Practice Economics

New Referral Distribution


Editor’s note: Second of a three-part series.

As I mentioned last month, there isn’t a proven best method to use when distributing new referrals among your group’s providers. The popular methods fall along a continuum of being focused on daily, or continuous, leveling of patient loads between providers (“load leveling”) at one end; at the other end of the continuum is having a doctor be “on” for all new referrals for a predetermined time period, and accepting that patient volumes might be uneven day to day but tend to even out over long periods.

There might not be any reason to change your group’s approach to patient assignment, but you should always be thinking about how your own methods might be changed or improved. I have shared (“Bigger Isn’t Always Better,” June 2009, p. 46) my concern that some groups invest far too much time in a morning load-leveling and handoff conference. Make sure your group is using only as much time as needed.

Small- to medium-sized groups can eliminate entirely the need for any such “air-traffic control” function if they assign all new referrals to a single doctor for specified periods of time. For example, from 7 a.m. to 3 p.m. today, all new referrals go to Dr. Glass, and from 3 p.m. to 11 p.m., they go to Dr. Cage.

Air-Traffic Controllers

Many large groups (e.g., more than 15 full-time equivalents) that assign patients to providers in sequence, like dealing a deck of cards, have a designated provider who holds the triage pager and serves as “air-traffic controller.” This person typically takes incoming calls about all new referrals, jots down the relevant clinical data, keeps track of which hospitalist is due to take the next patient, pages that person, and repeats the clinical information. As I’ve written before (“How to Hire and Use Clerical Staff,” June 2007, p. 73), many practices have found that during business hours, they can hand this role to a clerical person who simply takes down the name and phone number of the doctor making the referral, then pages that information to the hospitalist due to get the next patient. The hospitalist then calls and speaks directly with the referring doctor.

Small- to medium-sized groups can eliminate entirely the need for any such “air-traffic control” function if they assign all new referrals to a single doctor for specified periods of time. For example, from 7 a.m. to 3 p.m. today, all new referrals go to Dr. Glass, and from 3 p.m. to 11 p.m., they go to Dr. Cage.

Admitter-Rounder Duties

Many—maybe most?—large groups separate daytime admitter and rounder functions so that on any given day, a hospitalist does one but not both. The principal advantages of this approach are reducing the stress on, and possibly increasing the efficiency of, rounding doctors by shielding them from the unpredictable and time-consuming interruption of needing to admit a new patient. And a daytime doctor who only does admissions might be able to start seeing a patient in the ED more quickly than one who is busy making rounds.

Any increased availability of admitters to the ED could be offset by their lack of surge capacity leading to a bottleneck in ED throughput when there are many patients to admit at the same time and a limited number of admitters (often only one). Such a bottleneck would be much less likely if all daytime doctors (i.e., the rounders) were available to see admissions rather than just admitters.

Continuity of care suffers when a group has separate admitters and rounders, because no patients will be seen by the same doctor on the day of admission and the day following. This method requires a handoff from the day of admission to the next day. Such a handoff might be unavoidable for patients admitted during the night, but this doesn’t have to occur when patients are admitted during the daytime.

Who’s Seeing this Patient?

It seems to make sense to wait until each morning to distribute patients. That allows the practice to know just how many new patients there are, and they can be distributed according to complexity and whether a hospitalist has formed a previous relationship with that patient. But it means that no one at the hospital will know which hospitalist is caring for the patient until later in the morning. For example, if the radiologist is over-reading a study done during the night and finds something worthy of a phone call to the hospitalist, no one is sure who should get the call. A patient might develop hypoglycemia shortly after the hospitalist night shift is over, but the nurse doesn’t know which hospitalist to call.

And, perhaps most importantly, if patients aren’t distributed until the start of the day shift, the night hospitalist can’t tell the patient and family which hospitalist to expect the next morning. To test the significance of this issue, I conducted an experiment while working our group’s late-evening admitter shift. I concluded my visit with each admitted patient by explaining, “I am on-call for our group tonight, so I will be off recovering tomorrow. Therefore, I won’t see you again, but one of my partners will take over in the morning. Do you have any questions for me?” Every patient I admitted had the same question. “What is that doctor’s name?”

How does your group answer a patient who asks which hospitalist will be in the next day? If your method is load-leveling in the morning, then the best answer your night admitting doctor can give is probably to say: “I don’t know which of my partners will be in. There are several working tomorrow, and at the start of the day, they will divide up the patients who come in tonight depending on how busy each of them is. But all the doctors in our group are terrific and will take good care of you.”

I’m told the same thing when I get my hair cut: You’ll get whichever “hair artist” is up next. I put up with it at the hair place because it costs less than $15. But I still find it a little irritating. I’m sure all the barbers aren’t equally skilled or diligent, and I want the best one. (Maybe I shouldn’t care since there isn’t much that can be done with my hair.) I’m pretty sure patients feel the same way about which doctor they get. The public is convinced there is a wide variety in the quality of doctors, and they want a good one. If you have to tell them theirs is being assigned by lottery, they won’t be as happy than if you can provide the name and a little information about the doctor they can expect to see the next day.

When the patients I admit late last evening ask who would see them the next day, I’m glad when I can provide a name and a little more information. I say something like, “I won’t see you after tonight, but my partner, Dr. Shawn Lee, will be instead. That means you’re getting an upgrade! Not only is he a really nice guy, he’s voted one of Seattle’s best doctors every year. He’ll do a great job for you.”

To make this communication effective, the night doctor has to know which hospitalist takes over the next morning and has a list indicating which day doctor will get the first, second, third new patient, and so on, admitted during the night. This is possible if patients are assigned by a predetermined algorithm, or if the day doctors have their load-leveling meeting at the end of each day shift, rather than in the morning, to create a list telling the night doctor which day hospitalist he should admit the first and subsequent patents to. That way, the night doctor can write in the admitting orders at 1 a.m. “admit to Dr. X.” This eliminates confusion on the part of other hospital staff who need to know who to call about a patient after the start of the day shift.

Next month I will look at special circumstances, and some pros and cons of having an individual hospitalist take on the care of more patients at the beginning of consecutive day shifts, and exempting them from taking on new patients on the last day or two before rotating off. TH

Dr. Nelson has been a practicing hospitalist since 1988 and is co-founder and past president of SHM. He is a principal in Nelson Flores Hospital Medicine Consultants, a national hospitalist practice management consulting firm ( He is also course co-director and faculty for SHM’s “Best Practices in Managing a Hospital Medicine Program” course. This column represents his views and is not intended to reflect an official position of SHM.

Next Article:

   Comments ()