What a hospitalist is (and why your own doctor isn't at the hospital)

A hospitalist is a physician who practices only in the hospital. Most trained in internal medicine; some in family medicine or pediatrics. They admit patients, order the tests, adjust the medications, call in the specialists, write the daily notes, and decide when someone is well enough to leave. The field has a name of its own, hospital medicine, and a professional body, the Society of Hospital Medicine.

The schedule is the part that catches families off guard. Hospitalists commonly work in blocks, often a week on followed by a week off, a pattern schedulers call 7-on/7-off, and nights are usually covered by someone else entirely. The block is rarely explained at admission, so day eight arrives as a shock: a new face, a fresh read of the chart, and a family telling the whole story again from the beginning.

A writer at The Foglight described families "saying the same things over and over, each time to another new face." That is the honest shape of a long stay, and it is not a sign that anyone is doing the job badly. It is the schedule.

Until the late 1990s, a primary care doctor with admitting privileges would round on their own patients in the morning and then drive to the office. That model is largely gone. Many primary care practices no longer hold admitting privileges at all, and the ones that do often cannot make the arithmetic work: an hour of driving and rounding for two inpatients, against a clinic full of scheduled visits.

What replaced it is a doctor who never leaves. In a 2007 New England Journal of Medicine study of 76,926 patients at 45 hospitals, stays managed by hospitalists ran about 0.4 days shorter and $268 cheaper than those managed by general internists, with no difference in death in the hospital or readmission within 14 days (Lindenauer and colleagues, PMID 18094379). Against family physicians the differences were smaller and the cost gap was not significant.

In a 2017 JAMA Internal Medicine analysis of 560,651 Medicare hospitalizations at 4,535 hospitals, patients whose own primary care doctor stayed in charge had lower 30-day mortality than those handed to a hospitalist, 8.6 percent against 10.8 percent unadjusted. After adjustment the advantage narrowed but held, an odds ratio of 0.94, and those patients were also likelier to go home rather than to a facility (adjusted odds ratio 1.14). They paid for it with stays about 12 percent longer (Stevens and colleagues, JAMA Internal Medicine, 2017).

The authors' own reading is the sentence worth keeping: "longitudinal contact with a patient may translate into meaningful differences in care patterns and patient outcomes." Knowing someone for years is worth something measurable. The hospital does not have it. You do.

The phrase to ask for is attending of record. That is the physician who carries responsibility for the admission and whose decisions stick. In most adult medical stays the attending of record is a hospitalist, which is why "hospitalist vs attending" is a comparison that does not really exist. Attending is a role. Hospitalist is a place of practice. One person is usually both.

Ask it plainly at the nurses' station: "Who is the attending of record today, and how do you spell that?" Write it down. Then ask again tomorrow, because the answer changes more often than anyone warns you.

How often it changes is measurable. Across 617,680 Medicare admissions, in the first three hospital days alone, 49.6 percent of patients saw one hospitalist, 44.5 percent saw two, and 6.0 percent saw three (Goodwin and colleagues, JAMA Network Open, 2021). About half of families are already onto their second doctor before the third morning. That study found only small associations with drug toxic effects and none at all with length of stay.

The handoff itself has been studied on its own, and the picture is mostly reassuring with one edge to it. Across 597,288 Medicare hospitalizations, 30-day mortality was the same whether an admission fell near or far from a hospitalist's last working day, 10.6 percent against 10.6 percent (Farid and colleagues, JAMA Network Open, 2021;4(3):e213040). Among the sickest patients it ran slightly worse, 27.8 percent against 26.8 percent, an adjusted difference of about one percentage point. So the honest scope is this: no effect on length of stay, small drug-event associations, and a small mortality difference confined to the sickest patients. What changes most is what each new doctor knows.

The other half of the problem is that nobody volunteers the name. In a research letter from one academic medical center, 2,110 of 2,807 hospitalized patients, 75 percent, could not name any physician in charge of their care; of the roughly 700 who did offer a name, about 40 percent were correct (Arora and colleagues, Archives of Internal Medicine, 2009;169(2):199-201). If you do not know who your doctor is this morning, you are in the majority, and asking is the entire fix.

Hospitalist. Usually the attending of record for an adult medical admission. Rounds in the morning, writes the orders, calls the consults, decides on discharge.

Nocturnist. An overnight hospitalist. The doctor who admitted your mother at two in the morning is often not the doctor rounding at nine, and the nine o'clock doctor may be meeting her for the first time. Tell the story again anyway.

Resident or fellow. A physician in training, in a teaching hospital. They frequently spend the most time with you and know the case in the most detail, and their plan is supervised by an attending. Ask who that attending is.

Intensivist. A critical care physician who becomes the attending if the patient moves to the ICU. The hospitalist steps back at transfer in and forward again at transfer out.

NP or PA hospitalist. Nurse practitioners and physician assistants on the hospital medicine team, often carrying their own panel of patients. They can adjust the plan within the team's structure, and they are frequently the person you can actually reach.

Consulting specialist. Cardiology, nephrology, infectious disease. Each gives a recommendation on one problem. They do not run the admission, and their note is a suggestion the attending accepts or does not.

Case manager, discharge planner, social worker. Not physicians, and often the most useful people in the building. They own the machinery of leaving: rehab beds, home oxygen, insurance authorization, transport.

Nurse navigator. In cancer care especially, a nurse whose job is getting you through the system rather than treating you. When one is assigned, they are a fixed point of contact across a stay where everything else rotates.

The nurse is the routing layer, and there is no way around that. Hospitalists carry a pager or a secure phone, they are with other patients, and they answer the nurse. Asking the nurse to reach the doctor is the correct path, not a brush-off.

Ask the nurse two things at the start of every shift: roughly when the hospitalist rounds on this unit, and whether someone will call you when the doctor comes. Rounding windows are informal and they slip, so ask each morning rather than assuming yesterday's answer still holds. If you cannot be at the bedside, give a number that will be answered and ask for a call.

For anything bigger than one question, ask for a family meeting with the attending. That is a recognized request, and the Care Partner Project lists it among the things families should ask for outright. Say what you want covered so the right people get scheduled into the room.

Photograph the whiteboard in the room, sometimes called the care board, at every shift change. It usually carries the nurse's name, the attending's name, and the day's plan, and it is the cheapest record of who was responsible when. Write down every introduction and take a business card when one is offered. The Care Partner Project and VNS Health both build their bedside guidance around this kind of small, unglamorous documentation.

Keep one running log at the bedside: what changed, who said it, what you asked, what you were told. A printable daily log does the job. VNS Health also suggests asking nurses to read new drug orders aloud before giving them, which catches the medication changes that otherwise surface only in the discharge paperwork.

Often, not yet. Notification is not as automatic as families assume, and it is worth raising on the first day rather than the last.

There is a federal right here, and it is worth quoting exactly. Under the Medicare hospital Conditions of Participation, 42 CFR 482.13(b)(4), the patient has the right "to have a family member or representative of his or her choice and his or her own physician notified promptly of his or her admission to the hospital." The same section also gives the patient the right to participate in the development of the plan of care and to make informed decisions regarding that care.

Ask for it by name, at admission or as soon as you think of it: "We are exercising the right to have his own physician notified of this admission. Here is the practice and the number." Being specific about the doctor, the practice, and how to reach them removes the usual reason it does not happen.

If the person is with a concierge or small independent practice, call that office directly. Some will call the hospital team themselves, which is the fastest way to get the two sides talking.

One caveat worth knowing. No Medicare Condition of Participation requires the hospital to tell you the responsible physician's name, which is part of why so few patients can produce it. Having your own doctor notified is a right. Being told who is in charge is something you have to ask for.

Discharge is the handoff that decides whether the week survives. Clinicians call a handoff a sign-out, and the one that travels home with you is carried by a single document, the discharge summary, and the data on whether it actually arrives is unsettling.

A 2007 JAMA systematic review of 55 observational studies found direct communication between hospital physicians and primary care doctors in 3 to 20 percent of cases across the studies reviewed. A discharge summary was available at the first follow-up visit 12 to 34 percent of the time, rising to 51 to 77 percent by four weeks, which is after the visit that matters most.

Summaries often lacked diagnostic results (33 to 63 percent), and 90 to 92 percent omitted any record of what the patient and family had been counseled about. These deficits affected care in about a quarter of follow-up visits (Kripalani and colleagues, PMID 17327525). The same review put pending test results at discharge at 65 percent.

Then there are the results that come back after you are gone. On hospitalist services at two academic hospitals, 41 percent of 2,644 discharged patients had test results return after they left, and 9.4 percent of those results were judged potentially actionable (Roy and colleagues, Annals of Internal Medicine, 2005, PMID 16027454). Somebody has to catch those, and the person with the strongest incentive is standing in the room.

So ask for three specific things before you walk out. A printed copy of the discharge summary in your hand, not a promise that it will be sent. A written list of every test still pending, with the name of the person responsible for calling you with each result. And a named owner for every follow-up item. Our hospital discharge checklist walks the full list, and if the summary never appears you can request the records yourself under your right of access.

Rounds are short and the doctor is carrying a dozen patients. A written list gets more out of four minutes than a good memory gets out of ten. Five questions cover most days.

What is the working diagnosis right now, and what else is still on the list? What are we waiting on today, and when do those results come back? What has to be true before she can go home?

What changed on the medication list since admission, and why? And the one that matters most on a long stay: who is my doctor tomorrow, and will you be back?

Ask that last one every single day. It surfaces the block change before it happens instead of after, which is the difference between preparing a new doctor and being surprised by one. Our guide to questions to ask your doctor covers how to hold the floor long enough to get real answers.

Rotating doctors produce conflicting plans. When yesterday's hospitalist said one thing and today's says another, do not try to arbitrate it yourself. Name the contradiction out loud: "Dr. Reyes told us Tuesday that we were waiting on the culture before deciding about surgery. Can you tell me what changed?" Most of these turn out to be information gaps rather than real disagreements, and saying it plainly closes them.

When that does not work, every hospital has a patient advocate or patient representative, and states have long-term care ombudsman programs for facilities. That office exists to unstick exactly this kind of thing. Asking for them is not a hostile act, and both the Care Partner Project and VNS Health treat it as an ordinary step.

If a patient is visibly deteriorating and the response feels too slow, many hospitals have a Rapid Response Team; ask on day one whether this hospital lets a family member call it, and write the number down. You will probably never use it.

Night-time confusion deserves its own mention. New agitation, hallucinations, or sudden sleepy vagueness in an older patient is often a sign of hospital delirium and should be reported as a possible one, because it gets missed constantly. Report it to whoever is on, including the nocturnist at two in the morning. Do not save it for rounds.

And when the plan itself seems wrong rather than merely unclear, a second opinion is available during a hospital stay, not only in clinic. Our guide on how to get a second opinion covers how to ask for one without damaging the relationship with the team you still need.