6 things wrong with hospital medicine


Image result for hospitalist movement

In 2002, when I began my first hospitalist job, I was a dyed-in-the-wool hospital medicine convert, convinced that the transfer of inpatient care to true specialists in hospital medicine (hospitalists) would dramatically improve the quality and efficiency of inpatient care, increase patient satisfaction and decrease costs.

By 2008, I had developed serious doubts, which prompted me to publish an editorial in the Journal of Hospital Medicine, entitled “The Expanding or Shrinking Universe of the Hospitalist” (2008) that attempted to raise a red flag of concern about hospitalists, in general, failing to become “hospital medicine specialists” and instead accepting the inferior role of “triage shift workers.”

Now, in 2018, I believe it is more appropriate to raise a white flag of surrender. I could write a book on the topic, but briefly, here are the six pillars of what went wrong with hospital medicine, in my opinion.


First pillar. In the first decade of the hospitalist movement, private hospitalist management groups (and hospital-employed hospitalist groups) popped up quickly all over the country, jockeying aggressively for market share, and working with a simple equation: a hospitalist physician was a fixed cost, and his/her patient load (primarily) was revenue. So the larger the patient load per hospitalist, the greater the profit. Young hospitalist applicants — almost all fresh out of training, in debt, hungry for income and already accustomed to long hours of work — were easily lured to the hospitalist positions offering the highest salaries, which were logically accompanied by the highest patient loads. Rising salaries were repeatedly celebrated by hospitalist leaders as evidence of the growing value of hospitalists, whereas they were more likely a result of the above market forces.


Second pillar. The high workloads resulted, quite naturally, in hospitalists aligning themselves in ways that increased patient encounters but minimized effort, which largely meant deferring responsibility for patient care and clinical decisions to others; that is, primarily, a liberal use of specialist consultations. In my experience, hospitalist progress notes quickly evolved into something like this: “Acute kidney injury, per nephrology; Chest pain, per cardiology; Cellulitis, per infectious disease.” Next patient. Time-consuming tasks, like end-of-life care discussions, were whittled down to a single line: “Consult palliative care.” (One hospitalist colleague actually explained to me once how he strategically avoided patients whose families were currently in the room, since he had to see over 30 patients a day on weekends and couldn’t spare the time for any family discussions.) Obviously, this short-sighted approach to a new medical specialty was a death blow to almost all of the claimed benefits of the hospital medicine movement.


Third pillar. With hospitalists increasingly dominating inpatient care, hospital administrators found that they could use this captive group of young doctors to increase hospital revenue by raising the case-mix index with “proper documentation.” Whereas comprehensive documentation of one’s clinical findings and decision making is certainly an essential part of quality inpatient care, the unspoken goal of the hospitals was to push the case-mix index higher and higher. A troponin of 0.05 became an NSTEMI. A cough and temperature of 99.5 became sepsis or severe sepsis (if there was a slight creatinine bump or relative hypotension) — and why not add acute respiratory failure, if someone happened to catch a low oxygen saturation reading (from a malpositioned pulse oximeter). In a darkly comical twist, the risk management mantra that “if you don’t document it, it didn’t happen” was tragically flipped into its false corollary: “If you do document it, it did happen”; that is, “oxygen saturation dropped to 85 percent on room air,” “patient was in severe respiratory distress,” etc. Unfortunately, this gray area of potentially exaggerated documentation muddies the clinical communication between clinicians, not to mention issues of ethics and law.


Fourth pillar. In much the same way, hospitalists were placed in the center of “level of care” assignments; that is, observation status versus inpatient status. Specifically, if an inpatient stay could be justified, by a “good” hospitalist’s “improved” documentation, the hospital could increase revenue by two to three times over an observation stay. Hospitalists were given subtle encouragement to transform things like atypical chest pain, UTI, or tingling fingers into life-threatening conditions, requiring complex decision making, and fraught with numerous potentially serious complications, and absolutely requiring more than two midnights to evaluate and treat properly. Once again, the ideal of a careful and proper diagnosis, with an appropriate plan of care in an appropriate setting, was profaned. Clinical decision making often blurred into a form of hospitalist doublespeak which obscured the actual severity of illness to achieve desirable metrics, earn a bonus or negotiate a better contract next cycle.


Fifth pillar. In addition, utilization review nurses were pressing hospitalists to get fixed-DRG patients out of the hospital as quickly as possible, to increase profit margins and make room for more patients and more revenue. This rapid-fire inpatient management r encouraged “good” hospitalists to order a shotgun round of tests and consultations up front on their admitted patients, and ultimately led to a lot of unnecessary testing, and a lost reliance on a proper history and exam, serial assessments and a cognitive, algorithmic approach to diagnosis and treatment — all further diminishing the clinical acumen of highly-trained individuals who truly could have been, in a different world, hospital medicine “specialists.”


Sixth pillar. Quality measures, supposedly aimed at improving patient outcomes, were an additional blow, as they unfortunately led physicians to do things that were not consistent with good clinical judgment. For example, in a case I saw, a patient presented with an acute tonic-clonic seizure, and their lactic acid level was markedly elevated (of course, from the seizure); but they were treated for sepsis with a fluid bolus and broad-spectrum antibiotics, because if someone saw the lactate level, the case would “fall out.” Similarly, triple antibiotic regimens were inappropriately used for viral bronchitis because of a stated concern for health care-associated pneumonia. Basically, non-thinking was being promoted in the service of higher quality scores — not higher quality.


Although these pillars are surely not generalizable to all hospitalist programs, especially academic ones, the hospitalist movement as a whole is a perfect example of how administrative and market forces in health care can largely extinguish the incredible potential of a new specialty. And that’s sad.




Measuring What Really Matters


Image result for measuring

Not everything that is important for a person’s health can be measured and not everything that can be measured in health care is important to the average person.

For too long, value has been defined only for the benefit of regulators and purchasers. Our health care system is purpose-built to cater to their performance needs, oversight, and expectations, and as such has fostered the proliferation of all sorts of clinical quality measures by multiple organizations. The current state of quality measurement serves these audiences reasonably well.

However, the problem with evaluating quality using these tools is twofold. First as a physician, I still see too much variation in the technical quality of American health care. Second, clinical measures alone ignore how value is perceived through the eyes of those who actually use the delivery system. When we look at the highest users of health care – those with serious medical problems and functional limitations – we now have an abundance of technical measures for each condition on their problem list, and yet really no understanding of whether we are contributing to a person’s quality of life. Frankly, I care little about the fact that my 100-year-old grandmother has never had a screening colonoscopy, but I care mightily that no one seems responsible for her successful discharge and transition home after a bout of urosepsis.

We cannot improve what we do not measure…and it is time to start measuring health care from the vantage point of those needing care, not just for those who provide and pay for it. And if we are to achieve the dramatic improvements anticipated through new payment and service delivery models, the mushrooming of purely clinical measures must be thinned out to make room for a new generation of metrics that consider outcomes from the person’s perspective.


Performance measures put heat on hospitals


This year has seen a bevy of new programs and initiatives aimed at moving providers toward a value-based model of care.

Assessing Care Integration for Dual-Eligible Beneficiaries: A Review of Quality Measures Chosen by States in the Financial Alignment Initiative

Assessing Care Integration for Dual-Eligible Beneficiaries: A Review of Quality Measures Chosen by States in the Financial Alignment Initiative