CUSTOMER: Here’s one — nine pence.
DEAD PERSON: I’m not dead!
CUSTOMER: Nothing — here’s your nine pence.
DEAD PERSON: I’m not dead!
MORTICIAN: Here — he says he’s not dead!
CUSTOMER: Yes, he is.
MORTICIAN: He isn’t.
CUSTOMER: Well, he will be soon, he’s very ill.
DEAD PERSON: I’m getting better!
CUSTOMER: No, you’re not — you’ll be stone dead in a moment.

Monty Python and the Holy Grail

For some unexplained reason, people at work like to tell me of the positive interactions they have had with acupuncturists and chiropractors and others of that ilk. I must have a friendly face, but I keep checking my back for a “CAM me” sign.

One of the oncology nurses was telling me how she has chronic neck pain, and that she was skeptical about acupuncture, and would never recommend these therapies for one of her cancer patients, but she went to an acupuncturist, and by gosh and by golly if her pain wasn’t better, what do you think of that Mr. Skeptic?

Call me Dr. Skeptic, I replied. Show some respect for the dead.

It does make for an awkward conversation.

I cannot deny that she isn’t better. How can I argue that she doesn’t have decreased pain? She is the one who hurts and is the one who can best judge the degree of her discomfort.

“Nope. You are not better. Sorry. Wrong. You are still in the same amount of pain you were before.”

It is an untenable position.

Then there is the issue of causality. She hurt, she had acupuncture, and immediately afterwards she hurt less. How can I say that acupuncture did not cause her pain to go down? That experience cannot be trumped by arguments that there is no biologic plausibility for acupuncture, that clinical trials show no increased efficacy compared to sham acupuncture for pain relief.

“Sorry. Randomized clinical trials prove your pain is not better.”

There is zero credibility in any argument based on the science behind acupuncture efficacy when contrasted with the the effects of the moment in the person who has less pain.

I have had similar discussions over the years with other people on similar topics, and it usually ends with the person smiling at me like I am a slightly foolish child who just doesn’t understand that they are better. The intervention was effective. The proof is in the pudding, which is why I do not eat pudding.

How can you say a patient is better?

What do you mean by better. Does it depend on that the meaning of is is?

Measuring the response to a therapeutic intervention can have two components: the objective and the subjective.

The objective part is the simpler of the two. Tumor is smaller, the patient lives longer, the blood pressure is down. Stuff you can measure. In my world of Infectious Diseases (the legitimate ID), the objective component is often cure. If the infection is eradicated, the patient is better, right? My job is simple: me find bug, me kill bug, me go home. But better is not always so simple a concept.

A recent review of daptomycin for endocarditis suggested that 37% were cured, 27% were improved and 8% failed (the remainder were not evaluable) (1). Endocarditis is 100% fatal if not cured, so what good is being improved? No much. 27% were better, just not better enough. Improved isn’t good enough if the end result is death.

I have had the occasional HIV patient who have so much toxicity from the antiretroviral medications that the treatment is worse than the disease. The medications can stop the HIV replication and improve their T cell numbers, but they feel so bad that they have preferred to go off the medications, have their HIV progress and die. Did I make them better with the medications? Is it better to live longer in misery?

To give the most extreme example, I remember a case as a fellow where a lady took a medication for her toenail fungus and it was eradicated. No toenail fungus. She was better, right? But she got hepatitis from the medication and died from liver failure. Maybe not so much better.

I can cure your infection but in the process ruin your kidneys. Are you better? It depends on the disease. If the infection would have been fatal, then it may be a good trade off. If is a minor illness, then maybe not so good.

The subjective aspects of improvement are even harder to measure accurately.

Many psychological aspects into play with the subjective improvement of patients and it may not correlate with the objective findings. It brings up the question: is the patient better if they think they are better subjectively but not objectively better?

My favorite example that illustrates the difference between subjective and objective improvement is the study that looked at saw palmetto for prostate mediated urinary obstruction. As you get older, your prostate gets bigger, and it blocks urinary flow. Patients have increased frequency of urination and a decrease in urine flow. It takes longer to empty the bladder. These are all characteristics that can be measured with urodynamic studies.

Saw palmetto was alleged to improve the symptoms of a big prostate, so it was compared with placebo in a clinical trial. It didn’t work. No big surprise, as there was really no reason to suspect that it would be effective. Its failure as a therapeutic intervention was not the interesting result of the study.

They made the placebo brown and bitter, like saw palmetto, so it would be easy for both the placebo and treatment group to think they were in the treatment group.

For the first phase of the study EVERYONE received placebo. And everyone had a significant decrease in the subjective symptoms of the enlarged prostate. No objective change however. Their urodynamics and prostate size remained unchanged.

Then patients were randomized to placebo or saw palmetto and both groups had a further small decline in their subjective symptoms followed by a gradual increase in subjective symptoms such that by the end of the year both groups were back to baseline in terms of subjective symptoms.

At the time the patients were have subjective improvements, they were having objective measurements of their prostrate: changes size of prostate and how fast they could urinate. No change at all during the trial. Neither placebo nor treatment had objective improvement in their prostate.

Patents in the study thought there were getting a therapy and responded by thinking they got better. But they didn’t get better, right? Right?

This isn’t the only example where people think they are getting improvement when they are not. Patients randomized to either sham or real acupuncture (like there is a difference) had they same duration and severity of nausea and vomiting during cancer therapy, and both groups thought use of acupuncture lead to less nausea and vomiting and wanted acupuncture again (4). Their nausea wasn’t objectively better, but subjectively it was. They weren’t better. Were they?

It is an interesting question: If people are subjectively improved but objectively the same, are they better?

I don’t know. It depends in part on what you are treating. If the patient feels better but their cancer is progressing, then no. But pain? fatigue? depression? If the patient says they are better, then I suppose they are. Even if the therapy, is as best as one can tell, worthless magical thinking. Even if they are not better, they are better. Huh?

In part, this is due to the fact that some patients will have the result that is expected of them. People will think they are better and say they are better when, perhaps, they are not better, because it is expected of them. Some patients like to please their doctors, sort of a medical Stockholm syndrome.

The most amusing example of people reporting the experience that is expected for them is in Penn and Teller’s episode of Bullshit on Magnetism, where people were told they were being treated with magnets (they weren’t) but reported feeling effects from the ‘magnets’ none the less.

The most abhorrent example of people believing they are better when they are not are those who can walk again, for however short a period of time, after a faith healing. What happens in the audience of Benny Hinn is identical to what occurs in the office of the chiropractor, naturopath, homeopath and others of that ilk. When Benny’s supplicants fall over from the power of god, they are doing what is expected of them and what they want to do. It is a giant ideomotor effect. When a patient gets better from taking magic water or having their energies aligned, the same thing is happening. They are, in part, living up to the expectations of their ‘healer’. They are behaving as is expected of them and are better even though they are not.

I suppose that instead of relying on the patient, we could rely on the practitioner. Take the interpretation of the therapeutic response of out the hands of the patients and put it into the hands of the scientist. Problem is researchers, like everyone else, may see what they want to see or what they expect to see.

The archetype of this phenomena was the N-ray debacle for the turn of the century (5).

To quote from, because I am too lazy to re-write it,

“In 1903, Blondlot claimed he had generated N-rays using a hot wire inside an iron tube. The rays were detected by a calcium sulfide thread that glowed slightly in the dark when the rays were refracted through a 60-degree angle prism of aluminum. According to Blondlot, a narrow stream of N-rays was refracted through the prism and produced a spectrum on a field. The N-rays were reported to be invisible, except when viewed as they hit the treated thread. Blondlot moved the thread across the gap where the N-rays were thought to come through and when the thread was illuminated it was said to be due to N-rays.”

“Robert W. Wood of Johns Hopkins University to investigate Blondlot’s discovery. Wood suspected that N-rays were a delusion. To demonstrate such, he removed the prism from the N-ray detection device, unbeknownst to Blondlot or his assistant. Without the prism, the machine couldn’t work. Yet, when Blondlot’s assistant conducted the next experiment he found N-rays.”

The N-ray’s were figments of their imagination. N-rays didn’t exist. The bias of the researcher caused them to see what really wasn’t there. That kind of pernicious and unconscious bias makes all clinical studies done by those with a financial or intellectual investment to some degree suspect. Even what appear to be objective data may have unconscious subjective bias distorting the results. When I see an article that demonstrates daptomycin is non inferior to vancomycin for staphylococcus bacteremia, I find the results a little suspect even if published in the New England Journal of Medicine, since the study was funded by the makers of daptomycin. Unlike a homeopathic study done by homeopaths, at last I know that daptomycin has efficacy against S. aureus. But I have to wonder if the medication is as effective as is purported in the study.

One wonders how much of the positive published effects of ‘alternative medicine’ are N- rays redux with a coating of patient bias. Measured effects of ‘alternative medicine’ may have all the veracity of the N-rays, and when you remove the bias and expectation, you remove the prism, and the effects disappear. I suspect most, if not all, of alternative efficacy is due to bias and expectation, since the history of these modalities is that their effects are at the edge of detection and disappear once observer and patient bias are removed. Once neither the patient nor the researcher knows whether or not they are supposed to get better, they don’t. It is one of the reasons I remain profoundly skeptical of studies of alternative therapies that show positive results. Since the underlying therapy has all the validity of N-rays, can the results, even if positive, be more than self delusion?

Bias and expectation. They make the subjective aspects of response to medical intervention difficult to evaluate. When the patient reports improvement, can I trust it? When the researcher reports improvement, can I trust it? It is part of why all studies need to be understood in context from the foundations of the basic science up to the randomized placebo controlled clinical trial. When the patient says they are better or the researcher says the patient is better, are they better? Maybe, maybe not. Being an Ockams kind of guy, if the underlying intervention has zero biologic and physical plausibility, then the results are probably due to bias and expectation. That is the simple explanation.

I wonder if bias and expectation, both by the patient and researcher, accounts for most of the so called placebo effect. For objective criteria, placebo does nothing. For subjective criteria, placebo probably does nothing as well. Patients and clinician just think they are better when they are not. And the whole of non science based medicine is built on the foundation of convincing patients and ‘healers’ that the patients are better when in fact they are not.

If this entry sounds like I can’t make up my mind, I, can’t. Determining clinical improvement is not as simple as one would like. It has been said that the mark of an educated mind is the ability to carry two contradictory thoughts simultaneously. So maybe you can say that a patient isn’t better when they are. Or they are better when they are not.

The therapeutic modalities discussed in this blog can’t work because they violate all we know of the physical world. They don’t work because they have no effect on any of the underlying pathophysiology. When bias and expectation are completely removed from the intervention, the alleged effect of the ‘alternative’ therapy vanishes like an N-ray. When patients are made better by these interventions, they are not. Right? He says nervously in the dark, wondering if he is talking only to himself.

‘Alternative’ modalities will probably continue to persist no matter what science based medicine discovers. The fact that they are based on nonsense and are not effective will not prevent them from being perceived as effective. And if you think you are better, even if it is based on a lie, are you not better? The ethics of fooling yourself and your patients will probably be the topic of a future entry.

It is not as easy as one would like to determine if an intervention really, truly, honest to goodness makes someone better. It is why my conversation with the nurse was unconvincing for both of us. She perceives causality and improvement. She is better. I perceive bias and self-delusion. I know she is not better. Even though she is.

For now I will ask the question again. Its like a zen koan: if a tree falls in the forest and there is no one to hear it, does it make a sound? (Well, no, it doesn’t (3)). If the patient thinks they are better with a therapy when objectively they are not, are they better?


  1. Daptomycin in the treatment of patients with infective endocarditis: experience from a registry. Am J Med. 2007 Oct;120(10 Suppl 1):S28-33.
  2. Saw Palmetto for Benign Prostatic Hyperplasia. N Engl J Med 2006;354:557-66.
  3. The answer is no, it doesn’t not make a sound. This was definitely proven in an experiment by Bob and Ray over 40 years ago.
  4. Acupuncture Does Not Reduce Radiotherapy-induced Nausea, But Patients Believe It Does
  5. Blondlot and N-rays

Posted by Mark Crislip

Mark Crislip, MD has been a practicing Infectious Disease specialist in Portland, Oregon, since 1990. He is a founder and  the President of the Society for Science-Based Medicine where he blogs under the name sbmsdictator. He has been voted a US News and World Report best US doctor, best ID doctor in Portland Magazine multiple times, has multiple teaching awards and, most importantly,  the ‘Attending Most Likely To Tell It Like It Is’ by the medical residents at his hospital. His growing multi-media empire can be found at