A £20 SDR dongle, an antenna stuck to a kitchen window, and a copy of multimon-ng will still pull patient names, bed numbers, and clinical handover notes out of the air above most British hospitals. The protocol carrying them was standardised in 1981, under the chairmanship of the British Post Office, by a committee called the Post Office Code Standardisation Advisory Group. The acronym, POCSAG, is one of the longest-surviving pieces of pre- privatisation telecoms naming still in active operational use. Everything else the Post Office shipped that decade is in a museum or a skip. POCSAG is in a cardiac arrest team's belt clip.

Trend Micro mapped this in their 2016 Leaking Beeps report and TechCrunch confirmed the same pattern across NHS trusts in 2019. PageOne, the last remaining commercial paging network in the UK, quietly tells customers in their terms and conditions that messages can be intercepted, and that encrypted services are "available if required". Most trusts are not required. They run private networks, on private repeaters, broadcasting forty-year- old uncoded BCH(31,21) codewords into the same skies that carry end-to-end-encrypted everything else.

In 2019 Matt Hancock told the NHS to be off pagers for non- emergency use by the end of 2021. That deadline passed. Cambridge University Hospitals did manage a clean migration to an EHR- integrated secure messaging system, documented in BMJ Health & Care Informatics, and even there the bleep system was kept on standby rather than physically decommissioned. The implementer's report uses a phrase that should be carved into the side of every NHS digital strategy document: "costly obsolescence." Pagers cost more than the secure alternative and yet nobody can quite kill them.

The reasons are not sentimental. Hospital walls are thick enough to stop X-rays, which means they also block most mobile and WiFi signal. POCSAG runs slow and loud in the VHF band (UK allocations sit around 138 and 153 MHz), and it gets through concrete. The protocol is one-way and asynchronous, which means a paging transmitter can blanket multiple sites from a single high site without any handover logic. There is no app to crash, no battery management daemon to drain, no operating system update window. The receiver chirps. The carrier nurse runs.

Every proposed successor introduces failure modes the existing system does not have. TETRA, the encrypted radio standard the UK emergency services have spent the last decade trying to migrate away from, requires expensive handsets and a working core. App-based bleep replacements depend on the WiFi the X-ray-proof walls were designed to defeat. Smartphones get locked, lost, charged on the wrong ward, taken home in someone's pocket. The pager is a cheap, replaceable, one-purpose object that survives a drop into a sluice room. Hospitals know what they want from a critical messaging layer, and the 1981 spec, by accident, still describes it more honestly than anything sold to replace it.

The haunting is not that POCSAG persists. It is that the persistence is rational. A protocol designed before the Falklands War, before privatisation, before the web, before HIPAA, before GDPR, before the entire regulatory edifice that should have condemned it, is still load-bearing because it does one job reliably and because all of its replacements quietly do that job worse. The 2019 SDR demo was treated as a scandal. It should have been treated as a question. If a forty-year-old uncoded radio protocol is still the most dependable thing in the building, what exactly have we been building?

The same logic keeps the fax machine humming in radiology and oncology, and the same logic is why Capita can still hold the frequency the Home Office has been trying to replace for a decade. The ghost is not the technology. The ghost is the absence of anything better.

Sources: