We have previously seen that SCAM-use is associated with shorter survival of cancer patients. A new article now confirms this notion.
The investigators wanted to find out what patient characteristics are associated with use of SCAM for cancer and what is the association of SCAM with treatment adherence and survival. They thus compared the overall survival between patients with cancer receiving conventional treatments with or without SCAM and the adherence to treatment and characteristics of patients in both groups.
Their retrospective observational study used data from the National Cancer Database on 1 901 815 patients from 1500 Commission on Cancer–accredited centers across the United States who were diagnosed with nonmetastatic breast, prostate, lung, or colorectal cancer between January 1, 2004, and December 31, 2013. Patients were matched on age, clinical group stage, Charlson-Deyo comorbidity score, insurance type, race/ethnicity, year of diagnosis, and cancer type. Overall survival, adherence to treatment, and patient characteristics were the study endpoints.
The cohort comprised 1 901 815 patients with cancer (258 patients in the SCAM group and 1 901 557 patients in the control group). In the main analyses following matching, 258 patients were in the SCAM group, and 1032 patients were in the control group. Patients who chose SCAM did not have a longer delay to initiation of conventional therapies, but had higher refusal rates of surgery, radiotherapy, and hormone therapy. Use of SCAM was associated with poorer 5-year overall survival compared with no SCAM (82.2% [95% CI, 76.0%-87.0%] vs 86.6% [95% CI, 84.0%-88.9%]; P = .001) and was independently associated with greater risk of death (hazard ratio, 2.08; 95% CI, 1.50-2.90) in a multivariate model that did not include treatment delay or refusal. However, there was no significant association between SCAM and survival once treatment delay or refusal was included in the model.
The authors concluded that patients who received CM were more likely to refuse additional CCT, and had a higher risk of death. The results suggest that mortality risk associated with CM was mediated by the refusal of CCT.
This new evidence confirms previous papers: SCAM-use is associated with shorter survival of cancer patients. As it is based on a large sample size, its results are more compelling. They indicate that it is not SCAM per se, but the attitude of SCAM-users to conventional therapies that is the cause of the effect. As I have said and written hundreds of times: the most serious risk of SCAM is not a direct but an indirect one: the risk of neglecting effective therapies. Essentially, this means that better information targeted at vulnerable patients must be the way forward (one of the main ambitions of this blog, I hasten to add).
20 Responses to More evidence that SCAM-use is associated with shorter survival of cancer patients
Leave a Reply
This site uses Akismet to reduce spam. Learn how your comment data is processed.
1. Conventional cancer treatment gives you a 2.1% chance of surviving 5 years after the treatment. It drains your bank account. I’m not surprised that people choose alternative treatments. At least they have hope. What hope do they have with the conventional Big Three treatments?
2. I don’t think that we can expect radiologists to mention the other side of the coin regarding the success of conventional cancer treatments:
Conflict of Interest Disclosures: Dr Park reported receiving honoraria from Varian Medical Systems Inc and RadOncQuestions LLC. Dr Gross reported receiving research funding from 21st Century Oncology, Johnson and Johnson, Medtronic, and Pfizer. Dr Yu reported receiving research funding from 21st Century Oncology and serving as a consultant for Augmenix. No other disclosures were reported.
“Conventional cancer treatment gives you a 2.1% chance of surviving 5 years after the treatment.”
So where do the more than 80% 5-year survival data in the paper come from? Are the authors simply making up the numbers? Depending on the type of cancer, you’ll find very similar figures in lots of other studies. How about you tell us the source for your 2.1% survival datum. A video, perchance?
no, I think it was the Daily Mail
It is so interesting to note that Ernst previouslly KNEW about this truly awful survival rate of 2.1% after five years yet specifically chose to NOT mention it. THIS is why it is near impossible to believe anything Ernst says…because he always hides any potentially negative facts and figures on conventional medicine usage as well as any potentially positive information about CAM (or homeopathy). Clearly, this style of reporting is the exact opposite of a health scientific view.
1) I did not know this figure
2) I doubt its accuracy
BUT THANKS FOR PUTTING YOURSELF IN THE WRONG BY BLASTING OUT YET ANOTHER AD HOMINEM
Ad hom? Really!? What “name” did I call you? In actual fact, I was referring to a specific behavior…but heck, Ernst clearly sees boogeymen where none exist. It is further interesting that he seems to be able to DISH critique, but he certainly cannot take them. In the “world of Ernst,” is THIS also an ad hom?
oh stop foaming from the mouth Dana!
you said I knew these figure; as the discussion shows nobody knew them because they are false.
you also stated that “Ernst … always hides any potentially negative facts and figures.”
not ad hominem? really?
but thanks anyway for making such a fool of yourself.
There is big diff between describing a behavior and calling a person a “name.” Denial is a behavior that you just exhibited…is this an ad hom too? You’re getting so defensive these days. Such behavior typically suggests insecurity. I’m not surprised.
better defensive than wrong!
the 2.1% that you thought I knew and deliberately hid is wrong. as this is closer to the topic of this post, perhaps you want to comment on this issue?
“drains your bank account”. How? OK it might if you were an uninsured American, but not if you have a publicly funded health system.
Who has the financial axe to grind here? Do alternative practitioners give free treatment? Of course not, they have much to gain from making wild promises and ignoring evidence such as this. Is that ever admitted as a conflict of interest?
In fact, in most countries, it’s only the SCAM-providers who actually drain the bank accounts of patients.
I am wondering where you have got your statistic of conventional cancer treatment giving a 2.1% chance of surviving 5 years after treatment. Is it based on a review by Graeme Morgan and others, published in Clinical Oncology (the journal of the Royal College of Radiologists) in 2004? He concluded that the increase in 5-year survival attributable to chemotherapy in patients treated in the USA was 2.1%.
Coming back to Peter McAlpine’s comment, I have now had a chance to read the paper that seems to be the original source of the 2.1% 5-year survival statistic:
“The Contribution of Chemotherapy to 5-year Survival in Adult Malignancies”
Graeme Morgan, Robyn Ward, Michael Barton
Clinical Oncology 2004 (16) 549 – 560
Here is a link to the full text, though I am not sure whether it will work for non-subscribers as the publisher is Elsevier, who don’t make academic papers freely available online.
https://www.clinicaloncologyonline.net/article/S0936-6555(04)00222-5/fulltext
I have also found this link, which has the full paper in PDF format, though I would imagine that the owner of the Web page is in breach of copyright law:
https://www.chrisbeatcancer.com/wp-content/uploads/2011/12/contribution-of-chemotherapy-to-5-year-survival.pdf
As far as I can see the aim of the study was to shake up oncologists by drawing attention to the fact that their treatment may not be as effective as they think, and also to spark debate on where our priorities should be on research and treatment, given the costs involved.
Please note that he specifically looked at cytotoxic chemotherapy, not at radiotherapy, surgery or other treatment methods. Also, he only looked at the use of chemotherapy where the intention was to cure the patient. He was not looking at overall survival figures after cancer treatment, but instead he was trying to estimate the contribution to the figures that chemotherapy made.
He looked primarily at meta-analyses and overviews, rather than original research papers, and confined these to the period from 1st January 1990 to 1st January 2004. Since his end-point was 5-year survival, the most recently enrolled patient could not have been treated later than 1999. In practice, though, most clinical trials accrue patients over a period of several years and do not change the protocol during the recruitment period. Also, most reviews look at the totality of relevant data, not just the most recent, so they would be looking at research conducted much earlier. I think it would be fair to say, therefore, that he was analysing the survival of patients mostly treated in the 1980’s and early 1990’s, if not earlier still. This would certainly be in keeping with the dates of most of the 110 references.
He looked at data from cancer registries in Australia and in the US, which he analysed separately, to obtain the incidence of the 22 most important cancers in 1998. For each type of cancer, he then looked at the chemotherapy data to establish which subsets of the total would be eligible to receive chemotherapy (say, on the basis of the tumour stage, histological type etc.). He then looked at the results for that subtype to give the survival benefit attributable for chemotherapy (for instance a study may look at the addition of chemotherapy to surgery, and find an absolute 5-year survival benefit of 5%). He then multiplied the total number of cases of that cancer by the proportion being treated with curative intent who would be considered eligible to receive chemotherapy, and multiplied that by the percentage absolute survival benefit from chemotherapy. This gives an estimate of the number of patients with that type of cancer who are alive at five years who would not have have been alive without chemotherapy.
He then added up the totals for each of the 22 types of cancer examined (which between them account for 95% of total cases) and divided that by the total number of cancers registered.
Using data from the Australian Institue of Health and Welfare, he got 2.3%. From the American SEER database (Surveillance, Epidemiology and End Results) the figure was 2.1%.
There are some slightly complicated concepts here to get your head round if you are trying to follow what he is doing, but one very important thing to note was that the 2.1% was of the total number of cancers diagnosed, not the ones receiving chemotherapy, nor the ones where the chemotherapy was given with palliative intent (i.e. to improve quality of life or duration of survival in patients deemed incurable). Thus for melanoma he would include patients with a small mole removed surgically who never required any further treatment (which is most of them). Conversely, for testicular cancer, where in excess of 95% of advanced cancers are cured by chemotherapy, he included the majority with earlier-stage disease who were treated with orchidecomy alone or orchidectomy plus radiotherapy, to give 37.7% as the number benefiting from chemotherapy.
I can’t be bothered to go through his figures to calculate the number of patients benefitting from chemotherapy as a proportion of those that might actually be given it, but just a rough glance at the numbers suggests that it would be quite a bit mofre than 2.1%. And remember this paper was published 14 years ago, based on reviews from 1990 onwards, themselves based on studies in patients the majority of which were treated 30 – 40 years ago.
What about the importance of a small benefit? Suppose that the standard treatment for a cancer is surgery, and that surgery alone cures 60% of patients, and a study shows that the five-year survival is increased to 65% with the addition of chemotherapy, what does that mean in practice? The way I used to explain it to my patients is that on average, for every 20 people receiving chemotherapy, one would be alive in five years’ time who would otherwise be dead. Most of the rest have already been cured by their operation, so chemotherapy adds nothing. A minority will die regardless (arguably the chemotherapy may delay their recurrence and so keep them well for longer, but let’s ignore that). But for that one person in 20, chemotherapy makes a life-or-death difference.
When the data are explained that way, most cancer patients will opt for chemotherapy with an expected survival gain of 5% (in practice it is usually greater).
I should add that Graeme Morgan discloses that he was in receipt of educational grants from Varian and Astra-Zeneca.
Now this might be an interesting and valuable paper, reminding oncologists that what they do perhaps has less overall impact on cancer as a whole than maybe they think (since they confine their treatment to those expected to benefit from it). However, a quick Google search shows that it has been widely misunderstood and the conclusions quoted out of context. Mostly it seems to be given as a justification for declining conventional treatment.
And we should not forget how much cancer treatment has moved on. Better surgery with minimal-access techniques, new chemotherapy agents, better understanding of how chemotherapy works, better ways of controlling toxicity, using chemotherapy as a radiosensitiser, vastly better radiotherapy techniques (such as intensity-modulated radiotherapy and stereotactic radiotherapy), not to mention the deluge of targeted therapies hitting the clinic as a result of the insights that molecular biology has given us into what goes on inside cancer cells.
most helpful thanks;
do you happen to know realistic figures for the 5-year survival rates?
I have had a quick look at the Web page of the Institue for Cancer Research, which generally has good data presented in an easily readable way. They also let you drill down into more detailed views.
https://www.cancerresearchuk.org/health-professional/cancer-statistics/survival
Very roughly, the 5-year overall survival from all cancers in the UK is about 49% for men and 59% for women. This is a little below the European average, though there is quite a range from Bulgaria to Sweden.
thanks again
Could I assume that the patients who do not make the 5 year milestone may have died from everything from cancer to car wreck? Or is the death directly linked to cancer?
I think they were recording death due to cancer;
in any case, other causes of death would be the same in both groups.
jrkrideau:
In the paper referenced by Edzard, the authors were looking at all-cause mortality, not disease-specific mortality, so yes, it would have included everything from cancer to car wreck, including death attributable to treatment.
https://academic.oup.com/jnci/article/110/1/121/4064136
The survival figures given by the Institute of Cancer Research were adjusted to a standardised population in order to allow comparison between different time periods. They took account of changes in the age structure of the population, and also changes in the frequencies of occurrence of different cancer types (e.g. lung cancer is now more common than previously in women in England, buthas become less common in men, due to historical changes in smoking habits). They were also adjusted to exclude mortality from other causes. This is explained in more detail in one of the sources they reference:
https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(14)61396-9/fulltext
Graeme Morgan’s paper set out to calculate the proportion of all cancer patients (regardless of how they were treated) in which the survival at 5 years could be attributed directly to chemotherapy. His figures were extrapolated from trial data in multiple cancer sites which he applied to recorded cancer incidences.
It does get a bit confusing trying to compare survival figures when there are so many different ways of calculating them. It is also important to remember that not all studies are based on directly comparable populations. Finally, while survival figures from trials are useful in comparing the effectiveness of different treatments, caution should be exercised in applying these figures to individual cancer patients.
interesting article in THE TIMES on this subject:
https://www.thetimes.co.uk/article/andrew-wakefield-the-vaccine-scaremonger-whos-snared-supermodel-elle-macpherson-5tgzdth08
[sorry for the pay-wall]