Admission, Discharge and Readmission Policies - No Better Example Of Business Driven Pseudoscience

One of the recurrent themes of this blog is that the application of science to medicine, especially at the public policy level has plummeted over the past three decades.  That has been directly attributable to the influence of business on all levels of government.  One of the more pervasive themes is that the behavior of health professionals is best accomplished by incentivization.  In other words financially punish physicians to get them to change their behaviors that you don't want or financially reward the behaviors that you want them to produce.  I haven't looked at the scorecard lately, but my guess is that the punishments greatly outweigh the rewards.  Maybe my perspective has been skewed by working for an HMO for many years that considered it a reward if they "held back" part of your salary and then gave it to you if all of the physicians in your group met the desired productivity targets.  I am sure it took the MBAs a while to dream that one up.  The equation seems to be as simple as -  "OK here is what we want the goal to be.  We don't want to pay out any rewards anymore.  Let's just penalize people for not meeting the goal until eventually everyone is compliant with the goal."  There is probably no better example than the Medicare Hospital Readmissions Program.

A recent editorial in JAMA notes that the 2013 readmission rates for Medicare patients is about 18% within 30 days (1).  That is associated with a potential cost of $26 billion.  Since 2010 Congress has levied a 3% of Medicare reimbursement penalty on hospitals who have readmission rates that are considered too high.  The problem is that 80% of hospitals are being penalized are safety-net hospitals or those that have a disproportionate share of low income patients.  Those hospitals are more likely to be penalized all three years since the penalty started  and they are more likely to be the hospitals with the lowest operating margins.  The likelihood of penalty also correlates with the percentage of patients treated who are elderly and live with poverty or disability.  

The authors opine that hospitals should not be penalized "because of the demographic characteristics of their patients."  They point out that the evidence suggests that is exactly what is happening and they conclude:  “Targeting hospitals for penalties, even if indirectly, simply because those hospitals care for more poor people is not good policy”.  They use this as a foundation to build their argument for a proposed policy initiative – The Hospital Readmissions Program Accuracy and Accountability Act of 2014.  It builds in safeguards for hospitals treating patients from a disadvantaged socioeconomic status.   

The obvious problem with the authors’ logic here is that they seem to not realize that discrimination against patients of the lowest socioeconomic status has been institutionalized and occurring for decades.  The people I am referring to are those people with addictions and severe psychiatric problems.  The facts are clear.  For the past 30 years, even though psychiatric disabilities rank as some of the top 10 disabilities by any measure, they get a much smaller fraction of the health care dollar for care.  I have used the example of a middle-aged man or woman being hospitalized through the emergency department for acute chest pain.  I don’t know the fraction of those people who are discharged the next day.  But consider that basic scenario if the evaluation of chest pain turns out to be non-cardiogenic.  In the hospital where I have worked that generally means an evening on telemetry and serial troponins and either a stress echocardiogram the next day or an echocardiogram and a stress test.  Price tag about $25-30,000 for less than 48 hours in the hospital.  On the other hand,  let’s say a person has an exacerbation of an affective psychosis and is not able to function at home or has put themselves at risk.  The will be hospitalized in a very low tech psychiatric unit, the goal of which is to discharge them when they are no longer “dangerous” or to discharge them upon request if they cannot be held involuntarily.  Irrespective of the price tag for this care the best available data I have on the DRG reimbursement for this care is about $4,800 irrespective of length of stay.  The economic incentives all line up to rarely provide them with the discharge resources they require to maintain even a subsistence life style and remain stable enough to stay out of emergency departments or jails.  Furthermore in many cases, states previously charged patients a for a portion of their medication costs per month out of their disability income.  The direct and indirect costs incurred by patients and families with severe mental illness and addictions are a travesty of the highest magnitude.  The rationing mechanisms that have been in place for the past three decades have results in care that is subpar relative to any other medical specialty.  It has created an entire population to patients with chronic illnesses that are discriminated against.  The financing of care for them has set a number of perverse incentives that would seem to be more destabilizing such as an incentive for hospital discharge in order to beat the designated days in the diagnosis related group (DRG) and readmit them if necessary.  If the entire DRG incentivization for admissions and discharges is pseudoscientific sleight-of-hand based on very crude demographic variables - why would we expect readmissions policies to be any different?

The second dimension of this care is just how unscientific care based on demographic factors is in the first place.  I was previously in a practice where “consultants” who had never practiced medicine came in and commented on the “complexity” of our patients.  At the time I was caring for many patients who I knew would never be admitted to other general psychiatric units in any other hospital in the state due to their medical complexity.  The consultants concluded that my patients were no more complex than any other patients in the state even though they could not define the measures they used to make that determination.  Nobody mentioned the inherent conflict of interest when a pro-discharge administration hires consultants that agree with their world view - discharge patients as soon as possible.

In another scenario and on a committee, I asked if the demographic determined characteristics and time lines for treating community acquired pneumonia led to any differences in mortality or complications – and nobody knew.  The original Big Data approach in medicine looked at HEDIS variables.  Any practicing physician knows this is an incredibly crude approach that in many cases is meaningless.  There is no better example than saying that treating acute and chronic psychosis in a few days makes no difference in outcomes, when nobody knows the best treatment approach and practically no hospital screens for functional or cognitive capacity - two well known areas of psychiatric disability.  In the outpatient sphere, it is the equivalent of saying that 10 or 20 minutes three or four times a year with an emphasis on medications that are not likely being taken by the patient can possibly affect their real life outcome.

In the case of patients with addictions the treatment is more dire.  When a person using heroin, alcohol, and excessive amounts of benzodiazepines cannot get admitted for detoxification or they cannot get admitted for residential treatment, society and its representative governments at all levels are saying that this is a situation where we can ignore conditions that are clearly life-threatening and in many cases fatal.  We can ignore them because businesses and governments say that this is a collection of disabling and life-threatening diseases that we can ignore so that they can either make money or divert money to treat more socially acceptable life-threatening and disabling diseases.  

This is all a clear pattern of discrimination that not only affects the elderly but anyone with a psychiatric disability or addiction.  If the authors want to do something about that – I say let’s start by reversing over 30 years of discrimination against those with psychiatric and substance use problems that is clearly based on socioeconomics especially the lack of a vocal political constituency, very poor research based on demographic variables rather than complexity, and a lack of innovative research based on poor resource allocation.

George Dawson, MD, DFAPA

References:

1: Boozary AS, Manchin J 3rd, Wicker RF. The Medicare Hospital Readmissions Reduction Program: Time for Reform. JAMA. 2015 Jul 28;314(4):347-8. doi: 10.1001/jama.2015.6507. PubMed PMID: 26219049.

Attribution:

Photo by Mark Buckawicki (Own work) [CC0], via Wikimedia Commons.