hand captures since then add to the record but do not change that ratio, which describes the recovered back-catalogue only. Recovered posts are marked as sourced from the Internet Archive, not first-hand capture, and each is dated by its own stated issue date, not when we recovered it.

That figure is lower than the ~83% we published on 23 July 2026. The Internet Archive had captured some posts several times under different tracking query strings (?unlid=…), and we were counting each capture as a separate notice — which also made a handful of medicines look like repeat shortages when they had been notified once. We now count one notice per canonical post URL. The corrected figure is smaller and correct; see the corrections log.

The remaining ~24% are posts the Internet Archive never captured; a handful more are held but not yet machine-readable (older layouts with numeric dates our parser doesn't read yet — flagged, never guessed). So a pre-2021 total is still a floor, and 2021–2026 is now close to complete. Figures we later correct are logged on the corrections page.

Cadence & corrections

Sources are checked twice daily — morning and evening, UK time; the scheduler that runs the checks does not guarantee a fixed clock time, so rather than promise one, the homepage, the UK Shortage Index, the state-of-shortages summary and the press page each show when sources were last checked (currently 20 Oct 2026) and, separately, when the data last changed — distinct signals. Parsed data follows each check within the hour. Corrections are logged publicly on the corrections page when we get something wrong.

Who

Built and operated by Doctor Data Ltd — founded by Mustafa Ghafouri (MBChB, MSc), a medically qualified data scientist who worked on the MHRA's Regulatory Connect programme from December 2022 to September 2024 and who also built hospitalwaits.co.uk. ShortageSignal is not an official NHS or DHSC service.