Continuity Conundrum


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

In the two monthly columns preceding this one, I’ve provided an overview of some ways hospitalist groups distribute new referrals among the providers. This month, I’ll review things that cause some groups to make exceptions to their typical method of distributing patients, and turn from how patients are distributed over 24 hours to thoughts about how they might be assigned over the course of consecutive days worked by a doctor.

Equitable Exceptions

There are a number of reasons groups decide to depart from their typical method of assigning patients. These include:

  • “Bouncebacks”;
  • One hospitalist is at the cap, others aren’t;
  • Consult requested of a specific hospitalist;
  • Hospitalists with unique skills (e.g., ICU expertise); and
  • A patient “fires” the hospitalist.

There isn’t a standard “hospitalist way” of dealing with these issues, and each group will need to work out its own system. The most common of these issues is “bouncebacks.” Every group should try to have patients readmitted within three or four days of discharge go back to the discharging hospitalist. However, this proves difficult in many cases for several reasons, most commonly because the original discharging doctor might not be working when the patient returns.

The Alpha & Omega

Nearly every hospitalist practice makes some effort to maximize continuity between a single hospitalist and patient over the course of a hospital stay. But the effect of the method of patient assignment on continuity often is overlooked.

A reasonable way to think about or measure continuity is to estimate the portion of patients seen by the group that see the same hospitalist for each daytime visit over the course of their stay. (Assume that in most HM groups the same hospitalist can’t make both day and night visits over the course of the hospital stay. So, just for simplicity, I’ve intentionally left night visits, including an initial admission visit at night, out of the continuity calculation.) Plug the numbers for your practice into the formula (see Figure 1, right) and see what you get.

If a hospitalist always works seven consecutive day shifts (e.g., a seven-on/seven-off schedule) and the hospitalist’s patients have an average LOS of 4.2 days, then 54% of patients will see the same hospitalist for all daytime visits, and 46% will experience at least one handoff. (To keep things simple, I’m ignoring the effect on continuity of patients being admitted by an “admitter” or nocturnist who doesn’t see the patient subsequently.)

Changing the number of consecutive day shifts a hospitalist works has the most significant impact on continuity, but just how many consecutive days can one work routinely before fatigue and burnout—not too mention increased errors and decreased patient satisfaction—become a problem? (Many hospitalists make the mistake of trying to stuff what might be a reasonable annual workload into the smallest number of shifts possible with the goal of maximizing the number of days off. That means each worked day will be very busy, making it really hard to work many consecutive days. But you always have the option of titrating out that same annual workload over more days so that each day is less busy and it becomes easier to work more consecutive days.)

An often-overlooked way to improve continuity without having to work more consecutive day shifts is to have a hospitalist who is early in their series of worked days take on more new admissions and consults, and perhaps exempt that doctor from taking on new referrals for the last day or two he or she is on service. Eric Howell, MD, FHM, an SHM board member, calls this method “slam and dwindle.” This has been the approach I’ve experienced my whole career, and it is hard for me to imagine doing it any other way.

Here’s how it might work: Let’s say Dr. Petty always works seven consecutive day shifts, and on the first day he picks up a list of patients remaining from the doctor he’s replacing. To keep things simple, let’s assume he’s not in a large group, and during his first day of seven days on service he accepts and “keeps” all new referrals to the practice. On each successive day, he might assume the care of some new patients, but none on days six and seven. This means he takes on a disproportionately large number of new referrals at the beginning of his consecutive worked days, or “front-loads” new referrals. And because many of these patients will discharge before the end of his seven days and he takes on no new patients on days six and seven, his census will drop a lot before he rotates off, which in turn means there will be few patients who will have to get to know a new doctor on the first day Dr. Petty starts his seven-off schedule.

This system of patient distribution means continuity improved without requiring Dr. Petty to work more consecutive day shifts. Even though he works seven consecutive days and his average (or median) LOS is 4.2, as in the example above, his continuity will be much better than 54%. In fact, as many as 70% to 80% of Dr. Petty’s patients will see him for every daytime visit during their stay.

Figure 1:  HM Physician-Patient Continuity Calculation click for large version

click for large version

Other benefits of assigning more patients early and none late in a series of worked days are that on his last day of service, he will have more time to “tee up” patients for the next doctor, including preparing for patients anticipated to discharge the next day (e.g., dictate discharge summary, complete paperwork, etc.), and might be able to wrap up a little earlier that day. And when rotating back on service, he will pick up a small list of patients left by Dr. Tench, maybe fewer than eight, rather than the group’s average daily load of 15 patients per doctor, so he will have the capacity to admit a lot of patients that day.

I think there are three main reasons this isn’t a more common approach:

  1. Many HM groups just haven’t considered it.
  2. HM groups might have a schedule that has all doctors rotate off/on the same days each week. For example, all doctors rotate off on Tuesdays and are replaced by new doctors on Wednesday. That makes it impossible to exempt a doctor from taking on new referrals on the last day of service because all of the group’s doctors have their last day on Tuesday. These groups could stagger the day each doctor rotates off—one on Monday, one on Tuesday, and so on.
  3. Every doctor is so busy each day that it wouldn’t be feasible to exempt any individual doctor from taking on new patients, even if they are off the next day.

Despite the difficulties implementing a system of front-loading new referrals, I think most hospitalists would find that they like it. Because it reduces handoffs, it reduces, at least modestly, the group’s overall workload and probably benefits the group’s quality and patient satisfaction. 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 (www.nelsonflores.com). 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 ()