Aug 31 2012
Questioning Whether Psychotherapy and Support Groups Extend the Lives of Cancer Patients
You are currently browsing comments. If you would like to return to the full story, you can read the full entry here: “Questioning Whether Psychotherapy and Support Groups Extend the Lives of Cancer Patients”.
21 Responses to “Questioning Whether Psychotherapy and Support Groups Extend the Lives of Cancer Patients”
“repeatedly failed to show that psychological interventions were ineffective” – good job we don’t believe in Freudian slips here!
I know that this is a bit broader than the scope of the paper being critiqued here, but every once in a while, someone tries to argue that “stress kills” through everything from increased blood pressure to the body bottling up “poisonous stress products”. Usually, this is being cited either to try to convince someone to improve their work-life balance or to sell some form of relaxation product (drugs, biofeedback, spas, etc), but it seems like there’d be some basis for this. By having the patients attend therapy sessions, one would expect at least some improvement in overall health simply from reduced stress of having people to talk with, although I could also see this being masked by the large number of informal methods people have to relieve such stress.
I haven’t specifically studied people with cancer (yet) but I can tell you that for people living with HIV, depression, substance abuse, and just plain denial are definitely associated with non-adherence to antiretroviral therapy. People have to come to terms with the fact of their diagnosis and, to use a highly technical term, get their heads together, in order to cope effectively with the demands of self-care, which in this case mostly means taking pills very faithfully.
Many forms of cancer treatment also require fairly demanding — sometimes even more demanding — self-care behaviors, e.g. hematopoietic stem cell transplant. It doesn’t require any mystical mumbo jumbo to argue that there is a survival advantage associated with treatment of depression and perhaps motivationally focused behavioral counseling or other behavioral health interventions. I don’t know what the evidence shows about this when adherence is considered as a mediator, but I would consider the existence of an effect to be highly plausible. This is, obviously, a different argument but they could easily get mixed up.
This would seem to be evidence for Dr. Gorski’s “central dogma” that wishing it makes it so.
What happens to the thinking process of people who have advanced degrees and have likely completed coursework in statistics? Did they simply cram for the test and not really learn anything? I passed on an advanced degree because I did not think I could honestly pass a statistics class (being weak in math to begin with), but at least I am smart enough to read and heed the words of those who have successfully educated themselves.
You don’t have to know everything, but you do need to know who to listen to, think critically, and be a good skeptic.
“I passed on an advanced degree because I did not think I could honestly pass a statistics class . . .”
We call this the problem of self-efficacy. Believe me, two semesters of statistics really isn’t all that hard, at least not if it’s competently taught. In fact I have a two hour presentation that gets the basics across quite successfully for most people. Bringing more competence at causal inference to the masses would definitely make the world a better place. I am 100% sure that Janet Camp could do it if she tried.
That said, Gaussian statistics as commonly taught leaves a lot of room for erroneous thinking. SBM is getting more Bayesian all the time, I’m happy to say. Basically, that means that an observation that at first blush seems unlikely to arise by chance probably did arise by chance after all if it conflicts with what we already have reason to believe is likely. Our conclusions should be guided by the total weight of the evidence, not a single observation.
[...] Report: Cancer is now the leading cause of death in Iowa, Iowa City (KWWL) – Although cancer mortal…ontinue throughout the state, a report published recently indicates that cancer has surpassed heart disease to become the leading cause of death in Iowa Elizabeth Hurley on Breast Cancer Awareness [...]
[...] lung cancer that showed promise at ASCO 2012 in the next series of podcasts. Look for those soon. Here’s the last portion of Dr. Socinski’s presentation, which covers he last portion of Dr. Socinski’s presentation, which covers two important trials in small cell [...]
I renew my offer to publicly debate the hypothesis that psychotherapy promotes survival. I’m sure we could sell a lot of tickets and I would donate my proceeds to a charitable cause offering psychosocial interventions to improve the quality of life of under-served cancer patients and their caregivers.
We instinctively feel certain that psychosocial support should help quality of life, but how firm is the actual scientific evidence for that? How can reporting biases be factored out in studies of interventions where adequate blinding is virtually impossible? (This is a critical matter for other areas, such as with pragmatic studies of CAM methods, and in placebo research.)
@ pmoran,
“We instinctively feel certain that psychosocial support should help quality of life, but how firm is the actual scientific evidence for that?”
It is ironic that you should say this when you have been suggesting, without actual scientific evidence, that we should moderate our criticism of CAM because it fills the unmet needs of patients.
Harriet, nowhere, absolutely nowhere, have I suggested that CAM adequately “fills” unmet medical needs. I have been contending that they are a rarely mentioned major reason, if not THE main reason, why people turn to CAM.
I think you yourself have agreed that that CAM may be helpful for some through placebo and other non-specific influences. My question relates to how we can reliably measure such benefits, or prove that they don’t exist at all, if they are all illusion, as some skeptics claim.
To me this is important. It influences how I am prepared to talk about and react to CAM. (I have made it quite clear on many occasions that I don’t expect any form of CAM, apart from the odd herb, to have any significant effect upon the disease processes that provoke much CAM use. But can their use contribute to symptom relief? )
@pmoran,
“But can their use contribute to symptom relief?”
If that question is essentially different from suggesting that CAM helps fill unmet medical needs (partially if not “adequately”), I don’t see how.
@pmoran – first you said quality of life, now you are saying symptom relief. Which are you asking about?
I’m obviously not a psychologist, but I thought they had tests that were valididated to measure certain issues. Such as the Beck Depression Inventory. I don’t understand, are you calling into question the validity of ALL psychotherapy because it can’t be blinded to the same extent that drugs can?
Are you saying that most CAM research has the same level of proficiency as most pyschotherapy research?
Certainly, I know there is some junk psychotherapy research, but?
I have posed a question to Dr Coyne. I was not criticising psychotherapy or support groups, which I allow are very plausibly helpful for any kind of distressed person.
I truly want to know whether such obviously well-informed and rigorous researchers as he have any idea how to avoid reporting biases in studies that look at subjective outcomes but which are inevitably going to be imperfectly controlled. I think this is going to be the next major challenge for those of us involved in the “science vs CAM” debates, whether present parties realise it or not.
Harriet is massaging words and meanings and still selectively quoting. She has not objected to my comment that she herself accepts that CAM can help some people via non-specific influences, but she is trying to make that mean something else when I say something similar.
@pmoran,
I “accept that CAM can help some people via non-specific influences” in the sense that placebos can “help” people with subjective symptoms. I object to providers knowingly giving placebos to patients. I have no objection to patients trying CAM or any placebo as long as they are given unbiased information so they can give informed consent. I’ve said this many times in many ways. If I am massaging words and selectively quoting, I’m not the only one.
Isn’t it so funny how all of CAM only ever “treats” subjective symptoms, but never anything objective?
@pmoran – oh I see, I took your initial question as rhetorical…a question that is specifically meant to imply a particular conclusion, but it was an actual question.
I don’t think that was HH’s doing, I had the same interpretation before she responded, and then your switching between quality of life and pain measures confused me further.
As a laymen, Psychological research has always been fascinating to me…so many pitfalls to avoid. I’m glad that SBM found Dr. Coyne to write on the topic.
Here is one difference I see between good psychotherapy methods (there’s a lot of bad ones out there) and CAM methods. Good psychotherapy attempts to be targeted directly at the problem. Are you worrying too much? Let’s look for the most reliable methods to reduce worry. When successful, the patient ends up better equipped to deal with similar situations in the future and they become more self reliant.
A CAM method that “works” by distracting the patient from their ruminations on symptoms or by making the patient feel that their symptom has been explained so that the can move on…has not better equipped the patient and it sometimes makes the patient dependent upon the practitioner (such as with some supplements and chiropractic care)
For me, I’d rather learn cognitive techniques, then have possible solutions there in my head when I need them, then have to rely on some outside entity to distract me or provide closure each time I have trouble.
Thanks to all for the provocative and thoughtful comments my blog post has elicited. I’ll try to respond to them, both briefly here and in subsequent blogging. Call me on any points that I miss.
DugganSC raised issues about the role of stress in ill health and implicitly the possibility that stress reduction could slow the progression of cancer and improve its outcome. A lot of the psychosocial interventions for cancer patients assume that stress reduction will improve not only psychological outcomes, but physical ones. However the role of stress in cancer is not well-established. A major meta-analysis by Chida and colleagues claims a role for stress in cancer, but when one removes from consideration fraudulent data from the Eysenck and Grossarth-Maticek studies, the hazard ratio for psychological factors in incidence and negative outcomes in cancer drops to near 1.0.
I know this sounds like an outrageous assertion about fraudulent data, but you can read more in my critique in Nature Reviews Oncology (https://dl.dropbox.com/u/23608059/comment%20on%20Chidaf.PDF) and I will devote a future blog to the need to get fraudulent data out of the discussion of the stress cancer link. Much of the psychoneuroimmunology literature concerning cancer is simply wild speculation and unsubstantiated simplistic assertions about the role of immunology in the progression of cancer and of links between body and mind via immunology.
cervantes raised issues about psychological processes affecting physical health outcomes by way of adherence. I have no problem with this idea: it is readily testable and there are lots of relevant supportive data. The mechanism of adherence is observable and disconfirmable. My colleagues and I recently completed a trial of problem-solving training for adherence problems in HIV/AIDS and we affected not only patients’ problem-solving but their viral load. But we don’t assume there was anything mysterious going on there. Some of the nursing studies that David Spiegel cites as evidence of support processes affect in cancer outcome are more properly and more parsimonious explained by deliberate, targeted effects on adherence and symptom monitoring.
There are some aspects of cancer treatment that are more affected by adherence, such as infections and other complications following surgery. There are some cancers that also pose particular issues of adherence, but much of cancer treatment is just a matter of patients showing up and passively accepting the treatment that is offered and in so there are no prominent adherence issues to be solved.
pmoran raises issues about the strength of evidence for psychosocial interventions improving quality life. I have an embargoed critique of this literature that should be out shortly, but basically the literature attempting to demonstrate benefits of psychosocial intervention is disappointing. The first issue is the bulk of studies do not recruit patients with sufficient modifiable quality-of-life issues to register an improvement. Much of the literature concerns relatively high functioning early breast cancer patients. But these are the patients who come to support groups and seek psychosocial intervention, they are probably not there to reduce distress, but to get other less tangible and subjective benefits. The second issue is the trajectory of distress following a diagnosis of cancer. A diagnosis is unsettling, but not necessarily a mental health issue for most patients. Their distress will be elevated shortly after diagnosis, but within a few months returns to baseline levels. Psychosocial intervention needs to improve on this natural trajectory in order to be shown effective in clinical trials.
A small literature involves deliberately recruiting patients on the basis of their being distressed and effects are stronger, but there are too few such studies. Has I have discovered conducting such studies, most distressed patients do not view their distress as needing mental health interventions, or if they do, they already seeking them and they even have been in mental health treatment before the diagnosis of cancer.
pmoran’s comment elicited a discussion of CAM and the role of nonspecific influences. A very recent article in PLoS one clinical trials claims that CAM and more conventional interventions for depression and anxiety are indistinguishable in terms of differences with placebo treatments. But that article includes Irving Kirsch as a co-author and one can reasonably expect that this is what he would espouse without having bothered to conduct the review. I gave a quick look and the review is a mess, first because differences in the control groups and quality of CAM trials and second because Kirsch and his co-authors ignored that CAM does not differ significantly from active placebo control conditions, whereas psychosocial interventions do.
The broader and more troubling issue is that the evidence base for psychosocial interventions depends on no-treatment and waitlist controls for its strongest effects, and effects are weaker for comparisons with active placebos. In the latter category I would put the administration of pill placebos in clinical trials, conditions which carry a lot of support, structure, and attention. I think the big issue faced psychosocial interventions is whether there are specific effects associated with specific mechanisms or whether everything is nonspecific and simply a matter of paying attention to patients conveying positive expectations and providing support. The delivery of nonspecific interventions does not require much professional training and so there is much resistance to this interpretation.
Depression and anxiety, like pain, are highly subjective outcomes and so it is quite plausible that CAM affects these outcomes by providing a structured supportive interaction with a plausible rationale, at least to some patients. But so do sham CAM and so there is a difficulty showing any advantage of CAM over sham CAM for even these subjective outcomes. But here I am venturing into territory that others on the blog have more experience exploring.
Well, you’re pretty much putting a stake into the heart of that old “mind over matter” belief about health.
See the head of the Memorial Sloan Kettering CAM program weighing in, may surprise you, title says all. Barrie is sounding like David Gorski.
Cancer Quackery: The Persistent Popularity of Useless, Irrational ‘Alternative’ Treatments
Barrie R. Cassileth, IIan R. Yarett
This review provides a brief recap of the history of medical quackery and an overview of the various types of unproven or disproved cancer therapies popular now in the United States and elsewhere.
http://www.cancernetwork.com/integrative-oncology/content/article/10165/2093774
here…
Science-Based Medicine » Questioning Whether Psychotherapy and Support Groups Extend the Lives of Cancer Patients…
[...] http://www.sciencebasedmedicine.org/index.php/questioning-whether-psychotherapy-and-support-groups-e… – Science Based Medicine on positive attitudes/support/psychotherapy affect treatment outcomes [...]