Everyone wants to insure quality.  Quality is job #1.  Who can be against quality?  We should measure quality, report it, reward it, and publicize it.

But what is quality medical care?  Can we measure it easily?

Many argue that we can measure quality.  My friend and colleague, Bob Wachter, argues for quality measurement – On Quality Measurement, Babies, and Bathwater. 

As for me, I’ll keep critiquing bad measures and pointing out when new science emerges that changes how we should think about existing measures. But I’ll continue to support thoughtful implementation of transparency programs and experiments using P4P. From where I sit, of all our options to meet the mandates to improve quality and safety, tenaciously clinging to a Marcus Welbian (and demonstrably low quality) status quo or creating tests that can be passed by appearing to be working on improvement seem like two of the worst.

I have often written about the problems of quality measurement.  But given that I always have new readers, I will reiterate my position.

Imagine that you are picking a general internist.  What attributes would you desire?

As I consider my ideal internist, I would expect excellence is several domains:

1. Bedside (or chair side) manner
2. Appropriate time for my appointment
3. Accurate, complete history taking
4. Ability to perform physical examination
5. Understands lab test interpretation
6. Excellent medical knowledge
7. Diagnostic acumen
8. An understanding of the natural history of disease
9. Knows when to consult other physicians
10. Evidence based treatment of existing conditions

Now I may have not completed the list, but you get the idea.  Quality, if definable, is a multi-dimensional concept.  Yet most quality measurement (more accurately performance measurement) focuses just on dimension #10.  Now it gets even more complex.

As a physician I care for 100 patients this week.  25 of these patients have measurable treatments.  So in a report card, I am measured on only 1/4 of my patients.  Now it gets even more complex.

As a physician, I am often measured according to patient results (intermediate outcomes), such as HgbA1c, BP, LDL cholesterol.  So we have even another variable – the patient.  I am now responsible for the patient’s adherence.  Patient adherence has multiple influences. 

Now, I am not against an attempt to improve my measurable characteristics.  I try to provide the best possible care.  But we should use measurement on a personal level, and only when that measurement has proven to make a difference.

Wachter provides multiple examples of failed performance measures.  Those who favor performance measures explain the failures as errors in development.  Prominent advocates have argued that we do not need evidence if efficacy prior to adopting a performance measure.

This attitude would invoke laughter if it was not so dangerous.  The performance czars believe in measurement, and cannot imagine that it will not work.  I argue that there viewpoint is askew.  These bean counters do not really understand the complexity of medical care.  They live in a one dimensional world, and cannot imagine the other dimensions.

They forget Einsteins’s admonition – Not everything that can be counted counts, and not everything that counts can be counted.  They have not studied Onora O’Neill – In theory again the new culture of accountability and audit makes professionals and institutions more accountable for good performance. This is manifest in the rhetoric of improvement and rising standards, of efficiency gains and best practice, of respect for patients and pupils and employees. But beneath this admirable rhetoric the real focus is on performance indicators chosen for ease of measurement and control rather than because they measure accurately what the quality of performance is.

I will continue to rant against performance measurement except as a self improvement tool.  When we use performance measurement, we distort physician practice.  We provide too much focus on 1 dimension of quality, and thus the other dimensions must suffer. 

This issue will continue to anger me, because the lack of understanding is mind boggling.  I want to provide the highest quality care.  Numbers cannot capture or measure my successes or failures.

I received a comment today on my quality rant.

Even when the guidelines have retrospectively been shown to be flawed, where’s the evidence of patient harm? Did the harm outweigh the benefits? You should know better than to make an argument without these numbers.

Sounds like a straw-man argument to me. If you’ll actually take the time to talk with the folks at, say, the NQF or NCQA you might be pleasantly surprised.

First, please do not threaten – that is so unnecessary.  Dr. Groopman has quoted instances of harm.

Guidelines should be guidelines (see our JAMA editorial) not rules.  I have spoken to many quality leaders and was not pleasantly surprised.  The point that I have tried to make is that the conceptual framework is flawed. 

When I see a patient, I have a long list of tasks.  I must make the correct diagnosis (or diagnoses).  This requires a careful history, often a physical examination, and then appropriate diagnostic testing.  We teach the patient about the diagnoses; we assess their belief system and construct a therapeutic plan which fits their needs.  Sometimes they have a problem that has an associated guideline or performance indicator.  Often I try to satisfy the guideline, but sometimes the guideline really does not fit the patient.

Often we have to prioritize multiple problems, in which case the guidelines (and performance indicators) really do not help.

I have written about this often in the past.  Please reread this "classic" post – How do I judge physician excellence?

Finally, I will once again quote the British philosopher Onora O’Neill who wrote:

The new accountability is widely experienced not just as changing but I think as distorting the proper aims of professional practice and indeed as damaging professional pride and integrity. Much professional practice used to centre on interaction with those whom professionals serve: patients and pupils, students and families in need. Now there is less time to do this because everyone has to record the details of what they do and compile the evidence to protect themselves against the possibility not only of plausible, but of far-fetched complaints.

===========

In theory again the new culture of accountability and audit makes professionals and institutions more accountable for good performance. This is manifest in the rhetoric of improvement and rising standards, of efficiency gains and best practice, of respect for patients and pupils and employees. But beneath this admirable rhetoric the real focus is on performance indicators chosen for ease of measurement and control rather than because they measure accurately what the quality of performance is.
 

I cannot tell you how to measure quality, because quality is a complex amorphous concept.  I suppose there are no shortcuts to measuring quality, and simple check lists and audits do not measure quality.  What do they measure?  Perhaps my readers can tell me, because I do not think they really measure anything.

Regular readers know my opinion on P4P.  I have often expressed my concerns about calling something quality when in fact it only represents one of many dimensions of a complex concept called quality.

Dr. Groopman and colleague have an outstanding editorial which makes the case once again – Why ‘Quality’ Care Is Dangerous

These and other recent examples show why rigid and punitive rules to broadly standardize care for all patients often break down. Human beings are not uniform in their biology. A disease with many effects on multiple organs, like diabetes, acts differently in different people. Medicine is an imperfect science, and its study is also imperfect. Information evolves and changes. Rather than rigidity, flexibility is appropriate in applying evidence from clinical trials. To that end, a good doctor exercises sound clinical judgment by consulting expert guidelines and assessing ongoing research, but then decides what is quality care for the individual patient. And what is best sometimes deviates from the norms.

Yet too often quality metrics coerce doctors into rigid and ill-advised procedures. Orwell could have written about how the word "quality" became zealously defined by regulators, and then redefined with each change in consensus guidelines. And Kafka could detail the recent experience of a pediatrician featured in Vital Signs, the member publication of the Massachusetts Medical Society.

We have a vast bureaucracy which develops "quality measures."  This bureaucracy believes in a religion of performance measurement.  They believe that adherence to performance measures leads to better patient outcomes.  But Dr. Groopman offers several examples of the flawed reasoning that this religion provides.

I personally am agnostic about performance measures.  I agree with Dr. Groopman that the entire movement is causing us to focus on the wrong issues in medicine.  Given that we only have limited time to think and reason, this misplaced religious belief in performance measures must detract from the remaining issues in patient care.  We must continue to object to this incorrect health care strategy.

 For another nail in the coffin reader Dr. RW’s brilliant post – Leapfrog’s performance scores are meaningless

UnitedHealth and I.B.M. Test Health Care Plan

UnitedHealth will try giving doctors more authority and money than usual in return for closely monitoring their patients’ progress, even when patients go to specialists or require hospitalization. The insurer will also move away from paying doctors solely on the basis of how many services they provide, and will start rewarding them more for the overall quality of care patients receive.

The new approach, which is also being tested in various guises by other insurers around the country, is known as the “medical home” model of health care. Many experts hope it will prove one of the best ways to rein in the nation’s runaway medical costs, while making people healthier. The theory is that by providing a home base for patients and coordinating their treatment, doctors can improve care, prevent unnecessary visits to the emergency room, reduce hospitalizations and lower overall medical spending.

I have such mixed feelings about this experiment.  The medical home model does need evaluation, and hopefully will provide significant patient benefits.  But whenever I see that someone is going to measure and reward quality of care I shudder.

Why would I shudder?  What could be more desirable that high quality care?  How could anyone worry about that?

We should all worry about the measurement of health care quality.  As an investigator I participated in several "quality" studies during the 1990s.  These studies had several principles in common.  First, we defined a patient population who had a disease.  Then, we defined parameters of excellent care, while defining exclusions for the provision of such care.  Then we used structured data to measure physician quality.

In retrospect, I fear that in a very small way I contributed to a problem.  The problem is that we can measure things, and thus we can define quality.

But, as I have written often in the past, quality is a complex amorphous concept.  I have previously listed some parameters of quality medical care, and here rethink the problem.  Please consider these pieces of quality:

1. Making the correct diagnosis – readers know that I believe this to be the sine qua non of excellent medical care.  Measuring diagnostic accuracy remains a major problem.
2. Assessing the patient’s desires and wishes – we cannot provide high quality care unless our care matches our patient’s expectations.
3. Using the proper medications, in the proper doses, while avoiding drug interactions. 
4. Making priority decisions about each additional medication.  When patients have multiple diseases, sometimes we must put more effort into treating one rather than the others.
5. Providing comfort to our patients.
6. Appropriately asking subspecialists for their help.

I am certain that the readers will add a few more to this list.  I purposely do not include outcomes, because all I (or any other physician) can do is to provide the best evidenced based care.  Sometimes we do everything right, and the patient has a bad outcome.  As long as we have a supportable process, I cannot critique the physician.

Certainly for proceduralists, one could evaluate composite outcomes (for example, infection rate after appendectomy), but we would need adequate numbers to appropriately consider outcomes.

Will United Health emphasize all these dimensions or will they just look at HgbA1c levels and ACE inhibitor use?

The Potentially Fatal Consequences of Allowing Insurance Companies To Direct Health Care comes from the Liberal Values Blog. 

Humana thought they were following good medical practices in telling me this patient should be on ACE inhibitors and beta-blockers. Following their advice could have been fatal for the patient. Regardless of what insurance executives like Dave Snow and economists like Arnold Kling believe, treatment decisions must ultimately be made by physicians who have examined the patient and are monitoring the course of treatment.

So I am stuck at defining quality medical care.  I can define some aspects, but understand that the measurement process will distort medical practice, and possible not for the better.  If you pay me to worry about one thing, all other things will have a slight (or more) decrease in my attention. 

Readers know that I am obsessed with semantics.  I know that words are powerful.  Oft times words are labels.  Too often people use words to obfuscate meaning.  My hypothesis is that quality improvement is an obfuscating phrase.

The key here is the definition of quality.  How does one measure quality?  Opinion leaders have made the leap from performance measurement to quality.  We should work through the logic.

Researchers and "quality" gurus take guidelines and transform them into performance measures.  For example, the percentage of post myocardial infarction patients who are taking beta blockers or the percentage of type II diabetics with a HgbA1c < 7.  We have all read many performance measures.

These "quality experts" take well documented research findings and make the logical assumption that providing this "evidence based" care defines quality care.  For example, we know that lower HgbA1c correlates with better outcomes.  Therefore they assume that striving to lower the HgbA1c (and succeeding) defines higher quality care.

I submit that they need to study the philosophy of logic.  They are missing an important step.

The problem comes from the studies themselves.  Patients in studies are often much different from other patients.  As patients suffer more diseases, interpreting a performance measure on one disease must become more complex. 

"Quality improvement" implies that patients will have better outcomes.  Yet, few studies exist that show that improving performance measures really improves outcomes.

This model makes the unsubstantiated assumption that improving performance measures equates with improved outcomes, and thus we can label our efforts as quality improvement. 

We should be precise with our semantics.  We should call these programs performance measure improvement.  If we call them by the proper term, we would have much higher level discussions of this entire movement.  The movement, while well meaning, is increasing resource use without proving that we also are improving outcomes.  Perhaps we are; perhaps we are not.  We should not call these efforts quality improvement until and unless we have clear outcome data in standard patients.

I have been writing about the importance of accurate diagnosis throughout the 6+ years of this blog’s existence.  I frequently opine that the quality movement has ignored diagnostic accuracy because it is not easily measured.

Maggie Mahar tries to address this issue in two recent blog entries – The Silence Surrounding Diagnostic Errors; Part I and The Silence Surrounding Diagnostic Errors; Part II.  She does a credible reporting job, but misses a feel for the issue.  She has read articles, and can even cite data, but she has not experienced the diagnostic process.  She gets a B for effort.

I agree with her that diagnostic error deserves much more prominence.  Do a search on diagnosis on this blog, and you will find results galore.  I want to highlight my concluding piece of 2007 – db’s medrants 2007 considered – the year of quality – in which I said:

For me the most important concept is that the high quality physician thinks. Jerome Groopman wrote skillfully about this topic in his book – How Doctors Think? Here is one of many entries that I made relevant to that book – Crichton on Groopman

I believe that the big message of this book comes from the 3rd word of the title. Doctors must think. Thinking is our forte and our raison d’être. Patients come to us for answers. They expect us to figure out what the problem really is.

Unlike some of my colleagues I do not believe that we should emphasize management of chronic illness. Management of chronic illness follows from complete diagnosis. When I use the word diagnosis here I am referring to this definition: a determining or analysis of the cause or nature of a problem or situation.

Physicians make disease diagnoses, but we also diagnosis the patient’s social context, the impact of multiple diseases, and their belief systems. Designing a management strategy requires a complete diagnosis of the patient who has the disease(s).

I praise Groopman because he champions the role of physician thinking. I agree.

Now thinking takes time.  We can not do high quality medicine if we do not have and take enough time with each patient.  Quality health care requires time.  Yet our current reimbursement system incents physicians to keep visits short, and substitute testing or consultation for thinking.

Until our payment system values correct diagnosis and pays us to think, we will continue to have diagnostic errors.  Now some errors are almost unavoidable.  Diagnosis is the cornerstone of excellent medicine, and yet we have complex plans for giving bonuses for management tasks.  As Wachter said this week:

As the quality and safety movements gallop along, the need to fix Diagnostic Errors Exceptionalism grows more pressing. Until we do, we will face a fundamental problem: a hospital can be seen as a high quality organization – receiving awards for being a stellar performer and oodles of cash from P4P programs – if all of its “pneumonia” patients receive the correct antibiotics, all its “CHF” patients are prescribed ACE inhibitors, and all its “MI” patients get aspirin and beta blockers.

Even if every one of the diagnoses was wrong.

Now Maggie Mahar believes that one of the problems is the autopsy rate.  Most diagnostic errors do not immediately result in death.  This problem is much more pervasive.

We continue to have a health care payment system that rewards procedures – surgery, invasive cardiology, radiology, etc.  These are wonderful necessary adjuncts to great medicine.  But great medicine always follows from accurate, timely diagnosis.  We must reward the excellent detectives.

Allow me some time to consider my main messages over the last year. My blog has evolved significantly since starting in May 2002. Originally I focused on finding health related links. I would make some commentary, but rarely wrote essay style entries. Now I probably spend more time exploring ideas. This year I have focused my obsession on quality. I hope to synthesize my overall thoughts with this entry. Quality remains an elusive concept. We all desire high quality medicine. Many researchers believe that they can measure quality. Many insurers (including single payers like the UK) believe that they should measure quality. But quality may not be measurable. We can measure performance. We have many studies which decry physician’s performance on "quality measures." We could measure patient satisfaction. We could measure outcomes. I will assert that none of these activities can define quality. Quality is multidimensional. Physicians have multiple tasks, and address multiple concerns. If one tries to measure one or several performance measures, what is revealed represents a small part of quality. From my experience, a high quality physician excels in most of these attributes: diagnosis, doctor-patient communication (which includes eliciting information and explaining medicine to the patient), physical examination, ability to interpret laboratory tests, treating disease appropriately, asking for the proper consultations, balancing multiple diseases. For me the most important concept is that the high quality physician thinks. Jerome Groopman wrote skillfully about this topic in his book – How Doctors Think? Here is one of many entries that I made relevant to that book – Crichton on Groopman

I believe that the big message of this book comes from the 3rd word of the title. Doctors must think. Thinking is our forte and our raison d’être. Patients come to us for answers. They expect us to figure out what the problem really is. Unlike some of my colleagues I do not believe that we should emphasize management of chronic illness. Management of chronic illness follows from complete diagnosis. When I use the word diagnosis here I am referring to this definition: a determining or analysis of the cause or nature of a problem or situation. Physicians make disease diagnoses, but we also diagnosis the patient’s social context, the impact of multiple diseases, and their belief systems. Designing a management strategy requires a complete diagnosis of the patient who has the disease(s). I praise Groopman because he champions the role of physician thinking. I agree.

Now thinking takes time.  We can not do high quality medicine if we do not have and take enough time with each patient.  Quality health care requires time. Yet our current reimbursement system incents physicians to keep visits short, and substitute testing or consultation for thinking. I believe that the best way to save health care dollars is to invest in excellent primary care. Our best hope in 2008 comes from the retainer medicine movement.  While we probably will not adopt this model for all patients, we must learn from the model.  We must learn to optimize patient panel size, and allow physicians to deliver the care that patients’ desire.  We need to make telephone calls and emails a routine and desirable activity. I believe that well trained generalists , given appropriate time, can minimize unnecessary testing and consultation.  This will not happen unless we elevate their job financially and emotionally. 

We must abandon the gatekeeper image, and elevate the Happy Hospitalist’s more accurate conductor image. The most important attribute of the excellent generalist comes from a consideration of the whole patient.  We need the partialists – those who know a great deal about one organ, but too often they forget that the organ represents only one part of a greater organism. I have great hope as this year ends.  We have a new focus on defining quality.  Most physicians have rejected the P4P concept – and (as Dr. RW so nicely summarized recently) increasingly studies show that performance measurement does not lead to improved outcomes.

After 5′½ years of blogging, I have learned a great deal about expressing ideas.  My writing has improved.  I continue to enjoy the many comments that I receive.  I love engaging in debates with other bloggers.  Finally, I believe that my blog has some influence in the community of ideas.  I hope that I have influenced you, or at least made you think this year. Happy New Year!!!