A recent study with almost 2 million subjects evaluated the effectiveness of Complementary Medicine in fighting cancer. CM is that supposedly harmless stuff like yoga and essential oils and homeopathy taken in addition to standard, tested, genuine medicine — stuff that you’d think wouldn’t hurt (although it wouldn’t actually help, either, except maybe in your emotional well-being), except, ooops, it did.
Findings In this cohort study of 1 901 815 patients, use of complementary medicine varied by several factors and was associated with refusal of conventional cancer treatment, and with a 2-fold greater risk of death compared with patients who had no complementary medicine use.
Meaning Patients who received complementary medicine were more likely to refuse other conventional cancer treatment, and had a higher risk of death than no complementary medicine; however, this survival difference could be mediated by adherence to all recommended conventional cancer therapies.
That last paragraph is important: sure, aromatherapy isn’t going to harm you unless you use it as an excuse to avoid conventional treatments. And, unfortunately, from the statistics it seems that a lot of people were doing that, giving the overall group a 2-fold greater risk of dying. I think it’s important to note that this is a statistical assessment — supplementing your chemo with traditional Chinese medicine won’t kill you directly, it just puts you in a group that contains many members who will defy medical advice, and end up dead earlier.
I tried to poke a few holes in their conclusions, which is fairly easy to do in this kind of study, but the authors kept foiling me. One concern I had was that maybe their results were biased by the fact that people whose conventional treatments were failing were more likely to turn to desperate, unlikely treatments — so the results weren’t so much “CM causes people to neglect good treatments” as “failing treatments cause people to try CM”. They had an answer.
As patients receiving CM were more likely to be female, younger, more affluent, well educated, privately insured, and healthier, we hypothesize that our sample was biased in favor of greater survival for patients who used CM (vs no CM).
I guess it makes sense. If you’re intentionally taking a placebo, you probably think it is actually going to help you, and it’s that delusion that’s going to make you more willing to turn down effective, advantageous therapies, especially if they’re going to cause you more discomfort. One thing about CM is that it’s always mostly pleasant and doesn’t challenge the patient in any way. It may be doing harm by increasing complacency about a deadly disease.
How about it’s because CM is expensive, hard to find and really, really painful?
Susan Montgomery@1: Is CM hard to find ? As a possible counter-example — and whilst this is in France — I’ve given up on only doing business with pharmacies which do not stock homopathetic placebos. I do not recall even seeing a homopathetic-free pharmacy in years. I assume the homopathetic stuff is expensive, and being water, should normally be pain-free.
And I also presume the is a typo(? brain-fade?) for something like “CM is … rarely painful but SBM / EBM can be really really painful”.
(An alternative is a misunderstanding of “CM” — it is Complementary, not Conventional — Medicine; i.e., CM (as poopyhead noted) adds woo-woo “treatments” to EBM / SBM — albeit the “adds” seems to turn into “replaces” as per the OP. (SBM is Science-Based Medicine, EBM is Evidence-Based Medicine.))
Sorry, I meant “Conventional Medicine” Wish my laptop had a way of determining that I had insufficient caffiene to make comments.
Ok, thanks for clarifying !
My computer also lacks such a detector. The mildly deranged penguin can detect the problem, but whilst being beaten about the head with a lump of cheese by a feathered tuxedo smelling vaguely of herring is effective, I’m not all certain what it is effective at…
I wonder about two other hypotheses (which may, or may not, be covered in the article):
(1) Either or both of {cancer} {treatments for cancer} increases the patients’ vulnerability to impurities and/or doseage inconsistencies found in “natural” medicines not prepared under fully scientific supervision and thus results in less-favorable outcomes.
(2) The population of patients who willingly used CM is also skewed toward late DETECTION of cancer because its first reaction to apparent illness was CM and not medical intervention, meaning that their cancers themselves are not part of the same population of cancers subjected to the “conventional-medicine-only” chain. (Perhaps worse, those engaged with “woowoo” prior to the cancer diagnosis may have pressured their providers to perceived less-invasive treatment regimens with the specific intention of using CM — that is, despite best intentions of the study’s authors there really was use of CM as a substitute for “conventional medicine,” but not a complete substitute.)
tim minchin’s adage still holds:
“alternative medicine has either not been proved to work or been proved not to work. do you know what they call alternative medicine that’s been proved to work? medicine.”
I would be interested in reading the study. Would you please refer me to the site?
@7 rvoss
<blockquote=>I would be interested in reading the study. Would you please refer me to the site?
I thought the same thing. Link is below.
https://jamanetwork.com/journals/jamaoncology/fullarticle/2687972
From Cancer Patients Who Use Complementary Medicine (Refuse Conventional) Twice As Likely To Die, Says Study, the study is:
There does not seem to be a paywall, but if you want the PDF version, you do seem to have to register.
One thing not mentioned is the danger of drug interactions. A lot of herbal extracts, when interacting with specific cancer agents, can cause serious side effects, and given the serious pain meds most cancer patients are on, can be missed until too late.
I’ve been asked about supplements, herbs, and all that throughout my treatment. I don’t take any, so it’s not a problem for me, but any competent oncologist should be asking about that.
I do get massages several times a week after radiation, those are more than welcome, and help in a general sort of way, along with the specific for those of us also dealing with sciatica, which radiation makes worse. We’re registering for a 3 month exercise program, and yoga is often recommended, not for woo reasons, but because it’s an effective way to build flexibility again, which kinda goes out the window after 7 months of dealing with cancer, and yoga is an easy form of exercise to tailor to the limitations of cancer patients. Right now, what passes for my upper body strength is appalling, and strength training alone won’t work, because of nearly no flexibility there.
People who are prone to taking supplements and herbal remedies for everything tend to be very uninformed about cancer in general, and they are also often people who are big believers in just the right diet, etc. There’s still a perception that these are never harmful in any way, where cancer treatment is harmful.
There’s also a distrust of OTC meds. I was talking with a woman in radiation the other day, and we got to talking about the nightmarish acid reflux and nausea which sets in around infusion cycle 5 and never goes away. I mentioned that I now have a lav littered with Zantac, Prilosec, and Gaviscon. She mentioned how the recommended dose never works, and it doesn’t. I told her I found that taking 3 to 4 times that dose worked for me, and right away she was worrying about what that does. I’m sure it’s not all that good for you, but it’s what works, so that’s what you do. There’s a ton of that kind of thing in cancer treatment, nothing is ever easy or simple, and you have so many damn interactions and mini-overdoses to concern you, it’s not hard to see why some people desperately want to believe there’s an easier way.
Much as I despise complementary quackery and the results of the study are in the direction I would want, I do not find it convincing. To my eye, it looks like it could be data dredging (p-hacking.)
The database has roughly 1,900,000 cases, but only 258 (0.01%) are selected as having an unspecified amount and unspecified modality of CM combined with a number of other characteristics. 18% (roughly 50) died within 5 years. Pretty small numbers, compared to the impression you get from 1,900,000! Working with such a tiny portion of a sample could mean that some unknown selection bias invalidates the results.
I strongly doubt that only 0.01% of cancer cases use CM: I’d think the actual number was much higher but for some reason did not make it into the records. Whatever reason the number was low, because the excluded population might be quite different, the results are not conclusive.
And while I firmly believe essentially all CM is fake, the assumption that CM is CM and the lack of identification of the specific CM involved (unlikely the identification of the specific cancers) also makes this unconvincing: any proponent of one can blame the negative results on the others. And we do know that some CM is genuinely harmful (see birthwort, for example), so while there is discussion that the harms are from refusing medical treatments, they could also come from some CM directly.
And it’s easy to see how the conspiracy-minded could defend CM from this study by claiming that people who routinely use CM get less or weaker cancer and thus don’t even get onto the case list unless they have a particularly severe and intractable cancer. But that’s not a flaw in the study itself: you can always construct conspiracy theories.
I suspect that the same database could be “analyzed” by pro-CM authors and find the opposite results very easily. It’s just too weak a study to be really convincing.
has Orac discussed this? Curious before checking, he has a reputation of saying “complementary medicine” is neither.
@Caine, I’d get with doctor about the proton pump inhibitors and histamine h2 blockers, largely because they can lower your magnesium levels dangerously. My GERD is severe enough as to have caused extensive dental damage, my magnesium deficiency remained months after I stopped taking proton pump inhibitors (Prilosec) due to financial constraints. When I was hospitalized with a thyroid storm, my magnesium levels were dangerously low and required prompt replacement.
Alas, in the United States, our medicine, especially OTC medicines is rife with placebos. An example being phenylephrine, such as Sudafed PE, replacing the signature and ID required pseudoephedrine based (and effective) variety of decongestant. It was found that phenylephrine is precisely as effective as a placebo.
https://www.ncbi.nlm.nih.gov/pubmed/19441605
https://www.ncbi.nlm.nih.gov/pubmed/19230461
I’ve taken to discussing with doctor any medication that I take, counting OTC drugs. Herbals, I reserve for tea, for the most part and they’re imbibed for the simple pleasure of the taste, not a desired medicinal effect. For prescribed drugs, similar conversations are engaged in. As an example, doctor prescribed hydrocodone for pain caused by a herniated disc, he also had prescribed a muscle relaxer. Muscle relaxers are CNS depressants, that’s their nature and method of action. Opioids are CNS depressants by their very nature, as a side effect and via their mechanism of action. I demurred filling the muscle relaxer and experience proved me correct, as the low dosage of hydrocodone did stop muscle spasms in my calves and feet quite well.
Where some herbals may indeed have medicinal value, standard medicines were derived from them and have standardized effective doses. That makes dosing reliable and treatment effective. One Chinese traditional herbal showed good effects as an antimalarial drug, which was duly researched and developed into a highly effective antimalarial. This had come at a time when concern was rising that the 4-aminoquinoline class of drugs was showing decreased efficacy, due to the organism becoming tolerant of the drug. Now, we have artemisinin and additional drugs are being researched that are derived from the novel class.
https://www.pacificcollege.edu/news/blog/2014/07/04/chinese-herb-transformed-western-medicines-cure-malaria
Thus providing proof of astro @6’s quote.
While if it were a holiday weekend and I were constipated, I may boil some corn silk, the rest of the time I’d go with real medicine. While chamomile tea can and does lower my blood pressure, one amlodipine per day and half a metoprolol twice a day lowers it more predictably and reliably and that’s important when one has aortic dilation.
wzrd1:
Thanks for your concern, but it’s unnecessary. I certainly do not take such doses every day, or even three times a week. Phenergan has helped a great deal, as with me the nausea and acid are hand in hand. I have labs taken every single week, and have had for the last 7 going on 8 months. My levels of everything are constantly scrutinized (including magnesium), and my oncologists know everything they need to know.
As I said, everything is more complicated with cancer. The reflux is a result of a specific cytotoxin, and if it reaches a serious enough point, you simply get pushed back for a bit on treatment, or have your dose reduced. All that said, if I’m at home and can’t sleep because nausea/acid is so bad (and I’m not a fan of vomiting, especially all night), I do what’s necessary to get some sleep, then deal with it better in the following days. Believe me, there isn’t one aspect of my life or body which is not under constant scrutiny.
Oh, and as for pain meds, that’s not simple either. I’m in pain management with my pain clinic; and I have taken hydrocodone and cyclobenzaprine for ages. They work fine for me. Anyroad, the hydrocodone had to go, because the acetominophen was lowering my white count so much, I kept getting pushed back on chemo. So, morphine it is.
Honestly, you cannot compare any other illness with cancers of any type. The situation is completely different, and I hope you never get to experience how different. It’s as well to realize that each individual is different, and will not reliably conform to all meds. I have almost universal contrary reactions to everything. It’s a dicey business trying to act the expert.
@6 “and everything else is a nice bowl of soup and potpourri” Dara Ó Briain.
mikehuben@11 —
There’s no indication that there was any significance fishing in this study, and I think it’s important not to criticise it for flaws it doesn’t have. Significance fishing means running your data through lots of different different measures to find statistical significance, which will happen by chance when you include enough variables (even p<0.001 is easy to achieve if you churn through a ton of variables). Since the study used a small number of well-defined clinically relevant outcomes and risks, then the only way they could have done significance fishing is if they deliberately selected their cohorts to maximise the difference in outcomes — which would be outright fraud.
Having said that, there are some major flaws in this paper.
Firstly, as you noted, they only identified 0.1% of the cohort as CM users. This is ridiculously low. The fault is not really theirs — it stems from using deidentified data from a cancer registry, which means you can only test data already collected by other people who were not developing their data structure with this research question in mind. Which is why they identified cases who had been flagged with the label “Other-Unproven: Cancer treatments administered by nonmedical personnel”…which doesn’t even necessarily mean CM. If the patient had massage for lymphedema, for instance, it might have been tagged in this entry just because the person administering the treatment wasn’t a registered physiotherapist or therapeutic masseur. So here’s a problem from the get-go, and it’s not one they can get around because it’s hard-wired into their available dataset.
Secondly, and this really irks me, this is NOT a COHORT study despite the authors identifying it as such. In a cohort study, you recruit a cohort of people, find out their risk factors at recruitment and follow them over time to see if there is a difference in outcomes. A famous example is the British Doctors Study which ran from 1951 to 2001 (yes, fifty years!) and was one of the key pieces of evidence that smoking really did cause lung cancer and was unlikely to be mere correlation.
This study is in fact a NESTED CASE-CONTROL study. This has major implications for evidentiary power as case-control studies are below cohort studies on the hierarchy of evidence, largely because they are retrospective rather than prospective, but also because they rely on going back to identify risks after the outcome has been measured — sometimes years later — which is obviously fraught with confounding potential, or in this case relying on data collected at the time of registration by people who weren’t trying to answer this research question.
Now, overall I find the paper’s conclusion plausible as I have known people who delayed or avoided demonstrably effective conventional treatments in order to pursue CM. Steve Jobs is one famous example. So it makes sense. Also, when they accounted for people who undertook conventional treatment without delay, the increased risk of death became non-significant. The other thing that makes me think there is some validity to the paper is that two subgroups showed no difference with CM use: prostate and lung cancer…and in the time that the data was collected the conventional treatments for these cancers had poor outcomes, so delay or avoidance was less likely to affect overall survivability (I doubt the same result would happen in a study today as recent developments in oncology have shown extraordinary improvements).
So I guess the problem for me is that this paper aligns pretty well with what I think is a reasonable conclusion (although I suspect that a doubling of mortality is probably an excessive estimate of the real risk), but its basic design makes it impossible to add much evidentiary weight to my a priori opinion. If someone really wanted to find a better answer to this question, they should design a good case-control study (which would be quick and cheap) or if they can convince a funding body to pony up some major cash, they should do a proper cohort study and recruit people now and follow them for 5+ years. Cohort studies still can’t prove causation, but they’re a lot better than case-control studies.
@17 chrislawson
To me the take-home lesson was that if you follow medical treatment recommendations the use (or non-use) of CM doesn’t significantly affect outcome.
Self-correction: actually, despite my earlier statement, this is a cohort study rather than case-control study. I was thrown off because they used matching and retrospective analysis, which are usually found in case-control studies. The essential points about the study’s flaws remain unchanged.
Speaking of quackery and filth and gullible rubes being separated from their cash, check out the Great Tom Brady of the Deflatriots–er–Patriots, specifically his personal trainer/advisor/medical consultant/”body coach”, whatever. Alex Guerrero has a bit of a stinky history with the FTC involving some fraudulent cancer cure claims, a false claim of a medical doctor license, and general, filthy quackery like that. Brady even told the press very recently that he has “armor” (specifically, exactly his word) to pre-protect himself from concussions.
They have this bullshit TB12 “free” app that is free for a very short while, then $199/year or $19.95 a month; they sell magic supplements, magic “recovery sleepware,” other apparel (of course!), purified water with certain magic electrolytes, and other magic beans bullshit quackery.
The most infuriating part of this is Brady has more money than Allah and God combined and has no need to fleece any more rubes out of any cash, and this bullshit TB12 thing is specifically that. Rubes, money–opposite polar situation.
Cannot resist this stinger: Brady is a Drumpfh supporter . . . OF COURSE . . . and he aims his quackery precisely there.
Note that some generally harmless CAM (as well as some perfectly effective and usually harmless OTC drugs and some foods) can be deadly when used along with other medications. For example, St John’s wort and grapefruit juice both interfere with hepatic enzymes and make chemotherapy ineffective or excessively toxic. Vitamins can make anticoagulants ineffective or make you overdose on your prescribed dosage.
Final random point about CAM: Any medicine–natural or otherwise–that is potent enough to be effective has side effects. Sometimes the side effects can actually be useful or desired (as in a certain failed antihypertensive), but if a substance can alter your body’s chemistry, it should be considered to have potential risks as well as potential benefits. No matter whether it is all natural or made from pure elements by the maddest of Merck’s mad chemists.
chris61@18–
One of the findings of the study is that those who used CM but did not delay or refuse conventional treatment had similar outcomes. But as I said, the weaknesses in the study mean that you can’t really say that it adds much weight. Essentially, the study is not strong enough to change anyone’s mind.
One obvious example (which has been mentioned in comments above) is that it makes no distinction between types of CM — homeopathy is obviously extremely unlikely to have any direct deleterious effects while some “natural” herbal medicines can be dangerous even to healthy people, let alone those having cancer treatment.
Also, it ended up being a small study because they selected only 258 cases out of 1.9M in the registry, so when adjusted for treatment delay/refusal the risk of death was 39% higher than in the non-CM group, but this was not statistically significant. So if the study had been larger, it could have shown a significant difference. In other words, this study doesn’t even really clear CM because you’d want any big difference in outcome (and I happen to think 39% increase in mortality is a big difference!) to show up as a statistically significant signal.
chrislawson @ 17:
I think we are very much in agreement, with our reasons strongly overlapping.
However, you wrote: “There’s no indication that there was any significance fishing in this study”. Absence of evidence is not evidence of absence. I could envision that project as looking at a subset, realizing that there were confounding variables and then reducing the subset sequentially variable by variable until significant results were achieved. Limited by the decreasing sample size. Maybe that is not strictly significance fishing, but it is not a good practice. So I still think it could be data dredging: no proof, just a possibility.
The other problem with quack cures is their interference with real treatments. My brother-in-law is an oncologist and tells me that he always asks patients receiving chemotherapy if they are on any “natural” remedies. Apparently anti-oxidants in so-called superfoods can actually reduce the effectiveness of chemotherapy drugs and he has to adjust doses accordingly.There is also some evidence that a diet high in anti-oxidants can actually increase the risk of cancer.
chrislawson @22: “homeopathy is obviously extremely unlikely to have any direct deleterious effects”. Well, honest homeopathy, i.e., administration of pure water, is; but (at least as of several years ago, when I last looked at stuff on the drugstore shelves) some substances marketed as “homeopathic” not to have been diluted (or, not much)—I think I recall some echinacea preparation in that category (though on what grounds I decided it wasn’t diluted, I don’t remember…maybe because it was a powder?…or maybe I just assumed it claimed to be homeopathic because it was shelved that way?).
leerudolph@25, Yes, there are homeopathetic gunks which do contain active ingredients. As RationalWiki observes (references / citations at the link):
leerudolph@25–
I was trying to keep my comment short! But you’re right. CAM is largely self-regulated, which means you can’t always be sure of what you’re actually purchasing, especially if it’s not from a manufacturer with a valued brand name to protect.
This Australian review study (link to pdf) reports that 6-7% of CAM products don’t meet labelling requirements. The problems include:
— accidental substitution of one herb for another when there is confusion due to different common names in different regions (e.g. Podophyllum emodi instead of Gentiana)
— deliberate substitution of one herb for another when the labelled herb is unavailable or just cheaper (e.g. Aristolochia fangchi instead of Stephania tetrandra)
— undisclosed addition of ingredients from endangered animals (e.g. snow leopards)
— undisclosed addition of pharmaceuticals to produce the desired physiological effect (e.g. adding phenytoin to an anti-epilepsy product)
— undisclosed inclusion of heavy metals (too many examples!!!! — “arsenic, lead, mercury, copper, iron, thallium, cadmium, manganese, nickel, and zinc”)
— undisclosed inclusion of other toxic compounds (e.g. toad venom)
All of these examples are attached to cases of people suffering serious, sometimes fatal, side effects. (Except the snow leopards — that’s not dangerous to consumers, but to wildlife conservation.) I’m trying to dig up a recent paper that audited the complementary products on sale and found that some appallingly large percentage had NONE of the labelled ingredient in it — usually it was just ginger powder. Can’t find it at the moment, will post when I do.
@chrislawson: Not to mention that sometimes “alternative” medicine is effective because it contains unlabeled conventional medicine. For example, PC-SPES raised some interest for a while because it was an apparently effective “complementary medicine” for the treatment of prostate cancer. It turned out to be effective because it contained DES, a “conventional” medicine that can be helpful in some prostate cancers. Also, an NSAID and coumadin. All useful conventional medications, but not safe to take without monitoring in random doses.
Ahem, oops, you said that already. Sorry!
Imagine if Steve Jobs, who apparently had a low grade neuroendocrine tumor of the pancreas, easily cured with cancer didn’t try all his religious based therapy where we’d be now, imagine no religion…
I meant easily cured with surgery (as opposed to the usual nasty high grade adenocarcinoma we see in that organ)…imagine how much head and neck cancer, penile, vulvar, cervical and anal cancer wouldn’t happen if everyone got gardisil…
Well duh! Of course a science based study is going to find in favor of science based medicine instead of woo and snake oil.
dianne@28–
Yes, and even when CAMs have reasonable evidence for them — for instance St Johns Wort for depression — it’s almost certainly because of some pharmacologically active component — in St Johns Wort the reports on benefits and side effects make me feel that there is an as-yet-unidentified SSRI in it, the same class as Prozac.
The cam modalities that bother me the most are Homeopathy and Herbalism, for opposite reasons. Homeopathy (as everyone here knows) is simply water or sugar pills, and as such, useless unless you’re dehydrated, or suffering from hypoglycemia. Herbal remedies, on the other hand, might conceivably be effective — or they are bound to have some effect — but the effect if any is unproven, and the “Remedies” are untested, with the quantities of whatever active ingredient often being unknown, and the quality questionable. You might be taking St John’s Wort for depression, and not getting any of the supposed herb. There just aren’t any enforceable controls.
Why are we lumping yoga — which is basically a bunch of fancy stretching, something that is beneficial to the body and, perhaps, even the mind — in with Certified Woo?
Yoga ain’t gonna cure a damn thing, but it WILL increase your flexibility and strength!
Caine, thanks for the info about flexibility. It has been over 6 weeks since my cancer treatment ended and I started healing from the radiation burns and chemo infusions. I was surprised to discover how painful my right leg was when I did the slightest exercise or normal activity such as walking outside. Way less flexibility, I noticed. Gradually I have built up my tolerance and flexibility but the weird response to exercise still remains at a low level. My two infusions with chemo, which lasted 5 days each, were markedly different in body response, so I am lucky I did not get as many as you. The second infusion caused magnesium and potassium depletion and a long lasting lack of appetite, diarrhea, and general distaste for food. First infusion just caused me awful mouthsores. So it seems the reactions migrate to worse and longer lasting, as you say. I understand the scrutiny of everything that you detail. I have the same mindset. When I returned from vacation where we ate at restaurants for 5 days straight I had majorly swollen ankles. I worried it might be the chemo had done permanent damage to my kidneys. But I waited as I got back on my regular diet and it went away quickly. Maybe it made me less able to process salty foods but I think my kidneys are okay! Thinking of you all the time, wishing you the best. Glad I am now in the waiting stage to see if my cancer was cured by this nightmare treatment. A treatment that took all my focus and energy to comply with, so much so I did not even consider the yoga they offered or the facials, etc. It occupied all my attention. Took my family begging me to not quit during that last week. I trusted them and kept going. Glad I did.
wzrd1@13: Thank you thank you thank you for those citations! My personal experience of the “PE” decongestants is that they don’t work, but I’ve had people tell me this is just me being whiny. (Why, yes I am very whiny when you tell me I can’t buy the drug that actually does what it says because your computers are down or it’s too late or it’s Sunday.)
Now I have papers to show why I am throwing away the PE crap and will never buy it again. I will drive all over the city to find the 24 hour pharmacist before I bother with that crap again!