All Content

John Nelson: Fixing Complaints Between Primary-Care Physicians, Hospitalists Not Always Easy


John Nelson, MD, MHM
The hospitalist should have the final say regarding whether a patient is appropriate for direct admission, or whether it is best to stop in the ED first.

In the course of my work with hospitalist practices around the country, I end up speaking with a lot of primary-care physicians (PCPs) who refer patients to hospitalists. I nearly always hear the same three frustrations or complaints from them:

  1. “I’m not reliably notified when my patient is admitted or discharged.”
  2. “The hospitalists too often make unnecessary or unhelpful changes in patients’ chronic medicines, because they either never get an accurate home medicine list to begin with, or too liberally adjust chronic therapy that should be left to me.”
  3. “I wish the hospitalists were more open to directly admitting some patients from my office, to save the patient the stress and expenses of an unnecessary stop in the ED.”

I’ve listed them in ascending order of what I think is difficulty to fix. The first of these can be difficult but not impossible to fix, while the last one—direct admissions—is really tricky to “fix” to the satisfaction of both hospitalists and most PCPs.

Direct Admissions and HM Reluctance

When explaining why they resist direct admissions, most hospitalists raise concerns that I too share. They typically begin with an anecdote, often from years ago, of a patient the PCP described as stable, but was in extremis when arriving to the floor bed and required emergent transfer to the ICU. In fact, I suspect this has happened at least once or twice to nearly every hospitalist. Much to the frustration of PCPs, hospital leaders, and some patients, this concern has led a number of HM groups to adopt a policy of never accepting direct admissions. They insist that all patients are seen first in the ED, which typically means that the ED physician, rather than the hospitalist, is the first doctor the patient encounters at the hospital.

Other reasons cited for reluctance or refusal to accept direct admissions include the longer time required to get test results like blood work or chest X-rays when ordered from the floor versus the ED. And because the patient’s precise time of arrival can’t be known, it is tricky for some groups to determine in advance which hospitalist will be seeing the patient, resulting in a complicated handoff.

Some PCPs, especially those who have practiced for decades, might be remembering the rationale and process for admitting patients years ago and inappropriately request direct admission for a patient who might not even need the hospital. But while it seems clear this happens occasionally, hospitalists could have a bias, leading them to feel like it is a much more common problem than it really is.

All of these are legitimate concerns, though in most settings I don’t think they justify setting a firm rule of “no direct admissions.”

Dearth of Meaningful Data to Guide Policy

There are seemingly an endless number of studies about things like the effects of resident work-hours and the value of handoff communication, so the literature must be full of studies about direct admissions. Surely some of the risks are offset by improvements in safety and fewer handoffs (by eliminating the ED doctor). But sadly, there aren’t any studies to go on. I couldn’t find a single one. (If you know of one or more studies that directly examine direct admissions from PCP offices, please let me know.)

The Agency for Healthcare Research and Quality (AHRQ) has a 2008 case study titled “Is It Safe to Be Direct?” ( that describes and comments on a direct-admit case that didn’t go well, but it is an opinion piece without empiric data.1

The absence of research studies doesn’t stop a lot of people, including me, from expressing their opinions. Numerous articles and opinion pieces are available on the Internet. They generally summarize that despite having the same goal of safe and efficient patient care, PCPs and hospitalists often see direct admissions a little differently.

An Internet search of ”direct admission + hospitalist” turns up the practice website for a hospitalist group. I found several such sites that do accept direct admissions. Presumably, those hospitalist groups that refuse to accept direct admissions don’t advertise that on their website so don’t turn up in a search. This one is typical:

Hospitalists are also available to facilitate direct admissions to Beth Israel Deaconess Hospital-Needham, so that patients can avoid a trip through the emergency department. Please note that it is Beth Israel Deaconess Hospital-Needham policy that the patient needs to have been seen by the primary-care provider or specialist physician within the previous 24 hours to qualify for direct admission status.

Mass General Hospital for Children in Boston has posted a very detailed approach to direct admissions2 allowing them only from some PCP groups (presumably those in their system), and only when the patient has been seen in the office on the day of admission. And the hospitalist program at Johns Hopkins Hospital in Baltimore advertises its “VIP Direct-Admitting Service.”3


I’ve come to the following conclusions that I think most groups could follow, though I realize thoughtful people can see this differently.

  • Most hospitalist groups should not have a policy of refusing all direct admissions. They should thoughtfully listen every time a doctor calls asking to refer a patient directly from an office setting. And, at least some of the time, they should say yes.
  • You should more liberally accept direct admissions from PCPs you work with regularly. The better you know the PCP (i.e. have cared for many of that doctor’s patients), the more you can judge the risk the patient will arrive in a condition other than described.
  • Requiring that the patient be in the office at the time of the decision to accept the direct admission, or within the preceding 12 or 24 hours, is a good idea.
  • Work with your hospital to improve the speed of testing like blood work and X-rays done “on the floor” on new admissions so they’re resulted as quickly as in the ED. Consider notifying in advance the relevant department that you’ll likely be ordering a stat study as soon as the patient arrives. This is sort of like calling a restaurant to get in line for a table before you arrive.
  • The hospitalist should have the final say regarding whether a patient is appropriate for direct admission, or whether it is best to stop in the ED first. That is the case for all the practices I mentioned above. But don’t let this insulate you from the very real frustration suffered by PCPs and patients, should you unfairly refuse to allow it.

I don’t have any idea what might be an appropriate portion of direct admissions for a typical hospitalist practice; it’s probably no more than 1% or 2%. But I don’t think it should be zero.

Dr. Nelson has been a practicing hospitalist since 1988. He is co-founder and past president of SHM, and principal in Nelson Flores Hospital Medicine Consultants. He is course co-director for SHM’s “Best Practices in Managing a Hospital Medicine Program” course. Write to him at


  1. Kulkarni N, Williams M. Is it safe to be direct? Agency for Healthcare Research and Quality website. Available at: Accessed Feb. 2, 2013.
  2. MassGeneral Hospital for Children. Direct admit policy. MassGeneral Hospital for Children website. Available at: Accessed Feb. 2, 2013.
  3. Johns Hopkins Medicine. Hospitalists Introduce VIP Direct-Admitting Service. Johns Hopkins Medicine website. Available at: Accessed Feb. 2, 2013.

Next Article:

   Comments ()