# The Handoff *From the succession at Kartarpur and Khadur to the morning sign-out and the discharge* Kanwar Partap Singh Gill, MD · November 5, 2026 · Becoming a Doctor Medicine seldom loses information where it is made. A blood pressure is taken, a culture is drawn, a daughter mentions that her father has not been himself since Tuesday, and at the moment of its making each of these facts has a custodian: the person who saw it happen, who knows how it was obtained, how far it can be trusted and what it might mean. Afterwards the facts themselves are rarely destroyed. Records are durable, laboratories keep their results, and even the daughter's remark will usually survive somewhere, in a note or in a nurse's memory. What goes missing is subtler and more perishable than a fact, and it goes missing in a particular place. It is lost where one person's understanding has to become another's: at the change of shift, at the transfer from one ward to the next, at the door through which a patient leaves the hospital for a clinic down the road. Medicine calls that place the handoff. I have come to think it is where most of medicine's forgetting is done. A Roman poet found the image for passing things on in a race. Early in the second book of *De Rerum Natura*, explaining how the sum of things renews itself while each particular thing decays, Lucretius observes that some peoples grow and others dwindle, that the generations of living creatures change in a short space, *et quasi cursores vitai lampada tradunt*: and like runners they hand on the torch of life. The image comes from the torch races of the Greek cities, relays in which the prize went to the team whose flame reached the end still burning. The image has lasted because it puts the danger in exactly the right place. No runner covers the whole course, and the race is decided less by speed between the exchanges than by what happens at them. At the moment of passing two hands are on the torch at once, and that is where it is dropped, or gutters, or goes out. A hospital is a relay of this kind, run around the clock, and it is tempting to say that the patient's care is the torch. Before the end I will want to take that image back, because the patient is not the torch. The patient is the one person in the race who runs every stage. But Lucretius is right about where the danger lies. In medicine, knowledge is made in one pair of hands and dropped between two. ## Light and Body My family's roots are in [Khadur Sahib](../../roots/khadur-sahib/index.html), a small town in the Majha that owes its place in history to a succession. In 1539, at Kartarpur on the Ravi, Guru Nanak named the person who would follow him. He passed over both his sons, Sri Chand and Lakhmi Das, and chose a disciple called Lehna, who had come to him years before and stayed to serve. He gave him a new name, Angad, which tradition glosses as *my own limb*: ang, ਅੰਗ, is the Punjabi word for a limb or part of the body. The gaddi, the seat of the Guru (ਗੁਰੂ ਗੱਦੀ), passed in a ceremony remembered for its modesty, a coconut and five paise set before the disciple and the master bowing to him. Guru Angad then went to Khadur. There, tradition remembers, he gathered Guru Nanak's hymns and gave the Gurmukhi script its settled form, and there, before his death in 1552, he made his own succession. He too passed over his sons, and chose a disciple, Amar Das, by then an old man who had served him for years. In Guru Amar Das's time the hymns were gathered into the volumes now called the Goindval pothis; in 1604 Guru Arjan installed the Adi Granth at Amritsar; and in 1708, when the line of living Gurus came to an end, Guru Gobind Singh vested the guruship in the Guru Granth Sahib and in the Panth, the community of Sikhs. These are events in the history of a faith, and their meaning for Sikhs is spiritual. It is not mine to turn them into a lesson in hospital management, and I want to be exact about the narrow use I make of them. The oldest accounts of the first succession are songs. The bards Satta and Balwand, whose var in Raag Ramkali is preserved in the Guru Granth Sahib, sing of the canopy raised over Lehna's head and of the master bowing to his disciple while he still lived; and they say, in substance, that the light was the same and the way the same, and that only the body had changed. Bhai Gurdas, the scribe of the scripture's first compilation, tells it in nearly the same terms in his first var: Lehna installed in Guru Nanak's lifetime, light merged into light, the form altered. As the songs tell it, then, continuity did not depend on anything being explained to the successor. What continued was a single light that had taken another form, and no body of information had to pass between two minds for it to hold. That is why the history serves a physician as a contrast rather than as an analogy. A hospital is arranged the other way round. The body stays the same: the man in the fourth bed does not change at seven in the morning. What changes is the light that falls on him, the attention, knowledge and worry of the particular person responsible for him, and that light is never the same light. Medicine's successions are between different minds, which will not see the same thing in the same way however carefully they speak. Whatever continuity exists between them has to be built, a sentence at a time, in the few minutes the shift allows. A tradition that thought for more than a century and a half about continuity also left practices, and some of them are human and historical, open to anyone's notice. The first succession was made while the giver lived and in his presence, so that there was no interval in which the seat stood empty and no doubt about who had been chosen. It was made in public, and the giver's own gesture, bowing to the receiver, told everyone that the responsibility had moved. The receiver was given a name that bound him to the one before him. And the teaching was not left to memory: it was gathered, written in a script fitted to it, and finally given authority in the kept word and in the community that keeps it. Presence, acknowledgment, a bond of identity, a faithful record and a body of people around it. I do not offer these as a model for anything. I notice only that every clinical handoff I describe below fails, when it fails, for want of one or another of them. ## Seven in the Morning In most American teaching hospitals the day begins with a handoff at around seven, when the night team tells the day team what happened while it slept; in the evening the direction reverses. The number of these exchanges grew after July 2003, when the Accreditation Council for Graduate Medical Education set national limits on residents' hours: eighty a week averaged over four weeks, and [twenty-four hours of continuous duty](https://www.acgme.org/globalassets/dh_dutyhoursummary2003-04.pdf) with up to six more for continuity of care and education. My own residency, from 2007 to 2010, fell entirely under those rules. In 2011 the council went further and limited first-year residents to shifts of sixteen hours. In 2017, after a large trial in surgical training programmes had found flexible schedules no worse for patients than standard ones, it allowed first-year residents twenty-four hours again, with up to four more permitted for patient safety and education, including the work of handing patients over. I think the original limits were right in principle. A doctor in his twenty-eighth hour awake is not a safe custodian of anyone. But the arithmetic has a second column. Every hour taken off a shift is an hour in which the patient belongs to someone else, and every change of ownership requires a sign-out. A tired doctor who knew the patient is exchanged for a rested doctor who knows a summary. I read the extra hours of 2017 partly as a public admission that the summary has a cost. It is worth slowing down to ask what a handoff actually transfers, because the word suggests a single object passing from hand to hand, and there are at least five. There are facts: the diagnosis, the medicines, the result of this morning's scan. There is interpretation: what the departing doctor makes of the facts, how sure she is, which of them she trusts. There are anticipated problems: what might happen tonight, and what should be done if it does. There are unfinished tasks: the result still to be checked, the family still to be called. And there is responsibility itself, which is not information at all but a relation, the answer to the question *who will act*. A written list carries the first of these very well and the fourth tolerably. The other three it hardly carries at all. Imagine a hypothetical sign-out at the end of a long day. The departing resident has a dozen patients and a printed list. For one of them, an elderly man admitted two days earlier with pneumonia, the list reads: *improving on antibiotics, oxygen weaned, likely discharge tomorrow, follow up blood cultures.* Every word is true. What the list does not say is what the resident noticed at four in the afternoon and could not quite name: that the man was breathing a little faster than he had in the morning, that his daughter thought he was not himself, that he had not wanted his supper. None of it reached the threshold of a finding. All of it was the reason she looked in on him twice before leaving. The night doctor inherits a man who is getting better. If the night goes badly, he will meet the decline with no sense that it had begun, and he will spend its first minutes rediscovering what his colleague half knew. The list was accurate and it failed, and the failure was of a precise kind. It carried the facts and left behind the interpretation; it described a trajectory and withheld a doubt. A clean story is easier to tell at the end of a shift than a hesitant one, and the temptation, for a tired doctor who would like to seem competent in front of a peer, is to make an account sound more settled than the situation. Yet uncertainty is often the most important thing being handed over. A suspicion and an established finding are not interchangeable merely because they fit into the same sentence, and *improving on antibiotics* can carry either. Outside medicine everyone understands this. A woman leaving her house in a neighbour's care for a week can say that the boiler is old. Or she can say that it knocks, that the knocking is normal, that if it ever stops knocking her neighbour should ring this number, and that she honestly does not know what the small puddle by the back door means but would like to hear if it grows. The second version takes thirty seconds longer. It hands over judgement along with facts, and it admits what the person leaving does not know. ## The Crown on the Pillow Most writing about handoffs concerns the sender: what she should say, in what order, with what emphasis. The receiver is pictured as a container into which the account is poured. Shakespeare knew better, and the most searching scene I know about a handoff comes near the end of the fourth act of *Henry IV, Part 2*. The old king is dying. His crown lies on the pillow beside him. Prince Hal sits by the bed, sees that a downy feather at his father's lips does not stir, and concludes that he is dead. He speaks to the crown, *O polished perturbation! golden care!*, sets it on his own head, and walks out of the room. The king wakes, finds the crown gone, and has his son brought back. When Hal says he never thought to hear him speak again, the king answers with one of the most exact rebukes in English: *Thy wish was father, Harry, to that thought.* Then, after the son has explained himself and the two are reconciled, the real transfer begins. The king tells Hal what the crown is. *God knows, my son, / By what by-paths and indirect crook'd ways / I met this crown.* He describes the enemies it has made, the friends who helped him to it and may turn, and the policy he advises: to busy giddy minds with foreign quarrels, so that domestic ones are forgotten. And he hopes that his son will hold it more quietly than he did, *For all the soil of the achievement goes / With me into the earth.* There are two handoffs in that scene, and the first is a warning. Hal takes the object with the giver absent. He has the crown and nothing of what his father knows about it: not its history, not its dangers, not the contingencies the old man has spent a reign preparing for. Shakespeare stages this as a kind of theft, though Hal intended none, and the point is not Hal's ambition. To assume a responsibility without the account that comes with it is a presumption, however innocent. The second handoff, with the giver awake, is the one that matters, and it contains almost everything a good sign-out should. There is a judgement of severity, a summary, a list of threats, a plan for what to do if they arise. Then the receiver speaks. *You won it, wore it, kept it, gave it me; / Then plain and right must my possession be.* It is a synthesis, four verbs restating a reign, and it is the moment the transfer is complete. It is also, if one reads it closely, a small act of laundering. The father has just confessed the crooked ways; the son's summary turns them into a plain and right possession. The read-back that confirms the handoff has smoothed away the very doubt the giver was most anxious to pass on. And the doubt does not stay buried. In *Henry V*, on the night before Agincourt, the son prays, asking God not to think, not today, upon the fault his father made in compassing the crown. The soil did not go into the earth with the old king. It was handed over after all, unacknowledged, and it surfaced when it was least wanted. I find the scene more useful than most guidance because it shows that the receiver's work is real work, and can be done well or badly. Receiving is not acknowledging that a message arrived. It is testing whether the account fits what one can see; asking the question that the sender's fluency has made easy to skip; restating the plan in one's own words so that both people can hear where it does not match; saying plainly *I am not sure I agree*, when that is true; and, at the end, accepting particular tasks by name. When the Joint Commission made handoff communication a national patient safety goal in 2006, it required "a standardized approach to handoff communications, including an opportunity to ask and respond to questions." The receiver's voice was written into the rule. Hal's synthesis is a warning about what that voice can do when it confirms too readily. A good receiver restates the doubt along with the plan. *So he is better, but you were uneasy about his breathing this afternoon, and his daughter thinks he is not himself.* A summary that preserves the uneasiness is longer by one clause and more faithful by a great deal. There is a second temptation on the receiving side, which is to believe that fresh eyes automatically see more than tired ones. Sometimes they do. The departing doctor has stopped noticing what she has looked at all day, and the newcomer may see at once what has been sitting in plain view. But fresh eyes also lack history. They see the patient at one point in time, without the trajectory that would tell them whether that point is a good one. Each person holds part of the day, and neither holds the whole. The handoff is where two partial views can be made into a better one, and that happens only if both people speak. ## A Result With No Address The sign-out at least happens in a room, between two people who can see each other and who belong to the same team. The hospital discharge is a handoff whose sender and receiver often never meet, and whose courier is the patient. The difficulty is not only that documents go astray. Sometimes they do, and in a 2007 review in *JAMA* Sunil Kripalani and colleagues found that hospital physicians and primary care physicians communicated directly in only a small minority of discharges, and that a discharge summary was often missing at the first follow-up visit. But the subtler failure happens when every document is present and correct. A 2005 study of more than 2,600 patients discharged from two academic hospitals found that about four in ten went home with a test result still outstanding. A fact on its way, belonging, for the moment, to no one. Consider a hypothetical woman discharged on a Friday after three days in hospital with a fever that has settled. Two tests are still pending, one of them sent to an outside laboratory that takes a week. The discharge summary is a model of its kind. It lists both tests under *pending at discharge* and advises follow-up with her primary care physician in one week. Now follow the result. The physician who ordered it rotates off the hospital service on Sunday, and depending on how that hospital routes its results, the report may arrive in the queue of someone no longer caring for her, or of a team that has already filled the bed. Her primary care physician receives the summary, perhaps before the visit and perhaps after, reads *pending*, and reasonably assumes that the hospital will call if anything turns up. The hospital physician, equally reasonably, assumes that the primary care doctor will look for it at the visit. The woman assumes that no news is good news. The appointment is made for ten days rather than seven, or is not made at all, because the clinic's telephone line was busy on the Monday morning she tried it. Nobody has been careless. Each participant has made the most natural assumption available, and the assumptions interlock perfectly to leave the result without an owner. There is a difference between sending a task and knowing that someone has taken it up. A message can leave one person's queue, arrive in another's, and hang suspended between two reasonable beliefs. The sender believes it has been received. The receiver believes it is information rather than a request. Nothing has been lost in transit. What is missing is agreement about what the message asks someone to do. The discharge summary deserves to be read as a genre, with the attention one would give any other, because its conventions shape what it can say. It is written at the moment of departure, often by the most junior member of the team, at the hour of greatest hurry, and it is read days or weeks later by someone who cannot ask its author a single question. It has fixed parts: the reason for admission, the hospital course, the diagnoses, the medicines, the instructions. Its central section, the hospital course, is a chronicle told largely in the passive voice: the patient *was admitted*, *was started on*, *was transitioned to*, *was deemed stable*. In this grammar things happen to the patient, and no one in particular does them. Formally it is a comedy in the old sense, a story that ends with everyone going home, and its closing line, *discharged home in stable condition*, has the cadence of a resolution. Doubt is a poor fit for the form. It appears, if at all, in one small section near the end, *pending at discharge*, which is the place where the genre admits it is not finished. Look at the grammar of that section and the problem becomes visible. *Blood culture pending.* There is no verb attached to a person. *Follow up with primary care.* An imperative, addressed to whom? To the patient, who must arrange it? To the primary care physician, who was not in the room, has not been asked, and may not know the instruction exists? The summary assigns a task to someone who has not agreed to take it. That is an assignment, and an assignment is not yet a handoff. In the sign-out room the receiver can say *yes, I'll look for that at midnight*, and the saying makes the task his. Across the hospital wall the saying seldom happens, and the task remains grammatically present and practically homeless: recorded in a sentence, available to everyone, accepted by no one. A useful handoff has a receiving verb. Someone will review, call, arrange, reassess or explain. The verb needs a person or a defined role, and the person needs enough context to recognise the task as his. From the patient's side of the door, *the team knows* sounds reassuring. Inside, the team is several people with different schedules, different access and different responsibilities. The patient should not need to understand that machinery to know whom to call and when to expect an answer. ## The Runner Who Runs Every Stage Here I have to take back the torch. When the image of a relay is applied to medicine, the clinicians become the runners and the patient becomes the thing they carry. But the clinicians each run a stage, the night doctor from seven to seven, the hospital physician for three days, the primary care physician from the following week. The patient runs the whole course. She is present at every exchange, the only participant who is, and she is often the only one who carries the whole story from one stage to the next. Herodotus describes a relay of another kind. The Persian royal post, as he tells it, stationed men and horses a day's ride apart along the roads of the empire, each rider covering one stage before handing the dispatch to the next, and he compares the system to the torch race the Greeks ran in honour of Hephaestus. His praise of those couriers, that neither snow nor rain nor heat nor darkness kept them from their appointed stages, is carved, in an English rendering, across the front of New York's main post office. The Persian courier carried a sealed message. He did not need to understand it, and was perhaps better off not doing so; his task was to keep it moving and unopened. At discharge the patient is often given the courier's role: carry these papers to your doctor, take this list to your pharmacy, tell them at the clinic what happened here. But she is not a courier, because the message is about her, and because, at the end of every relay, she is the one who must act on it. She goes home with a list of medicines that may not match the bottles in her kitchen: some new, some stopped, some changed in dose without the change being obvious from the label. She will be asked to tell the story of her admission again, at the pharmacy, at the clinic, perhaps to a relative who drives her, and each telling will be a translation from what she was told, in the hurry of the morning she left, into what she understood. Whether she was told in a language she reads comfortably, whether she has a telephone that is answered and a way to reach an appointment across town, whether anyone checked what she had understood: these are questions about her particular circumstances, and they will be answered differently for different people. They are also, in each case, conditions on which the handoff depends. It would be wrong to paint the patient as helpless. Many patients, and the families who sit with them, are the most continuous record the system has. A daughter who has kept every discharge paper for five years knows things no single clinician knows. A man who has lived with his illness for decades understands its rhythms better than a doctor meeting him for the first time. Patients are capable of carrying a great deal across a transition. The trouble is that transitions are often designed on the assumption that they will, without anyone checking what they have been given to carry. Some of the most careful work on discharge has started from exactly this point. A randomised trial at Boston Medical Center, published in 2009 under the name of the Re-Engineered Discharge, treated the patient as a receiver in her own right: a nurse prepared a written plan for life after hospital, reviewed it with the patient, asked her to explain it back in her own words, and a pharmacist followed up by telephone. Patients in that trial returned to hospital less often in the following month. The method is not complicated. Its premise is the important part. The person going home is not a sealed envelope. She is a reader of her own story, and the discharge succeeds only if she has been allowed to read it. ## Five Letters and What Stands Behind Them There is a handoff method with unusually good evidence behind it, and it is worth reading closely, partly because it is so often summarised badly. In 2014 Amy Starmer and colleagues reported in the *New England Journal of Medicine* on I-PASS, introduced in the paediatric residency programmes of nine hospitals. The letters stand for illness severity, patient summary, action list, situation awareness and contingency planning, and synthesis by receiver. Across [10,740 admissions](https://pubmed.ncbi.nlm.nih.gov/25372088/), medical errors fell from 24.5 to 18.8 per hundred admissions, a reduction of 23 percent, and preventable adverse events from 4.7 to 3.3, a reduction of 30 percent. Adverse events judged unpreventable did not change, which is roughly what one would hope to see if the improvement came from better communication rather than from some general change in the patients. Six of the nine sites showed a significant reduction on their own. The spoken handoffs took about the same time as before, two and a half minutes per patient, and the residents' time with patients and families did not fall. Those are striking numbers, and they are easy to misread. The study compared periods before and after the programme; it was not a randomised trial. The programme was not a mnemonic. It was a mnemonic for spoken and written handoffs together with training in communication, a programme in which faculty observed handoffs and gave feedback, and a campaign to keep the changes going once the novelty had worn off. Five letters printed on a card are the visible part of an effort that changed what senior doctors watched for, what trainees practised and what an institution rewarded. A hospital that adopts the letters without the observation and the culture has adopted a card. It would be unreasonable to expect the same results from it, and the study does not claim them for every setting. Structured handoffs have a characteristic way of failing. A structure can turn conversation into recitation. Once there is a template, people fill it; the boxes fill faster than the thinking does, and a complete-looking account can be more falsely reassuring than an incomplete one, because its completeness suggests that someone has considered everything. Templates also grow. Every adverse event prompts a new field, every field takes a few seconds, and the handoff that was meant to transmit judgement becomes a list of everything that might conceivably matter, in which the one thing that does is harder to find. Anyone who has read a modern electronic note knows how documentation can swell while saying less. The I-PASS results answer part of this objection: the handoffs did not get longer. But the deeper answer lies in what the letters ask for. Two of them, situation awareness with contingency planning, and synthesis by the receiver, are not fields of information at all. They are prompts for judgement and for interaction. *If this happens, do that* requires the sender to imagine the night ahead, which is precisely the work the accurate but failing list did not do. The receiver's synthesis requires the receiver to speak. A form of this kind works the way a strict verse form does. The fourteen lines of a sonnet do not write the poem, and a poet who merely fills them produces something worse than prose. What the form gives is a pressure toward selection: it insists that something be left out, and so it forces a decision about what matters. A good handoff structure does the same. It is valuable to the degree that it makes the sender choose, makes the receiver answer, and makes someone own each task, and it fails to the degree that it lets people fill it in without doing any of the three. Even a perfect handoff, delivered and received with care, will sometimes carry an interpretation that turns out to be wrong. That is not a failure of the method. It belongs to the nature of what is being handed over. ## When the Next Reader Disagrees Suppose the night doctor in the earlier example, having heard the full account, the improvement and the unease, examines the man at midnight and concludes that the day team's reading was wrong: the pneumonia was never the whole story, and something else is going on. What does continuity require of him now? It does not require loyalty to the plan. A handoff transmits a plan in order that it may be carried forward intelligently, which includes the possibility of changing it. The receiver who sticks to a plan he believes mistaken, because it was clearly transmitted, has confused fidelity to his colleague with fidelity to the patient. Medicine's word *continuity* can mislead here. What should continue is the responsibility and the thread of reasoning, so that the next person can see why the plan changed. The plan itself is meant to be revised. A good handoff makes revision easier, because the receiver knows what the sender was thinking and can tell whether a new finding contradicts the plan or merely complicates it. What revision requires, and seldom gets, is a return leg. Most of medicine's relays run in one direction. The night doctor changes the plan and tells the next morning's team, but the evening resident who handed over may never learn that her unease was justified, or that her reading was wrong. The hospital physician who wrote the discharge summary rarely hears whether the pending result was ever seen. Senders seldom learn what their handoffs failed to carry, and so the same omissions recur, made by people who were never told they had made them. A relay that never reports back to its earlier runners cannot improve. One of the quieter merits of the morning handoff, when it works, is that it is a conversation between people who will meet again the following evening, when the direction reverses. The discharge has no such rhythm. Its senders and receivers may never meet at all. There is a humility in this that I would want any physician to keep. To hand over is to accept that someone else will continue the work differently, and may see what one missed. To receive is to accept that one's own fresh reading may be the mistaken one, and that the colleague who looked worried at four in the afternoon may have been seeing something real. Neither person owns the truth about the patient. They own, for a time, the responsibility for looking for it. ## A Sentence With an Addressee I have argued in [The Primary Care Office After the EHR](../the-primary-care-office-after-the-ehr/index.html) that software may soon hold the thread across these transitions: noticing the pending result, matching the medicine list, reminding someone that an appointment never happened. I hope it does, and I would welcome a great deal of that help. But a system that keeps track of every fact can still leave the question of ownership exactly where the discharge summary leaves it, recorded and unaccepted. It will be safe to rely on such help only if it makes plain which human has taken up the problem, and whether that person said yes. The test I would put to any handoff, then, is a simple one, and harder to pass than it sounds. When the exchange is over, does the next person know what is known, what is merely suspected, what is still pending, and what responsibility he has accepted? If the answer to the last part is unclear, the handoff has not happened, however complete the documents. Every handoff contains an unfinished sentence, which is the reason for the handoff in the first place. A patient's story is always unfinished when the shift ends or the door opens, and the work of the exchange is to make sure the sentence has an addressee: a named person who has heard it, said it back, and agreed to write the next line. I return to Khadur, and to the narrow thing I allowed myself to take from it. The tradition tells of a successor chosen in the giver's lifetime, acknowledged in public by the giver's own bow, given a name that joined him to the one before, and entrusted with a teaching gathered carefully enough to outlast them both. What the tradition means by the light that passed is its own, and not a matter for a physician's comparison. But the human gestures around that passage have something to say to anyone whose work is to hand things on. Be present when it happens. Make it plain who now holds the responsibility. Keep the record faithfully. And let the one who receives say, aloud and in his own words, that he has received it. A patient should not have to begin existing again because a different doctor has come on duty. Lucretius's runners hand on the torch of life; ours hand on something smaller and more fragile, a partial understanding of a single life, which is always at its most vulnerable in the moment when two hands hold it at once. ## Evidence and Editorial Note **Documented findings.** Public biography: family roots in Khadur Sahib; PGY-1 at UAMS, 2007–2008; PGY-2 and PGY-3 with UCSF Fresno at Community Regional Medical Center, 2008–2010; practice in California since 2010. Guru Nanak's choice of Lehna over his sons Sri Chand and Lakhmi Das in 1539 at Kartarpur, the name Angad glossed as "my own limb," the ceremony of a coconut and five paise, Guru Angad's move to Khadur, his gathering of Guru Nanak's hymns and formalising of Gurmukhi, and his choice of Amar Das over his own sons before his death at Khadur in 1552: standard Sikh historical reference sources (Wikipedia, "Guru Angad" and "Guru Gaddi"). The Goindval pothis compiled in Guru Amar Das's time, the installation of the Adi Granth in 1604, and the vesting of the guruship in the Guru Granth Sahib and the Panth in 1708: standard Sikh historical reference sources. The var of Satta and Balwand in Raag Ramkali in the Guru Granth Sahib and Bhai Gurdas's first var as accounts of the succession: described and paraphrased, not quoted. Lucretius, *De Rerum Natura* II.79 ("et quasi cursores vitai lampada tradunt"): standard text. Herodotus, *Histories* VIII.98, on the Persian royal post and its comparison with the Greek torch race for Hephaestus, and the paraphrase of his praise inscribed on New York's main post office building: standard text and general knowledge. Shakespeare, *Henry IV, Part 2*, Act 4 (the crown scene) and *Henry V*, Act 4 (the king's prayer before Agincourt): standard texts. ACGME duty-hour standards effective July 2003 (80 hours weekly averaged over four weeks; 24 hours continuous duty plus up to six for continuity of care and education): ACGME summary (linked). The 2011 sixteen-hour limit for first-year residents and the 2017 revision permitting 24 hours plus up to four for patient-safety activities, including transitions of care, and education, after the FIRST trial in general surgery programmes (Bilimoria KY et al., *N Engl J Med* 2016): general knowledge of the ACGME Common Program Requirements and the published trial; Starmer AJ et al., "Changes in Medical Errors after Implementation of a Handoff Program," *N Engl J Med* 2014;371:1803–1812 (prospective intervention study at nine hospitals; 10,740 admissions; errors 24.5 to 18.8 and preventable adverse events 4.7 to 3.3 per 100 admissions; nonpreventable adverse events unchanged; six of nine sites with significant reductions; oral handoff 2.4 versus 2.5 minutes per patient; no significant change in resident workflow; components including the mnemonic for oral and written handoffs, communication training, faculty development and observation, and a sustainability campaign): PubMed abstract (linked). The Joint Commission's 2006 requirement for a standardized approach to handoff communications including an opportunity to ask and respond to questions: Relias Media, reporting the 2006 National Patient Safety Goals. Kripalani S et al., *JAMA* 2007;297:831–841 (infrequent direct communication; summaries often unavailable at the first post-discharge visit): PubMed abstract. Roy CL et al., *Ann Intern Med* 2005;143:121–128 (2,644 patients at two hospitals; 41% with results pending at discharge): AHRQ PSNet summary and the *Harvard Gazette*, July 2005. Jack BW et al., "A Reengineered Hospital Discharge Program to Decrease Rehospitalization," *Ann Intern Med* 2009 (randomized trial at Boston Medical Center; nurse discharge advocate, after-hospital care plan, teach-back, pharmacist telephone follow-up; fewer hospital return visits within 30 days): general knowledge of the published trial, stated without figures. **Vendor representations.** None. Software is discussed only in general terms. **Recollection and tradition.** The author's training dates and sites are his published biography; no particular sign-out, handoff format, colleague or patient from his own training or practice is described. The successions of 1539 and 1552 and the events of 1604 and 1708 are Sikh history and tradition as given in standard sources; the doctrine of one light in successive bodies is stated as the tradition's understanding, not as the author's claim about what all Sikhs believe. The elderly man with pneumonia, the woman discharged on a Friday, and the house and boiler are hypothetical illustrations, introduced as such; none describes a real person or event. **Analysis and forecast.** The use of Sikh succession as a contrast rather than an analogy, and the reading of its human practices (presence, acknowledgment, name, record, community), are the author's personal interpretation and are distinct from the spiritual meaning of those events for Sikhs. The readings of Lucretius, Herodotus and Shakespeare, the five components of a handoff, the reading of the discharge summary as a genre, the "receiving verb," the comparison with verse form, the view that the 2003 limits were right in principle but multiplied handoffs, and the remarks on software are the author's interpretation and opinion as of November 6, 2026. **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. ## Related reading - [Where Does This Fact Live?](https://personal.kpsgill.com/essays/where-does-this-fact-live/) - [The Last Day of Residency](https://personal.kpsgill.com/essays/the-last-day-of-residency/) - [The Empty Waiting Room at 7 AM](https://personal.kpsgill.com/essays/the-empty-waiting-room-at-7-am/)