Statistical Independence is not necessary or sufficient for causality (and is evidence based medicine really faith based at the bottom?)

Daniel Brotman's article in the Jan 24, 2005 issue of the Archives of Internal Medicine(vol. 165 p138-145) makes important points regarding the concept of "independent risk factors". Basically,he and his coauthors remind the readers that: statistical independence does not mean causality, is context dependent ( ie in that particular data set) and risk factors may be causal even if not statistically independent. Independence is a statistical concept relying on a particular statistical model.

I remember discussing the role of elevated triglyceride values in the context of heart attack risk and downplaying its significance because I had read triglycerides were not an independent risk factor. Now, or course, clinical studies have shown the opposite. The point is that a risk factor can be "significant" i.e. important whether or not a medical publication's analysis indicates that is an "independent risk factor"

Articles like this one are important antidotes to the faith that we tend to have in the black box magical output of multivariate analysis. Few physicians have plowed through the pen and paper process of doing a multivariate analysis or even understand generally what it all about.I don't claim to. That exercise might give one a real sense of what is being done and perhaps how small variations in data input can alter the answer- changing an independent risk into one that is not and vice versa.

In regard to heart disease, the authors assert that as more variables are linked to disease, no study will be able to properly model all the risk factors to enable them to say that X is an independent risk factor. This problem of residual confounding limits medicine's search for the causes and might make us more circumspect when we make pronouncements to patients about what causes what and what we should do about it.

William Barrett In his book "Illusion of Technique"(Anchor Books, 1979) says that Logic is the only modern science that has shown its own limits by showing the limits of formal systems(through the work of Godel and others). We might tend to forget when we read " X is an independent risk factor for disease Y" that we are dealing with "provisional conclusions " extracted from "fragmentary" data. Some have make a distinction between "faith based" medicine with "evidence based medicine". Considering the faith required to believe the output of mysterious mathematical models about which most physicians readers are ignorant, this distinction begins to fade away.Maybe we need a medical version of Godel's theorem as an antidote to hubris or faith based believe in technique. I eagerly await more articles,such as Brotman's,pointing out the limits of medicine's knowledge-gathering techniques.

The New Medical Paternalism

The paternalism that was- to varying degrees- part of medical practice has largely been removed by the priority of patient autonomy and the need for informed consent. Medical ethicists seemed to agree that autonomy was important leaving little or no place for paternalism. But the paternalism as we knew it (the physician deciding what was good for the individual patient and not necessarily involving the patient in the decision) is being reborn in a much more dangerous type paternalism.

I referred to this as the "New Paternalism" in a letter to the editor of the JAMA rebutting the position of Dr. David Eddy who advocates a medical practice based on the good of the group and not of the individual patient. (JAMA April 5, 1995, vol 272, no 13 pg.996) .

The abrogation of the individual patients' welfare to the collective good is implicit in the [new] medical professionalism as articulated by the Charter of Medical Professionalism which was published in both the Lancet and The Annals of Internal Medicine. A cogent critique of this "Charter"by Dr. Jerome C. Arnett, Jr. can be found in the Medical Sentinel 2002;7 (2) 56-57. and I will not attempt to improve on his comments. Internists who have read the Annals publication of the Charter should read Arnett's paper to see what their professional organization is advocating.

What I find particularly frightening is the unabashed, explicit suggestions for implementation of this approach in the series of articles by Eddy in reply to which I wrote the 1995 letter.
Eddy (Eddy DM.Rationing resources while improving quality.JAMA 1994;272;817 824) said we must "help patient understand the consequences of a limited resource pool and the need to be fair" and he continued " in most cases the explanation might be "Well, for patients like you the appropriate approach is ...".
Dr. Robert L. Wears in his letter to JAMA said " patients in general (society) must agree to this return to a beneficent paternalism"as at least he seemed to believe that before someone would be sacrificed to the collective good, he should agree to do it. Of course "society" being a figure of speech cannot agree to anything, only individual patients could agree or disagree.
As arguably bad as the old paternalism may have been At least an individual doctor decided what was best for the individual patient.With the new paternalism the medical elite versed in the ways of outcome analysis and cost benefit analysis will decide what is good for all the patients while the individual doctor and the individual patient's assessment and values are no longer relevant let alone determinative.
Similar views to those expressed here were published- in the same general time frame as the letters to JAMA- by Chervenk and McCullough in an article in J.Clin.Ethics 1995;6:320323 titled "The threat of the new managed practice of medicine to patients'autonomy."

A Randomized Clinical Trial may tell us what to do... or not

Since it is believed that testosterone is involved in prostate cancer it seemed reasonable-biologically plausible-that a drug that inhibits the conversion of testosterone to dihydrotestosterone could decrease prostate cancer risk. So the Prostate Cancer Prevention Trial was instituted.18,000 men studied for seven years with the treatment group taking finasteride. Happily, there did seem to be a decrease in prostate cancer (a 24.8%decrease in prevalence in the seven year period of the trial) in the group given 5 mg. finasteride daily. However, there was an increased risk of high grade cancers in the finasteride group.37% of the tumors in the treatment group were Gleason grade 7 or higher versus 22% in the placebo group. The increase in high grade tumors was, of course,an unexpected finding.A decreased prostate cancer risk was expected-or hoped for- and occured. But now what advice would you give someone in regard to taking finasteride? Previously we could say finasteride could decrease prostate size and might decrease cancer risk.But now ...
"We have a drug that will decrease your risk of prostate cancer a bit but if you do get prostate cancer it may be a more malignant kind". Not a great offer.
For those who think RCTs give us clinical truth (and sometimes they probably do) the PCPT is a good example of how RCTs can make the clinical waters even more opaque.
Urologic science is undaunted. A new trial is planned (or possibly now underway). The REDUCE trial (J Urol.2004 oct:172 (4 pt 1) 1314-7) will enroll 8,000 men to take either placebo or a different 5 alpha reductase inhibitor,dutasteride for a projected 4 years.After the PCPT results what would you be able to say to eligible trial participants that would induce them to sign up? Well you do get free prostate biopsies and maybe this time there will be no surprises.

Decisions, "Society" and Thomas Sowell's Insights

Relatively few books provide insights so powerful and useful that they can change your way of looking at things. Thomas Sowell's book, "Knowledge and Decisions" ( Basic books, 1996) is such a book.

Consider the first sentence in Chapter 2 "despite the fashionable practice of personifying "society" as a decider and actor, decision making in the real world can be understood only in the context of the actual decision making units (DMUs- my abbreviation) that exist, and the specific, respective sets of constraints and incentives within which each operate".
He emphasized that the term society is a metaphor."There is no one called"society" that decides anything." He continues to say that sometimes this metaphor is used to shift the decision making from smaller and more numerous units to a single national wide unit. No DMU controls a whole society except to varying degrees in a totalitarian state.
Some may consider the government as the institutional personification of "society". Of this Sowell says "...the diversities,conflicts and disparate incentives and constraints which make "society" a meaningless abstraction as a decision making unit also make government a fragmentary aggregation of decision makers."

Discussion about various issues, e.g. "distribution" of health care, overlooks-through ignorance or intention- the fact that the most basic decision is who makes decisions and under what constraints and feedback.

Medical Journal articles abound with comments such as "society can or cannot afford this or that" . The things to remember there are: 1) society is not a decision making unit 2) issues that are addressed by the "society" obfuscation need to be thought of in terms of the incentives and constraints of the various DMUs potentially involved and 3) Government is not society (articles frequently propose or at least imply a governmental solution to the issue at hand).

The "Moral Interconnectedness" of Medical Care

Brennan (Brennan TA.An ethical perspective on health care insurance reform.Am J Law Med 1993:19:37-74 ) has asserted that a physician's decision about one patient affects all other patients.On first analysis this statement sees incomprehensible.. How, for example, does my prescribing a low dose diuretic to a person with hypertension -in any real existential way-affect a person three thousand miles away with a sore throat.Presumably we are not talking about some type of strange quantum physics effect where we are told somehow there is action at a distance.Why would we not assert -making just about as much sense-that a car repair decision in Atlanta affects all car owners?
He speaks of the moral interconnectedness of medical care. This cannot be taken literally so it must be intended in a metaphorical way. If metaphorical is it of the same genus as "it takes a village to raise a child" and "we are all in this together" and "global village" comments.? What is the purpose of evoking such a metaphor?

If one accepts these notions as something more than an absurd figure of speech,then the following proposal is not surprising. This is the proposal of "social justice" as a fundamental principle for the [new] medical professionalism as revealed in the Feb 5, 2002 annals of internal medicine, vol. 136, number 3, p 243-246.

Hayek's critique of the concept of "social justice" is worth serious consideration as physicians are now being told that it is a fundamental element of their professionalism.
The following are some of his analysis of this concept as described by Michael Novak. (Firstthings.com)
The term is vague and typically undefined."It is allowed to float...as if everyone will recognize an instance of it when it appears."
It can become an instrument of ideological intimidation for the purpose of bringing about some change desired by the speaker.

"Social justice" in the medical professionalism charter is "defined" as calling upon the profession to promote a fair distribution of health care resources. What remains undefined is "fair distribution" and the means to enforce this distribution and over what group of people should this distribution occur.

Only one published letter to the editor challenged the basic premises of the charter's authors.
When one reads the letter and the author's reply it is clear no reconciliation of views is possible.
Dr. Jerome C. Arnett Jr. wrote that individual rights and group rights are mutually exclusive;the authors reply "we strongly disagree". Arnett wrote than-other than under socialized medicine-the provision of services to one patient does not diminish that available to other.The authors reply "We also disagree". However, to his comment " How...is it ethical for a group such as the Medical Professionalism Project to weaken our code of ethics to promote a political agenda...?"The authors said only the charter is not a code of ethics. Apparently they choose to not deny that they had a political agenda.
When one talks about social justice, doesn't one always have an agenda ? And is not talk about "moral interconnectiveness" etc an attempt to build a base for such an agenda , however vacuous such a base may be?

The Internet, the wrong kind of vertigo and Hickam's Dictum

A vignette in which Hickam's Dictum trumps Occam's Razor as a patient labeled as having Meniere's Disease discovers-thanks to Google- she really had a different kind of vertigo.
A 43 year old lady developed aural fullness, tinnitus and unilateral hearing loss and is told she had 2 of the 3 criteria for the diagnosis of Meniere's Disease. Within a few weeks she developed vertigo ( now she had the third symptom) which was precipitated by head movement and lasted only a few minutes. Researching her condition on the web she learned that the typical vertigo of Meniere is episodic and may last from hours to days and the type of vertigo she experienced seemed more like that described as BPPV. She told an audiologist of the symptoms and was treated with the Eply manuever. She had been free of vertigo since that time.
Even though she had the classic triad of Meniere's she may have actually had Meniere's without vertigo-people write of such an entity- plus BPPV.
Had she not researched her condition no one might have tried the otolith repositioning manuever.
Hickam ( if there really was such a person) purportedly said " A patient can have as many diagnoses as he darn well wants", Occam's Razor notwithstanding. ( Hickam's contribution to medical diagnostic lore is a bit obscure-I only got two relevant hits on a Google search on 1/12/05)

Can "Society" Afford so many Cell phones?

Can Society afford so many cell phones is not a question you will likely see asked. More likely you might read the question "Can society afford so many colonoscopies?", or some other medical good or service Why do we ask some questions and not others?. Concern about cost and cost benefits analysis (CBA) are common now in medical publications.

CBAs are typically cast in terms of the cost to "society". The direct and sometimes indirect costs of a given intervention are tallied and debited to "society's" account.
Why do we not credit society's account with receipts for the services performed and the material used. One man's costs is another's revenue. Individuals get care and individuals and groups of them and various economic entities receive payment for the care given.
"Society" is not some third payer in the game. Society is not some super entity. Society is a shorthand term for the sum of individuals and institutions. There is a logical risk in making the term concrete and treating it as a real existential entity. Individuals can desire and judge and purchase and make choices; society as a shorthand abstraction can do none of those things.
Why are costs decried in this particular area of economic activity? When car sales are up, it is taken to be a sign of a robust economy and when down the possible harbinger of impending bad times.
Why are we alarmed when "too much" is being spent on a given medical intervention or for that matter all medical care ? Why is buying iPods good and buying flu shots not?
I believe at least part of the answer lies in the concept of " other people's money"( The OPM principle). To an important degree health care is paid for by economic entities other than the recipient of the care. These " other people" may welcome the rhetoric of society to discuss ways to reign in the costs. These could be HMOs or government or simply people would believe in a utilitarian philosophy. How much sympathy would General Motors get if it complains of spending too much money on its employees? Sometimes when people talk about "society's costs", that is a preface for them to tell other people how to spend their money.