Medical Practices Not Supported by Science

Introduction


For most of recorded history, medical practitioners administered treatments and procedures to patients that were based on tradition, superstition, and incorrect assumptions about health. In their book Trick or Treatment, Ernst and Singh outline the history of the practice of bleeding and other traditional treatments. With the advent of scientific clinical trials, we no longer use bleeding as a treatment except in rare instances where it is sometimes warranted (like hemochromatosis).

Remember, clinical trials are meant to control for many things that are involved in treatment except for the treatment itself. Control groups are designed to contain the elements that may bias one's perception of efficacy but without the treatment itself. If there are no significant differences between treatment groups and control groups, then we say that the treatment does not work. Placebos and shams are simply null placeholders for the actual treatments.

However, many common practices today are supported by surprisingly little evidence. In some cases, clinical science shows that they do not work any better than placebo or sham treatments. In other words, they do not work. Yet, our biases toward these treatments may be so strong, some of us may be reluctant to let them go. We discussed in the EBM vs. SBM section that treatments should ideally hold some plausibility from basic science (or at least not violate basic science) and demonstrate effectiveness in controlled trials. 

The items listed below are examples of how critical thinking plays a role in the practice of medicine. We can exercise our skeptical tools. The list is certainly not exhaustive, but it demonstrates recent or modern practices that seem plausible from a basic science perspective, but fail in clinical trials. As they seemed plausible before the clinical trials, they did not really represent 'extraordinary claims'. As these claims are empirically testable, they pass our Hume's Fork and Falsification criteria for science.  However, the skeptical tool of Occam's Razor will lead us to reject them as treatments. Claims that are not significantly more likely to be true than the null hypothesis should be rejected. 

These items serve as evidence that we can be easily fooled.


Internal Mammary Artery Ligation for Angina Pectoris


This is a classic example. To our credit, it is not practiced any longer but serves to prove the point.

From 1939 through 1960, patients with heart disease and chest pain were sometimes treated with this procedure. It seemed to work often enough that doctors continued to do it. The idea was that angina (chest pain due to blocked blood flow through the heart arteries), could be treated if more blood was diverted to the heart arteries. To achieve this, surgeons would tie off the left internal mammary artery, thereby forcing more blood to flow to the coronary artery. This is akin to closing off an exit on a highway, thereby forcing twice as many vehicles onto another exit. After surgery,  most of the patients reported improvement.

However, in the late 1950's, the technique was put to the test. Patients were randomized and blinded to get the actual procedure or to get a sham procedure. The patients in the sham group were treated just like the treatment group in every way, except they did not have their internal mammary arteries ligated. They awoke from anesthesia with a superficial wound on their chest, complete with stitches. They received the same post-operative care. 

When the trial was over, both groups had statistically similar outcomes. Patients could not tell the difference. Both groups reported similar rates of angina and nitroglycerin use after the procedures. Internal Mammary Artery Ligation does not work for controlling angina, even if we thought it did. The null hypothesis requires no new assumptions and Occam's Razor leads us to reject internal mammary artery ligation as a treatment for angina. 

Since surgical and medical treatments carry risks, they should be abandoned if they cannot prove their effectiveness over placebo or sham treatments. Internal mammary artery ligation is no longer performed for the treatment of angina.


Arthroscopic Surgery for Osteoarthritis of the Knee


Osteoarthritis of the knee causes a tremendous amount of pain and suffering. Once the condition is sufficiently advanced, little can be done medically. Total knee replacement is effective but is rather involved. For decades, the less drastic surgical procedure of arthroscopic surgery was considered to be a viable option. 

Essentially, the surgeon inserted a camera into the knee joint, shaved rough edges down until smooth with a catheter device and then flushed the knee clean with fluid. The procedure seemed to help about half of the patients. The procedure made intuitive sense and the evidence for benefit was based on experience. 

In 2002, Mosely et.al. published a study in the New England Journal of Medicine that tested this claim. In the study, patients with osteoarthritis were randomized to either the above procedure or a sham procedure (similar to the patients in the internal mammary artery ligation study above). The sham group was treated the same as the procedure group in every way except that they did not actually receive the arthroscopic surgery. They had the same preoperative care, the same anesthesia and the same external knee scars as the patients who actually had the surgery. They were treated the same postoperatively as well (pain meds, nursing care, physical therapy, etc).

After the study was unblinded, the results were surprising. There were no differences between the surgical or the sham groups by any relevant measurement (functionality, pain medication use, quality of life).  The authors concluded, "If the efficacy of arthroscopic lavage or débridement in patients with osteoarthritis of the knee is no greater than that of placebo surgery, the billions of dollars spent on such procedures annually might be put to better use."

Understandably, the results were not well received by many orthopedic surgeons who perform this procedure. The procedure seemed to work and they were used to getting positive feedback. The cognitive bias that may be at work here is the confirmation bias. There were some criticisms about the methodology of the study, inclusion criteria, etc. The procedure was so well ingrained as effective, the claim that 'it doesn't work' seemed to be the more 'extraordinary claim'. However, the evidence was obtained through proper scientific methodology. The evidence had the proper amount of 'extraordinary' required to support the claim.

The study was reproduced by Kirkley et.al. in 2008. They were careful to address some of the methodological criticisms of the earlier study.  Their conclusion echoed the previous one. "Arthroscopic surgery for osteoarthritis of the knee provides no additional benefit to optimized physical and medical therapy." 

Any perceived benefits in the sham group represent the null hypothesis. Since there were no significant differences between the groups, we do not need the assumption that arthroscopic surgery works to explain the findings. The null hypothesis always carries less assumptions. Occam's Razor leads us to favor the null hypothesis and reject arthroscopic surgery for osteoarthritis of the knee in most situations.


Antidepressants and Depressed Mood


Antidepressant medications (ADM) have been used effectively for decades. They have been well studied against placebo in patients with severe depression. The commercial success and safety of the newer ADMs led to their widespread use, particularly since the 1980's. According to PharmacyTimes.com, in 2010, Lexapro (TM) and Sertraline (generic for Zoloft) accounted for over 46 million prescriptions written in the U.S. alone. Patients who may not meet the criteria for severe depression on depression scales have been put on these medications due to complaints of common stress, anxiety and feeling sad over life's everyday misfortunes. To the prescribing doctor, these medications seem to work, as patients will commonly report improvement over time. Sometimes they worsen again, and the dosage is increased or the medication changed. Again, patients seem to improve over time.

The published literature on antidepressants looked pretty good for ADMs. However, a 2008 New England Journal of Medicine analysis revealed a disturbing finding. 74 studies on 12 different ADMs covering 12,564 patients were submitted to the FDA. Of the 74 studies, 33 of them were negative. Twenty-two of these negative studies were not published, and 11 of the negative studies were published, but were spun to look positive. The NEJM article states, "According to the published literature, it appeared that 94% of the trials conducted were positive. By contrast, the FDA analysis showed that 51% were positive. Separate meta-analyses of the FDA and journal data sets showed that the increase in effect size ranged from 11 to 69% for individual drugs and was 32% overall." (see Publication Bias)

Since depressed mood and anxieties come and go with life's twists and turns, it was unclear if ADMs were actually helping people with less severe depression over what would be expected from the placebo effect and the natural variation of moods over time. In 2010, results of a meta-analysis were published in JAMA comparing ADMs with placebo over the measurable range of depression severity using the Hamilton Depression Rating Score (HDRS). On this scale, a score of 0-7 is considered normal and values over 20 are considered moderately severe.

The results raised more than a few eyebrows.

For patients with depression scores less than 25, there were no significant differences found between ADMs and placebo. That is, the majority of people taking antidepressant medications are likely not benefiting from the drugs. The study confirmed earlier results in that it showed significant benefit for people with severe depression. However, for everyone else, these medications are not warranted. They have significant costs and are not without side effects. Skeptical doctors should be aware of results such as these and turn the skeptical eye back toward their own practices.
 

So far, antidepressants are still widely used. Patients come to primary care doctors and psychiatrists looking for them. Family members and friends of sad patients encourage them. They work for severe depression. For most people, antidepressants are a placebo security blanket. 


Vitamin E Supplementation


In the 1990's, some observational studies supported the idea that supplementation with Vitamin E and other 'antioxidants' could lower the risk of heart disease, heart attacks and strokes. Some claimed that it could lower the risk of certain cancers. These claims were plausible from a basic science perspective. Free radicals are molecules that can damage cells and potentially make LDL cholesterol more atherogenic. Free radicals occur naturally from breathing and exercise. Antioxidants, such as Vitamin E, were thought to lower free radicals and decrease atherosclerosis and perhaps even some cancers.

Randomized controlled trials were lacking, but it seemed prudent for doctors to recommend Vitamin E to their patients who had risk factors for heart disease. After all, it was thought at least to do no harm. Because of years of these recommendations in the 90's, it became common knowledge to the general public that Vitamin E is good for you. Today, it is widely sold and promoted to prevent heart attacks and prostate cancer.

However, the results of double blinded, randomized, controlled trials were completed in the 2000's. Contrary to hopes and beliefs, Vitamin E supplementation was found to be of no benefit in the prevention of heart attacks, strokes or cancer. 

In 2000, the H.O.P.E. trial  was published in The New England Journal of Medicine. It followed nearly 10,000 patients for 4.5 years. The authors concluded, "In patients at high risk for cardiovascular events, treatment with vitamin E for a mean of 4.5 years had no apparent effect on cardiovascular outcomes."

In 2005, a follow-up study of the H.O.P.E. trial patients was published in JAMA. The study authors concluded, "In patients with vascular disease or diabetes mellitus, long-term vitamin E supplementation does not prevent cancer or major cardiovascular events and may increase the risk for heart failure."

In 2011, results of the SELECT trial were published in JAMA. Over 35,000 men in 3 countries were randomized and blinded to receive Vitamin E and Selenium or placebo over 3 years. This study was designed to establish the effect of Vitamin E and Selenium on the risk of prostate cancer. The authors concluded, "Dietary supplementation with vitamin E significantly increased the risk of prostate cancer among healthy men."

A 2014 randomized controlled study of male VA patients with mild to moderate Alzheimer's Disease. All patients were already on donepezil ("Aricept"). Vitamin E therapy was found to have a slightly statistically significant decrease in the rate of decline in functional activities compared to placebo. However, a group of patients receiving memantine ("Namenda") and Vitamin E showed no better decrease in decline than memantine alone. Also, Vitamin E did not show an effect on cognitive decline compared to placebo. This study is being touted as evidence for Vitamin E for use in Alzheimer's patients, but the results are not impressive in this relatively small study.


Despite the disconfirming evidence for the use of Vitamin E for the prevention of heart disease, Vitamin E is still widely promoted by doctors and healthcare providers who are unaware of the evidence and by promoters of pseudoscience who deny the evidence. The "Optimal Health Resource" blog, riddled with vitamin advertisements, currently is claiming "Vitamin E is an essential nutrient that is rapidly gaining status as a powerful agent in the fight against heart disease, cancer, dementia and immune health."


    



Cough Medicines for Colds

Just about all parents deal with children with respiratory infections who cough, especially at night. We naturally want to help them, and to silence their coughs. Doctors want to help too. Consequently we buy a lot of cough medicines every year. According to Harvard Health Publications, as of 2006, Americans spent 3.5 billion dollars annually on over-the-counter cough remedies. But do they work? Like vitamin E, they have some basic science plausibility. But basic science is not enough. Skeptical doctors also need clinical science. Until recently, these cough medicines have not been adequately tested clinically.

Over-the-counter cough medications are made from a limited number of drugs in various combinations. Dextromethorphan is a central nervous system depressant. It is the ingredient in cough medicines labelled "cough suppressant".  Guaifenesin hydrates mucous to promote mucous break-up so that it can be mobilized. It is the ingredient in medicines labelled "expectorant". Diphenhydramine and chlorpheniramine are first generation anti-histamines. They are effective against allergic reactions. Pseudoephedrine is not a 'cough medicine', but is commonly found in combinations with the others. Pseudoephedrine is labelled as a 'decongestant' and is a stimulant similar to adrenaline. It may help congestion temporarily by decreasing blood flow to the nose. Codeine is a narcotic and a central nervous system depressant. It is only available by prescription and has been labelled as a 'cough suppressant'. 

In 1991, a Finnish study called into question the effectiveness of dextromethorphan in managing cough in children. Other studies followed. In 2002, a systematic review in the the Archives of Disease in Childhood concluded, "Based on the available evidence from a small number of studies, we cannot recommend OTC cough medicines as a first line treatment for children with acute cough.

Other studies and reviews came to the same conclusion. 

The 2010 Cochrane systematic review of the literature looking at all of the decent studies on OTC cough medicines has failed to find any benefit in adults or children. It covered studies that looked at a total of 3421 adults and 616 children. The authors conclude, "There is no good evidence for or against the effectiveness of OTC medicines in acute cough."

The studies that looked at the individual cough medicines are equally unimpressive.

Dextromethorphan and codeine were not more effective than placebo (Taylor et.al., J. Pediatrics, 1993). Dextromethorphan failed again in a trial on infants in 2006 (Merenstein et.al., Archives of Pediatrics and Adolescent Medicine),  and in 2007 (Paul et.al., Archives of Pediatrics and Adolescent Medicine),

Dextromethorphan and diphenhydramine were no more effective than placebo (Paul et.al., Pediatrics, 2004).

Guaifenesin has little data to support its use. In 2009, Storms et.al. reviewed the literature and found, "The use of mucoactive agents is controversial due to limited data and equivocal efficacy in available studies."  The 2010 Cochrane review states, "Two (adult) trials compared the expectorant guaifenesin with placebo; one indicated significant benefit whereas the other did not." No studies in children were of sufficient quality to be included. 

Two decades of studies and reviews have failed to support the use of cough medicines for acute respiratory infections in children or in adults. Yet doctors continue to recommend them, companies bombard us with advertisements to promote them, and parents and patients continue to use them. Their use epitomizes the placebo effect. Their effect is consistent with doing nothing. The pertinent logical fallacy at work is post hoc ergo propter hoc.

In the setting of a common viral upper respiratory infection, people seek care or begin taking cough medication when symptoms are at their worst. Typically, this is at about day 3 (
JAMA. 1967;202(6):494-500). At that point, anything may seem to work as the symptoms begin to improve at that point anyway. This apparent success leads to the incorrect beliefs that cough medications, decongestants and antibiotics are helpful (and even required) to recover from a cold (see Regressive Fallacy).

Skeptical doctors can be very effective in managing patients with upper respiratory infections. Patients will likely feel just as good about their care (or even better) if doctors take the extra time required to educate their patients and guide them properly through their symptoms. Compassion and validation are the best forms of the 'placebo effect'.


Vitamin C for Colds and Cancer

Vitamin C (ascorbic acid) is an essential dietary vitamin needed for the production of the connective tissue protein collagen. People with significant vitamin C deficiency develop 'scurvy', a serious condition characterized by bleeding and loss of function of multiple organ systems. The need for regular consumption of vitamin C is not controversial. It seems plausible that taking higher doses of  vitamin c may help prevent or treat other diseases.

Linus Pauling was a chemist and is considered to be one of the greatest contributors to modern chemistry. He won 2 unshared Nobel prizes for chemistry and peace. In the fields of theoretical chemistry and quantum mechanics, his expertise made him a relevant authority in these areas. Later in his life, he became convinced that high daily doses of vitamin C could prevent colds and cancer. He practiced what he preached and took high doses of vitamin C daily. He claimed to experience less colds. He published books on the subject (Vitamin C and the Common Cold, Vitamin C, the Common Cold and the Flu, Cancer and Vitamin C, and How to Live Longer and Feel Better). Since his first book in 1970, it has become common for people and some doctors to use high doses of vitamin C at the first signs of a cold and daily for prevention. Many think that there is real science behind the practice.

Studies have been done. From 1967 to present, no double-blinded, placebo controlled trial has revealed any evidence of vitamin C's ability to prevent the common cold. Studies and reviews of studies flourished in the 1970's. Some studies reported a possible benefit in reducing the duration of colds, others did not. 

In the 1979 review, Coulehan concluded, "Large-scale clinical trials of the effect of vitamin C on the common cold conducted since Pauling's pronouncement on its efficacy in 1970 have produced conflicting results. Benefits shown have been modest, and positive findings have not been reproducible. Reported benefits probably are due to statistical artifact, minor vitamin-induced physiologic changes, the placebo effect, or a combination of these."

Studies continued to be done into the 2000's. It is a popular thing to study. Vitamin C clearly did not prevent the common cold, but some studies still suggested a possible decrease in symptom duration with doses over 3000 mg a daily (that would be six 500 mg tablets a day). In 2001, a decently performed RCT was done to look at "mega-doses" of vitamin C starting at the onset of colds. In this report, Audera et.al. concluded, "Doses of vitamin C in excess of 1 g daily taken shortly after onset of a cold did not reduce the duration or severity of cold symptoms in healthy adult volunteers when compared with a vitamin C dose less than the minimum recommended daily intake."

The 2007 meta-analysis by Douglas et.al. looked at over 30 trials and over 11,000 people. Their conclusion was, "The failure of vitamin C supplementation to reduce the incidence of colds in the normal population indicates that routine mega-dose prophylaxis is not rationally justified for community use."

Pauling's claim that oral vitamin C is good for common colds is not significantly different than the null hypothesis. Occam's Razor leads us to reject the claim. But what about cancer? Pauling claimed in his later book that vitamin C can be good for cancer patients. This claim is more difficult to test. In fact, it is more difficult to define because cancer is not any one disease. It is a collection of diseases with a common theme (one's own cells multiply unregulated such that they harm the body, often fatally). What does science say about vitamin C and cancer?

Not too much, actually. In the 2008 JNCI RCT on the effect of vitamins C, E and beta carotene, Lin et.al. looked at over 8000 women randomly assigned to vitamin supplementation regimens. They were determined to be cancer free at the start of the study and followed for over 9 years. The vitamin C group took 500 mg a day. They concluded, "Supplementation with vitamin C, vitamin E, or beta carotene offers no overall benefits in the primary prevention of total cancer incidence or cancer mortality. "  Another study pursued the claim that antioxidant vitamins can decrease cancer risk (this one looked at beta carotene). Here the authors speculated that, "These inverse associations may be the result of carotenoid measurements’ function as a marker of a healthier lifestyle (higher fruit and vegetable consumption) or of residual confounding by smoking."  We can speculate that vitamin C may have a similar perceived benefit for the same reasons. This demonstrates the skeptical tool of Correlation/ Causation

The American Cancer Society's website states, " Although high doses of vitamin C have been suggested as a cancer treatment, the available evidence from clinical trials has not shown any benefit."

But Pauling stated that oral doses cannot achieve the blood level needed to fight cancer. He claimed that only i.v. vitamin C can do the job. Well, some research was done on this as well. 

The practice of using high dose vitamin C appears relatively widespread and generally safe. A small study in breast cancer patients reported, "Complementary treatment of breast cancer patients with i.v. vitamin C was shown to be a well tolerated optimization of standard tumour-destructive therapies, reducing quality of life-related side-effects."

The 2008 report from the Annals of Oncology states, "High-dose i.v. ascorbic acid was well tolerated but failed to demonstrate anticancer activity when administered to patients with previously treated advanced malignancies."

So currently, there is no convincing evidence that vitamin C (oral or i.v.) prevents or treats cancer better than placebo. Although it is safe, the claim appears unsupported by science. Maybe someday a benefit will be found, but the evidence does not warrant aggressively defending the practice. Like cough medicines, vitamin C's placebo effect is reinforced by the post hoc ergo propter hoc logical fallacy and confirmation bias.

Linus Pauling was a great chemist and a true authority in theoretical chemistry and quantum mechanics. However, his expertise in these fields does not grant him expertise in medical science, especially clinical science. The relevant logical fallacy here is appeal to authority.


Fenofibrate for Preventing Heart Attacks in Diabetics

It is has been well established that high 'bad cholesterol' (LDL) and low 'good cholesterol' (HDL) are risk factors for cardiovascular disease. Treating very high LDL levels with a 'statin' medication has been established as an effective way of preventing heart attacks and other complications. Other major risk factors include diabetes, high blood pressure, smoking and family history. Studies show that having type 2 diabetes may raise one's risk of heart attack to about the same level as a non-diabetic with a previous heart attack. 

Triglycerides are fatty acids that can be measured in the blood. They are included with cholesterol in most common laboratory panels. Many people with heart disease have high triglycerides as well as high cholesterol. Very high triglycerides are known to damage the pancreas. It seems like common sense that treating people to lower the triglycerides should lower their risk of heart attacks. Fibrates are drugs that are most effective at lowering triglycerides. They can also help to raise HDL to a small degree. In the U.S., fenofibrate is the most commonly used fibrate. Worldwide sales of fenofibrate in 2008 topped $17 billion U.S. dollars, and over $14 billion in 2009. Clearly, it has been the common wisdom that fenofibrate should be used to prevent cardiovascular events.

Diabetics tend to have high triglycerides for many of the same reasons that they have high sugars. High triglyceride levels are a sign of insulin resistance. As above, diabetics are at particularly high risk for heart attacks. In 2010, the results from the huge ACCORD Trial were published. In this double-blinded trial, over 5000 at-risk diabetics who were on statins were randomly assigned to also use fenofibrate or placebo. Since diabetics are particularly at risk for heart attacks, the benefits of fenofibrate should be most notable in this group. They were followed for an average of 4.7 years. The results and conclusions were surprising. 

Although the fenofibrate group had measurably lower triglycerides than the control group, the two groups did not differ in the number of cardiovascular events (heart attack or stroke). It made no positive difference. There were more side effects in the fenofibrate group, albeit a small number. A significant number of women in the study had to drop out due to declining kidney function.

There was a slight benefit that can be seen on the forest plot in the subset of patients with very high triglycerides and very low HDL, but the error bar overlapped with the line of insignificance. The plot seems to indicate that it is likely the effect of the low HDL that accounts for the statistical significance toward benefit in this subgroup rather than the high triglyceride. The overall benefit among diabetics was equivalent to null. 
Again, Occam's Razor should lead us away from using fenofibrate as a treatment to prevent heart attacks. As with arthroscopic surgery for osteoarthritis of the knee among orthopedic surgeons, primary care doctors and cardiologists are still utilizing fenofibrate for this purpose. Habits and confirmation bias are hard to overcome. Another skeptical tool to use here is correlation is not necessarily causation. Diabetes can lead to heart disease (and in turn, heart attacks). Diabetes is associated with high triglycerides. Therefore, heart disease and high triglycerides are significantly correlated. This easily has lead us to believe that triglycerides significantly cause heart disease. Cum hoc, ergo propter hoc.

It is difficult to take someone off of a medication that, on paper, seems to be working so well. However, in light of the new extraordinary evidence against their use, the claim that fenofibrate can prevent heart attacks is now the extraordinary claim. The burden of proof has shifted.

The ACCORD Trial was summarized in Journal Watch. There the author stated, " The burden of proof is firmly on advocates of this drug to justify the cost and risk to patients."  


Twelve Step Programs for Substance Abuse

Do 12 step programs like Alcoholics Anonymous and Narcotics Anonymous actually work? Dr. Harriet Hall reviewed this problem in 2009. She points to the Cochrane Review that found no “definitive” evidence for their overall efficacy over essentially doing nothing outside of routine medical advice. However, there certainly are many individuals in the “it worked for me camp”. Perhaps the real answer to the question is: “it depends”.

It would seem that those who voluntarily seek out the help of 12 step programs may be of the mindset that would do best in such programs. There are strong religious undertones in such programs which would appeal to some but not others. Perhaps the more important question is one of legal ethics: should we sentence people with substance abuse problems to such program?

Here is where the evidence seems to suggest the worthlessness of such an intervention. Randomized trials do not support their effectiveness, perhaps because randomization eliminates the self-selection bias. Sentencing would seem more akin to randomizing. It would seem to follow that sentencing one to a 12 step program would be similarly null, or even harmful, in its effects.


Banning Pacifiers in the Postpartum Wards

Pacifier use in the newborn period seems like a trivial issue to tackle on this site. However, the point here is to demonstrate some key skeptical tools and also to highlight how easily one can fall prey to type 1 errors when a belief-in influences a belief-that.

Breastfeeding offers many health and psychological benefits to infants and mothers. These benefits have been well documented. It has become the standard of care to promote breastfeeding to new parents. Hospitals have developed policies and teams of lactation consultants to educate and facilitate healthy breastfeeding. In a trend to be "baby friendly", many hospitals have adopted UNICEF’s 10 steps to successful breastfeeding.

Step 9 of 10 states, “Give no artificial teats or pacifiers (also called dummies or soothers) to breastfeeding infants.

The American Academy of Pediatrics 2012 guidelines also call for a ban of pacifiers with healthy babies in the hospital: “...pacifier use in the hospital in the neonatal period should be limited to specific medical indications…Mothers of healthy term breastfed infants should be instructed to delay pacifier use until breastfeeding is well-established, usually about 3 to 4 weeks after birth.

The notion that pacifier use inhibits breast feeding is very plausible. It deserves a Bayesian prior probability of at least 0.5. The evidence supporting the idea comes from several papers published in the 1990’s through the early 2000’s.

In 1995, a Brazilian longitudinal study in Pediatrics concluded, “Pacifier use is highly correlated with early weaning, even after controlling for possible confounders. Until it is determined if pacifier use is causally related to weaning or is a marker for other undetermined causes, pacifier use probably should not be recommended for breast-fed infants.

In 2003, a randomized controlled trial in Pediatrics compared 4 groups of about 170 babies each. These groups were classified as bottle/early pacifiers, cup-fed/early pacifiers, bottle/late pacifiers and cup/late pacifiers. Each group received some type of supplemental feeding according to design. They concluded, “Pacifier use in the neonatal period was detrimental to exclusive and overall breastfeeding. These findings support recommendations to avoid exposing breastfed infants to artificial nipples in the neonatal period.” It is recognized that supplemental formula is an impediment to breastfeeding. The use of pacifiers among exclusively breast fed infants was not evaluated. The findings do not support the conclusion.

Thus, the policy to effectively ban pacifiers has been adopted in many hospitals. There is a correlation between pacifier use and poor breast feeding. At this point, skeptical doctors should be recalling one of the basic tools of skepticism: correlation is not necessarily causation. This tool was considered in the 1997 article in Pediatrics which pondered, “Pacifiers are related to a shorter duration of breastfeeding. However, it is unclear whether this association is causal, because confounding, reverse causality, and self-selection of mothers may play a role.” The authors concluded, “Pacifiers may be an effective weaning mechanism used by mothers who have explicit or implicit difficulties in breastfeeding, but they are much less likely to affect infants whose mothers are confident about nursing. “ They further predicted, “Breastfeeding promotion campaigns aimed specifically at reducing pacifier use will fail unless they also help women face the challenges of nursing and address their anxieties.

The UNICEF recommendations for prohibiting pacifiers in the newborn period was challenged with the results of a 1997 European Journal of Pediatrics randomized controlled trial that looked specifically at pacifier use in the first 5 days after birth. The authors concluded, “In our study population fluid supplements offered by bottle with or without the use of pacifiers during the first 5 days of life were not associated with a lower frequency or shorter duration of breastfeeding during the first 6 months of life.

The Cochrane Review of the literature concluded, “Pacifier use in healthy term breastfeeding infants, started from birth or after lactation is established, did not significantly affect the prevalence or duration of exclusive and partial breastfeeding up to four months of age. However, evidence to assess the short-term breastfeeding difficulties faced by mothers and long-term effect of pacifiers on infants' health is lacking.”

A 2009 comprehensive review of the literature available in the English language was published in Archives of Pediatric and Adolescent Medicine. Among the observational, cohort studies, there was a trend toward a correlation between early pacifier use and shortened breastfeeding. However, when the available prospective, randomized controlled trials were analyzed, the correlation vanished. The authors concluded, “[T]he highest level of evidence does not support an adverse relationship between pacifier use and breastfeeding duration or exclusivity. The association between shortened duration of breastfeeding and pacifier use in observational studies likely reflects a number of other complex factors, such as breastfeeding difficulties or intent to wean.

A 2012 study presented at the Pediatric Academic Societies meeting in Boston (from the Oregon Health and Science University) looked at 2000 infants over a one year period. It was found that infants that were not permitted pacifiers were more likely to fail at exclusive breast feeding. Over that period, the pacifier restriction was in place. Exclusive breast feeding actually fell from 79 to 68%.

As with so many ideas in medical science, apparent causal effects are shown to not exist when the power of the studies increase and the controls are tightened to eliminate selection and confirmation biases. Many mothers have difficulty with breastfeeding for a variety of reasons. It is no surprise that these mothers tend to use pacifiers more often than do mothers who have less problems and more confidence. The correlation is not causation. The new UNICEF-based hospital restrictions on pacifiers will likely fail because the pacifiers are not the cause of the poor breastfeeding. Poor breastfeeding is the cause of higher pacifier use.

In the meantime, we are likely increasing the stress on new mothers by enforcing a no pacifier policy, one which will likely increase the level of guilt among moms who resort to pacifier use due to legitimate causes of poor breastfeeding.

Those intimately involved with lactation counselling in hospitals appear to have an a priori belief in the idea of breastfeeding without artificial nippling. Such a belief-in may stem from the naturalistic fallacy and other convictions. However, we have policies against pacifier use despite the contradictory “highest level of evidence”. Thus, the belief-in has become a belief-that which is not supported by science.



Conclusion


Science is a self correcting process. Recall that ideas come from inductive reasoning. We prefer that our ideas are induced from established science. Coming up with plausible ideas is only half of medical science. If we put plausible ideas into practice without testing them deductively, we often will get into trouble. The real problem is with ideas that are set into practice for long periods of time without testing. Our biases and natural tendency toward logical fallacies allow us to become very attached to these practices. In some cases, the practices may define us as professionals.

Again, the above list is not complete by any means. The topics serve as examples of many skeptical concepts in action. The list will likely grow over time. There will always be medical practices unsupported by science.

When faced with the hard reality of properly obtained, negative studies, it is natural for many to become denialists. Many will still cling to the favored practices. Vitamin C is still given for colds. Cough medicines are actively recommended. Arthroscopic knee surgery is still performed (and avidly defended by some). Fenofibrate sales will likely still be in the billions for years to come. Being skeptical doctors means that we follow the evidence, even if we may not (at first) like where it leads. True compassion is guided by truth.


John Byrne, MD





References and Links

Dimond, E, C Frederick Kittle, and James E Crockett. "Comparison of internal mammary artery ligation and sham operation for angina pectoris." The American journal of cardiology 5.4 (1960): 483-486.
<http://www.ajconline.org/article/0002-9149(60)90105-3/abstract>

"Sham Surgery Returns as a Research Tool - The New York Times." 
<http://www.nytimes.com/library/review/042599surgery-ethics-review.html>

Moseley, J Bruce et al. "A controlled trial of arthroscopic surgery for osteoarthritis of the knee." New England Journal of Medicine 347.2 (2002): 81-88.
<http://www.nejm.org/doi/full/10.1056/NEJMoa013259>

Kirkley, Alexandra et al. "A randomized trial of arthroscopic surgery for osteoarthritis of the knee." New England Journal of Medicine 359.11 (2008): 1097-1107.
<http://www.nejm.org/doi/full/10.1056/NEJMoa0708333>

Fournier, Jay C et al. "Antidepressant drug effects and depression severity." JAMA: The Journal of the American Medical Association 303.1 (2010): 47-53.
<http://jama.ama-assn.org/content/303/1/47.short?home>

Turner, Erick H et al. "Selective publication of antidepressant trials and its influence on apparent efficacy." New England Journal of Medicine 358.3 (2008): 252-260.
<http://www.nejm.org/doi/full/10.1056/NEJMsa065779>

Fournier, Jay C et al. "Antidepressant drug effects and depression severity." JAMA: The Journal of the American Medical Association 303.1 (2010): 47-53.
<http://jama.ama-assn.org/content/303/1/47.full>

"Top 200 Drugs in The us MarkeT by - Pharmacy Times." 2011. 
<http://www.pharmacytimes.com/media/pdf/26babf64401fed80734dbf8dce471e26.pdf>

Hemilä, Harri, Elizabeth Chalker, and Bob Douglas. "Vitamin C for preventing and treating the common cold." Cochrane Database of Systematic Reviews 3 (2007).
<http://summaries.cochrane.org/CD000980/vitamin-c-for-preventing-and-treating-the-common-cold>

Hemilä, Harri, Elizabeth Chalker, and Bob Douglas. "Vitamin C for preventing and treating the common cold." Cochrane Database of Systematic Reviews 3 (2007).
<http://www.ncbi.nlm.nih.gov/pubmed/15495002>

Chalmers, TC. "Effects of ascorbic acid on the common cold. An evaluation of the ..." 1975. 
<http://www.ncbi.nlm.nih.gov/pubmed/1092164>

Dykes, MH. "Ascorbic acid and the common cold. Evaluation of its efficacy and ..." 1975. 
<http://www.ncbi.nlm.nih.gov/pubmed/1089817>

Coulehan, JL. "Ascorbic acid and the common cold: reviewing the evidence." 1979. 
<http://www.ncbi.nlm.nih.gov/pubmed/382147>

Karlowski, TR. "Ascorbic acid for the common cold. A prophylactic and therapeutic trial." 1975. 
<http://www.ncbi.nlm.nih.gov/pubmed/163386>

Audera, C. "Mega-dose vitamin C in treatment of the common cold: a ..." 2001. 
<http://www.ncbi.nlm.nih.gov/pubmed/11700812>

Hemilä, Harri, Elizabeth Chalker, and Bob Douglas. "Vitamin C for preventing and treating the common cold." Cochrane Database of Systematic Reviews 3 (2007).
<http://www.ncbi.nlm.nih.gov/pubmed/17636648>

Lin, Jennifer et al. "Vitamins C and E and beta carotene supplementation and cancer risk: a randomized controlled trial." Journal of the National Cancer Institute 101.1 (2009): 14-23.
<http://jnci.oxfordjournals.org/content/101/1/14.short>

Cabanillas, Fernando. "vitamin c and cancer: what can we conclude-1,609 Patients and 33 Years later?." PR Health Sciences Journal 29.3 (2010).
<http://www.ncbi.nlm.nih.gov/pubmed/20799507>

"Vitamin C - American Cancer Society." 2010. <http://www.cancer.org/Treatment/TreatmentsandSideEffects/ComplementaryandAlternativeMedicine/HerbsVitaminsandMinerals/vitamin-c>

VOLLBRACHT, CLAUDIA et al. "Intravenous vitamin C administration improves quality of life in breast cancer patients during chemo-/radiotherapy and aftercare: results of a retrospective, multicentre, epidemiological cohort study in Germany." in vivo 25.6 (2011): 983-990.
<http://www.ncbi.nlm.nih.gov/pubmed/22021693>

Padayatty, Sebastian J et al. "Vitamin C: intravenous use by complementary and alternative medicine practitioners and adverse effects." PLoS One 5.7 (2010): e11414.
<http://www.ncbi.nlm.nih.gov/pubmed/20628650>

Hoffer, LJ et al. "Phase I clinical trial of iv ascorbic acid in advanced malignancy." Annals of oncology 19.11 (2008): 1969-1974.
<http://www.ncbi.nlm.nih.gov/pubmed/18544557>

"Does Lowering Homocysteine Prevent ... - Quackwatch." 
<http://www.quackwatch.com/03HealthPromotion/homocysteine.html>

Yusuf, Salim et al. "Vitamin E supplementation and cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators." The New England journal of medicine 342.3 (2000): 154.
<http://www.ncbi.nlm.nih.gov/pubmed/10639540>

Lonn, Eva et al. "Effects of long-term vitamin E supplementation on cardiovascular events and cancer: a randomized controlled trial." JAMA: the journal of the American Medical Association 293.11 (2005): 1338.
<http://www.ncbi.nlm.nih.gov/pubmed/15769967>

Klein, Eric A et al. "Vitamin E and the risk of prostate cancer." JAMA: the journal of the American Medical Association 306.14 (2011): 1549-1556.
<http://jama.ama-assn.org/content/306/14/1549.full?ijkey=ifuY4wQu5RGts&keytype=ref&siteid=amajnls>

Lin, Jennifer et al. "Vitamins C and E and beta carotene supplementation and cancer risk: a randomized controlled trial." Journal of the National Cancer Institute 101.1 (2009): 14-23.
<http://jnci.oxfordjournals.org/content/101/1/14.short>

Paul, IM. "Abstract - Pediatrics." 2004.
<http://pediatrics.aappublications.org/content/114/1/e85.abstract>

Cough Medicines - Harvard Health Publications - Harvard University." 2009. 
<http://www.health.harvard.edu/press_releases/cough-medicines>

Schroeder, K, and T Fahey. "Should we advise parents to administer over the counter cough medicines for acute cough? Systematic review of randomised controlled trials." Archives of disease in childhood 86.3 (2002): 170-175.
<http://adc.bmj.com/content/86/3/170.abstract?ijkey=1dd6c631f0aa0d2c6c69ae0a584bf24d882194f2&keytype2=tf_ipsecsha>

Schroeder, K, and T Fahey. "Over‐the‐counter medications for acute cough in children and adults in ambulatory settings." The Cochrane Library (2007).
<http://www.ncbi.nlm.nih.gov/pubmed/15495019>

Smith MH, Feldman W. Over-the-counter Cold Medications: A Critical Review of Clinical Trials Between 1950 and 1991.JAMA. 1993;269(17):2258-2263. doi:10.1001/jama.1993.03500170088039.
<http://jama.jamanetwork.com/article.aspx?articleid=405816>

Smith, SM, K Schroeder, and T Fahey. "Over-the-counter (OTC) medications for acute cough in children and adults in ambulatory settings." (2010).
<http://summaries.cochrane.org/CD001831/over-the-counter-otc-medications-for-acute-cough-in-children-and-adults-in-ambulatory-settings>

Taylor, JA. "Efficacy of cough suppressants in children." 1993.
<http://www.ncbi.nlm.nih.gov/pubmed/8496765>

Paul, Ian M et al. "Effect of honey, dextromethorphan, and no treatment on nocturnal cough and sleep quality for coughing children and their parents." Archives of pediatrics & adolescent medicine 161.12 (2007): 1140.
<http://archpedi.ama-assn.org/cgi/content/abstract/161/12/1140>

Gwaltney Jr, Jack M et al. "Rhinovirus infections in an industrial population." JAMA: the journal of the American Medical Association 202.6 (1967): 494-500.
<http://jama.ama-assn.org/content/202/6/494.abstract>

"ACCORD, Questions and Answers,- NHLBI, NIH - National Heart ..." 
<http://www.nhlbi.nih.gov/health/prof/heart/other/accord/q_a.htm>

"The ACCORD Lipid Study: Fenofibrate Doesn't Help - Cardiology." 2010. 
<http://cardiology.jwatch.org/cgi/content/full/2010/314/1>

Barros, F. C., Victoria, C. G., Semer, T. C., Filho, S. T., Tomasi, E., & Weiderpass, E. (1995). Use of pacifiers is associated with decreased breastfeeding duration. Pediatrics, 95(4), 497-499. 

Ferri, M. "Alcoholics Anonymous and other 12-step programmes for alcohol ..." 2006.

<http://www.ncbi.nlm.nih.gov/pubmed/16856072>


Howard, C. R., Howard, F. M., Lanphear, B., deBlieck, E. A., Eberly, S., & Lawrence, R. A. (1999). The effects of early pacifier use on breastfeeding duration. Pediatrics, 103(3), 33-38.

Howard, C. R., Howard, F. M., Lanphear, B., Eberly, S., deBlieck, E. A., Oakes, D., & Lawrence, R. (2003). Randomized clinical trial of pacifier use and bottle-feeding or cup feeding and their effect on breastfeeding. Pediatrics, 111(3), 511-518.
<http://pediatrics.aappublications.org/content/111/3/511.abstract>

Kramer M. S., Barr, R. G., Dagenais, S., Yang, H., Jones, P., Ciofani, L. & Jane F. (2001). Pacifier use, early weaning, and cry/fuss behavior: A randomized controlled trial. Journal of the American Medical Association, 286(3), 322-326.

Lawrence, R. (1987). The management of lactation as a physiologic process. Clinics in Perinatology, 14(1), 1-10. [Abstract]
Niemela, M., Uhari, M., & Mottonen, M. (1995). A pacifier increases the risk of recurrent acute otitis media in children in day care centers. Pediatrics, 96(5), 884-888.<
http://www.ncbi.nlm.nih.gov/pubmed/7478830?dopt=Citation>

Righard, L. & Alade, M. (1997). Breastfeeding and the use of pacifiers. Birth, 24(2), 116-120.
<http://www.ncbi.nlm.nih.gov/pubmed/9271978?dopt=Books>

Victora, C. G., Behague, D. P., Barros, F. C., Olinto, M. T. A., & Weiderpass, E. (1997). Pacifier use and short breastfeeding duration: Cause, consequence, or coincidence? Pediatrics, 99(3), 445-453.

Jaafar SH, Jahanfar S, Angolkar M, Ho JJ. Pacifier use versus no pacifier use in breastfeeding term infants for increasing duration of breastfeeding. Cochrane Database of Systematic Reviews 2011, Issue 3.

Haffner, Steven M et al. "Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction." New England journal of medicine 339.4 (1998): 229-234.
<http://www.ncbi.nlm.nih.gov/pubmed/9673301>

"Baby-Friendly Hospital Initiative - Unicef." 2002. 
<http://www.unicef.org/programme/breastfeeding/baby.htm>

"Breastfeeding Hospital Policy Recommendation Policy #7." 2008. 
<http://www.cdph.ca.gov/HealthInfo/healthyliving/childfamily/Pages/BFP-MdlHospToolkitPolicy7.aspx>

"New York State Model Hospital Breastfeeding Policy ..." 2011. <http://www.health.ny.gov/community/pregnancy/breastfeeding/docs/model_hospital_breastfeeding_policy_implementation_guide.pdf>

Howard, Cynthia R et al. "Randomized clinical trial of pacifier use and bottle-feeding or cupfeeding and their effect on breastfeeding." Pediatrics 111.3 (2003): 511-518.
<http://www.ncbi.nlm.nih.gov/pubmed/12612229>

Jaafar, SH et al. "Effect of pacifier use on duration of breastfeeding in full-term infants." (2012).
<http://summaries.cochrane.org/CD007202/effect-of-pacifier-use-on-duration-of-breastfeeding-in-full-term-infants>

Schubiger, G, U Schwarz, and O Tönz. "UNICEF/WHO baby-friendly hospital initiative: does the use of bottles and pacifiers in the neonatal nursery prevent successful breastfeeding?." European journal of pediatrics 156.11 (1997): 874-877.
<http://www.ncbi.nlm.nih.gov/pubmed/9392404>

O'Connor, Nina R et al. "Pacifiers and breastfeeding: a systematic review." Archives of pediatrics & adolescent medicine 163.4 (2009): 378.
<http://www.ncbi.nlm.nih.gov/pubmed/19349568>

Victora, Cesar Gomes et al. "Pacifier use and short breastfeeding duration: cause, consequence, or coincidence?." Pediatrics 99.3 (1997): 445-453.
<http://www.ncbi.nlm.nih.gov/pubmed/9041303>

"Pacifiers May Encourage Breastfeeding in Newborns - ABC News." 
<http://abcnews.go.com/Health/w_ParentingResource/pacifiers-encourage-breastfeeding-newborns/story?id=16229351>