Friday, March 30, 2007

Still more cogent critcism of P4P and guidelines that ignore risk adjustment

Two important analyses are discussed in the March 30, 2007 posting on Health Care Renewal. They deal with the fantastically, overhyped goal for hemoglobin A1c which is given a thoughtful and critical once-over as is the 4 hour rush-to-give antibiotics for patients diagnosed with community acquired pneumonia (CAP).

Dr. Roy Poses references and discusses two articles regarding these topics and raises serious questions regarding the dangerously simplistic foundational premises behind P4P.

Poses says

"... developing performances measures that will truly benefit patients will require...understanding of the clinical context,statistical analysis of the available research data and careful balancing of the benefits,harms and costs."

and calls for a detailed disclosure of all conflicts of interests of those involved in the guideline development processes.

A major problem that was discussed was the typical lack of risk adjustment in the guideline movement. For example, in regard to the glycoselated hemoglobin target, the one size fits all target is mandated for patients included even though they posses clinical characteristics which excluded them from the clinical trials which demonstrated the risk reduction benefit for lowering the A1c. In this group of patients, e.g. the elderly and very elderly and those with multiple other illnesses, the risk of hypoglycemic reactions may well negate any unproved theoretical benefit in long term cardiovascular and microvascular disease reduction. But that type of decision has to be made by a thoughtful physician in consultation with the patient taking into account the individual clinical context and the patient's values and preferences, none of which necessarily have anything to do with guidelines.

The further the decision making is removed from the rapidly- becoming- marginalized physician-patient decision making unit the more likely it is that harm will result and reasoned decisions will be pre-empted by treatment mandates that ignore the individual details that embody the patient.The overarching principles -even if they may not be driven by one special interest or another- of the guideline authors and advocates need not reflect the values and preferences of the patient.

More and more reasoned, thougtful critiques of P4P appear in the medical literature but the movement shows little signs of slowing down. I hope we are not seeing another instance of " the dogs barking and the caravan moving on" but I am afraid we are.

Does ACP's proposal for Medical Home ignore the reality of how patients run their lives?

I recently wrote about how P4P is very unlikely to work based on the way that patients run their lives and see various doctors. Exercising their freedom to choose whatever physician they want, patients may see numerous physicians within a year making it, in many instances, impossible to administratively "assign" a given doctor to a given patient.

The American College of Physicians (ACP) has proposed what they call The Medical Home and suggested that physicians be paid on the basis of a defined bundle of services and provide "patient centered care" through a "qualified medical home" in which a physician accepts overall responsibility for the care of the patient and leads a team that will provide integrated,comprehensive care.

If Mr Jones sees Dr. White, a primary care physician who recommends a treatment plan according to the very latest ADA guidelines, and then decides to see instead Dr.Black who recommends something else, who is "responsible" for his care ? What is Dr. White to do when Jones does not return for his 4 week followup,what does he do if Mr.Jones decided not to get his colonoscopy according to the latest American Cancer Society, or his eye exam according to ADA guidelines, or did not return for his flu shot which was recommended, etc etc. If later Mr. Jones injured his knee while bowling and went to see an orthopedist that his brother had recommended without letting Dr. White integrate his care and an "unnecessary" ( by some Ottawa decision rule or whatever) MRI and then arthroscopy were done, will Dr. White be responsible for misuse of "scarce medical resources"?

Here's the thing. As long as people are free to see whatever doctor or doctors they want and are free to conform with or ignore medical advice and take medicine or not the, ACP's suggested medical home ( and as mentioned before, P4P) will not work. If it could work without coercion, I would like to hear how. Does it make any sense outside of a single-payer-single tier system? Maybe that is what ACP wants.

Maybe Mr. Jones would have to sign a contract to adhere to the doc's plan or maybe there will come a time when a entity (government or a government sanctioned insurance agency) will assign Mr.Jones to a primary care doctor and permission wil have to be obtained to change doctors.

Thursday, March 29, 2007

Guidelines-the mantra "treat to goal" re-shaping medical thought?

Two things stimulated the thoughts that follow.

The first was comments made by Dr. Jerome Groopman in the first chapter of his new book " How Doctors Think" when he opined that clinical algorithms (I tend to place algorithms and guidelines in the same conceptual box) fall apart when physicians need to think outside the box and (I would add) when the framing up of an ambiguous clinical situation needs to be done. His concern was that current house officers are so guideline and algorithm oriented that perhaps that the outside the box,creative, independent thinking region of their brains may be at risk for atrophy.

The second was observations from a "CME" dinner talk about diabetes (one that was not given by Joslin).It seemed that all of the questions hinged on how to get a patient to "goal". It did not seem to be at all about tailoring the treatment to the patient or really treating the patient at all but it was about the getting the numbers right. I wondered what even happened to old house office motto of " treat the patient not the numbers".

It seems to be all about number now. the numbers of the statistics of the clinical trial, the numbers that committees set for goals that should be achieved,the number of hours in which antibiotics must be given to reach a quality goal for treatment of community acquired pneumonia,the number of minutes a physician is allotted per patient, the likelihood ratio regarding a diagnostic test. Groopman said it well:

" Today's reliance on evidence-based medicine risk having the doctor choose passively,solely by the numbers. Statistics cannot substitute for the human being before you; statistics embody averages, not individuals."

Only a thinking physician in collaboration with the patient can determine if the "best therapy from a clinical trial fits a patient's particular needs and values"

Observing house officer on round lead Groopman to the following conclusion:

" ...the next generation of doctors was being conditioned to function like a well programmed computer that operates within a strict binary framework."

Is that valid?

The emphasis on guidelines and algorithms fits in nicely with those whose vision of medical care includes a major role for "mid-level" practitioners.(This is a interesting term.Note we do not say lower level practitioners.Who would want their medical problem assessed or treated by a lower level practitioner? If PAs and NPs are mid level, for what type of practitioner would we use the term "low level"practitioner?).

The guideline-algorithm mentality also may suggest the idea of a physician as an exchangeable commodity.Dr. Stuart Henochowicz , whose whose interesting blog is MEDVIEWS, defends the position that doctors are not commodities in a recent Medscape interview.

Tuesday, March 27, 2007

The COURAGE trial-fewer PCIs down the road for stable angina

The biggest news from the American College of Cardiology meeting in New Orleans(March 2007) has to be the results of the COURAGE trial. This trial compared optimal medical therapy with the combination of medical therapy plus PCI for stable angina patients and medical therapy did just as good. This lead to headlines across the news channels proclaiming that many angioplasties are unnecessary.Full text of the report and a commentary is available from the New England Journal web site for free.

It is important to look at who was not included in this trial:

Patients with very severe angina (CCS stage iv-meaning rest angina or angina brought on by any physical activity)
Patients with very positive stress tests (meaning st changes in stage 1 and/or hypotension brought on by the exercise)
ejection fraction less than 30%
history of revascularization in the previous six months.

Patients had to have at least one epicardial coronary artery with 70% stenosis plus objective evidence of myocardial ischemia to be eligible for the trial.

The study showed no difference in mortality or myocardial infarction in the 4.6 years of observation and was consistent with earlier,smaller trials that suggested no mortality benefit to PCI but perhaps greater relief of anginal symptoms.

I found the following sentence helpful in trying to understand what may be the relevant pathophysiology:

"Unstable plaques that lead to myocardial infarction are not necessarily severely stenotic and severely stenotic plaques are not necessarily unstable ."

While the severely stenotic plaques may be a marker for the unstable variety being present elsewhere in the coronary circulation it is the latter that typically causes the myocardial infarction or acute coronary syndrome so that dilating the former may relieve angina symptoms but is not likely to prevent infarcts.On the other hand, medical therapy is thought to treat all of the coronary arterial tree and may make unstable plaques less likely to cause trouble.

PCIs will be continued to be recommended for many acute coronary syndromes and for anginal patients whose symptoms resist medical therapy but there are bound to be fewer PCIs done for stable anginal patients because of this trial.

Sunday, March 25, 2007

How doctors think and the importance of "check your premises"

Jerome Groopman's new book, "How Doctors Think" has attracted a bit of media attention replete with a favorable write-up in Time magazine.

Groopman discusses a topic of considerable personal interest and one that I have blogged about now and again. The topic of how cognitive mechanisms such as heuristics serve us well most of the time but can also lead us down a very wrong end is well is illustrated by Groopman with clinical histories, sometimes of his own medical adventures and misadventures. Placing the abstract concepts with real life clinical vignettes give them limbic valence.

Pattern recognition seems to be very important in the diagnostic acumen of experts and a well functioning pattern recognition system which develops over time and with experience seems to be a major distinguishing feature of the expert from the novice. It is useful but not perfect based as it is based on generalizations and abstractions. Generalizations are generally right but not always- so that even with optimal functioning of pattern recognition mistakes can occur.

This pattern recognition system ( a.k.a heuristics, or mental shortcuts) while serving the expert well comes with a dark side, one aspect of which is the phenomenon of "premature closure". This refers to making a diagnosis or an initial impression and then shutting out consideration that the diagnosis may be wrong even as contradictory evidence accumulates.

To mitigate this premature closure,using Groopman's words, the physician needs to:

" repeatedly factor into the analysis the possibility that he is wrong" ... "cogent medical judgments meld first impressions-gestalt-with deliberate analysis." and the physician:

"should be schooled in heuristics-in the power and necessity of shortcuts and in their pitfalls and dangers."

Good diagnosticians need to frequently remind themselves to reconsider the diagnosis as events unfold and test data comes in play asking What else could it be? What have we missed? What is the worst thing it could be?

Basically what we are talking about was nailed by Any Rand when she had her super-heroes say:

"Whenever you think you are facing a contradiction, check your premises, you will find that one of them is wrong." Atlas Shrugged.

For physicians who might want a shorter version ( Dr. Groopman's stated aim was to write a book for layman's consumption) of the key issues with cognitive errors in medical diagnosis, here is an excellent article by Dr. Donald Redelmeier.

Tuesday, March 20, 2007

Dueling ethical views in Archives Internal Medicine re: "tiered" health care systems

In the March 12,2007 issue of the Archives of Internal Medicine, we find two differing views on the ethical status of a tiered medical care system. A one tiered system is exemplified by the Canadian system in which patients are forbidden to purchase care outside of the governmental sanctioned system. ( At least in Canada, they can still medically escape to the U.S.)In England, a private medical insurance scheme thrives, existing along side of the National Health Service.

Dr. Ezekiel Emanuel and Benjamin Krohmal argue in their article that a tiered system (i.e. one with more than one tier) is "just in principle and ..just in practice". They justify their position in part by reference to Rawls principles of justice, the first of which is that "all have the right to the greatest individual liberties compatible with the same liberties for others." Since some medical goods or services must go without public funding, justice protects the liberty of individuals to pay for medical options that the public (government) fails to provide.

In a companion editorial , Dr. Allan S. Brett takes a different point of view. He argues that a second tier may not be necessary and is problematic from an ethical point of view. He says that if we eliminated administrative waste and unnecessary medical interventions then people could then receive comprehensive medical care without a second back up , paid -by the-patient- tier. Since he also recommends a single payer system-i.e government-the "if" in eliminating administrative waste become a very big "if" and then we face the issue of who decides what is "unnecessary".

Further, he suggests that "effective evidence-based interventions should be included in the first tier, and ineffective or unproven interventions should not be provided in any tier."

His use of the term "should not be provided" glosses over what he is saying.What he is saying is that purchasing services not provided in the first tier will not be allowed, i.e will be made illegal. Phrasing this in the more candid way makes his claim that his view is not antithetical to American values empty.

Dr. Brett does not elaborate on what the process would be to decide which interventions are or are not effective but one can assume that would not be done by the individual physician in consultation with the individual patient.Those decisions would be made by committees appointed by the one party payer (which Dr. Brett favors) either directly or indirectly. Ironically, the egalitarian one tier patient system would establish a two tier physician system in which the higher tier would be the elite physicians who decide what should be "provided" and in the lower tier will be the rest of the physicians who will explain to patients what they will be not allowed to have.

Friday, March 16, 2007

More data showing why P4P will not work-planners undaunted

A recent (free without subscription) article in the NEJM presents data that indicate that the reality of medical care in the U.S. is such that the P4P will not work. Many Medicare patients- and I am sure this applies to non-Medicare patients as well- do not have their medical care provided by a single practitioner.

One problem that policy-wonk, medical planners have is they are dealing with people who live in a free country who typically exercise their choice about a lot of things including who they see when they are ill. If a person is not happy with their doctor-for whatever reason-they change doctors and sometimes they change doctors even if they are happy with the current physician.

In the course of a year a so a person may see several different doctors. This poses a problem for the P4P advocates-namely who should the patient be "assigned to" in terms of the rewards or the penalties of the pay for performance system. The study authors doubt how effective P4P will be in improving quality-for those of you who believe that is about quality. They also suggest maybe we need to assign patients to a given doctor.

The authors may want to review section 1802 of the Medicare bill which said:

" Any individual entitled to insurance benefits under this title may obtain medical services from any institution,agency or person qualified to participate."

Although we can find little comfort in that assurance since the government stopped paying any attention to section 1801 of that bill long ago:

"Nothing in this tittle shall be construed to authorize any federal official or employee to exercise any supervision or control over the practice of medicine..."

Of course, the way that medical care is actually"delivered" is only an inconvenient truth to the planners-something that needs to be restructured to better fit their plans for what everyone should do. That pesky freedom thing does have a way of interfering with central planing.The companion editorial-by Karen Davis, a PhD economist and currently President of the Commonwealth Foundation- suggests ways to changes things so P4P could work. Somehow, we could encourage doctors and patients to participate in a system more like that in which P4P might work and encourage the growth of large integrated systems. Then those systems could reap the rewards of a P4P arrangement. She then admits that P4P at best would be a intermediate step-an interim program- until the day when their bigger and better plan could be realized.Stay turned for details of what that might be.

Wednesday, March 14, 2007

Momentum builds for use of aromatase inhibitors in estrogen positive breast cancer

A recently published clinical trial provides more reason to prefer the aromatase inhibitors to tamoxifen in the treatment of estrogen positive breast cancer.

There are currently three aromatase inhibitors (AIs) approved for breast cancer treatment in hormone positive breast cancer:

Anastrazole (Arimidex)
Letrozole (Femara)
Exemestane (Aromasin)

Here is a NCI website with an annotated bibliography of some of the major clinical trials involving these drugs. The role of AIs seems most clear in high risk patients (large tumors,positive lymph nodes,higher grade histological changes) while the more severe estrogen depletion side effects makes its putative superiority over tamoxifen less clear in those breast cancer patients with low risk of recurrence.

Probably more than any other internal medicine sub specialists, oncologists have been obliged to plow through and attempt to master the deluge of clinical trials and try and determine how to translate those studies into the real life, day to day, medical management of patients with cancer. Two months ago, I had the opportunity to attend a 2 hour seminar on treatment of breast cancer given by a local private practice oncologist and by a local academic oncologists. The mastery of the trial data and their advice regarding practical application to treatment decisions that these two physicians demonstrated reminded me of one of the reasons that some of us went into internal medicine -the pleasure and sense of accomplishment of mastery (or as close as one can come to that with the moving target of medical advances and the accumulation of data contradicting older data) of complicated ideas and multiple facts and putting it all together into the care of sick patients.

Wednesday, March 07, 2007

Internists doing fewer traditonal internist procedures

The results of a survey sent to 2500 general internist members of the American College of Physicians and replied to by 1389 indicate that internists are doing fewer of the procedures that traditionally had been done by that speciality.This report was published in a recent issue of The Annals of Internal Medicine.

The authors compared the results with those of a survey done in 1986. Here are some of the data.

In 1986, 66% did thoracentesis versus 23 % in 2004, for abdominal paracentesis-60% dropped to 23%, bone marrow aspiration from 27% to 8 %, flex sig dropped from 42% to 20%, insertion of central line 39% to 16%.Interestingly skin biopsies increased as did cryosurgery for skin lesions . Gram stain of the sputum decreased from 50% to 5%.

The increased numbers of IM sub-specialists,tighter credentialing procedures at hospitals,the advent and proliferation of hospitalists were some of the reasons suggested by the authors behind the changes.

Of interest to me was the companion article-an editorial from the American Board of Internal Medicine (ABIM)- regarding what procedures should be taught to physicians in training to become internists.The authors distinguished between the types of procedures residents would be taught to do versus those that they would be taught about.In regard to the latter, we are told that for certain procedures, only what the ABIM refers to as "cognitive competence" will be required. This interesting pedagogic, linguistic construct apparently means that the resident must be shown to be competent in discussing the indications,contraindications, complications etc related to the procedure and be able to obtain an informed consent for the procedure-apparently that someone else will do. Included in the set of procedures are several that many of us older internists consider basic internist activities: thoracentesis, abdominal paracentesis, naso gastric intubation and lumbar puncture.

Internists in training will be required to actually learn how to do ( as opposed to learn how to talk about) the following: venous and arterial access,pap test, endocervical culture and the basic elements of ACLS including intubation.

As best, I could determine from web sources, family practice residency programs continue to teach a wide variety of procedures including those that internists now only have to be able to talk about including thoracentesis, lumbar puncture and abdominal paracentesis.

Thursday, March 01, 2007

Communication between Hospitalist and primary care physicians

It has been so long since I functioned as a officist and a hospitalist-a role known at the time as an internist- that I read with considerable interest an article in JAMA that concluded the current state of information transfer between hospital docs and PCPs was clearly in need of improvement.

Our pulmonary group enjoyed a large referral practice from a fairly wide geographical area of our state and it was thought to be very important to do three things when a patient was discharged from the hospital: 1) give the patient a summary of his/her medications and treatment plan, 2)call the referring physician and 3) write the referring physician and send him a copy of the discharge summary.

The JAMA article extracted data from numerous observational studies and present the reader with typical number of percentages too large to really keep up with. However, here are some:

Direct communication between the hospital physicians and the PCPs occurred only 3-20% of the time. Only 12-34 % of the time was the discharge summary available to the PCP at the time of the first discharge visit. In 2-40% of the time the discharge summary did not have information about the discharge medications.

While this article puts some worrisome numbers to one aspect of the current paradigm shift of the hospitalist phenomenon, there is more insight and enjoyable reading found in a brilliant 1999 NEJM article by Dr.Farrin A. Marian, to whom I believe we must give credit for the term "officist".He defines officist as a general internist who sees patients only in the office as long they remain relatively healthy.

The economic forces at work when I was a part time hospitalist gave us incentives to maintain close and mutually beneficial relationships with our referring physicians part of which was being sure they were kept well informed about what transpired with their patients. The economic forces now at work at least in some practice setttings may be a bit different.

I have been giving thought to the question "why do we have hospitalists anyway?" I am leaning to the conclusion that this is the result of physician fees price controls put into place in 1992 for Medicare patients and the controls placed on hospital charges for Medicare patients in efforts to control the rising cost of medical care for the elderly. Hopefully more on that later.