Essentials of Complementary and Alternative Medicine (June 1999)

W
HAT
 P
OPULATION
 W
AS
 S
TUDIED
?
If quality evidence is found, the next most important step is to ask whether the population studied is reasonably similar to the patient at hand. Research involves 
attempting to bring together homogeneous groups of patients to control for, as far as possible, extraneous and confounding variables that might influence the 
outcomes. These study groups are created through a series of inclusion and exclusion criteria that may result in the study findings being applicable to only a very 
narrow range of patients. The challenge for the physician is to identify whether the patient group reported in the study is similar enough to warrant its application to his 
or her patient. Although this matching is somewhat subjective, practitioners can compare at least five areas of how a study has been conducted to help judge this 
match. These areas are:
1. Age of the study population.
2. Gender of the study population.
3. A setting similar to the patient's own (e.g., primary, secondary, or tertiary care center).
4. A culture similar to the patient's own (e.g., Western, Eastern, developing, industrialized).
5. A population in which the accepted diagnostic criteria for the condition are similar to that used in that population's system of medicine (e.g., do patients in a 
United States study meet the accepted criteria for diagnosis of diseases such as osteoarthritis or congestive heart failure?).
Other factors, of course, such as the severity of the illness and the magnitude of the patient's desire for treatment, must also be considered in the decision. If the 
characteristics of the population studied are not close to that of the patient, then it is unlikely that even valid data can be reasonably applied, because insufficient 
evidence exists for the use of that CAM therapy in that patient population. However, if the match is close, then the data provide an appropriate body of evidence for 
moving forward with a therapeutic trial.
Efficiency of EBM
One of the concerns many physicians have about practicing EBM is the time it takes to search and evaluate the literature, apply the quality criteria, and decide 
whether the data apply to the patient. Skills for efficient use of the current medical literature—rapid access, accurate screening, skilled quality assessment, and 
application—are usually not taught in medical school or residency. With practice, however, the clinician can rapidly screen and assess the literature for the likelihood 
that an evidence-based decision can be made. It is possible to make the majority of clinical decisions using an evidence-based process (
19
). Obviously, a separate 
complete literature search and evaluation procedure is not needed for every patient who comes into the office. Only in circumstances in which an alternative practice 
is already being used or in which an alternative practice is sought for the first time would such a procedure be needed. Once this procedure is done for the first few 
patients, this information can often be used for subsequent patients who have similar problems and are using similar CAM practices. If the study assessment criteria 
mentioned previously and summarized in 
Figure 5.1
 are not met anywhere along the decision path, then the patient should be informed that there is insufficient quality 
research to make an evidence-based decision about the practice. If, however, the physician finds adequate evidence, a therapeutic trial with proper monitoring may 
be warranted. The reader is referred to 
Chapter 2
, 
Table 2.4
, for guidelines on how to follow patients using CAM practices. Assurance of adequate training of the 
practitioner and quality of the products involved must be considered. The reader is referred to 
Chapter 2
 and to the descriptions of the main CAM systems in Part III of 
this book.
F
IGURE
 5.1. Decision tree for evidence-based complementary and alternative medicine.
a
 See 
Appendix B
 and 
Table 5.3
.

b
 See 
Table 5.1
 for quality criteria.
c
 See 
Chapter 2
 for management guidelines.
Balancing Belief, Plausibility, and Evidence in Medical Decisions
Although EBM is necessary to make sound decisions, the practitioner must be aware of several items requiring caution.
Randomized controlled clinical trials (RCTs) assume an important presupposition: namely, that the participants do not have any strong preference for one of the 
treatments offered. This presupposition generally is met when the RCT is an evaluation of a new conventional drug. In this case, the drug and its effects are usually 
unknown, and there are no strong inclinations among practitioners or patients. In addition, an RCT is the only way a drug can be marketed.
In CAM treatments, however, there are usually strong preferences, either by the patient, the doctor, or both, and the treatments are already on the market and in use. 
If the a priori probability is high for a patient (e.g., he or she has a strong belief in and preference for the treatment), then this patient is probably different from typical 
patients in clinical trials. Furthermore, some forms of CAM—mainly, traditional healing systems like acupuncture, homeopathy, and folk medicine—rely on a long 
history of uncontrolled experience, which lends high credibility to its application in the eyes of those practitioners. If this high prior probability on the side of the 
practitioner is met with a high prior probability for the patient—that is, when a believing patient meets with a charismatic doctor who is convinced of what he or she 
practices—then strong nonspecific effects may occur. This effect may not have been seen in RCTs and meta-analyses. Many of the effects of CAM may be 
nonspecific or placebo. Nevertheless, from the clinical (not scientific) perspective, if there is a high prior probability on the side of the patient and practitioner, a CAM 
practice may trigger self-healing responses even in the face of missing or negative evidence.
EBM is currently the best way to make decisions about general effectiveness and the specific efficacy of CAM and conventional medicine alike. However, in 
day-to-day medical practice, it may be unclear to what extent this evidence applies to individual patients and to what extent nonspecific effects are responsible for 
healing.
Most patients and physicians agree that within the boundaries of ethical practice, as outlined in 
Chapter 3
, the power of these effects should be harnessed as much 
as possible for the welfare of the patient. In this sense, the prior belief system of patient and practitioner may often turn out to be paramount and can lead to the 
effectiveness of otherwise ineffective treatments (
24
).
Given this situation, how should the goal of scientifically identifying therapeutic options be balanced against the clinical need to maximize optimal healing responses? 
The authors suggest that the physician explicitly consider plausibility and belief in the therapy of both patient and practitioner.
In conventional medicine, plausibility is usually implicitly accepted by both patient and physician, but this is not the case for CAM. Belief in therapeutic agents, even 
those with specific effects, has long been known to affect outcome; high belief and expectation enhance positive outcomes, and low belief and expectation interfere 
with positive outcomes (
25
, 
26
, 
27
 and 
28
).
A physician may believe that a CAM practice has incredibly low plausibility (e.g., faith healing, homeopathy). If a practice is far outside the belief systems of both the 
physician and patient, permitting such a practice is ethically unacceptable. However, the patient may have a strong belief in the therapy and find it completely 
acceptable. This  prior probability of belief by the physician and patient should be considered in the decision to use or allow the practice to go forward. If both the 
physician and the patient can reasonably believe in the plausibility or potential benefit of the practice, and a physician has found good evidence to support such a 
belief, then a therapeutic trial is warranted. If, however, the patient has only marginal belief and the plausibility for the practitioner is extremely low, then, even in the 
face of clinical evidence, discouragement from the practice or referral elsewhere for treatment is appropriate. In some circumstances, the patient may have a strong 
belief in the practice, whereas the physician may find it extremely unbelievable. In those situations, the physician must work with the patient to decide the best action, 
and referral elsewhere may be the best option. In conventional medicine, the disparity between plausibility and belief is rarely addressed because it is assumed that 
both the patient and physician believe in the plausibility of the practices in which they engage. Such assumptions should be explored to make sure that they are true 
for CAM practices.
Is the Diagnostic System Working?
In some cases, a conventional Western diagnosis may not be useful for the treatment and management of a patient. This occurs most often for chronic conditions with 
vague or unknown etiologies characterized by subjective symptoms and lack of general well-being. Examples include chronic fatigue syndrome, fibromyalgia, chronic 
idiopathic urticaria, and functional and psychosomatic problems. If the conventional diagnosis does not help for improving the patient's condition, the clinician may 
want to consider complete evaluation with a complementary and/or alternative system. Because these systems use different diagnostic classifications and approaches 
(e.g., deficient kidney  chi in Traditional Chinese medicine,  pitta imbalance in Ayurvedic medicine, or the sepia syndrome in homeopathy), obtaining an assessment by 
one of these alternative systems may prove useful. The reverse situation may also occur. A patient may undergo diagnosis and treatment based on a CAM system 
with little effect when the condition could be managed simply and effectively with a conventional medicine approach. The physician should be alert to both 
possibilities.
Because CAM systems are generally not studied scientifically using their own diagnostic classifications, it is rare to find research evidence using such classifications. 
In such cases, a professional consultation may be needed, and the guidelines outlined in 
Chapter 2
 should be followed. A competent complementary practitioner 
should be able to inform both the physician and patient whether the diagnostic category of the alternative system is clear and is likely to be useful.
For example, a patient with chronic idiopathic urticaria and fatigue was not helped after evaluation and treatment from an allergist, psychiatrist, nutritionist, 
dermatologist, and general practitioner. Conventional diagnostic workups were negative and therapeutic trials with antihistamines, antidepressants, steroids, and 
other treatments were ineffective or produced unacceptable side effects. The patient was sent for a complete homeopathic evaluation, which showed a clear 
homeopathic diagnostic category. Homeopathic treatment was initiated and resulted in rapid and permanent resolution of both problems.
Sometimes, however, vague symptoms are simply vague symptoms, and both conventional and alternative systems cannot make sense of them. In cases when the 
alternative diagnostic classification is not clear, a therapeutic trial is not warranted, and the physician should consult guidelines previously mentioned (
Chapter 2
).
SUMMARY OF STEPS
Figure 5.1
 summarizes the approach to evidence-based CAM. The steps (or criteria) are:
Step 1
Is the patient already using or wanting to use a CAM approach, or is an alternative sought (see 
Chapter 2
, 
Table 2.2
 and 
Table 2.3
)?
Step 2
Is the practice inexpensive and unlikely to produce direct (toxic) adverse effects?
Step 3
Is there evidence for this practice from randomized controlled trials or observational and outcomes studies?
Step 4
Does the quality of the studies meet the minimum quality criteria (see 
Table 5.1
)?
Step 5
Is the population in these studies similar to the patient at hand?
Step 6
Is the belief and rationale for the therapy acceptable to both patient and physician?

Step 7
If yes to all the above, consider a therapeutic trial provided:
A A quality product or procedure by a competent practitioner can be obtained (see 
Chapter 2
, 
Table 2.4
 and 
Table 2.6
), and
B The patient can be monitored while undergoing the treatment (see 
Chapter 2
, 
Table 2.4
).
Step 8
Also consider if a new diagnostic assessment by a CAM system is in order.
By following these steps, the physician can increase the likelihood that decisions about CAM practice are based on research evidence; or, if there is no research 
evidence, the practitioner can be clear that clinical decisions are based on other criteria (e.g., opinion, clinical judgment).
EXAMPLES OF EVIDENCE-BASED APPROACHES TO CAM
The following case examples help illustrate the use of an evidenced-base approach.
CASE 1: ALLERGIES
A 23-year-old woman comes to her practitioner for the treatment of seasonal allergies. She is in good health and has normally taken standard antihistamines and 
decongestants for her condition. However, these agents make her sleepy, so when she came across a homeopathic allergy remedy in the drug store, she tried it for 
several days. It seemed to help.
Step 1. Now she would like your opinion on whether she should continue with this preparation.
Steps 2 and 7A. The product is a combination of very low doses of plant extracts and is said to be manufactured according to the standards of the United States 
Homeopathic Pharmacopoeia–a product for which the Food and Drug Administration does provide some regulation and oversight.
Step 3. A search of the available databases comes up with two meta-analyses of randomized controlled trials and several other randomized controlled trials on the 
treatment of allergies with homeopathic preparations (
15
, 
17
).
The effect size for these studies shows that approximately 75% of patients will experience clinically significant relief, but the preparations in the studies are not exactly 
the same as the ones in the product she has found. No significant adverse effects have been reported from these dilutions, nor are any expected given the extremely 
low dose.
Step 4. A review of two controlled trials with the minimum quality criteria mentioned earlier in the chapter shows that these studies meet these standards.
Step 5. The populations in the studies are varied enough to be similar to this patient's situation.
Step 7B. The patient is very happy to have the physician monitor the treatment and switch back to conventional treatment if the alternative practice proves 
inadequate.
Step 6. However, the physician believes that the plausibility of homeopathy's effect is very unlikely.
Thus, all of the steps in establishing the evidence criteria are fulfilled except for step 6 (plausibility), with some disparity of belief between the patient and practitioner. 
In addition, the product is not exactly the same as that studied, so there is no direct evidence for this exact product on this condition. The physician must then decide 
with the patient the proper course of action (referral, limited therapeutic trial with reevaluation, etc.).
CASE 2: DEPRESSION
A 45-year-old woman suffering from depression in combination with a dependent-depressive personality disorder seeks help, primarily psychotherapy. History reveals 
a series of losses in her past: her grandmother, who was very kind and important to her, had died when she was 5. Her father, whom she admired and who doted on 
her, died when she was 12. From the time her father died, she had taken on considerable responsibility for herself and her younger sister. When she was in her 
teenage years, she started dating a young man from her neighborhood with whom she fell in love. Because he was a Protestant and she was a Catholic, a 
relationship seemed impossible, given the strong religious background of both families. After several years, the young man was diagnosed with leukemia and soon 
died. This was the same disease the patient's father had died of. Shortly afterward, she married a close friend of her deceased boyfriend who courted her but whom 
she did not really love. Recently, a son of a good friend of hers had committed suicide, which started the patient's suicidal impulses and brought her to therapy. From 
reading popular books she had taken a fancy for homeopathy and inquired whether her problems could benefit from taking homeopathic remedies.
Step 1. Under no circumstances is the patient willing to take any conventional psychotropic medication or consider seeing a psychiatrist. Verbal psychotherapy in a 
general psychodynamic framework with behavioral elements is initiated, but the patient persistently requests homeopathic treatment.
Step 2. The risk of adverse effects is low if the treatment is embedded in a general psychotherapeutic environment in which the patient is not alone with her 
experiences. If such a patient is left unmonitored with only CAM intervention, this approach would be very risky. Direct risk of adverse effects in this situation is low 
and costs are negligible, but refusal of potentially effective conventional treatment by the patient is of concern.
Step 3. There is no evidence from clinical trials or observational studies about the effectiveness of homeopathy in depression. One could consider the herb 
Hypericum (St. John's wort), a phytotherapeutic alternative that has been studied and seems to be effective as a mild antidepressive, but this is also refused by the 
patient. It is also doubtful whether Hypericum alone is sufficient because it is used only for mild depression. The patient has been reading case histories in the 
homeopathic literature about positive effects of homeopathic treatment in depression. She thinks that her case would be clear to a homeopath because of depression 
and sadness after multiple losses, which she thinks is a key homeopathic diagnostic symptom.
Step 4. The evidence for a homeopathic intervention is weak. There are no trials and no formal observational studies, only single observations and case reports that 
are very likely due to placebo and nonspecific effects—something to which depression often responds.
Step 5. It is difficult to say whether the people reported on in the case studies the patient reads are comparable with the patient here. Usually, there is not enough 
information in the case studies to make a determination.
Step 6. The evidence certainly is not enough for relying solely on a homeopathic intervention. Given the patient's refusal of other therapies and her high belief in the 
system, nonspecific effects might be enhanced if a confident practitioner is found and careful follow-up with psychotherapy is maintained.
Step 7. There is no way of starting a therapeutic trial unless a responsible homeopathic practitioner is found, is willing to work with the physician (see 
Chapter 2
), and 
uses standard homeopathic products overseen by the FDA. The physician also might consider this patient noncompliant and refuse further therapy. However, given 
the patient's propensity for alternative medicine, this may place the patient at increased risk (albeit of her own choosing).
Step 8. The diagnosis is clear. If the patient is refusing psychotropic medication (which she does), psychotherapy is the only alternative left. If the circumstances in 

step 7 for referral and follow-up exist, it is possible that additional benefit from adding a therapeutic trial of homeopathy may occur.
Follow-up: A homeopathic treatment is initiated in parallel to continuous psychotherapy. Over the course of 1 year, this brought out a variety of destructive impulses, 
sadness and mourning—feelings she had not allowed herself before. In the course of this treatment, the homeopathic practitioner says that a decisive homeopathic 
symptom surfaces. The patient relates the single fear she is most ashamed of–losing her financial and economic security and becoming poor. However, this fear is 
irrational (the family is quite rich) and is correlated as a key symptom to the homeopathic remedy Aurum, made from gold. Homeopathic treatment is initiated and 
gradually the patient loses her self-hate and self-destructive impulses. Psychotherapy seems to progress better. When therapy was terminated, the patient said that 
her major gain was that she could never think of harming herself any more.
It is not clear in this case whether homeopathy in itself was helpful. Subjectively it was, but it is doubtful whether the effect would have been visible without ongoing 
psychotherapy. Certainly, the evidence for a homeopathic treatment alone was weak and the decision to proceed was not determined by convincing evidence. If the 
prior probability (step 6) had been low either on the side of the patient or on the side of the therapist, proceeding with CAM therapy could not be recommended. Had 
the prior probability been low on the side of the patient, homeopathy or any alternative treatment could have sustained a false hope, diverting the patient from the 
need to seek definitive treatment. If the prior probability had been low on the side of the therapist, it is doubtful whether he or she could have integrated both 
perspectives, which again could interfere with an effective therapeutic process.
CASE 3: SMOKING CESSATION
A 32-year-old woman who has smoked a pack of cigarettes a day for 15 years saw an advertisement for a smoking cessation clinic that used acupuncture. She comes 
to her physician's office seeking advice about this program, which claims to have great results. When she visited the smoking cessation clinic, there was a single 
practitioner with an acupuncture degree from a Midwestern institution. The practitioner claimed that lots of research proved that acupuncture is highly effective in 
helping people stop smoking. The practitioner gave the patient some research articles—a randomized controlled trial written up in an Italian cardiology journal 
claiming 60% effectiveness, and a French article showing that acupuncture was effective in helping to stop smoking (
29
). He said that in his clinical experience, 80% 
of people that he treated were able to stop smoking.
The treatment consists of ear acupuncture 3 times a week for 2 months, followed by a single monthly treatment for 1 year. Also, a small “acu-ball” is taped to points on 
the patient's ear for several hours daily. The cost is $35 per visit (plus $5 for the “acu-ball"). The practitioner requests 2 months' payment at the start, yet gives a 
discount if it is all paid upfront ($240 at the first session for eight sessions). He said this prepayment helps patients become motivated to complete the first eight 
treatments. The patient would like to try this treatment because she really wants to stop smoking and has failed at previous attempts. She has tried smoking cessation 
classes and nicotine gum, but her nonsmoking status lasts only a few days or weeks. She is enthusiastic about the acupuncture and feels it may help her. However, 
before spending the money, she wanted to know if her physician thought it was a good idea.