Health for life

Check out the three page article on "U.S. Health Cares Future" in the Marketplace section of the December 12 Wall Street Journal. Using personal stories of a doc, hospital CEO, insurance executive, human resources manager and a patient, the news piece portrays the blurring business lines between insurers, buyers and providers.  Mainstream readers of the Journal are likely to think the topic is both timely and novel.  Regular readers of the Disease Management Care Blog learned about his months ago.

Who are these five canaries in the health care coal mine, these bellwethers of the insurance business, these oracles of management and what are they telling us?

1. Dr. McCullough, a salaried physician with 28% of his income contingent on quality and satisfaction.  Some measures were imposed by the local Blues plan, which was passed through to the him by his employer.

Message: Purchaser and buyer control of physician reimbursement is already big and its growing.

2. Jim Taylor, a hospital CEO who cannot buy an electronic record system unless he merges with two other hospital systems.  If the merger is approved, the hospital will also be able to take on "warranty-style" payments from insurers.

Message: "Bigger is better" for capital-constrained hospitals.

3. Chris Day, an Aetna executive who got an Arizona health system to share insurance risk.  The main sticking point was the two-way mutual sharing of internal cost and contracting data.

Message: If insurers are willing to share internal pricing data, they must really mean it and think its an important success factor.

4.Robert Jacobs, the HR person, who linked about $10 per week of employees health insurance premiums to healthy behaviors (like tobacco) and quality test results (like blood cholesterol levels).

Message:  "Dont just stand there," say the employers, "do something."

5. Louis Kandor, an 86 year old man with advanced diabetes, who is being visited by a nurse who, in turn, is employed by a care management service provider under contract by his Medicare Advantage insurer.

Message: One key to mitigating risk for every insurer (except fee-for-service Medicare) is to use nurse-led care management.

While the Journal article doesnt spell it out, the DMCB believes the anecdotes can be distilled down into three useful insights:

1. Stakeholders are scrambling to demonstrate measurable outcomes to an increasingly educated and  skeptical public.  Thats the basis for physician pay-for-performance and premium surcharges.

2) Sharing proprietary insurance data is important.  Is information the secret ingredient that was lacking during the similar - and mostly unsuccessful - insurer-provider collaborations back in the 1990s?  Well see. 

3) For those hospitals that cannot or will not take risk, the next best answer is to merge.  That will mean economies of scale, access to capital and negotiating leverage.

This way to consolidation....

Disease Management Care Blog readers may recall that Medicares Hospital Readmissions Reduction Program was among the many provisions of the Affordable Care Act. 

According to the finalized regulations, if a hospitals readmission rate within 30 days for heart attack, heart failure or pneumonia exceeds an established norm (using three years of data based on a minimum of 25 patients with a statistical risk adjustment to account for co-morbid conditions), that hospitals Medicare payment rates will be reduced for all discharges in the following year. The reduction, depending on the excess rate, can go from zero (readmissions meet the norm) to a maximum of 1% (the hospital penalty results in payment of only 99% of the applicable fee schedule).

Now, Kaiser Health News has just looked at the numbers and calculates that, thanks to the HRRP,over 2000 hospitals will forgo close to $300 million. According to KHN, 278 hospitals - including some household names - will achieve the dubious distinction of a full 1% reduction.  You can check out how your local hospital will likely fare here.

While readmissions themselves are a significant problem, the approach used by the HRRP has its own set of under-appreciated methodologic challenges (as noted here and here). Now that hospitals are about to get battered by a well-meaning if flawed payment system, your DMCB raises one more red flag:

This will drive hospital consolidation.

That may well be one intent of the law. Cheesecake Factory logic tells us that large hospital systems have the intellectual and capital resources to systematize care, apply best practices, reduce variation and maximize outcomes.  Rather than weep for those hospitals that are losing income, Washingtons policymakers are probably hoping that the losers have one more reason to join forces with the bigger, smarter and more efficient hospitals or systems nearby or in the next state (especially the ones with a smartly run disease management program).

Yet, whether or not hospital consolidation alone would make a palpable difference in cost or quality remains to be seen (as indicated here and here). What could happen instead is the rise of too-big-to-fail, politically connected and market-dominant health care systems.

Well see.

Image from NIHSeniorHealth

"I  think I need to go back to the hospital....."

The Disease Management Care Blog remembers her very well. She had come to the United States after escaping on foot from post-war Hungary.  She liked to have sliced pears and cognac for lunch. As her health began to fail, her stubbornness began to exceed her self-sufficiency, leading to multiple hospitalizations. One time, despite making sure she had a post-discharge outpatient appointment, a fresh pile of prescriptions and home visiting nurse services, her illness flared and she had to come back to hospital.

She was not only my patient, but she represented one of Medicares dreaded readmission statistics.

By now, DMCB readers know that CMS, buoyed by its value-based purchasing program, has targeted readmissions by reducing payment levels to hospitals that fall outside the expected norm. Threatened by the loss of income, it’s assumed that hospitals will respond by developing higher quality discharge planning and care programs that keep patients from having to come back.

An important part of reducing readmissions is to identify those patients that are at greatest risk. That would help on two levels:

1) if a hospital had more than its fair share of patients at risk, it could argue that an increased number of readmissions is the result of a sicker patient population and not quality of care. As a result, the hospital could be held "harmless;

2) hospitals would be able to focus extra care resources on those patients who are spotted early as likely to come back, thereby reducing the readmission rate.

In other words, patients like the lady from Hungary would not necessarily lead to a cut in hospital payment rates and, for example, she could be proactively given extra care, such as a doctor appointment within 48 hours, a week’s supply of free medications and twice a day home nurse visits.

Which is why this just published JAMA article "Risk Prediction Models for Hospital Readmission" by Devan Kansagara, Honora Englander, Amanda Salanitro David Kagen, Cecelia Theobald, Michele Freeman and Sunil Kripalani is important.  The authors set out to see what the evidence-based published scientific literature had to say about predicting readmissions.   They filtered thousands of erences, reviewed 286 publications found 30 rigorous studies that described 26 models. 

To the DMCBs delight, the authors applied a “c statistic” to the 30 publications to assess a wide variety of retrospective and concurrent prediction methodologies using a host of data inputs such as age, gender and past diagnoses. According to this article, the c (or "concordance") statistic measures how well a test can predict the presence or absence of a "condition" which, in this case, was being readmitted to the hospital.  One way to think of this is the likelihood of correctly identifying a condition when there are two people, one with it and one without it.  If the likelihood is 50%, thats no better than random guessing.  If its 100%, thats perfect.  By the way,if this sounds a lot like the area under the receiver operator curve, youre right.

And what did the all-seeing "c statistic" say?  All of the published models had disappointingly similar levels of performance that ranged between the extremes of .52 to .83 with most in the .50 to .7 range. Whats more, only one study examined the most important question of all: is it possible to find patients with preventable readmissions?

What does the DMCB think?

1) This may be another area where national health policy has gotten out in front of the scientific evidence.  If we cant reliably assess or predict readmissions with sufficient accuracy, there is a distinct likelihood that statistical variation, not quality of care, will lead to some hospitals being victimized by CMS with lower payment rates. Whats more, if hospitals cant tell which patients are likely to come back, how are they supposed to target their expensive care management programs at those who are most likely to benefit?

2) There are undoubtedly some proprietary predictive models that havent been reported in the literature that claim to have higher levels of accuracy.  Yet, without the scrutiny of successful peer-reviewed publication, itd be difficult to believe that theyre really any better than the mainstream published range of .5 to .7.  The next time the DMCB runs into one of these outfits, its going to ask about the "c statistic" and if they havent published their results, why not.

3) Last but not least, while the hospital payment rates are being held hostage by CMS, its the doctors that are making the call on readmissions based on the best interest of their patient.  The c statistic suggests that that will be the most important determinant in the readmission rate.

"Maybe you should go back to the hospital!"

Ask most wonks - especially ones who never took care of a patient - about "readmissions," and, after quoting this article, these health policy Urkels will tell you that returning to a hospital is the poster-child of all that ails U.S. medical care. Providers who cant get it right the first time, they say, are not only giving slipshod care, but are double dipping because their mistakes generate even more fat fees the second time around.

"Balderdash!" says the Disease Management Care Blog. Many Medicare inpatients are so sick that its a miracle that they get to go home in the first place.  Keeping patients in the hospital can be more life-threatening than the home environment and, when things dont get well after a discharge, its often more a function of social support than medical skill. 

That doesnt mean that CMS is going to listen to docs and back off of its Hospital Readmissions Reduction Program (HRRP). Using risk-adjusted actuarial projections, every U.S. hospital will be prone to a possible payment reduction if their observed rate of readmissions for heart attack, heart failure, and pneumonia exceeds the expected rate. Based on those projections, approximately two thirds of hospitals could be penalized.

Writing in the New England Journal of Medicine, Karen Joynt and Ashish Jha point out that hospitals are concerned because 1) readmissions fall outside of their control and 2) the actuarial projections are imperfect.  As a result, hospitals that care for the most fragile and socioeconomically disadvantaged are at risk for paying more than their fair share of CMSs $280 million claw-back penalty. 

The NEJM authors recommend three modifications to CMS HRRP:

1. Include patients socioeconomic status in any risk adjustment modeling. One easy-to-obtain modifier, for example, could be whether the patient is on Supplemental Security Income.  Patients on SSI are less able to cope, which is why they quality for the program in the first place.

2. Include hospitals mortality rates in any risk adjustment modeling.  Hospitals with special expertise are less likely to have borderline patients die on their inpatient services, which means theyll have their more than their fair share of fragile survivors.

3. Limit the penalty to readmissions that occur within hours or days of a discharge, instead of the current problematic policy of counting any readmission that occurs within 30 days.  It makes sense to believe that a premature discharge or slipshod discharge planning is at fault if the patient returns within 3 days instead of three weeks.

Since its unlikely that HRRP program is going away, the DMCB agrees with the three recommendations.  In the meantime, it also suggests:

1. CMS should be held accountable by Congress to execute well on the program,

2) Claims analytics - possibly using a "Big Data" approach - should be applied to Medicare claims to examine whether hospitals are turning to two potential options to undermine the program:

a) gaming the system by altering how they "code" the billing for their readmission patients, or

b) accepting the penalty because of favorable income from readmissions.

Image from Wikipedia

Dr. Deborah Wexler and her colleagues at Massachusetts General Hospital in Boston screened 695 adult patients admitted during 11 days in 2006.

How many of those had an HbA1c greater than 6.1% (indicating probable diabetes) but had no diagnosis?

"Nearly one in five adult patients admitted to a large general hospital had unrecognized probable diabetes, based on elevated HbA1c levels."

That rate, the authors said, was "roughly 5-fold higher than in the general outpatient population."

Also, random blood sugar tests performed during the patients hospital stay were found to be poorly predictive of diabetes. In this case, sampling blood glucose with a meter was not as effective as assessing blood glucose over the prior few months, which an HbA1c test can measure.

Unfortunately, only 15% of those with elevated HbA1c were diagnosed within a year, pointing to a area where more attentive follow-up could prevent diabetes-related complications.

Dr. Wexlers research was published in the November issue of the Journal of Clinical Endocrinology & Metabolism:
Prevalence Of Elevated Hemoglobin A1c Among Patients Admitted To The Hospital Without A Diagnosis Of Diabetes
________

Im baaaaack!

When persons are admitted to a hospital, insurers payment rates are based on the diagnosis, not the number of days in the hospital (known as a "length of stay").  As a result, once the admission is triggered, the hospital has important economic incentive to discharge the patient as quickly as possible.  The Disease Management Care Blogs physician colleagues used to refer to this as "treat, then street."

Unfortunately, discharging patients too soon can result in readmissions.  Thats why the DMCB has agreed with others that diagnosis-based payment systems and a policy of "no pay" for readmissions were working at cross purposes.  Unified bundled payment approaches like this seem to be a good start.

But thats all theoretical.  Whats the science have to say?

Peter Kaboli and colleagues looked at the push-pull relationship between diagnosis-based payment incentives  and the likelihood of readmissions in a scientific paper just published in the Annals of Internal Medicine. 

The authors used the U.S. Veterans Administration (VA) Hospitals "Patient Treatment Files" to examine length of stay versus readmissions in 129 VA hospitals.  The sample consisted of over 4 million admissions and readmissions (defined as within 30 days and not involving another institution) from 1997 to 2010. The mean age started out at 63.8 years and increased to 65.5 years, while the proportion of persons aged 85 years or older increased from 2.5% to 8.8%. Over the years, admissions also grew more complicated with a higher rate of co-morbid conditions, such as diseases of the kidney (from 5% to 16%).

As length of stay went down, readmissions should have gone up, right?

The answer was yes and no.

Yes, if the data were trended over time: Over the 14 year period of observation, the number of days in the hospital (length of stay or LOS) decreased from 6.0 days to 4.3 days.  Yet, as LOS decreased, readmissions also decreased from 16.6% to 15.2%. 

The decreases held up when the LOS was risk-adjusted for hospital and patient characteristics.  There was also no increase in mortality rates

No, if hospitals were compared to each other:  Hospitals with risk-adjusted low lengths of stay had higher readmission rates compared to their average peers.  In that group, each day of saved LOS was associated with a 6% increased rate of 30-day readmissions.

It gets even more complicated.  As the LOS increased beyond the average, each additional day in the hospital was associated with a 3% increased rate of 30-day readmissions.

What should the DMCB learn from these data?  Keeping in mind that the VA is not necessarily generalizable to the typical community medical center,

1. Over 14 years of worth of VA data for 129 hospitals suggest it is possible to have your cake (a lower LOS) and eat it too (lower readmissions).  Thats the good news.

2. While overall performance improved over the years, between hospital comparisons showed there is a "U" shaped relationship between days in the hospital and the likelihood of readmission.  The DMCB agrees with the authors: premature discharge before the patient is ready is associated with an 6% per day readmission rate, while patients who are very sick and have to stay a few extra days in the hospital are also at risk to the tune of 3% per day.  Thats the sobering news.

What are the implications?

Overzealous efforts to discharge patients can backfire with readmissions.  It appears theres an optimum length of stay that minimizes, but will never eliminate, readmissions.

Patients who do go home "too soon" or need extra days in the hospital appear to be at special risk.  Accountable care organizations and population health management service providers should use this information to target patients at special risk of "treat, street and... repeat."