Halbridge Infirmary after midnight, when 1,140 pagers remain the most reliable way of finding anyone in the building. Engraving by Bram Keller
At 2:51 on a Tuesday morning in October, Dr. Amara Lowe was in the basement of Halbridge Infirmary, two floors below ground and behind a meter of concrete, watching a patient slide into a CT scanner. She was carrying two devices. The first was a hospital smartphone, issued to her six weeks earlier as part of a pilot, which had shown no signal since she stepped out of the elevator. The second was a black plastic pager the size of a matchbox, running on an AA battery she had last changed 19 days before. The pager went off. It made a sound like a small, furious bird, and its screen showed a single word and a ward: CRASH 7B.
Lowe handed the scan over to the radiographer, took the stairs two at a time, and was on Ward 7B, three floors up and a long corridor away, 94 seconds later. The patient, a man in his seventies whose heart had stopped, survived the night. The smartphone, which was supposed to have received the same alert, buzzed with it at 2:54, by which time she had been running the resuscitation for more than a minute. “The phone’s fine,” she said afterward in the doctors’ mess, in the tone of someone being scrupulously fair. “The phone is lovely. The phone is just not where I am.”
Daniel Okafor is the magazine’s technology correspondent, covering infrastructure, old software, and the people paid to keep both running, from hospital pagers to a payroll system older than most of its maintainers. Before journalism he spent a decade as a systems administrator in the signaling office of the Corran Regional Railway, an experience he turned into the book Nothing Has Changed Since Tuesday. He lives in Corran, still carries a pager out of professional sympathy, and keeps a green-screen terminal on his desk that he insists is load-bearing.
Payroll, pensions and the water bills of a small country still run on code older than the engineers maintaining it. Meet the night shift that keeps it alive.
A farmer, a broken sensor and a dealer two hundred miles away. Inside the quiet fight over who is allowed to fix modern machines.
Nadia Rahman·14 min read
This article is free to read. You have 3 of 3 free articles left this month.Subscribe
The phone is lovely. The phone is just not where I am.
— Dr. Amara Lowe
Pagers are, by most measures, obsolete. The wide-area paging networks that once served doctors, plumbers, and drug dealers alike have been switched off almost everywhere; the country’s last commercial one closed in 2022. A smartphone can do everything a pager does and a great deal more. Yet on any given night at Halbridge Infirmary, 1,140 pagers are clipped to belts, scrubs, and lanyards, and the hospital has been trying, and failing, to get rid of them since 2017. The reasons have less to do with nostalgia than with what a hospital at night actually needs from a machine.
A machine that only shouts
A hospital paging system is about as simple as radio gets. When the switchboard sends a page, three transmitters on the infirmary’s roofs broadcast it at a low frequency that slips through stone and concrete in a way the signals carrying phone calls and wireless data do not. Every pager in the building hears every message and ignores the ones not addressed to it. Nothing has to know where anyone is. There is no handshake between device and network, no login, no app, no update to install at four in the morning. The pager listens, beeps, and shows a number. That is all it can do, and it is all it has to do.
Dennis Fulford has carried one for all 34 years he has worked as a porter at Halbridge Infirmary, and he understands the building as only a porter can. The infirmary is three hospitals pressed against one another—a stone block from 1888, a concrete tower from 1966, and a glass east wing from 2012—joined underground by a tunnel nearly half a kilometer long, which carries pipes, linen carts, the dead, and Fulford. He can list the places where a mobile phone gives up: the dog-leg in the tunnel under the old chapel, the lead-lined rooms of radiology, the mortuary, the boiler house, two elevators in the tower, and the whole of the pathology basement. The pager works in all of them. “It’s never once let me down,” he said. “I’ve let it down. I’ve dropped it down a toilet twice. Different toilets.”
Battery life is the other half of it. A pager runs for about three weeks on a single AA battery, and when it does run down, a replacement sits in a drawer at every nurses’ station. A smartphone running a clinical messaging app, hunting constantly for a signal in a building that seems designed to block one, needs charging at least once a shift and often twice. During the hospital’s first smartphone pilot, in 2017, the most common single cause of missed messages was a phone that had simply gone flat in the small hours, which is when the fewest people are on duty and each of them counts the most.*
Three attempts and counting
Callum Reyes, the hospital’s head of clinical communications, has led two of the three attempts to retire the pagers, and he describes them with the even-tempered weariness of a man recounting a long renovation. The first, in 2017, put smartphones on four wards for six months. Messages were delayed or lost; phones died; nurses who could not tell whether a doctor had received a message sent it again, then rang the switchboard to have the doctor paged anyway. By the end of the pilot, those four wards were generating more pages per night than before it began.
The second attempt, in 2021, fixed most of the technical faults. It brought a better network, charging racks on every ward, and phones shared between shifts, signed in and out like keys. It foundered on other things. In its first six months, 93 phones went missing. Wiping them with alcohol between patients, as infection control required, clouded the screens within a year. Unlocking one through surgical gloves proved close to impossible. And one night in 2022, an automatic software update restarted every phone on the network at 3:10 a.m.; they came back one by one over the following 20 minutes. “Every problem was solvable,” Reyes said. “The trouble is that the pager doesn’t have any of them, so every solution is pure cost.”
The third plan, approved in 2024, is less ambitious and more likely to work. Smartphones will take over routine messages, which are most of them: a nurse asking a doctor to review a blood result, a ward asking for a porter. Pagers will be kept for cardiac arrests, for major incidents, and for the moments when everything else has failed. The date for switching off the last of them has slipped, over eight years, from 2019 to 2023 to 2026, and now reads “2028, subject to review.” Reyes no longer says the year aloud without the qualifier.
What the beep is bad at
None of this makes the pager a good device, and the junior doctors who carry the most of them have a long list of complaints, nearly all of them fair. A page tells you only that someone wants you, and a number to ring. It does not say who, or why, or how urgently; to find out, a doctor has to find a phone, dial the extension, and often wait while someone goes to fetch the nurse who sent it. A survey of Halbridge’s junior doctors in 2023 found that they spent an average of 47 minutes per night shift answering pages, much of it on hold.† Lowe once received 31 pages in a single hour, four of them about the same patient, from three different people.
Pagers are also, by modern standards, wide open. Their messages travel unencrypted, and anyone with a cheap scanner in the car park can read them. The hospital’s answer is a rule that pages carry no names, only numbers and ward codes, which keeps patients private and makes the pages even less informative. And the devices are aging out of existence. Their manufacturer stopped making the model in 2019. In 2023, Halbridge bought 400 secondhand pagers from a hospital in Westmere that had completed its own switch to smartphones, and keeps them in a cupboard as spares.
Gail Prentice, who has run the infirmary’s night switchboard for 19 years, sends about 600 pages a night and sees the argument from the middle of it. Her desk has a red button under a hinged plastic flap. Lifting the flap and pressing the button sends a crash call to all nine members of the resuscitation team at once, with its own five-note tone; by her estimate, she has pressed it more than 4,000 times. The new system will let nurses message doctors directly, which should take most of the routine traffic off her desk, and she is all for it. She would still like to keep the button. “When I press that, I know it’s gone,” she said. “I don’t have to hope.”
That, more than batteries or basements, is the pager’s real advantage. It is not that it does more. It is that it does so little that almost nothing is left to go wrong, and everyone who carries one knows precisely what it will and will not do. A smartphone is a better tool in nearly every way that can be written into a procurement document. What it cannot yet offer is that kind of certainty, earned over decades, about the one message that matters.
At 7:30 a.m., at the end of her shift, Lowe plugged the smartphone into a rack in the doctors’ mess, beside two dozen others blinking red. The pager she unclipped from her scrubs and handed, still warm, to the doctor taking over from her. He did not check the battery. Nobody ever does.
Notes
*Halbridge Infirmary, Clinical Messaging Pilot: Evaluation Report (2018). Across four wards and six months, 1.8 percent of messages sent to smartphones arrived more than two minutes late and 0.4 percent did not arrive at all. Flat batteries accounted for 38 percent of the failures, more than any other cause; gaps in network coverage accounted for 29 percent. ↩
†Halbridge Infirmary junior doctors’ forum, night-shift workload survey (2023), based on 112 responses and a sample of 40 logged shifts. The 47-minute figure covers the time from a page arriving to the end of the resulting call, and excludes crash calls. ↩