American law defines a medical device by its purpose. The Federal Food, Drug, and Cosmetic Act calls a device “an instrument, apparatus, implement, machine, contrivance,” or a similar article, “intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, treatment, or prevention of disease.” The definition turns on one word, intended, and by that word the fax machine was never a medical device. Nobody designed it to diagnose or treat anything. It was an office machine, sold to law firms, newspapers and shipping agents as readily as to clinics, and it sat in corridors beside the photocopier.
Yet consider what passed through it. Laboratory values on which doses were adjusted. Referral letters asking a specialist to see someone soon. Consultation notes, prescriptions, imaging reports, instructions for a patient being moved from a hospital to a nursing home. If an instrument is judged by what it does rather than what it was meant for, the fax machine was for decades among the most consequential instruments in medicine, and one of the very few that nobody designed or tested for the clinical work it did. I do not suggest it should have been regulated as a device. The title is a provocation with a narrower point: a machine that carries clinical meaning ought to be judged by clinical results, and when it is, most of what deserves credit or blame turns out not to be the machine.
This essay looks at how clinical information moved in 2002, the year I completed my MBBS in Amritsar. That date is the only autobiography here. The essay describes no office, hospital or fax tray I knew. It describes the channels as they are documented across health systems generally, and it follows one document through them, which is invented. The question it asks is the one the machine could not answer: when a page arrives, has anything been communicated?
A Page With a Job #
Begin with the task, because the machine is only interesting in relation to it. Imagine, as a hypothetical, a primary care physician who has seen a patient with a new problem that needs a specialist’s opinion within a few weeks. She writes a referral letter, attaches two recent results, and sends the bundle to a specialist’s office across town.
What does she need to happen? She needs the specialist’s office to know who the patient is and how to reach her. She needs someone there to understand the question she is asking and how urgent it is, so that the appointment is neither delayed nor needlessly rushed. She needs the specialist, when he sees the patient, to have the results in front of him so that the tests are not repeated. And she needs, eventually, a reply: what the specialist found, what he recommends, and who is to do what next.
What does the receiving office need? To identify the patient among thousands, using whatever identifiers the sending office chose to put on the page. To recognise the document for what it is, a referral, and not a result, a request for records or an advertisement. To judge its urgency from words written by a stranger. To put it in front of someone who will schedule the visit and in front of the specialist on the day. And to know where to send the answer.
Laid out like this, the referral has at least six stations: it must be sent, received, matched to a patient, read and judged, acted on, and answered. Each is a place where it can fail, and only the first two belong to the machine. The page’s job was never to arrive. It was to change what several people did over the following weeks.
A Picture of a Page #
The idea of sending a picture down a wire is older than the telephone. A Scottish inventor named Alexander Bain received a British patent in May 1843 for what he called an electric printing telegraph, a mechanism for reproducing marks at a distance. In the 1860s the Italian physicist Giovanni Caselli’s pantelegraph carried handwriting and drawings commercially between Paris and Lyon, about a decade before anyone made a telephone call. For a century afterwards facsimile transmission was a specialist’s tool, used by newspapers for photographs and by weather services for their charts. Xerox’s machines of the mid-1960s brought it into offices, slowly; one of the early desk models took about six minutes to send a page.
What made the fax ubiquitous was not a machine but a standard. In 1980 the international telecommunications body then called the CCITT adopted a recommendation, known as T.4, for what became Group 3 fax, and with it any compliant machine could talk to any other, whoever had made it, in well under a minute a page. Through the 1980s and 1990s fax machines spread into nearly every office that had a telephone line. By 2002, in North America and much of Europe, a clinic without one would have been as strange as a clinic without a telephone.
The mechanism matters for everything that follows. A fax machine does not read. It scans the page into a grid of tiny black and white dots, compresses the long runs of white that make up most of any document, and sends the result as tones down a telephone line; the machine at the other end reverses the process and prints a copy. The word facsimile comes from the Latin for make alike, and that is all the machine does. It makes a likeness of marks. It has no idea whether the marks are a laboratory value, a signature or a lunch menu. Everything that makes the page a clinical document, the patient’s name, the date, the question, the urgency, exists only for the human being who looks at the copy.
That indifference to meaning was the fax’s weakness and the secret of its success. Because it did not care what the page said, it required no agreement about what pages should say. Two offices that shared no software, no vocabulary, no patient numbers and no contract could exchange documents the moment one of them knew the other’s telephone number. A handwritten note, a printed report, a signed consent and a drawing travelled equally well. In a field that has spent forty years and enormous sums trying to make its computer systems exchange information, the humble fax solved the exchange problem in 1980 by declining to understand anything. It was, in its way, the most successful interoperability standard medicine ever adopted, and it succeeded because it asked so little.
What the Report Said #
When a fax was sent, the sending machine could print a small report: the date and time, the number dialled, the number of pages, and a single word, usually OK. For a busy office this report was precious. It was evidence, which could be stapled to the original and filed, that the referral had gone.
It is worth being exact about what that word proved. It proved that a machine answering at the number dialled had accepted the pages over the line. It did not prove that the number was the right one; a misdialled digit produces a perfectly confident OK from the wrong office. It did not prove that the pages printed legibly at the other end, or at all, if the receiving machine had run out of paper or toner and was storing them in a memory that might later be cleared. It did not prove that the right person collected them, that they were matched to the right patient, that anyone read them, or that anyone did anything. The report confirmed the first of the six stations. It was silent about the other five, and its brisk OK made the silence easy to overlook.
Shakespeare built the catastrophe of Romeo and Juliet on a message that failed without a report. Friar Laurence, who has given Juliet a potion that will make her seem dead, sends a letter to Romeo in Mantua explaining the plan, and in the fifth act we learn its fate. The letter never arrives. Friar John, who was to carry it, went first to find another friar who was visiting the sick, and “the searchers of the town, / Suspecting that we both were in a house / Where the infectious pestilence did reign, / Seal’d up the doors, and would not let us forth.” He could not deliver the letter, nor find anyone to carry it, “So fearful were they of infection.” He brings it back: “here it is again.” By then Romeo’s servant has already ridden to Mantua with the news he believes, that Juliet is dead, and Romeo has asked him the question that should have saved the play. “Hast thou no letters to me from the Friar?” “No, my good lord.” “No matter,” says Romeo, and sends for horses.
A medical reader can hardly miss that the letter was stopped by a public-health measure, a house sealed by searchers who feared plague. But the structure of the failure is what interests me. The sender had no way of knowing his message had failed until the messenger walked back through his door, too late. The receiver noticed that a message he expected had not come, asked about it, and dismissed the absence. A transmission report would have rescued the first half of that sequence: Friar Laurence would have known within a minute that his letter had not gone and could have tried again, as he does when he says, too late, “I will write again to Mantua.” That fast confirmation of failure was the fax’s real contribution to safety, and it should be credited. What no report could have rescued was the second half. Romeo’s “No matter” is a receiver’s decision about an absence, and no machine at either end can make that decision well on his behalf.
ਪਹੁੰਚ #
Punjabi has a word that holds the whole problem. ਪਹੁੰਚ (pahunch), which Punjabi shares with Hindi and Urdu, means arrival and reach in ordinary speech: to have pahunch somewhere is to have access to it, and a person with pahunch is someone with connections who can get things done. Platts’s dictionary of 1884 lists its senses in a sequence that reads like a description of a referral’s progress: “Arrival; reach; access, admittance; attainment, comprehension; sagacity, penetration; capacity, power; — receipt, acknowledgment.” In one word, the language runs from a thing arriving, to a mind grasping it, to an answer that says it was received.
The fax’s OK reports pahunch in its first sense only. Everything after arrival in Platts’s list is human work, and most of it was done by people whose job titles said clerical.
Follow the hypothetical referral into the receiving office. It comes out of the machine, perhaps interleaved with pages from three other senders that arrived in the same quarter of an hour. Someone has to take the stack from the tray and separate it into documents, which is easy when each has a cover sheet and numbered pages and surprisingly hard when it does not. Someone has to decide what each document is: this is a referral, this a result, this a request for records, this a pharmacy query, this an advertisement for office supplies. Someone has to identify the patient, using the name and date of birth the sender wrote, since the sender’s own patient number means nothing here. Someone has to judge urgency, from the words of a doctor they have never met, and decide whether this page goes to the bottom of the scheduling pile or to a nurse now. Someone has to put it where the specialist will see it on the right day, and in a paper office that meant filing it in a folder, with all the hazards described in What Paper Charts Did Terribly. Later, in offices with electronic records, someone had to scan it back into a computer as an image and label it, by document type and date, so that it could be found; a fax labelled wrongly at that moment becomes the note on page forty, present and unfindable, as The Consultant’s Note on Page Forty describes.
There is a line in Romeo and Juliet that belongs at every fax tray. Early in the play Capulet hands a servant a list of guests to invite to his feast, and the servant, who cannot read, goes into the street and asks the first gentleman he meets: “I pray, sir, can you read?” The gentleman is Romeo, who reads the list aloud, learns of the feast, and goes. The play’s first document is delivered to a carrier who cannot read it and routed, by that accident, to the one man who should never have seen it. Every fax that arrived in a clinic posed the servant’s question to someone: can you read this, and do you know whose it is and where it should go? The machine never asked. The people at the tray asked it all day, and the safety of the whole channel rested on how well they answered. The flagship essay on this site, The Primary Care Office After the EHR, describes such staff as human integration software between systems that do not talk to each other. In 2002 they were the integration, and the fax machine got the credit.
The Cover Sheet #
Because the machine carried no information about what it was carrying, the people who used it invented a genre to supply some. The fax cover sheet is one of the most widely written documents of the late twentieth century and one of the least read, and it rewards a moment’s attention, because it is a record of everything the channel lacked.
Look at its fields. To, with a name and sometimes a department: the machine delivered to a number, not a person, so the sheet had to say who the person was. From, with a return number: the machine’s report told the sender that the page had gone, but nothing told the receiver where to send a question. Pages, including this one: the machine did not know when a document ended, so the sheet told the receiver how many pages to expect, which was the only defence against the missing middle described below. Re, or Subject, and often a box to tick for urgent: the machine treated every page alike, so the sheet had to carry priority. In clinical offices the sheet often grew a further line, a patient’s name or identifier, and then a dilemma, because a name on a cover sheet is visible to everyone who passes the tray.
At the foot of most clinical cover sheets sat a paragraph in small print. Its wording varied, but its substance was nearly universal: the information that follows is confidential and intended only for the named recipient; if you have received it in error, notify the sender and destroy the pages. Read as literature, it is an extraordinary paragraph. It is a message addressed to the wrong reader, written in advance, by a sender who knows that some fraction of what she sends will go astray and is asking a stranger to repair the failure. No other clinical document I know is written in advance for the reader it should never reach. The confidentiality notice was the channel’s confession, printed on every transmission.
The cover sheet was a good piece of improvised design, and like most improvisations it worked only as well as the people filling it in. A sheet with every field completed turned a picture of a page into something close to an addressed message. A sheet left blank, or a stack sent with no sheet at all, returned the receiving office to the servant’s question at the tray. What the sheet could never do was report back. It travelled one way, like the page it introduced.
Five Ways a Page Goes Wrong #
The failures of fax communication had different origins and called for different remedies, and lumping them together as the fax problem has made them harder to fix. At least five can be distinguished.
The first is the wrong destination. A number is misdialled, or an old number is used for an office that has moved, or a speed-dial button is pressed one row too low. The page arrives, with its OK, at a stranger’s machine, where it is a privacy breach and a lost message at once. The United States health privacy regulator’s own guidance, discussed below, lists precautions against exactly this: confirming a number before faxing to one not regularly used, and programming frequently used numbers into the machine to avoid misdirecting information. A safeguard is a portrait of a hazard: nobody writes down precautions against a danger nobody expects.
The second is the missing page. A three-page referral arrives as pages one and three because the second jammed in the sending machine or was left on the glass. If the pages are numbered 2 of 3, the receiver can notice; if they are not, the missing middle may never be suspected, and the letter that reaches the specialist will read coherently enough to conceal that half its argument is gone.
The third is poor quality, and here there is documented harm. In 2000 J. V. Linden of the New York State Department of Health reported three serious errors in transfusion medicine that had occurred within three months, all arising from the misinterpretation of laboratory results transmitted by fax. Two units of blood that should never have been released were released, and one unit mislabelled as to blood group was taken into a hospital’s inventory. Most blood collected in the state was tested by reference laboratories, and their results were often faxed. The causes identified were distortion during transmission, misreading, and failure to notice a separate report that some results were still pending. Her recommendation was modest and exact: results that are going to be faxed should be made clear. A machine that reproduces marks reproduces faint marks faintly, and a reader who expects a familiar layout reads what he expects.
The fourth is duplication. A sender who hears nothing sends the referral again; a result is faxed and then also posted; an amended report is faxed as a new document while the original is already filed. Now there are two versions in the receiving office, and nothing on either says which is current or that the other exists. Duplication looks like the opposite of loss, and it produces the same uncertainty.
The fifth, and in my view the most underrated, is the unmatched identifier. The fax carries whatever identifiers the sender chose to write, and the receiver must match them to his own records. A name may be spelled differently in the two offices; Punjabi names transliterated into English are a rich source of such variation, with one family name written several ways, or Kaur recorded as a surname by one clerk and a middle name by another. Dates are worse. A date of birth written 03/04/1960 is the third of April in India and Britain and the fourth of March in the United States, and a page that crossed between those conventions could match the wrong patient, or none. A page matched to no one is filed nowhere; a page matched to the wrong person is filed somewhere dangerous.
These are five different failures with five different remedies: verified numbers, numbered pages, clear layouts, version labels and shared identifiers. None of them is a remedy for the machine.
Other Roads in 2002 #
The fax was the fastest written channel in 2002, but it was one road among several, and an honest account of how information moved that year has to include the others.
The post carried most letters between doctors in most countries. It was slow, measured in days, but it was sealed, and a letter in an envelope addressed to a named physician had a privacy and a sense of occasion that a page on a shared tray lacked. Inside large hospitals, internal mail and couriers moved notes and films between departments, with all the virtues and vices of any system that depends on a person walking a route. The telephone remained, as it still is, the channel best at conveying urgency and worst at leaving a record. A call could convey alarm in a tone of voice that no page could carry, and it could be answered with a question. But unless someone wrote down what was said, a telephone conversation left no trace, and the details of a result read aloud could be misheard by a tired listener as easily as a faint fax could be misread. Electronic mail existed and was spreading quickly in offices, but it was not yet an ordinary clinical channel, and most clinicians would have hesitated to send a patient’s results by it.
And then there was the patient. In much of the world in 2002, including many parts of India, the most reliable carrier of a person’s medical history was often the person. Outpatient records were commonly in the patient’s own keeping, as prescriptions, reports and discharge cards kept at home and brought to each new doctor. India’s national rules of professional conduct for doctors, issued by the Medical Council of India in the spring of 2002, required, among much else, that a request for medical records from a patient or an authorised attendant be answered with the documents within 72 hours. That was a right of access, and an important one. It was also, in practice, a design for how records travelled between institutions: on request, to the patient, who carried them onward. The Handoff, published here earlier, considers the patient made into a courier at discharge and argues that she should be treated as a reader of her own story rather than as a sealed envelope. Here I want to add a distinction that the year 2002 makes visible.
There is a difference between a patient-held record by default and a patient-held record by design. By default, it is a bag of papers that accumulates because no institution keeps a complete account, and its completeness depends on the patient’s care and luck. By design, it is a document that clinicians are expected to read and write in, made to last, in a form everyone recognises. Japan introduced a maternal and child health handbook, carried by the family, in the late 1940s; in Britain, parents are given their child’s personal health record, the “red book,” at or soon after birth; in many countries women carry their own maternity notes. The evidence on the last of these is modest but interesting. A Cochrane review updated in 2015 found four small trials involving 1,176 women. Women who carried their own notes were more likely to feel in control of their care; in the two trials that reported it, there was no significant difference in notes lost or left at home; and women who had carried their notes were more likely than the others to want to carry them again. The reviewers also reported more operative deliveries among the women carrying their notes, in the one trial that measured them, and listed this among the possible harms; they judged the trials small and the evidence insufficient. Nobody should build a policy on four small trials. But they suggest that a record in the patient’s hands is not inherently a failure of the system. It becomes one when it is the only channel and nobody designed it to be.
Why the Machine Stayed #
It is easy, and by now customary, to laugh at the fax. A technology older than the personal computer was still carrying clinical documents long after most industries had abandoned it, and the sight of a modern hospital printing a page so that it could be scanned back into a computer has an obvious comedy. In December 2018 the government in England told the National Health Service there that it would be banned from buying fax machines from the following month, and must phase them out entirely by 31 March 2020, replacing them with secure email; earlier that year the Royal College of Surgeons had found nearly 9,000 fax machines in use across the NHS in England. The headline in the BBC’s report called them “absurd.”
Ridicule obscures what the fax actually solved, and a fair account should give the machine its due before judging it. The fax required a telephone line, a machine and a number. It required no shared software, no common patient identifier, no agreement about vocabulary or format, no contract between institutions and no technical staff at either end. It carried handwriting, signatures and drawings. Its standard dated from 1980, and almost every fax machine in the world complied with it. A small clinic could exchange documents with a large hospital, an insurer, a pharmacy and a laboratory on the day it opened, without negotiating with any of them.
Every proposed replacement had to overcome the same obstacle, which is the obstacle of any network. A new channel is useful only when the person you need to reach has adopted it too. The first office to abandon the fax loses contact with everyone who has not, and so every office has a reason to keep its machine until all the others have given up theirs. Costs and benefits are also unevenly shared: an office that invests in sending structured electronic documents bears the expense, while much of the benefit, in cleaner matching and fewer misreadings, goes to the receiver. Under those conditions the persistence of the fax is not evidence of stubbornness or stupidity. It is evidence that the fax was the lowest common denominator of a fragmented system, and the lowest common denominator is precisely what connects strangers.
This is why I think the ridicule is aimed at the wrong target. The fax was not the cause of fragmented communication. It was the symptom that made fragmentation tolerable, and in making it tolerable it reduced the pressure to cure it. A replacement that is only faster than the fax will fail, because speed was never the fax’s problem. A replacement worth having must match the fax’s universality, its willingness to connect any office to any other, and add the three things the fax never carried: structure that a machine can read, identity that both ends share, and an acknowledgment that comes from a person rather than a machine.
The Return Leg #
A referral is a question, and a question is not finished until it is answered. The return leg of the hypothetical referral, the specialist’s reply to the referring physician, travelled in 2002 by the same channels the referral had used, and it met the same hazards in reverse: a reply faxed to a general number, matched by a clerk to the right patient or not, filed in a folder or placed in front of the doctor who asked. But it also met a hazard of its own, which was that nobody was necessarily waiting for it.
The careful offices of the paper era solved this with a humble tool, a referral log: a book or a sheet in which each referral sent was written down, with the date and the specialist, and a column left open for the date the reply came back. The open column was the whole point. It turned the absence of a reply from a silence into a visible gap, and a clerk or a nurse who looked down the column once a week could see which questions had gone unanswered and telephone to find out why. It was the paper answer to Romeo’s “No matter,” a system for noticing that an expected letter had not come and treating the absence as a matter after all.
The log required no technology and a good deal of discipline, which is the profile of most effective safety measures. Its weakness was that it depended on someone looking at it, and its strength was that it put the burden of noticing absence on the sender, who was the one person certain to know that a reply was due. Neither the fax machine nor the post could do that. A channel can carry an answer. Only a person, or a system designed by people for the purpose, can notice that an answer has not been carried.
Privacy Without Slogans #
Privacy is the subject on which talk about the fax is most often careless, in both directions. It is common to hear that fax is secure, or that it is compliant, or that it is a breach waiting to happen. None of these is true as a general statement, because privacy belongs to the way a channel is used rather than to the channel.
In the United States, 2002 was a year of preparation. The federal Privacy Rule under the Health Insurance Portability and Accountability Act had been published in December 2000; its final modifications were published on 14 August 2002; and most covered entities had to comply by 14 April 2003. The regulator’s guidance on faxes, which it has kept in place, is notable for its moderation. It says that the rule allows health care providers to share protected health information for treatment without the patient’s authorisation, by telephone, fax, email or otherwise, as long as they use reasonable safeguards. Its examples are ordinary: a laboratory faxing a patient’s test results to a physician; a hospital faxing a patient’s care instructions to a nursing home. Its safeguards, described earlier, are just as ordinary. The guidance neither blesses the machine nor condemns it. It asks for care in proportion to the channel.
The same moderation applies to the hazards. A fax tray in a shared corridor is an exposure; so is a misdialled number; so is a printed page left on a desk. A machine in a locked room, sending to verified numbers, with pages collected promptly by the person responsible, is a reasonable channel for many purposes. In India, the 2002 regulations placed the duty of confidentiality on the physician, in language that What Paper Charts Did Terribly quotes. The duty did not change with the medium. Only the places where it could be breached changed.
A Changed Mind #
Return to the hypothetical referral and ask what it would mean for it to have been communicated. Not that it arrived: the report said OK within a minute of sending. Communication happened, if it happened, some days later, when a scheduler understood that the patient needed to be seen within weeks and booked her accordingly, when the specialist read the question and answered that question rather than another, and when a reply came back that told the referring physician what to do next. Each of those was a change in what a person knew and did. None of them was visible to the machine.
If the fax machine is to be called a medical device, then it should be judged as one, by what happened to patients. By that standard its record is mixed and, more importantly, not really its own. Where clinics had careful people at the tray, numbered pages, clear layouts and verified numbers, the fax carried a great deal of medicine well, and quickly, between institutions that had no other way to talk. Where those things were missing, it delivered pages with perfect reliability to places where they could not be understood. The machine’s performance was the performance of the people and procedures around it, and that is the lesson to carry into whatever replaces it. A new channel should be judged not by the speed at which it moves a document but by the probability that the document reaches pahunch in all of Platts’s senses: arrival, comprehension, acknowledgment.
Friar Laurence, in the end, did not leave everything to a second letter. Learning that his first had failed, he said he would write again to Mantua, but he also sent for a crowbar and went to the tomb himself, because Juliet would wake within three hours and someone who knew the plan had to be there when she did. He arrived too late. The play does not offer that as a lesson, but medicine might. Some messages are routine and can travel by any channel that is reliable enough. A few carry so much that they need a person who understands them to see that they have been understood. The skill, in 2002 and now, lies in knowing which is which, and in building systems that tell the sender, as the fax never could, not only that the page arrived but that someone read it and said so.
Evidence and Editorial Note #
Documented findings. Public biography relied on: MBBS, Sri Guru Ram Das Institute of Medical Sciences and Research, Amritsar, 2002; Punjabi as first language. The statutory definition of “device,” quoted: 21 U.S.C. § 321(h)(1), Legal Information Institute text. Alexander Bain’s British patent of 27 May 1843 for an “Electric Printing Telegraph”; Giovanni Caselli’s pantelegraph and the commercial Paris–Lyon service of 1865; Xerox’s office machines of 1964–1966, one taking about six minutes a page; the CCITT’s 1980 Recommendation T.4 and Group 3 interoperability; the scanning, compression and tone transmission of a bitmap image; fax ubiquity in offices in the 1980s and 1990s: Wikipedia, “Fax,” and the sources it cites. The NHS in England barred from buying fax machines from January 2019 and required to phase them out by 31 March 2020 in favour of secure email; the Royal College of Surgeons’ finding of nearly 9,000 machines in use; the headline word “absurd”: BBC News, “NHS told to ditch ‘absurd’ fax machines,” 9 December 2018. Linden JV, “The release of unsuitable units through misinterpretation of laboratory results transmitted by facsimile,” Transfusion 2000;40:435–438 (three serious errors within three months in New York State; two unsuitable units released and one ABO-mislabelled unit taken into inventory; reference laboratories often faxing results; causes including distortion, misreading and failure to note a separate report of pending results): PubMed abstract, PMID 10773055. HIPAA Privacy Rule published 28 December 2000, final modifications 14 August 2002, compliance by 14 April 2003 for most covered entities; the HHS answer on sharing information for treatment by fax, its examples and its safeguards (confirming numbers, pre-programming frequently used numbers): US Department of Health and Human Services web pages on the Privacy Rule and its frequently asked questions. Indian Medical Council (Professional Conduct, Etiquette and Ethics) Regulations, 2002, notified 11 March 2002, clause 1.3.2 (requests for records acknowledged and documents issued within 72 hours): text published by the National Medical Commission, incorporating amendments to 2016. Brown HC et al., “Giving women their own case notes to carry during pregnancy,” Cochrane Database Syst Rev 2015, CD002856 (four trials, 1,176 women; greater sense of control; no significant difference in notes lost or left at home in two trials; more operative deliveries in the note-carrying arm of the one trial reporting it; preference to carry notes again; trials small): PubMed abstract, PMID 26465209. Japan’s maternal and child health handbook from the late 1940s, and the UK Personal Child Health Record (“red book”) given to parents at or soon after birth: Wikipedia, “Personal Child Health Record,” and the sources it cites. Shakespeare, Romeo and Juliet, Act 1 scene 2 and Act 5 scenes 1–2, quoted: Project Gutenberg text. ਪਹੁੰਚ (pahunch) and its quoted senses: Platts, A Dictionary of Urdu, Classical Hindi, and English (1884), via the Digital Dictionaries of South Asia, with ordinary Punjabi usage. The Latin origin of facsimile; the usual fields of fax cover sheets and the common form of their confidentiality notices; the use of referral logs in paper-based offices; the general state of post, couriers, telephone and email as clinical channels around 2002; patient-held outpatient files in much of India; differing day-month conventions in written dates; the network obstacle to replacing a shared channel: standard reference knowledge, stated in general terms. Vendor representations. None. Xerox is named only as a historical maker of early office fax machines; no current product is named or assessed.
Recollection and tradition. The only biographical fact used is the public one that the author completed his MBBS in 2002; no office, hospital, fax machine, colleague, patient or incident is recalled. The referral from a primary care physician to a specialist, followed through its stations, is a hypothetical illustration, introduced as such. Remarks on the transliteration of Punjabi names, on patient-held files, on cover sheets and on referral logs describe general practice, not recollections.
Analysis and forecast. The reading of the device definition as a provocation; the six stations of a referral; the argument that the fax succeeded as an interoperability standard because it ignored meaning; the readings of Romeo and Juliet and of pahunch; the reading of the fax cover sheet and its confidentiality notice as a record of what the channel lacked; the five failure types and their remedies; the referral log as a means of noticing absent replies; the distinction between patient-held records by default and by design; the account of the fax’s persistence as a network effect and a symptom of fragmentation; and the criteria proposed for a replacement are the author’s interpretation as of November 7, 2026. They are not legal or compliance guidance. Independence. This essay is not sponsored by any institution, company, health system or technology developer. It is personal reflection, not medical advice; readers seeking care should contact their own physician’s office.
About the Author #
Kanwar Partap Singh Gill, MD, traces his family roots to Khadur Sahib in Punjab’s Majha region and grew up in Amritsar. He attended Spring Dale School and completed pre-medical studies at Khalsa College in 1994–1996. He earned his MBBS from Sri Guru Ram Das Institute of Medical Sciences and Research in Amritsar in 2002.
He completed PGY-1 at the University of Arkansas for Medical Sciences in Little Rock in 2007–2008, and PGY-2 and PGY-3 with UCSF Fresno at Community Regional Medical Center from 2008 through 2010. He has practiced medicine in California since 2010.
His career has crossed four distinct information eras: predominantly paper-based medicine, early and fragmented electronic records, the integrated-EHR era, and the emerging age of generative and agentic artificial intelligence.