Love to a fault: How the best of intentions is hurting care for Americans who live in rural areas

Ensuring that Americans who live in rural areas have access to health care has always been a policy priority.  In healthcare, where nearly every policy decision seems contentious and partisan, there has been widespread, bipartisan support for helping providers who work in rural areas.  The hallmark of the policy effort has been the Critical Access Hospital (CAH) program– and new evidence from our latest paper in the Journal of the American Medical Association suggests that our approach needs rethinking.  In our desire to help providers that care for Americans living in rural areas, we may have forgotten a key lesson: it’s not about access to care.  It’s about access to high-quality care.  And on that policy goal, we’re not doing a very good job.

A little background will be helpful.  In the 1980s and 1990s, a large number of rural hospitals closed as the number of people living in rural areas declined and Medicare’s Prospective Payment System made it more difficult for some hospitals to manage their costs.  A series of policy efforts culminated in Congress creating the Critical Access Hospital program as part of the Balanced Budget Act of 1997.  The goals of the program were simple: provide cost-based reimbursement so that hospitals that were in isolated areas could become financially stable and provide “critical access” to the millions of Americans living in these areas.  Congress created specific criteria to receive a CAH designation: hospitals had to have 25 or fewer acute-care beds and had to be at least 35 miles from the nearest facility (or 15 miles if one needed to cross mountains or rivers).  By many accounts, the program was a “success” – rural hospital closures fell as many institutions joined the program.  There was widespread consensus that the program had worked.

Despite this success, there were two important problems in the legislation, and the way it was executed, that laid the groundwork for the difficulties of today.  First, Congress tried to show “flexibility” by allowing states to waive the distance requirement – and waive it they did.  The number of hospitals designated as CAH increased by nearly 50% over the 2000s.  By 2010, there were nearly 1,300 CAHs, many of them in suburbs or even urban areas, just miles down the road from a large institution with no mountains or rivers to impede travel.  Congress finally stepped in to close the loophole, but in many ways, it was too late.  Nearly 1 in 4 acute-care American hospitals today has the “CAH” designation.

The excessive use of the CAH designation may seem to primarily be a money problem.  We know, based on MedPAC data, that hospitals that get the CAH designation get paid substantially more than they would otherwise – and that in the early years of the program, CAH cost growth substantially outpaced comparison hospitals.  Having nearly 1,300 hospitals with the CAH designation is not great for managing cost growth – “cost-based” reimbursement is not the ideal way to incentivize organizations to pay attention to efficiency.  But the problem is much bigger than that – the real issue is that we have very little information about the quality of care that these institutions provide – and that’s the topic of our latest JAMA paper.

The biggest problem with the CAH program is that policymakers consistently exclude these hospitals from national quality improvement initiatives.  All U.S. hospitals have to collect data on process measures and report them to CMS – except the Critical Access Hospitals.  Most U.S. hospitals now have their performance on outcomes measures (such as mortality) or utilization measures (such as readmissions) publicly reported through Hospital Compare.  Critical Access Hospitals?  Not so much.  Even value-based purchasing, the new pay-for-performance program that CMS is running, has mainly exempted CAHs.  The arguments for exempting CAHs are numerous, including the notion that it may be too hard for these organizations to collect and manage quality data, and that they should spend their limited resources elsewhere.  There are some practical issues, of course, such as small sample sizes, but those issues are manageable – and there are plenty of small hospitals with small sample sizes that continue to collect data and report them to CMS.

So, what’s the impact of our efforts to protect these hospitals from the quality initiatives?  The data are coming in, and it’s not good news.  In our latest paper in JAMA, we find that a decade ago, mortality rates for AMI, CHF and pneumonia were pretty comparable at CAHs and non-CAHs (especially when you adjust for things like hospital size, case volume, etc.).  Over the next 9 years, however, there was a clear separation: while most hospitals improved, CAHs got worse.  By 2010, a patient arriving at a CAH with an acute MI had a 33% higher chance of dying than a comparable patient elsewhere.  Even in our matched analysis, where we compared CAHs to other small, rural hospitals that were not in the CAH program – we see similar findings:  CAHs and other small, rural hospitals started off comparable in 2002 – but by 2010, the CAHs were substantially worse than other small rural hospitals.  Other studies, including some of our prior work, have found similar results.

So what’s the take-home?  First and foremost, we have to remember that many of these hospitals do provide critically important access to care for many Americans.  However, the program – built on the best of intentions, needs updating.  There are three things that policymakers ought to do right away.

First – eliminate the Critical Access designation for any hospital that does not meet the original criteria.  In 2010, nearly 17% of CAHs were not in small towns or rural areas. Some are even in urban locations.  They are not “critical access” hospitals in any sense of the notion — and should have their designation revoked (unless they can make a very compelling case otherwise).  Second, these hospitals should be participating in all national quality improvement initiatives.  Even though they may have small sample sizes, the very act of collecting the data and reporting it is likely to be helpful.  The Institute of Medicine said as much back in 2004:  “The committee emphasizes that rural providers should not be excluded from public reporting initiatives. Public disclosure and eventually pay-for-performance payment methods…are potentially powerful incentives for encouraging improvements in quality. Rural providers, like urban, will benefit from these external levers for change.”  Nearly a decade later, there is still time to heed the IOM’s call.

Finally, we need a more proactive approach to helping these hospitals manage their sickest patients.  Hospital care has become extremely complex and the challenges for CAHs will only get steeper the longer we wait.  The defenders of the current program will argue that it is not fair to expect CAHs to manage an acute MI patient as well as a big academic center can.  But if we are being patient-centered, that is the wrong argument.  The right argument is that we can, and should, do better as a health care system for our rural populations.  The question now is – how can we help these hospitals manage their acute MI patients better?  How can we help them manage their pneumonia or stroke patients better? Here, there is a tremendous role for technology and for policy.

Every one of these hospitals should be equipped with telemedicine – and should have a formal partnership with a bigger institution.  Medicare can provide extra payments to larger hospitals that agree to partner – but tie part of those payments to outcomes for CAH patients.  In exchange for extra resources, these larger hospitals should provide technical assistance around quality measurement and improvement.  Further, we need to encourage CAHs to transfer their sickest patients to these partner institutions when additional care is warranted (amazingly, the proportion of patients who are getting transferred from CAHs to other hospitals has fallen by 25% over the past decade) and ensure that once the patient is stabilized, they are transferred back to the CAH.  These partnerships do exist among some subset of Critical Access Hospitals, but clearly not enough.  They need to be universal and a key part of the program.

Over 16 years ago, we embarked on a national effort to save rural hospitals – and closures of rural hospitals have declined precipitously.  Preserving access to hospital care is a good thing.  Yet, the program clearly has also gone astray in important ways:  there are too many hospitals with the CAH designation that do not need it – and it’s wasting taxpayer money.  Even more importantly, our effort to shield these hospitals from the difficulties of participating in national quality improvement efforts may have been well intentioned, but it has not benefited those Americans who count on CAHs for their hospital care.  We can surely do better.

3 thoughts on “Love to a fault: How the best of intentions is hurting care for Americans who live in rural areas

  1. Pingback: Chart of the day: Can helping hospitals harm patients? | The Incidental Economist

  2. My group is now providing coverage at one of these hospitals. We have set up strict guidelines on sending out sicker patients, and the hospital has gone along quite well. However, what I have been told is that the hospital gets higher overall reimbursements if it can document that it is treating sicker patients. If that is true, then I can see why CAH places dont want to transfer.

    Steve

  3. Great point — indeed, on the outpatient side, many FQHCs report a lot of clinical data. My thoughts are more about inpatient care, where CAHs are not required to report to the Hospital Compare program and they don’t participate in national pay-for-performance efforts.

Leave a Reply