To discuss the role of complementary and alternative medicine (CAM) in family practice we must of necessity address several theoretical issues - the nature of science, the prevailing philosophy/ideology in Western medicine, evidence-based medicine and the ethics of the patient-physician relationship. However, even more important in debating the role of CAM in family practice is the following question - is this an academic discussion of abstract concepts and ideologies occurring in a clinical vacuum or are we discussing therapies which can have profound effects, both positive and negative, on our very real, very alive patients? As practicing physicians we have a moral obligation to become involved with CAM because it can and does profoundly affect the lives of our patients. The question then becomes how should physicians practicing both the science and art of medicine become involved with CAM and to what extent? 

The Nature of Science
Francis Bacon (1561-1626) and Rene Descartes (1596-1650) were two contemporaries who through their writings helped shape the process of scientific inquiry as we now know it. At the beginning of book II of the Novum Organum, written in 1620, Bacon states that "It is the task and purpose of human knowledge to discover the form of a given nature, or its true specific difference, or nature-engendering nature, or source of emanation."1 Bacon's use of the word form is clarified later - "For when I speak of forms I mean nothing but those laws and definitions of pure actuality, which govern and constitute any simple nature, such as heat, light, weight, in every kind of material and subject that is capable of receiving them. Therefore the form of heat or the form of light are the same things as the law of heat or the law of light."2 In other words, nature is to be understood by the simplest building blocks of knowledge - the incontrovertible truths (pure actualities) which are based on the objective, measurable, non-negotiable properties of matter and energy. 

Descartes, in Rules for the Direction of the Mind, written in 1628, enumerates several rules for methodically discovering the truths of nature. Rule II states that "We should be concerned only with those objects regarding which our minds seem capable of obtaining certain and indubitable knowledge. And so, in accordance with this rule, we reject all knowledge which is merely probable, and judge that only those things should be believed which are perfectly known, and about which we can have no doubts. Now from all of this it is to be concluded, not that arithmetic and geometry are the only subjects to be studied, but only that in seeking the correct path to truth we should be concerned with nothing about which we cannot have a certainty equal to that of the demonstrations of arithmetic and geometry." In discussing Rule VII Descartes says that "It is only concerning genuinely simple and absolute matters that we can have absolute knowledge." Later in his commentary on Rule XII he adds "All human knowledge consists of this one thing, that we perceive how these simple natures combine to produce other things." 

The emphasis on the pure actuality of "forms" by Bacon and absolute truths of "simple natures" by Descartes was revolutionary at that time. Scientific inquiry had been subject to the prevailing dogmas of theologians, philosophers and institutions. Bacon and Descartes effectively liberated rational inquiry from the confines of dogma to pursue verifiable truths. The theologians and philosophers were now the ones who were constrained - theirs being the domain of the non-verifiable. While this was a welcome and necessary change which allowed for the advancement of science, over the centuries rational analysis has evolved into scientific reductionism. Taking the logic of Bacon and Descartes (absolute knowledge of the simplest components to understand the whole) to ridiculous extreme leads us to conclude that humans are little more than complex machines composed of the sum of their parts. Bacon and Descartes probably never anticipated that their thoughts would be taken to the extreme of justifying the separation of mind and body and the consequences of that separation.

We now know, incontrovertibly, that our thoughts affect our physiology3 , through neuropeptides and the hypothalamic-pituitary-adrenal axis4 and consequently that the separation of mind and body is an example of bad science based on the faulty assumptions of scientific reductionism.

Medical Philosophy/Ideology 
The Western biomedical model is based on scientific reductionism. The body is treated as a mechanical system with the annoying tendency of responding to the whims of an intangible mind. Therefore to study the workings of the human body we have devised tools that attempt to eliminate the confounding effects of the unruly mind. 

The tool of choice for some time has been the double-blinded, placebo-controlled study. This tool has several assumptions and limitations. The primary assumption is that all extraneous variables can be controlled. If we were studying robots, genetically controlled strains of mice or lobotomized human clones, this might be true. However, since we are dealing with people, each with their unique variations in genetics, biochemistry, physiology and personality this tool will fail to detect clinically significant outcomes in certain subsets of the studied population. For instance if 20% of a population responds to a therapy but we won't recognize an effect unless there is more than a 30% ("placebo") response we have overlooked a valuable therapy. In effect we are guilty of scientific discrimination against that minority subset of the population - perhaps the basis of some future ERA suit! In fact studies show that the placebo effect5 can be as high as 70-90%6 , much more than we had previously assumed and in many cases stronger than the "scientifically proven" therapies we prescribe. The logical conclusion then is that any study which fails to prequalify and stratify patients on the basis of their optimism or pessimism will be saddled with uninterpretable data. Patients may be blinded as to whether or not they have received a placebo or "the real thing" but if they are optimists they will expect the real thing and the pessimists will expect the sugar pill regardless of what they actually received. In effect, by not quantifying the mind set of the patients being studied, we have potentially allowed a 70-90% error to be introduced into our data. 

Aside from the huge flaws of treating all patients as if they were identical physiologically and not taking into account their behavioral mindset, another major limitation of the double-blinded, placebo-controlled study is that it is incapable of studying individualized therapies as opposed to generic, one-size-fits-all therapies. A perfect example of this is a study on the effectiveness of acupuncture in asthma.7 In true blinded fashion the patients did not know whether they received stimulation of real or sham acupuncture points. All patients who received stimulation of real acupuncture points had the exact same points stimulated. Traditional Chinese Medicine (TCM), from which acupuncture is derived, recognizes at least seven major types of "asthma" with many minor variants. The choice of acupuncture points to be stimulated in TCM depends on the individual and the unique pattern that they demonstrate. This study is therefore not only an insult to anyone who has the slightest knowledge of TCM, its conclusions are invalid because it applied a reductionistic tool to a non-reductionistic therapy. 

Clearly the philosophical and ideological underpinnings of Western medicine are flawed. Constitutionally we may all have the same rights but we are not all equal when it comes to matters of physiology and attitude. While we desperately need science to advance the frontiers of medical knowledge our philosophy needs to more closely approximate that of Hippocrates (460-377 BC) who said "It is more important to know what sort of person has a disease than to know what sort of disease a person has." 

Evidence-Based Medicine
Despite limitations in the scientific method, as we now know it, and the limitations of scientific tools, as we now use them, there are things we can be sure of. We continue to seek these incontrovertible building blocks of clinical knowledge even though we are assailed with estimates8 and reports9 that only 15-30% of medical interventions are supported by solid scientific evidence. 

The recent medical literature is rife with examples of widely used medical therapies which have not been convincingly substantiated by our own standards. Two hundred years of using digoxin for heart failure was finally "validated" in a 1996 study of 7,788 patients reported at the annual meeting of the American College of Cardiology.10 The study showed a 14% reduction in death rate due to worsening heart failure and a 25% reduction in death rate of first hospitalizations for heart failure. The same study also uncovered a 12% increase in death due to presumed arrhythmias and a 26% increase in deaths from presumed myocardial infarctions! This hardly seems like a ringing endorsement! The routine treatment of acute maxillary sinusitis11 and acute otitis media12 with antibiotics has not been proven to consistently improve outcomes. Even the common practice of treating fever with antipyretics has been brought into question with evidence suggesting this may increase the duration and/or severity of certain infections.13 

If we were to only use therapies proven beyond a shadow of doubt we truly would have very little to offer our patients. Just because many of the therapies haven't been proven doesn't mean they aren't effective and don't have an important clinical role. Unproved is not the same as disproved. We need to remind ourselves of this fact when we talk about the use of CAM therapies. Neither we nor our patients can afford to do nothing for the next 200 years while we wait for academicians to try to prove the obvious, or not so obvious. In the real-life gap between practice and proof we need to curiously inquire, keenly observe and cautiously balance the risks and benefits of all therapies while we humbly admit we don't have all the answers. To quote Sir William Osler, "In seeking absolute truth, we aim at the unattainable and must be content with finding broken portions." 

Evidence-based medicine is the remedy for Western medical hubris and frees us from the prevailing dogmas and philosophies just as rational analysis did for Bacon and Descartes. CAM and Western medicine must now both face the same difficult challenge of proving what is thought to be true. Since Western medicine has defined the rules of evidence, albeit erroneously, we have somewhat of a head start. 

The question is if the claims of CAM are proven by our rules of evidence - what do we do with that information? Some would argue that when it is proven it ceases to be CAM and it is now simply conventional medicine. This is akin to the child who says about toys - "If it's good, it's mine. If it's broken, it's yours. If you fix it, it's mine." However, more often than not, the reaction is to ignore or criticize the research - imposing a higher standard of proof than we require of ourselves. A case in point is homeopathy which sticks in the craw of most physicians. Despite some well done14 studies15 and a meta-analysis16 showing that there are instances in which the response to homeopathic remedies cannot be explained away by the placebo effect most physicians resort to the argument that it can't possibly be true because this would defy the laws of physics. I would argue which laws of physics? The mechanical Newtonian physics of last century, which are relied on heavily in the reductionistic biomedical model, or the quantum physics of this century which have been all but ignored by Western medicine? Perhaps we should question our assumptions. The third reaction often seen among physicians confronted with evidence of the efficacy of CAM is to say "That's nice. I agree," and then go on practicing medicine as always, as if this proven therapy had no clinical relevance. This is, for the most part, the reaction of physicians to the inarguable evidence in favor of mind-body medicine. Placebo is still treated as a four letter word in many medical circles. After all real doctors cure patients despite their beliefs. While the placebo effect is the researcher's foe it should be the clinician's friend! As the old ad campaign used to say, "A mind is a terrible thing to waste." And yet, this is precisely what we do when we systematically treat our patients as if what they think were entirely irrelevant!

Evidence-based medicine allows us to pursue truth vigorously while admitting that the majority of what we do in day to day clinical practice has not been proven. This should not leave us feeling impotent, but rather humbled and excited by the vast frontier of knowledge awaiting our exploration. When confronted with credible evidence of the efficacy of CAM therapies we cannot legitimately hide behind the facade of science, change the rules of evidence or willfully ignore the evidence as if it were irrelevant. 

We have to play by the rules! 

Patient-Physician Relations
Ethics Several studies have documented the widespread use of CAM by patients. The most often quoted study on this subject is that of Dr. David Eisenberg, et al.17 In that study one in three Americans had used some form of unconventional therapy - loosely defined as "medical interventions not taught widely at U.S. medical schools or generally available at U.S. hospitals." The other even looser, but perhaps more illustrative, definition is "US," (the scientific ones), and "THEM," (the unscientific ones). As unfair and ridiculous as this derogatory definition is, it is nonetheless how many physicians interpret and react to the Eisenberg article. With this all too prevalent attitude it is no surprise that the study showed 72% of patients who were using CAM did not tell their physicians that they were doing so. This creates a potentially dangerous situation where the right hand doesn't know what the left hand is doing. 

The use of CAM by patients is on the increase with a recent Gallup poll documenting a 70% increase in the use of herbal medicines in the last year alone. Is there evidence of efficacy for herbal medicines and other CAM therapies? Absolutely. Is there a potential for danger when herbs or other CAM therapies are used inappropriately? Unquestionably. Can there be interactions between CAM therapies and Western medical therapies? Routinely. Is there a lot of questionable and sometimes dangerous information about CAM in the popular press and on the Internet? Yes. Are there some ill-informed, dogmatic, unscientific and perhaps unscrupulous CAM practitioners? Unfortunately. Even more unfortunate is the fact that this last statement could just as easily be applied to medical doctors.

It is a grave ethical error to not talk to our patients about the use of CAM. If we talk down to patients on this subject with an "us versus them" attitude we will drive them into the subterranean 72% who don't tell their physicians about what CAM therapies they are using. The latter ethical error would be worse than the former. The ethics of the patient-physician relationship dictate that we need to pull our heads out of the sand with respect to CAM and have at least some minimal level of involvement in this arena.

"Passive" Involvement
The first and most basic level of involvement which cannot even be questioned is that we need to ask our patients about their use of CAM. Health practices that at least one in three Americans are using should automatically be a routine part of the medical history. How else will we find out that the man with prostate cancer is taking bovine orchic extract, that is stimulating his tumor, on the advice of a CAM practitioner? How else will we find out about the milk thistle a patient on digitalis is taking which is causing his drug levels to be subtherapeutic? How else will we find out that a breast cancer patient is hemorrhaging from the pau dÕarco she is taking along with her coumadin? We must ask. If we expect to get an honest answer we must ask non-judgmentally or we might as well not bother asking.

The next level of involvement is having a basic understanding of the more common CAM therapies. Once we have asked patients about their use of CAM the dialogue that ensues should help patients make wise choices in their use of CAM. When the patient starts talking about kava, uva ursi, applied kinesiology and "leaky gut syndrome" will we have a blank stare or will we give impromptu advice about something we know nothing about? In either case the patient is not helped and we lose credibility. With the widespread use of CAM we should go out of our way to learn the potential risks and benefits of these therapies in order to properly advise patients on this important subject.

"Active" Involvement 
A more active, though indirect, involvement in CAM by family practice physicians would be to actually recommend or refer patients for CAM therapies in situations where there is a reasonable likelihood that they would respond favorably with minimal potential for harm. The recent NIH panel reporting that acupuncture is beneficial for a variety of chronic conditions is a case in point. I have repeatedly seen skilled acupuncturists work wonders with fibromyalgia, headaches, and many other conditions which have not responded to decades of aggressive medical interventions. On the other hand potential complications of unskilled acupuncture include infection, pneumothorax, or an embedded foreign body, among others. If we recommend CAM to our patients we should also help them find reputable practitioners who are competent, recognize their limits and feel comfortable collaborating with physicians. If a patient opts to use some form of CAM at our recommendation we should monitor the effectiveness of the therapy as objectively as possible. Much research is needed in CAM and careful observation by physicians trained in honest scientific inquiry can provide useful information in this endeavor. We shouldn't complain about the lack of research if we are unwilling to do our part. 

The most active level of involvement would be for family practice physicians to actually use CAM in their practices. A prerequisite for including a given therapy, CAM or conventional, in our practices should be scientific evidence for that therapy. As mentioned in the article introducing this topic there is research available for many CAM therapies. The quantity and quality of research can unquestionably be improved on. The question we must answer is what degree of evidence is necessary to justify inclusion in a medical practice? Perhaps we should consider the advice of Sir William Osler to "be not the first nor the last" with respect to new therapies. The last physicians to adopt a therapy run the risk of withholding valuable treatments from their patients for the sake of absolute, incontrovertible proof. On the other hand physicians who adopt new therapies with little or no evidence of safety and efficacy may waste their patients time and money and risk their health for the sake of what is trendy. 

Aside from scientific evidence another criteria for inclusion in a medical practice is practicality. Is the therapy simple, straight forward and easily incorporated in our practices or is it complex and disruptive if it is to be practiced proficiently. An example of a CAM approach that can be easily incorporated in a medical practice is the therapeutic use of herbs, vitamins and nutritional supplements. On the other hand to do justice to the practice of acupuncture a physician would need substantially more training and would also need to devote a significant part of his or her practice to acupuncture - taking away time from other activities. Can physicians perform acupuncture? Absolutely, if they choose to devote a substantial part of their practice to do it well. Otherwise, it is probably in the best interest of patients that we refer them to a practitioner who is an expert rather than us dabbling. 

Conclusion
The historical antagonism of physicians toward CAM is based as much on ideology and philosophy as on science. Evidence-based medicine teaches us humility rather than allowing us to continue our scientific posturing. Evidence-based medicine also opens up infinite possibilities for scientific inquiry and frees us from the constraints of the dogma of scientific reductionism. We need to carefully choose the right tools to reflect what is being studied. If the evidence dictates, we need to be ready to question our assumptions and expand our horizons. If we make the rules we must also be willing to play by them.

Above and beyond any theoretical considerations, the widespread use of CAM obliges us to both ask and intelligently advise patients about CAM so that they can make good health choices. We also have the opportunity, based on scientific evidence, to expand our therapeutic armamentarium to include therapies which can help improve the quality of our patients' lives when conventional tools have failed to do so. Monitoring our patients' responses to CAM interventions can only serve to further expand our knowledge base 

"The secret of caring for the patient is to care for the patient." This quote, from Dr Francis Peabody Weld, summarizes why family practice physicians need to be involved, at some level, with CAM. We need to care what they are doing outside our offices that can impact their health. In this respect, physician involvement with CAM represents caring for the patient. It is just plain, good medicine.