Friday, October 15, 2010

Hospitalists as extensivists

I had heard something about this, but couldn't find it. A colleague here finally tracked it down. The story is about Caremore, a California based organization. Hospitalists generally are internal medicine doctors based in the hospital; but here they care for frail elderly members at high risk of hospital admission or readmission in skilled nursing facilities and in outpatient settings both before and after a hospital stay. Here's an article on the AHRQ Innovations Exchange website.

An excerpt:

A Medicare Advantage plan expanded the role of its employed hospitalists, using them to continue following and caring for recently discharged members until their condition stabilizes, as well as other members at high risk of a hospital admission. Known as "extensivists" and supported by sophisticated information technology (IT) systems, these physicians generally split their time between the hospital, where they round on a small group of members each day, and an outpatient clinic, where they see recently discharged members and other members at high risk of an admission. Once or twice a week, these physicians also see members in SNFs.

The results:

The program reduced readmission rates and has led to low LOS (lengths of stay) and to below-average inpatient utilization in a high-acuity population.

Is this worth considering more broadly? What are the conditions for success? I welcome your thoughts.


Anonymous said...

These folks by definition may not be hospitalists. They spend most days, at least 50% of their time, in outpatient setting, and another 1-2 days/week with half day in a SNF. Perhaps "extensivist" is the right word.

The issue of employment, they work for MCO, and not hospital or physician group interesting--therefore not exportable to all venues.

High risk clinic seems novel, ie, PCP or NP can refer "high risk" patients to this setting for eval. However, if one is to embrace PCMH model, one could see this as an alternate take ("chronic care extenders"), or an abandonment of that ideal. Essentially, rather than managing the high risk condition within the sphere of the PCP or NP, the patient is shuttled elsewhere.

This is likely one approach that works in one setting, but again, reproducibility not a sure thing if tried elsewhere: many payer and setting intangibles.


Unknown said...

Just one note about hospitalists; we are not limited to internists. There are also pediatric hospitalists, and "adult" hospitalists who are trained in family medicine.
The key feature is expertise and commitment to the inpatient settings with attention to a systems approach to quality improvement. Your example of hospitalists getting involved with care transitions by providing transitional care outside of the hospital is an example of looking at a problem creatively, examining root causes, and addressing solutions from a systems standpoint.

Anonymous said...

I agree with some of Brad's comments regarding whether this concept supports or abrogates the idea of the medical home - except that if the hospitalist has seen the patient inhouse, s/he would be more familiar with the current high-risk medical problems than the PCP.

I think this concept would fit very well into the framework of an ACO, where the ACO has an interest in reducing readmissions and coordinating inpatient/outpatientcare. The global payment presumably would encompass such activity.

My question is, does an M.D. provide that much extra value to justify the cost of these home visits as opposed to a NP - perhaps for just the first one or two visits to establish the parameters of care.

nonlocal MD

jonmcrawford said...

This is basically what all Managed Care tries to do via case management resources, follow the patient regardless of setting, a luxury (?) that hospitalists would not normally have, however having someone on the ground in physical interaction with the patient is much better. We're considering doing something similar for the Medicaid population we serve, in partnership with some of the major hospital systems in our area. We anticipate big gains there also, although a large part of that may also be the population (in contrast to the study you reference with Medicare patients), where just getting ahold of some of these members is a major challenge.

Barry Carol said...

I think this approach makes a lot of sense for elderly patients with multiple co-morbidities, especially for those who are also in a long term care facility. Nursing homes are often quick to send patients to the hospital as the path of least resistance. To the extent that this use of “extensivists” can mitigate that, it’s a good thing.

That all said, insurers, as payers and assumers of actuarial risk, have plenty of economic incentive to keep patients healthier, nip problems in the bud, and generally reduce utilization of healthcare services, especially hospital based care. For hospitals like BIDMC, you would likely lose revenue under a fee for service payment model. As part of an ACO or a global payment model or at least if you assumed some significant amount of actuarial risk, this approach would make economic sense for hospitals as well, in my opinion. Under fee for service, by contrast, you get paid more for doing more and for treating more patients.

Anonymous said...

As a patient that is well cared for by my PCP and very fortunate to be admitted to her when hospitalized, having a hospitalist involved in my care is not the answer. The best way to provide continuity of care, decrease re-admission rate and decrease adverse medical events is to return to a model where all primary care providers admit to themselves. Why and how has this model changed?

I do believe hospitalists are invaluable and work well for patients that do not have a primary care provider or have providers that are in the community setting and or for patients that are admitted to an institution their PCP might not be part of. This isn’t to say that hospitalists aren’t capable of caring for all but time and time again, continuity of care with the same primary care provider has proven best outcomes.

Perhaps a study can be done comparing several different outcomes of providers that admit to themselves versus those who use hospitalists. Even a retrospective study could be performed.

Keith said...

Transitions from hospital to nursing home are often handled very poorly, with the hospitalist not alway communicating information to the PCP (and visa-versa). Very often, the first knowledge the PCP gets of his patient being at another facility is the call from the nursing home to verify the orders, often with little knowledge of what transpired in the hospital. To have the hospitalist engage in this transfer process actuallly makes more sense. This is primarily why some of us prefer not to use hospitalists since you lose critical knowledge of the pertinent issues and data that need follow up unless you are engaged in the patients hospital care throughout the hospitilization. Pouring over hospital records can become tedious and time consuming, and is not currently a compensable duty, which is why it is often not done adequately.

This example again addresses the problems asociated with transfers of care. My strong opinion is that the more transfers you have to make, the more opportunities for errors. The best time to make these transitions will be when things are stable; not at the end of a critical illness requiring hospital care. Thus it makes good sense that if the hospitalist is going to care for the patient, he carry it through till near full recovery. As other have pointed out, this is problematic given the current employment of hospitalists often by hospitals, and there ability to work in other settings.

I found the best solution is for hospitals to maintain acute care SNFs on their campus where the patient could be easily followed by the PCP or hospitalist after discharge from acute care. There is easier availibility for specialty consult, diagnostic testing, and most importantly, continuity of care with the medical team. The hospital where I work had such a unit which worked wonderfully, but it was closed due to changes in SNF compensation some years ago. I know many patients who benefitted from having this transition to such an on campus facility for those that are in a more frail condition at the end of their acute care episode.

76 Degrees in San Diego said...

Fascinating! So, the family physician, internist, or pediatrician that sees her/his inpatients is the following:
"primary care physician, hospitalist, extensivist, and "SNF-ist". Do I have that right?

Unknown said...

Third generation hospital case management programs are distinguished from their more traditional UR/DCP models by partnering case managers with physicians to manage progression-of-care across the continuum. In integrated delivery systems, this may operatinalize as a single consistent case mgr resource providing f/u in the community post acute setting. In community hospitals, the introduction of 'transition' case managers are serving the same purpose albeit with seamless hand-offs to assure continuity. So far, studies are evidencing clinical and economic benefits.
This is just one of the reasons that the role of the hospital case mgmr is transitioning so quickly in hospitals across the country.

Anonymous said...

The extensivist that covers acute care and rehabilitation together will still have to avoid sloppy communication when the transition to longer-term follow up occurs. And then the receiving physician/provider will be at a bigger disadvantage, having been left out of more of the process. Because of this, and because hospitalists are no better at being in more than one place at one time than office-based docs are, the extensivist model is often a lot less helpful in practice than it first seems on paper.
Paul, many of your posts discuss root-cause analysis and team assessments, importantly including patient and community points of view in the process. I agree with comments above that midlevel providers or case managers may be a more important means for improving transitions than spreading the doctors around. And how many patient voices would chime in, "I can't read that handwritten prescription," or "All I get is voicemail when I try to call for an appointment"?

Nate said...

This is very interesting- it is an innovative use of hospitalists and as an attempt to bridge the gap that has arisen between inpatient and outpatient care, it is a welcome improvement for patients. This sort of care makes sense for people who are on the verge of hospitalization, particularly as we move toward capitation and ACOs.

I wonder how truly innovative it is though- it could also be seen as a step back from the separate roles that have developed for outpatient/PCP and inpatient/hospitalist internists. Smaller patient panels for these providers makes sense and using them for this patient populations is smart- I expect to see many places try similar approaches. Thanks for sharing.

Anonymous said...

Caremore does very well in Whittier California and it puts its competitors to shame in many circumstances with their very keen yet understated interest in the bottom line - profit. They are starting to be seen in at one at least one of its contracted hospitals and many private physician offices as being quite predatory. They are at all local events promoting their plans and are finding a large base of patients wanting away from their current plans as Caremore advertises no co-pays and will provide transportation in most circumstances.

Some of the other ways that it maximizes profit may be in the way that they deal with the end of life issues. Patients graduating from hospice are not uncommon.

Anonymous said...

Hmm, doesn't sound "predatory." It just sounds like good service to patients.

Anonymous said...

Interesting idea. As others have mentioned, this constitutes a blurring of the hospitalist/PCP distinction. There has always been a tradeoff between the benefits of the hospitalist movement (enhanced inpatient availability/expertise) and its costs (increased fragmentation between the inpatient and outpatient settings, not to mention the brain drain on outpatient primary care introduced by the cross-subsidized income differential between inpatient and outpatient cognitive services).

Having hospitalists do more outpatient work changes the nature of the fragmentation. It's no longer inpatient/outpatient. It's outpatient/outpatient, with the fragmentation being between the "outpatient hospitalist" and the PCP, assuming the patient has one.

How long will the outpatient hospitalist stay involved in the discharged patient's care? A cynic would say "long enough to outlast the window for CMS readmission penalties...and not a moment longer."

Douglas Allen, MD., MMM. said...

I was the author of these two pieces on the CareMore model. I'm excited by the many comments already posted and wish to add my own comments surrounding some of the questions being asked.

I'll start with the one on hospitalists.

Why hospitalists versus the more traditional process of the PCP following his/her own patients upon admission?

The question of “why hospitalists”, was a very heated one in California 18 years ago or so. But time after time the question was answered. PCPs being distracted by outpatient work, unable to round on patients multiple times throughout the day as necessary, unable to respond to the rapidly changing health status of their patient(s), relied increasingly on specialists to manage their patient. And the specialists did not (I am generalizing here) think of themselves as coordinators of care. They treated their body part and maybe the body part next to it (heart-lungs, for example). Being reimbursed for volume and intensity, procedures and long hospital stays were a natural consequence. Hospitalists were utilized by pre-paid insurance agencies and medical groups, to focus on optimizing the efficiency of the inpatient experience. One example of how they did this was assuring outpatient workups were done in the outpatient setting rather than the high risk, high cost inpatient setting. Another was regular interaction with specialists, clarifying their plan of diagnosis and treatment, often times questioning the use of high cost imaging or drug treatment when alternative higher value options might be more appropriate. They were also the ones looking at the hospital episode from a higher level, determining what alternatives to an inpatient setting might be more appropriate, while becoming ever more familiar with these alternatives to a greater extent than the primary care physician. The lower lengths of stay were important from another perspective. It was not until much later that Institute of Medicine and others began to quantify the risks of staying in the hospital, with such works as To Err is Human. Back then there was just an intuitive understanding that staying longer than is necessary in the hospital could lead to poor outcomes.

And hospitalists worked. Their involvement brought down lengths of stay at first from 5.5 on up for Medicare patients, to 4.5 days. Then the next generation of hospitalists reduced lengths of stay to an even greater extent, achieving 3.2 days on average. But there were potential pitfalls to such a narrow focus, which led to the creation of alternatives to pure hospitalist systems. And to new metrics to focus on, such as readmission rates. But more on that later.

Douglas Allen, MD., MMM.

Anonymous said...

Good afternoon,

The link to the article at the top of this post has changed. The new link is