The Hospital Transitions Emperor Has No Clothes

When reading international comparisons of health system performance, a cautious eye never hurts. I do not believe elaboration is required, but each country approaches measurement a tad differently, and we unwittingly omit meaningful population characteristics using available variables when examining outcomes. The nuances embedded within cultures do not express themselves readily, even to keen eyed investigators.

However, data don’t hurt, and every few months The Commonwealth Fund releases something we can and should chew on. A recent release, A Survey of Primary Care Doctors in Ten Countries, contained interesting findings on a topic we all scrutinize.

Most readmission initiatives such as Project BOOST and RED, all emphasize the need for provider follow-up shortly after discharge (less than a week). Most observers, including myself, anticipate inclusion of notification time in future CMS report cards and value-based purchasing programs. Therefore, postulating a reference point and what others label as acceptable serves as a locus as we reshape the discharge checklist.

Before I highlight the study findings and continue on the above theme however, I wish to use an additional one as a bridge to assist in my message.

If we had to choose one diagnosis with high costs and recidivism rates, with already established chronic care management programs, and a CMS imprimatur for value enhancement, the pick is CHF.

Conveniently, Hospital to Home (H2H), a program sponsored by the ACC and IHI, concentrates on the aforementioned. I cite the initiative because of its size, number of participants (594), and connection to CHF.

A recent study in JACC surveyed the same H2H hospitals for interventions implemented to reduce CHF readmits. The  paper used a questionnaire, looking at years 2009-10, and had a response rate of 90%.  Big study. Big players. Motivated group.

How did these facilities perform concerning primary care follow up?  Take a look:

I highlighted relevant questions, but note other similar themed queries within the box.

If you anticipated superior results from (presumably inspired) participants, middling is all you got.   With the expectation the bar should be close to 100%, or exceed above average levels in this advantaged group, these practitioners, and by proxy, us, have a ways to go. Excellence exists in some institutions, but the performance norm falls short of our potential.

Incidentally, subpar achievement should not surprise us, as publications, JAMA here from 2007 for example, demonstrate an ongoing deficit in transferring information from hospital to community.

The Commonweath study then, similarly inspected primary care practices and explored if discharge notification occurred, and if so, at 48 hours or two weeks post patient release. Keep in mind, the study cohort included only European nations (and Canada). Also, consider the often superior IT integration these countries possess, and their focus on primary care. The dependability of these systems supersede ours in the connectivity domain, and occasionally American physicians (begrudgingly) look eastward for instruction.

With that in mind, note the findings:

Here is a more comprehensive look at the results:

Observe the U.S. systems’ ability in notifying providers of discharge. Even allowing for some measurement inaccuracies, we lag.

However, the differences between countries belie a greater fact, and an important one, and my post speaks to this finding. Look at the results, and observe how poor all countries perform. Why?

Apparently, no leading-edge nation completes care transitions adequately. And given the large sample size, including every intervention availed to a modern health system (targeted education, lures of bonuses, use of technology, and peer pressure), we collectively fall short. Think of the legions of docs, hospitals, and cultures included in the mix. The talent pool, with modesty, does not get deeper.

I suspect, self-directed providers, even with technology and care team support–without system redundancies–will flounder.  An organization must build layers of protection to ensure we do not.  Expecting otherwise will lead to more failure, and the data above serves as a reminder.  Moreover, the frenetic nature of the hospital and office, along with a focus on coding for record keeping and bills–not for clinical info exchange for the purposes of care, also contribute to the status quo. The “arrangement” worked splendidly forty years ago, but no longer.

Because of the multifactorial nature of the discharge dysfunction, some countries get some things right, but not the sum of it, and consequently, we have a collective outcome fail.

Preparing a discharge summary entails time and support.  Our institutions require money and knowledge to transform the latter–and both are in short supply, here and abroad. However, until we automatize, auto-populate, e-send, fax, distribute, and utilize trained personnel to assist in managing transitions, we will underachieve.  Administrators and medical directors mistakenly believe information transfer entails one person, the provider. They are incorrect.

As discharge communication goes, our system serves us like a bicycle serves a fish: less than choice.  We bear some responsibility, but so do our hospitals.  Our institutions cannot plead the fifth, because the results do not allow the option.

Like procedure bundles and checklists, institutions must endeavor to reinforce our efforts and avoid beholding transitions as an n=1 activity.   Short of this, our patients will receive short shrift and we will perpetually underperform.  Care transitions, and provider to provider communication, does take a village.  Hopefully, these studies and others will aid in their construction.