The Inspections Will Continue Until the Quality Improves...

Not-so-long ago, on this continent, one of us was Chief of Staff at a local Columbia Hospital when the Joint Commission for Accreditation of Hospitals (JCAHO or here, JC) appeared for a scheduled survey. Surveys were a little different then; they were announced long in advance, and, if a hospital did exceptionally well, they would receive accreditation with commendation; sort of a gold star. RFI's (recommendation for Improvement) were called  "Type 1" recommendations, and were taken as show-stoppers by all parties, requiring follow-up to ensure correction/compliance. Enough Type 1s and you lost your accreditation, which meant financial ruin. At that time, the Columbia hospital chain owned over 300 hospitals, and had convinced the Joint Commission that their operating practices were so different from other hospitals that they needed their own dedicated surveyors. During the inspection at issue, the physician on that surveying team insisted on a type 1 recommendation based upon his sense that medical staff leadership was marginalized by administration (or some such- the actual wording was never shared with the clinicians). Subsequently, administration announced that this type 1 recommendation had been "rescinded" and that the hospital had received "Accreditation with Commendation." Rumor circulated that said physician was henceforth dis-invited from surveying Columbia hospitals...

The JC was once a consensus building organization that followed the lead of its subscribers and accomplished a great deal. When Congress conveyed "Shall Deem" authority for medicare certification upon the JC, it wrested control of the JC from it’s subscribers and transformed it into a regulatory limb of congress and CMS. While the JC is undeniably under the direction of honorable people, their efforts have become distorted by the politics of money and power which surround them. It's no secret around hospitals that practitioners see no connection between the Joint Commission process and quality of care. Indeed, every practitioner understands that some of the worst hospitals attain JC accreditation almost effortlessly, while some of the best struggle to maintain their certification. Historically, JC inspection was centered on physical plant and policy/procedure. Dreadful care was fine, as long as the policy and procedure manual was up to date and concordant with the most recent guidance. Joint Commission accreditation was, and is, a high stakes game, and unfavorable decisions are very likely to be contested in court (or with the threat of litigation). Consequently, JC regulatory activity has progressively focused on inspection activities that can withstand such litigation. This trajectory has relentlessly uncoupled Joint Commission inspection and accreditation from even a remote relationship to quality of clinical care.  High scores and necessary accreditation have become contingent upon putting up a temporary facade of strict compliance, which frequently obstructs, rather than enhances, care. For the past two decades, the closest thing to a Potemkin village in the American culture has been a hospital preparing for a JC survey.

Since clinicians are primarily focused on care of individual patients, incentives are significantly misaligned between those clinicians and facility administrators. Those administrators have the unenviable task of reconciling the absolute need for accreditation (upon which most insurer reimbursement depends) with the uncooperative ambivalence or even hostility of clinicians who see the survey as a "Chinese fire drill" disrupting routines of care, and diverting attention from more relevant clinical concerns. This is exacerbated by the JC's frequent utilization of tin-eared apparatchiks as surveyors.

Over the past decade, other certifying organizations have arisen to compete with the Joint Commission. In response, the JC has attempted to diversify its portfolio, embracing clinical quality and safety as meaningful additions to its mission. The Joint Commission has had limited success with these endeavors. Why? Because, with the mindset of inspectors, they are constitutionally incapable of this transformation. Just as putting on a white coat does not make you a clinician, declaration of intent does not transform the JC from an inspecting organization into a quality/safety organization. There is an enormous amount to know about both, and the Joint Commission has struggled (along with all of health care) in even understanding where the state of the art currently resides. There is perhaps no better example of this struggle than the Joint Commission’s Sentinel Event policy, which has been in force for more than a decade, been through multiple revisions, and has generated almost no meaningful reporting. Why? Because, even with its Sentinel Event policy, institutions feel threatened by the JC. Thus, the only events reported are those unlikely to generate any regulatory interest. Almost always, the first hospital discussion of a sentinel event is one that  justifies classification of the particular event as not reportable to the Joint Commission. As a result, in a world filled with sentinel events, the Joint Commission’s database has failed to capture most of them.  Of all of the entities in health care, none is currently better positioned than the Joint Commission to study, analyze, and learn from such events and to widely distribute the lessons learned. This lost opportunity is staggering. Example? Wrong side surgery.

Wrong side surgery happens very rarely; but, in a country of 300 million people, it happens regularly. The JC is determined to change this, and has developed its ‘Final Verification’ Protocol to extinguish the problem. The outcome? Absolutely no measurable change. None. Zero. Why? Because they didn’t go out and study how such failures occur. This would have required a different sort of approach; not a regulatory focus, but an investigative one; an approach that requires intellectual resources, specialists, outside expertise, and the insight that mere proscription is insufficient. It could be done given the will and vision, but it would require major transformation of the JC culture. Wrong side/site surgery happens because it is very tricky to prevent 100% of the time. Preventing it requires more than a mandate to fill out a form (indeed, only people disconnected from bedside care could imagine that this could be effective).

The irony is this: the Joint Commission has worked hard to develop a comprehensive database that catalogs such sentinel events but has not developed an appropriate infrastructure to understand how such events happen; they are not process savvy. This lack of understanding is the root cause of the failure of final verification. Thank goodness that the National Transportation Safety Board (NTSB) does not take a similar approach to aviation accidents. This is important. For thirty years, healthcare  quality efforts have been primarily modelled on the manufacturing industry; Deming, six sigma, Total Quality Management. That's what the consultants have been selling, that's what the healthcare business-people have been buying. Wrong model. That's a production-oriented measurement philosophy.  That's not what the NTSB does; their model is based on an intimate understanding of process, and how it fails in specific instances. The NTSB is deep with Human Factors Engineers, and the first object of attention in any flight mishap is the recorder; the detailed process record stored in a virtually impregnable, beacon alerting box. The healthcare environment needs something besides "widgets-off-the-line" thinking to help it improve the very difficult business of providing care to patients, and at present, that necessary something is not to be found within the Joint Commission, nor do they appear to be heading in a promising direction. But, as the saying goes, "you can't beat something with nothing."

Fortunately, the University HealthSystem Consortium is just such a something. As the name implies, UHC is a group of academic health systems collaborating to advance systems of care that make clinical and economic sense, guided by data and ongoing experience. They are attempting to elevate care through careful understanding of the processes involved in the provision of bedside care, and by helping institutions deal with the daunting logistical effort required to support that care. In this effort, they are enlisting the help and input of participants at all levels in the care chain. All of this stuff is hard; much harder than it appears to outsiders, who imagine that caregivers should instinctively know what to do. Ask any clinician how they define quality, for instance, and you are likely to get the Potter Stewart answer; " I know it when I see it." (Justice Stewart was referring to pornography, but never mind that...) Although it is valuable, it's not sufficient  to drive improvement. The truth is that the state of the art is elusive, variable, and continually evolving in ways that are difficult to perceive or explain. In UHC, everyone participating, a self-select group, sees the floor, and is trying to get further away from it. Amongst academic practitioners, UHC participation carries far more weight than Joint Commission accreditation.

Quality, like the proverbial elephant, has a radically different feel depending upon which blind man you are and what part of the elephant you are touching. If quality were easy to understand or measure, very little would have been published about it, and no one would have been able to build a career in healthcare founded upon it. For a quick introduction, here are three resources. The original framework for discussions of medical quality was exhaustively laid out by Donabedian in 1966 in the Milbank Quarterly. From a modern population perspective, Berwick et al have distilled two decades of original work by them and others into a nice summary, branded as "The Triple Aim." From the individual patient perspective, a particularly pragmatic definition can be found in a document published by the AHRQ. Since what constitutes quality in healthcare remains a matter of discussion and dispute, is it any wonder that quality improvement is a difficult issue?

Quality improvement is at its core a translational activity. It imports ideas from other domains, maps them to the terrain of clinical experience, and tries to find a better, safer, cheaper, path through the jungle of clinical medicine. The current state of the science of quality improvement in healthcare is woefully incomplete. Our understanding is more akin to Aristotle's primitive notions of earth elements than it is a systematic understanding; we have neither attained the insightful scope of the theory of evolution, nor the immense power and detail of molecular biology. Most good ideas for quality are doomed to fail, mostly for reasons that are only obvious in hindsight (there are a few prescient practitioners who can see these failures prospectively). Progress on quality is arduous, and indeed does require the kind of deep understanding (or serendipity) required for progress elsewhere in medicine. Perhaps the biggest obstacle to progress in quality is that very few people, inside or outside medicine, truly understand and believe this. Until the actual process of how to investigate quality and its improvement undergo fundamental intellectual advancement, our efforts will be inefficient and disappointing.

In the meantime, the inspections will continue until quality improves.

This post was co-authored by Mike O'Connor and Mitch Keamy.

Comments

[James Goldberg]

IF YOU THINK THE JOINT COMMISSION GIVES A DAMM ABOUT QUALITY, YOU ARE OUT OF YOUR MINDS....HENCE CONGRESS HAS JUST REVOKED THEIR DEEMING AUTHORITY.

THE JOINT COMMISSION HAS ONE THING IN MIND....EXTORTION.

JAMES GOLDBERG
SAN FRANCISCO
415 686 7139

[mkeamy]

I don't believe extortion is a JCAHO motive; but I do believe that their 40 year monopoly on "deeming authority" has resulted in a certain complacency. My understanding of the change you describe comes from this blog;

http://msspnexus.blogs.com/mspblog/2008/07/joint-commission-loses-unique-deeming-authority-provided-by-law.html

and suggests that regular, recurring accountability will now be a feature of their relationship to Congress. Is this good or bad? Well, since the status quo is not particularly inspiring, change ought to be good, right? On the other hand, this will bind JCAHO more tightly to the whims of political power, further decreasing what little autonomy they possess, and the move opens the field to new players willing to do the political bidding of whoever is in charge of meting out contracts; we have seen how such ideological patronage contracting has served the country in the last years...

[Scott Hodson]

Many American health systems are significantly underinvested in quality management Infrastructure, Process, and Organization. To achieve breakthrough improvements in quality, patient safety, and resource utilization hospitals and health systems must develop a "world class" quality management foundation that includes:

Strategy: including a clear linkage of quality and patient safety to the organizational strategy and a Board-driven imperative to achieve quality goals.

Infrastructure: incorporating effective quality management technology, EMR and physician order entry, evidence based care development tools and methodologies, and quality performance metrics and monitoring technology that enables "real time" information.

Process: including concurrent intervention, the ability to identify key quality performance "gaps," and performance improvement tools and methodologies to effectively eliminate quality issues.

Organization: providing sufficient number and quality of human resources to deliver quality planning and management leadership, adequate informatics management, effective evidence based care and physician order set development, performance improvement activity, and accredition planning to stay "survey ready every day."

Culture: where a passion for quality and patient safety is embedded throughout the delivery system and leaders are incented to achieve aggressive quality improvement goals.

My firm has assisted a number of progressive health systems to achieve such a foundation, and to develop truly World Class Quality.

[cory]

Your argument that the Deming model is not particularly applicable to medical care makes sense - but he and his disciples do make some important points to learn from. And one of them is defining quality.

Every discussion like this invariably brings up the definition of "quality". Then invariably we meander off into how hard it is to define (you guys did it in this otherwise sharp discussion) .

Deming and Frederick Taylor before him defined "quality" in a uniform and useful way - and one perfectly applicable to our discussion.
Quality is essentially conformance to standards. Makes sense. In this respect the JCAH is right- that's what they look for, conformance to standards.

The problems with the JCAH you identify, and problems they are -the Potemkin Village reference is perfect - is not their demand for conformance to standards but the standards they set up have very little to do with clinical care.

This is the problem we face - not defining quality but identifying the standards we desire. One of the reasons you anesthesiologists have been more successful than other specialties is your ability to identify standards for certain important clinical events- e.g. misplaced intubations. But again there is more of a uniformity in certain OR regimens than in other clinical areas (not to diminsh your efforts). So it is easier to define the numbers you have, and the ones you want to reach for events that are actually important.

It's not always easy in clinical care to identify standards -it is doable and should be done but unfortunately hospitals and clinicians don't keep internal statistics about clinical events very well. (Example - ask anyone in the hospital what the overall mortality rate in their hospital is within a half a percent. Of course by itself it doesn't convey much information but it's a basic statistic of what we do - I'm willing to bet no one you ask, clinician or administrator will know. NOT ONE PERSON will know the overall mortality rate in their hospital. Seems unbelievable but in my years I never met one person who knew, unless they were guessing. Ask them how they know if they give you a figure).

You might say that's not important. I would disagree based on my experience but regardless, that's just an example to show how primitive our knowledge is. Without simple internal statistics, how can anyone identify standards they should conform to? Case-mix and practice variability make it hard to identify standards for clinical care but we will not get very far with any nonuniform procedure without much better statistical reporting from every hospital about what they do.

In essence it's not defining quality that's the problem - it's identifying the standards that go into actually attaining quality. At that point the model you describe will actually work in getting us where we should be.

[Mitch Keamy]

Cory and Scott; each of you, in your admirable idealism, misses the gist of the argument. Cory, the reason we didn't discuss the nitty-gritty of what constitutes quality is because the argument is about the need to address techniques of quality improvement. I have no doubt that a list of standards could (and should) be established (no wrong sided surgery is just such a standard), but as the saying goes, "if wishes were horses, beggars would ride." But imagining that articulating that list would eventuate in improvement is optimistic; progress towards this goal will require the acquisition of basic understanding in healthcare quality improvement that is much more advanced in air transport, and which requires a tremendous intellectual commitment; not just the application of administrative tools and imperatives. Or, put another way, knowing where we want to go is not the same as being there. We just don't have the ruby slippers. Existing TQM-like technique may tell us how we are progressing, but they won't tell us how to get there; it's like a swim parent shouting "swim faster, swim faster" to their kid, wwithout any idea how he/she might accomplish that goal.

And Scott, while your finely crafted words reveal an obvious commitment to the goal of accomplishing higher quality, I just don't believe that the technology exists at this time in healthcare quality management to realize that lofty (and profitable) ambition. Understand that by technology, I don't mean computers and software; I mean procedural and behavioral approaches based upon a deep understanding of the process factors that lead to errors and suboptimal quality in healthcare. Just compare NTSB work product with healthcare system quality interventions. The difference is striking. They have it right. We don't. That's why commercial air travel is so darn safe, and healthcare is not. Yet.