If the American principle of voting can best be described as "one man, one vote" (leaving aside that it ain't just men), I'd say there's a similar, though less well defined and understood, principle that applies to hospital medicine: one patient, one hour. That is, to do good medicine in a hospital-based setting, it takes a general internist, on average, about one hour to take care of a patient.
This may come as a surprise to anyone who has been hospitalized and seen a doc for about five minutes each day, but that hour applies to all the tasks required in caring for that patient. A doc's gotta review the vital signs, the med sheets, the nursing notes, the social worker and physical therapist's assessments, look up the daily labs, check the consultants' notes, and then come before the patient, examine them, answer their questions, talk to family members, set up a plan for them, and then write a note in the chart. Of course, individual styles may vary: I tend to enjoy spending time talking with patients and their families, but that slows me down, and I'm not highly efficient to begin with. Moreover, not every patient every day requires an hour, as the patient with uncomplicated cellulitis needs IV antibiotics, not a huge re-evaluation on a daily basis. But based on my experience and discussions with lots of hospital docs, the one-hour-one-patient "rule" is a pretty good predictor of the quality of medicine. Cut that amount substantially, and sooner or later a doc will make a mistake, whether it's an oversight in drug-drug interactions, a missed lab value, or a misunderstanding with a family about prognosis because a meeting between doc and family didn't take place.
So this past weekend, when I took to covering the "floor" patients as part of a moonlighting gig at a local community hospital, I was given a list of 14 patients at 7 am. I wasn't surprised very much when I left the hospital almost exactly 14 hours later. It took about a half-hour to divvy up the patients initially, and I took about ten minutes to wolf down a lunch, but the average time I spent was just under an hour for each patient, and by the time I headed home I felt I had a decent handle on what was going on with them.
Two particulars about this experience bear mention, however.
First is that a "census" of 14 is, based on what the full-time floor docs tell me, on the lower end of the spectrum, with typical numbers in the 18-20 range, sometimes higher. The reason for this is money, money, money. The docs at this particular hospital work for a for-profit company (as do I when I moonlight for them), a company whose stock is publicly traded and for whom profits are by definition their lifeblood. While I'm not making any comments on how the company is run and how they try to maximize profits and simultaneously provide high-quality care, I can give you an idea about the numbers.
In the northeast, where I work and where salaries are a little higher than in other parts of the country, a typical hospitalist (i.e. a hospital-based doc, though more on "doc" anon) costs a company around $250K per year when you add benefits, malpractice insurance, and administrative costs to the salary, which at least where I work is a touch under $200K. When you factor in reimbursements from patient care, all it takes is an accountant to figure out how many staff are needed to see so many patients on average and estimate the census size required for the company to make a profit at a given hospital. Since the company's reason for existence is profits, they are always trying to push the envelope with patient census, and shooting for a census of 20 (or higher!) keeps the company in the black. This article from KevinMD features the musings of a doc who once routinely had to care for forty patients a day--and he casually notes that he currently has a daily census of "maybe around 20".
Now, a non-profit hospital running its own hospitalist program also has a bottom line and still has to think about having its revenue stream cover its costs. But there's one difference, and it's a huge one: the need for profit. I haven't done a lit search to see if there are any articles looking at this, but I'd be willing to bet more than $5 that if you surveyed the average patient census of hospitalists working for non-profit hospitals versus for-profit companies, you would find a statistically significant larger census in the latter group. By how much, I don't know. But what I do know is that if you move an average census above 12-14, and if you ever move it above 18, you simply can't be a decent doctor. It can't be done.
(As this piece notes, two studies have shown that as census numbers increase, face-to-face patient time does not decrease, but the critical behind-the-scenes work of "documentation, writing order, and communicating with nurses and primary care physicians" does. For those interested in reading the primary academic literature on staffing requirements, you can see this article from 1999--it assumed an average census of about ten in making its calculations.)
The second item of note is that, while I was slogging away seeing my patients, I saw a brief verbal altercation between two of the younger hospitalists, neither of whom I knew. I ran this past one of the hospitalists with whom I've worked for many years and he shrugged. "Oh, X is mad because Y just left yesterday at noon and turned her pager off," he said. "Noon?!" I responded. To leave at noon, this doc saw fifteen patients in around five hours. That's twenty minutes per patient--and she was meeting all of these patients for the first time. That means she looked through the chart, reviewed the medications, saw the patient, wrote a note, and (a theoretical conjecture--I doubt it really happened) communicated with family members...all that in less than the length of a sitcom. Which is an appropriate comparison, since all one can do when confronted with such negligence is laugh. My heart goes out to this doc's patients and families. I hope she remains the exception in our profession.