The Hospitalist as Bed Czar: Indispensability, But At What Cost?

In last week’s Annals of Internal Medicine, Eric Howell and colleagues describe an innovative experiment in which the hospitalists at Johns Hopkins Bayview became the institution’s bed czars. It worked.

So should my program and yours take this one on?

Hopkins Bayview is a 335-bed teaching hospital affiliated with Johns Hopkins. The Chief of Medicine, David Hellmann, is an old friend and a gem, a graceful and eloquent man who is constantly looking for improvement opportunities. Under his guidance, several years ago the hospitalist group, led by Howell, agreed to become the medical center’s “Active Bed Managers” for medical patients. The ED sees 54,000 patients a year, and admits about one-quarter of them, three-quarters of these to Medicine.

If you looked up “Thankless Task” in the dictionary, you might see “Active Bed Manager.” So how did they do this?  And why?

One hospitalist at a time serves on the ABM service, in 12-hour shifts. During this shift, the hospitalist has no other responsibilities, freeing him or her up to act as a full-time air traffic controller for all medical patients. This involves keeping up to speed on the bed status of all medical, step-down, and intensive care units, “prediversion” round in the ICU, evaluating (by phone or in person) all new admissions, expediting ED-to-floor transfers, and sundry other tasks. After a few years of doing this during the days, in 2006 they began providing ABM around the clock, 365 days a year. When all hell breaks loose, the ABM hospitalist notifies the “Bed Manager” – Eric or another senior hospitalist leader – who has the authority to activate resources or knock heads to free up beds or expedite transfers.

The results were truly impressive. ED length of stay for admitted patients fell by 98 minutes (458 minutes in control period to 360 minutes after the intervention), a tremendous improvement, particularly when multiplied by 10,000 patients a year. The time that the ED was on full divert – which costs the hospital both money and good will (and probably costs a few lives as well, as patients are shunted to less appropriate or more far flung hospitals) – went down by a staggering 87% (from 31% to 4% of the time)!

I spoke to Howell last week to find out more, since I was reasonably sure that I – and my fellow hospitalist leaders around the country – would receive “why don’t you do this?” calls from our CMOs within minutes of the publication of these results. “I watched for years as the hospital tried to improve throughput and stay off ambulance diversion,” he told me. “Nothing worked, but we knew that we could help fix this. After a while, we decided that it was worth trying.”

A short fiscal primer for those of you who don’t traffic in DRGs and bed-days-per-thousand: Hospitals that run full spend staggering amounts of money on efforts to improve throughput. They hire consultants (which never works, but their PowerPoint presentations are pretty to look at), they tweak admission criteria, they shop eBay to buy second-hand electronic tote boards discarded by the Hyatt. These interventions rarely make a significant dent, because to make a real impact you need someone to make scores of tough, contentious decisions in real-time, preferably someone with the negotiation skills of Richard Holbrooke.

Most hospital ultimately throw up their hands and solve the problem of throughput by – you guessed it – building more beds, at a cost these days of 1-2 million dollars per bed, depending on whether you have to meet earthquake standards (the cost is even higher for ICU beds). But hospitals can’t afford to leave their bed shortage problem unsolved – not just because they need to dis-impact the ED, but more importantly (for the bottom line) because they need to free up beds upstairs for elective surgeries. Canceling such surgeries because of bed shortages is intensely expensive and demoralizing to the C-suite folks. Plus it makes the surgeons very unhappy, a bad job retention strategy for most COOs.

I wanted to know how the ABM intervention had affected Howell’s hospitalists’ relationships – with the ED, the nurses, and the residents. He told me this:

“All relationships got better. The ED loves us – the ED chief sits in medical board meetings and asks for more hospitalists. The ICUs like us, maybe love us, because we got rid of ambulance diversion. The nursing supervisor loves us, because we help them enforce stuff, or can override policy if needed (when common sense dictates). The residents? First they were reluctant, now they love it.  But it does put the hospitalist in the middle of resolving conflicts between two house officers, house officers and the ED, sometimes house officers and the nurses…”

This intervention can’t be done on the cheap: having dedicated hospitalists on this service 24-7-365 (not performing billable activities) would likely require about 4-6 FTEs-worth of hospitalists, or close to a million dollars a year (Eric and I didn’t get into the precise numbers at Hopkins Bayview, but the math is pretty straightforward). And, in order to motivate Eric’s group to do this, the hospital anted up some additional salary support for both rank-and-file hospitalists (who saw an increase in academic “protected” time) and for leadership positions. The latter was particularly important, since the junior hospitalists were instructed to bump issues to a senior hospitalist leader (the “Bed Manager”) when the disputes got too difficult or new resources were required. At first, this was just Eric and one colleague who were always on call for this purpose; by the end, four leaders were sharing this difficult but crucial role. 

Finally, I asked Eric – given what must have been Too-Numerous-To-Count political challenges – whether he was glad he did it. I also asked how he’d rank this intervention against alternative uses of the same dollars (such as surgical co-management or proceduralist services), most of which would cause less loss of hair and gastric mucosa. He responded this way:



“Yes, I am glad I did it. It put my group on the map at Johns Hopkins. Before hospitalists were largely considered “non-essential” by other faculty. Now they see us as equals, because we fixed something that they could not… for years. Also the hospital LOVES us; the president introduces me as the man who runs the entire hospital (not true but flattering)…”

I’m going to give this intervention a very high degree of difficulty – in the Hospitalist Olympics, I’d rank it as a reverse one-and-a-half somersault with three-and-a-half twists, with a good chance for a Belly Flop if it isn’t skillfully executed. In other words, Active Bed Management is not for the faint of heart, nor something to take on if you have staffing challenges elsewhere.

In part because of that, although you might get a warm and fuzzy feeling about improving throughput and decreasing diversion for your hospital, there is no way a group should take on this role simply to have its costs recouped. If you do ABM and see Eric’s results, you have created several millions of dollars of value for the average hospital (and many hundreds of thousands for the surgeons), and some of this needs to be allocated to the hospitalist program itself, in the form of more protected time, higher salaries, or other items on its wish list.

But my premise from the moment this field began was that hospitalists – because of their near-universal dependence on outside (usually medical center) support – had to constantly be looking for opportunities to add value. Particularly in tough economic times, the opposite of being Indispensable is being Dispensable. That’s not a good thing to be right now.

I haven’t told my group this (or perhaps I just did) but, assuming we have sufficient staffing, I think we should begin looking at ABM in the not-so-distant future, probably starting with a daytime service to see whether it is do-able before taking on the much more challenging task of nighttime coverage.

The American Hospital Association just released its 2008 estimates, and the number of hospitalists is now pegged at 27,000, which makes the field larger than cardiology or emergency medicine – the largest non-primary care field in Internal Medicine, and the fastest growing field in the history of medicine. This is staggering (next time, please remind me to trademark a term when I coin it), and owes to the fact that when most docs are running in the other direction, hospitalists step up to the plate and fix problems that need fixin’.

So a shout out to Eric Howell and the Hopkins Bayview crew for adding one more arrow to the hospitalist Quiver of Indispensability.

6 Responses to “The Hospitalist as Bed Czar: Indispensability, But At What Cost?”

  1. DZA December 12, 2008 at 10:58 am #

    somewhere a can is missing it’s worms…

  2. Annie December 12, 2008 at 2:10 pm #

    I was an actual – erg – throughput director – for a not-too-dissimilar institution (a little larger, but essentially the same services). I elected not to use my nursing credentials and to approach the position (newly created) from a generalist’s point of view. I think either a nurse or physician can do this, but that the keys to why it hadn’t previously been effective are lurking in the responses from Eric Howell:

    All relationships got better. The ED loves us – the ED chief sits in medical board meetings and asks for more hospitalists. The ICUs like us, maybe love us, because we got rid of ambulance diversion. The nursing supervisor loves us, because we help them enforce stuff, or can override policy if needed (when common sense dictates). The residents? First they were reluctant, now they love it. But it does put the hospitalist in the middle of resolving conflicts between two house officers, house officers and the ED, sometimes house officers and the nurses…”

    The bolded statements are where I found significant successes, as well. And they are where the fundamental root problems lie.

    It’s the hospitalists’ collegial support and independent physicians’ influence which can override institutional employment dictates. Nurses by and large, do not enjoy this protection. Indeed, the nursing leaders are themselves employees and express fealty to the senior administration over their nurse colleagues.

    Nursing middle managers and directors most often practice in a vertiable no-man’s land where they are not supported by clinical (bedside) nurses because of their direct/control/hire/fire aspects of their positions and they enjoy no senior administrative support, being at-will employees. They are not covered by NLRB so-called protections. THey are often competing for scarce institutional resources – budgets, nurses, ancillary staff, desirable patient populations, and thus, they even compete against one another.

    This divide and conquer strategy cannibalizes nurses. There is risk that it could occur to hospitalists in these positions, as well, so consider this a heads up.

    I think that the way to address patient placement issues is to have a radical and fundamental change in how nurses and physicians practice. My recommendation is for both to form collaborative self-governed professional practice groups in which they s/elect their own practice and administrative leaders who are accountable to them, devise their own clinical/academic/research career advancement mechanisms, and then contract directly with groups of patients, businesses and patient care institutions to provide professional services.

    In so doing, they would be able to recapture usurped non-professional third party control over professional autonomy and authority, and they would have a much higher stake and clout in negotiating safe work places and acceptable patient case loads.

    In my own case, I straddled hot political potatoes in which patients (think highly reimbursed) got on service vs off service beds, which electives were brought in while patients boarded in the ED, and which physicians’ whims were always accommodated regardless of the madness and mayhem in-house. I was loved unless I was viciously hated. There wasn’t much middle ground.

    I had daily pleas from physician directors of ICUs and medical units (think low reimbursement) to find slack in the system somewhere. That usually stopped where nurse staffing stopped. Again, it’s professional nursing which is the deciding capacity factor, even more so than physical beds, in the majority of cases (at least at my situation and in the literature I read).

    But the key there is to build a professional nursing staff which retains professional autonomy and authority, which has robust nursing education and research, and which protects nurses from employment separation or retaliation when they make unpopular with administration decisions to advance safe patient case loads and patient safety. That translates into nursing managers who are supported when they speak “truth to power”. And that is woefully missing entirely in the vast majority of hospitals and medical centers.

    I know first hand and will never work again in my field for just this reason. Administrators are vicious in very real and tangible ways to nurses who do this. My experience is the norm – it’s just that no one is studying the cohort of nurses who have been ostracized, segregated and denied employment for this. They range from a past president of the National League for Nursing, to many nursing faculty, to nursing managers and administrators and the uncounted clinical nurses who decide that risking patient safety on a routine basis with no ability to mitigate it is intolerable.

    As more physicians practice as employees, expect them to rise up in horror when they, too, are exposed to this disconnect.

    Finally, I found great improvements by bringing all stakeholders to the table and giving them power to act. This included putting a housekeeping supervisor in a workspace next to a registration worker. As requests for bed cleaning arrived, the reg and housekeeping people worked jointly in real time to assign workers by patient priority and by geography. Prior to that, EVERY request was placed as a STAT clean, and therefore, no bed was given priority. The reduced housekeeping staff simply started at one end and worked their way through the institution. The new process allowed for deployment which didn’t make workers back track and make extra trips, and it addressed the real needs of having beds ready when the patients needed them. Everyone loved it, and the turnaround for clean beds on evenings and nights dropped by over 30 minutes the first week it rolled out.

    As your colleague said, the ED loved me, the inpatient nursing unit staff loved me, and the physician unit directors who previously had little voice loved me. The housekeeping staff loved me. I walked around and listened to staff, physician and patient/families every day – on all shifts, and I showed up on holidays, and weekends, too.

    When there was administrative support, it worked. When there wasn’t, it failed spectacularly.

    Thanks so much for writing this. if anyone wants to chat further, I wrote about patient flow strategies on my blog (link to ED patient management meta post at my name)

  3. menoalittle December 13, 2008 at 2:12 am #

    Bob,

    Thank you for the summary. Leave it to the Hopkins to innovate such complexity. Annie shared important insight of her experience, important for all. The penetrance of HIT at the hospitals is a curiosity and whether the communication and improved throughput Bayview was facilitated by electronic communications, electronic ordering, and digital records (or was it traditional face to face)? On that related topic and associated patient safety, the JC issued a Sentinel Event Alert yesterday:

    http://www.jointcommission.org/SentinelEvents/SentinelEventAlert/sea_42.htm

    http://hcrenewal.blogspot.com/

    Best regards,

    Menoalittle

  4. JHBMC Hospitalist December 13, 2008 at 4:14 pm #

    Thank you for the thoughtful comments by everyone. At JH Bayview, there were no HIT interventions to help with ABM (active bed management). It was all face-to-face, or by telephone. We are developing an “e-triage” program, that will further enhance communication by making the ABM process web based (and can be seen by everyone) process, rather than the current paper process.

    Many thanks
    Eric Howell

  5. BHarte December 16, 2008 at 2:28 am #

    At Cleveland Clinic we do something a bit analogous called the “quarterback” who is involved with bed placement as well as hospital transfers, in-house ICU transfers, and ED admissions and who among other duties, is authorized to place patients on subspecialty services (CC still has subspecialty services). While I would not ever say that this is an enjoyable way to spend a day (or night, we do it 24 hours a day) it has put us in an unusual and favorable postition within the hospital because it has put us in a position where we are not just recognized for clinical excellence — which, let’s face it, is tough to get recognition for in comparison to cardiologists, urologists, etc — but also for our systems-based thinking and recognition of the “big picture.” We are also willing to make — and stand by — tough decisions. And, what’s more, it is a boost to us generalists to be able to drive clinicial and bed decisions when we feel patient care is best served, instead of having them dictated to us by subspecialists.

    Finally, it has put us in the spotlight further becuase some of our docs have taken the ball and run with it in terms of further systems improvements, working with the existing clinical and IT systems to ensure safe handoffs and patient tracking.

  6. jodigirl December 27, 2008 at 4:07 pm #

    I agree with the kudos to JH Bayview. Good research and article about the effect of hospitalists on ED movement. And I think we all get the point that the hospitalist program model can increase a hospital’s bottom line.

    But, what was missing from the JH Bayview study is the same thing that I find missing from all discussions of the hospitalist program model – patient safety and continuity of care.

    Appreciating the fact that hospitalists are the rain-makers of the hospital system does not answer the question of whether hospital resource managers (which are an important part of the hospitalist program model) *really* coordinate appropriate post dc care? Do “turf wars” between the admitting (clinic) physician and the hospitalist? Or, do the clinic docs really want to hand-off care and all decision making on a hospitalist who effectively just met the patient and lacks the wealth of pre admission knowledge of the patient’s medical hx?

Leave a Reply