I was a hospitalist before the word was fashionable. It was July 1988 and I had just completed a 3 year post residency stint at a community mental health center as part of a public health service scholarship payback. For one of those years I commuted another 300 miles to keep a community hospital psychiatric unit open. I headed for the hospital where I did my rotating internship in Internal Medicine, Pediatrics and Neurology. It was the only real metropolitan inpatient treatment setting I had known at that point. In my residency program, the interns were split up into two groups and each group worked at one of the major county hospitals in the Twin Cities. It was a unique setting at that time because psychiatrists provided almost all of the medical coverage. They had to be able to diagnose and treat a lot of common medical problems, write for all of the patient's medications, attend to acute medical problems and do the appropriate diagnosis and triage. I had a wide range of medical problems admitted directly to me ranging from gunshot wounds to delirium. Any psychiatrist working in these conditions realizes that the term "medically stable" is a relative one. I had many patients admitted to my service with severe medical problems only because they also had a severe psychiatric disorder and were symptomatic at the time. In many cases I had to rapidly assess them and transfer to medicine or an intensive care setting.
I had excellent back up by consultants and many of them to this day are some of the best physicians I have ever seen. But they really did not want to hear from me unless I had a very specific probable diagnosis and most of the evaluation was done. There are not too many places in psychiatry where jobs like that exist anymore. If anyone asks me about similar positions - I actively discourage them from accepting a similar job. With this arrangement the work is far too long and all of the medical care is provided for free - psychiatrists do not get any extra credit for it.
In those days there were six of us covering 3- 20 bed wards, five days a week. The ads for psychiatrists these days often speak of "psychiatric hospitalists" - but every one of them specified no medical coverage. They also tend to leave out the part that it is basically a rapid triage and discharge position and the job is to either maintain or cooperate with high discharge rates. The only thing they have in common with the Internists and Family Physicians who have come to be designated as hospitalists is that they work 7 days on and 7 days off. A schedule that very few people question.
I naturally picked up this week's copy of the New England Journal of Medicine to see what the two perspective pieces on hospitalists (1,2) had to say. I was also interested because my brother is an Internist and over the years we have discussed the issue at length. The initial essay by Wachter and Goldman documents the rapid rise of hospitalist care as a medical specialty. Since 2003 the number of hospitalists has increased 5-fold to 50,000. That makes hospitalists the largest speciality within Internal Medicine. They cite the growth of managed care, Medicare DRG payments, and possible evidence as reasons for the growth of the field. I am always skeptical of the term efficiency especially when it is combined with the term quality. I guess it is difficult for some people to accept the fact that managed care and Medicare DRG payments are rationing mechanisms that are tied to quality only by the tenuous thread of government and healthcare company rhetoric and advertising. The other critical question is efficiency for who? It certainly is more efficient to administer a group of physicians who work 7 days on and 7 days off and happen to all be in the same chain of command. It is a lot easier to get them to accept the role of rationing care in the interest of the hospital or health care group than the patient's personal physician who may see their part of their role as patient advocacy.
The authors have an interesting take on the deficiencies of the model. They talk about the 7- days-on, 7-day-off model as implying that during the off period the physician is literally off and suggests that time might be better spent contributing to key institutional programs. To me - this schedule seems more conducive to burnout and anyone who works it needs the off time to fully recover. I have never seen a study on the cognitive efficiency during the 7-days-on, but my conversations with hospitalists suggests that by day 6 it starts to plummet. With hospitalists supplanting specialists and subspecialists as inpatient attendings they suggest that trainees have less exposure to basic and translational science. Although not stated in the article, the model involves eliminating whole blocks of specialty care. I worked at a hospital where an entire Neurology service was eliminated by hospitalist care. When I questioned that decision I was told: "We have an Internist who is interested in strokes." Changing neurologists from attendings to consultants with hospitalists as the primary physicians for neurological problems changes the entire nature of care. It also changes the associated nursing care when staff have no ongoing interest in the care of complex neurology patients. The authors also note that hospitalists do not seem to have focused on investigating common inpatient illnesses. They suggest possible remedies - but these seem like major problems that will only get worse with the increasing business rather than academic emphasis in medicine.
Gunderman points out that as opposed to the usual delineators of speciality care - patient age, physician skillset and body system hospitalists are delineated only by patient location. He doesn't make it explicit but what is the relationship between location and his list of putative benefits? Looking at length of stay for example - that could logically follow as a concentrated effort in the location, but is that a clinical effort or an administrative one? He points out that the increasing number of hospitalists per se, cannot be taken as evidence of benefit and that perverse incentives exist. I agree with the most perverse being the low reimbursement incentive for high volume practice. Seeing complex inpatients with a high frequency of initial and discharge assessments may reduce the volume necessary for productivity demands. When I was a psychiatric hospitalist, this dimension was manipulated in a number of ways. I was initially told, I was responsible for a set number of inpatient beds. At some point there was a great deal of pressure for me to start running outpatient clinics because they would be more "interesting" than just seeing inpatients. I resisted that and had significant leverage because nobody else wanted to do my job. I eventually did run a Geriatric Psychiatry and Memory Disorder Clinic for many years while continuing inpatient work. That clinic was eventually closed by administrators because they claimed our productivity was not high enough to work with a nurse. The neurologist and I needed all of the collateral data that she collected to do our work. The expectation was that we would see complex dementia patients and do everything that the nurse in our clinic did - so we closed. In over two decades of political wrangling around inpatient productivity the current consensus is that covering 10-12 inpatient beds is a reasonable approach. At one point I was covering 20 beds with the help of an excellent physician assistant but at the cost of doing no teaching.
The critical aspect of Gunderman's thesis is his emphasis on the physician-patient relationship exemplified by this sentence:
"The true core of good medicine is not an institution but a relationship - a relationship between two human beings."
He points out that physicians being affiliated with institutions creates significant conflicts of interest, isolates hospital staff from the rest of the medical community and that naturally leads to less expertise in the entire community. It also creates the illusion that an institution rather than the relationship is the core of medical care and it is not. Government-business constructs like Accountable Care Organizations have a similar effect. I have experienced this first hand many times as I dealt with the iterations of hospitalists consulting on my patients. In one case I talked with a young hospitalist about a patient with Type 2 diabetes mellitus. The patient had a trace of renal insufficiency and was on metformin - a medication that is risky in that context. The hospitalist advised me to call the primary care Internist taking care of the patient because "He has been doing it a long time and probably knows more about it than I do." In addition to the relationship - there is clear expertise associated with caring for people with multiple complex medical problems for years in an outpatient setting - compared to a few days as an inpatient. The medical industrial complex does not adequately value that expertise.
I think that there is room for hospitalists and psychiatric hospitalists. They have to be focused on the needs of both the patient and the patient's outpatient physician. There have to be clear goals for the hospitalization and one of those goals is what the patient's personal physician would like to see accomplished. Since making the transition to strictly outpatient care - it is clear that the hospitalists no matter who they might be don't have much control over who gets admitted to the hospital and what happens there. They are having less to say about when a person is discharged. This is probably more true for psychiatry than medicine and it results in a large number of psychiatric outpatients not being able to access needed care.
And I can't help but notice that inpatient hospital medicine is still a far better resource than inpatient hospital psychiatry.
George Dawson, MD, DFAPA
1: Wachter RM, Goldman L. Zero to 50,000 - The 20th Anniversary of the Hospitalist. N Engl J Med. 2016 Aug 10. [Epub ahead of print] PubMed PMID:27508924.