February 14, 2008
Why Is NEJM Bad-Mouthing Prevention?
The lead Perspective in today's New England Journal of Medicine takes aim at candidates (all but John McCain among those who remain in the race) who place prevention near the top of their health care priorities list. Some prevention strategies do save lives at no or low cost, they write, but "sweeping statements about the cost-saving potential of prevention are overreaching." Further, "broad generalizations made by many presidential candidates can be misleading" because "the vast majority (of prevention interventions) reviewed in the health economics literature do not."
The article, lead authored by Joshua T. Cohen from the Tufts University Center for the Evaluation of Value and Risk in Health, concludes that "careful analysis of the costs and benefits of specific interventions, rather than broad generalizations, is critical." Finally, the authors state that careful analysis should also be applied to treatments as well as prevention to "encourage the appropriate delivery of efficient interventions."
What's the data to support this skeptical view about the potential for prevention? They analyzed 599 articles in the health care economics literature that looked at various health care and prevention interventions for their cost-effectiveness, with the standard measure being the cost to save one quality-adjusted life year or QALY. Those articles analyzed 1,500 interventions, of which just 279 involved prevention strategies like smoking cessation or prophylactic aspirin use to prevent heart attacks.
The authors then stratefied them within each group for their cost-saving potential, and compared the two groups. So about 19 percent of the prevention strategies saved money right off the bat, compared to 18 percent of the treatment strategies. Another 48 percent of prevention strategies saved lives for less than $50,000 per QALY compared to 50 percent of treatment strategies.
Is this really six of one, half dozen of the other, which is the essence of their conclusion?
An analyst who used their perspective and looked at the raw numbers a different way could have drawn this conclusion:
Among health promotion strategies analyzed in the economics literature, there are four times as many medical treatments being offered patients today that are not cost effective (using a commonly accepted benchmark of $50,000 per QALY as a cutoff point for cost-effectiveness) as there are prevention strategies that are not cost effective. Indeed, over 80 percent of treatment interventions cost more than they save, just as over 80 percent of prevention interventions cost more than they save.
Why didn't they make that the emphasis of their article?
Moreover, their stratification tells us nothing about what is currently being deployed by the health care system. Only that tells the public and health care analysts what potential each strategy has to save lives and at what cost.
For example: only 40 percent of men over 40 and women over 50 at risk of a heart attack regularly take aspirin as a prevention strategy. Getting that up to 90 percent would save 45,000 lives annually at no cost to the health care system, according to the Partnership for Prevention, a business-backed health care reform coalition.
On the other hand, cholesterol screening for men over 35 and women over 45 to reduce heart attack risk is already at 80 percent of the population. The nation is currently spending about $20 billion a year on drugs to reduce cholesterol that translates into a cost to the health care system of somewhere between $15,000 and $50,000 for every QALY saved.
Both are below the $50,000 benchmark. But which has the greater potential for the least cost?
Moreover, I would like to know what were the interventions in each group that fell above the $50,000 per QALY benchmark. How many are being widely deployed? The study would have been truly enlightening if they had added this third variable to the analysis: how many patients are being offered each treatment strategy, whether prevention or treatment. Then, we could have analyzed the overall cost-effectiveness of our health care system as it currently exists and come up with some rough estimates for the true potential of prevention.
I can't know if it affected his thinking, but Dr. Cohen through his prior work at the Harvard Center for Risk Analysis has conducted research for numerous industrial groups, including housing (lead in children), telecom (cell phones and driving) and food (mercury and fish). The Tufts Center where he works has received unrestricted educational grants from numerous drug companies.
Whatever accounts for his "perspective" on these issues, the fact is that another analyst using the very numbers contained in the studies under review could have drawn a very different conclusion, one that wouldn't have started by raising questions about the candidates who are providing a wonderful service to the American people by reminding them that prevention -- both at the personal and health care system level -- should be near the top of the list of priorities when it comes to thinking about the nation's health.
Posted by gooznews at February 14, 2008 08:13 AMI did not get the sense the author was doing any "bad mouthing." These commentary pieces (they are not studies) are not the end all of what needs to be said on a given subject. They are allowed limited space (I can say that as a dedicated reader) and are meant to be thought provoking. Many of the issues you raised above are valid, but I certainly thought about them as I read the commentary. From my perspective, I thought the essay did a fine job in raising some interesting points. Again, not definitive, but it gave excellent food for thought.
Posted by: Anon at February 14, 2008 01:29 PMI am objecting to the tone. The presumption was that the candidates, and presumably other advocates of a prevention first strategy, are naively calling for blind support for all prevention interventions without regard to cost. That's not what candidate or prevention advocates are saying.
Moreover, the headline read: "Does Preventive Care Save Money?" As I pointed out in the post, the data could just have easily been interpreted to read: "Does Health Care As Delivered by the Nation's 700,000 Physicians Save Money?" The answer, in 80 percent of both cases, is no. So why just pick on the one?
Posted by: Merrill at February 14, 2008 03:31 PMPost was totally benign, reread, did not get a sense of any venom. Regardless, I apologize if I cast a bad light. I like your blog and have followed for a while--no harm, no foul.
Dont think I disagree, just pointing out that commentary was provocative, and regardless of how you spin, if that is the right word (maybe not), I did not sense a partison or ideological agenda. I am pretty good with stats, have some training in this area, and can critically appraise a study with the best of them. I just saw it for what it was, ie, the tagline, "prevention all good," needs to be viewed with caution.
I will let you have the last word, and thanks again for a great blog.
Posted by: Anon at February 14, 2008 04:39 PMthe article seemed rather cosmic and, from a political perspective, simplistic. from a civilian perspectives, things appear simpler. if there are preventive strategies that are cost effective, let's do them. If others aren't, let's not. same standard with medical procedures. hard to fault candidates for saying it would be preferable to keep folks healthy rather than caring for them when they become sick. their view may be less nuanced than those of the experts (and properly so), but at least they, unlike the experts, start with the knowledge that the odds are heavily against any serious change next year.
Posted by: jim jaffe at February 16, 2008 09:15 AMI have very little grounding in statistics but a lot in politics. When a candidate, like Huckabee, harps on prevention, he is blaming the victim for being fat, for smoking, for being sedentary instead of looking at the whole health system.
When a centrist candidate speaks of prevention, it is a crowd-pleasing cliche that dodges detailed exegesis.
The conclusion one derives from looking at statistics, or even conducting a particular study, depends very much on the researcher’s “perspective” (interest, funding, etc.) Is it the paradigm of “health” or of “care.” If it is health, prevention assumes great importance, if it is care, it is the opposite. To keep the needed balance of healthcare, there do need to be two camps of researchers, commentators, philosophers, politicians, etc.
There is one preventive step that is so cost effective and so efficacious that a society should always measure any interventions against it. It is, you might have suspected it, the value of physical activity. Below is an abstract of a paper that I wrote last year looking at the impact of physical activity on colon cancer. The findings were more than surprising: physical activity and colon cancer are engaged in a reciprocal relationship. As one goes down, the other rises and vice versa. The reason is that physical activity exerts its positive influence in all of the currently known colon cancer causing categories; and without side effects. It is unlikely that a pill can ever duplicate that. For those readers who are interested in the whole article, the reference is below; the article is in an open-access journal.
Colon Cancer and Physical Activity: A Content Analysis of Reciprocal Relationship.
Clinical Medicine: Oncology 2007:1 81–89
http://la-press.com/cr_data/files/f_CMO-1-Janecka-et-al_437.pdf
Background: Colon cancer is among the leading causes of cancer mortality and its incidence is increasing worldwide. This is true in spite of broad basic research into colon cancer while, concurrently, physical activity has been shown to offer significant preventive potential. This background led to the formulation of the following research questions:
• Why is physical activity so effective in decreasing the incidence of colon cancer?
• Is there a common denominator to colon cancer and physical activity, which has a reciprocal function?
• Knowing the potential for public health impact of physical activity on colon cancer, has physical activity-colon cancer relationship been in the forefront of research efforts?
Methods: Content analysis of archival literature has been carried out on census of 32,822 message units, extracted from the National Library of Medicine and its PubMed database. The following search terms were used: colon cancer, physical activity, melatonin, age/genetics, diet (obesity, vitamin D, calcium), immunity/inflammation, and bioactive substances incorporating insulin-like growth factor 1, interleukins, and prostaglandins. The research timeframe for each category began with the first article published and ended with the last one printed in 2005.
Results/Conclusions: The effectiveness of physical activity in decreasing the incidence of colon cancer is likely the result of its biologic activity within not one or two but all of the major known colon cancer etiologies, demonstrating a powerful reciprocal relationship. Melatonin is identified as a plausible common denominator of colon cancer and physical activity. The greatest volume of publications deals with colon cancer and genetics. A significant societal health care impact could be achieved by adopting physical activity as a major cancer control strategy.