Bone Loss at a loss
September 10, 2009 9:05 AM Subscribe
Up to 270 women with osteopenia would have to be treated with drugs for three years so that one of them could avoid a single vertebral fracture. Millions of women worldwide, have been told they have osteopenia and should take drugs to prevent bone loss. Drugs like Fosamax, Boniva and Actonel. But now many health public health experts say it's a case of disease-mongering. The WHO has stepped into the fray with an online diagnostic tool only to stir up its own controversy.
Im sure the drugmakers are not complaining.
But I also wonder what is going on in the minds of the prescribers - I'm sure some of them get a cut of the profits, but I would think that others are saying to themselves "better safe than sorry" - it's better that more suffer the extra cost and side effects of these drugs instead of one person being diagnosed with this disease.
Our family doctor orders rounds and rounds of tests at the slightest cough. It's likely paranoia on the MD's part. But some of my family members appreciate it - "You see, our doctor orders rounds and rounds of tests, and you see what they found out? If only Johnny had our doctor, then maybe they would have caught his disease sooner..."
I admit, it's a difficult ethical issue for me. How would I feel saying "No, you don't have osteopenia" to my patients, only to find out that I've made a mistake and a few of them actually have?
By prescribing a drug, I am doing the most that I can (and at the same time, have washed my hands of all responsibility - if you get the disease now, it's the drug's fault, not mine).
It's easy for me to judge, but then again, the only thing my mistakes kill are the plants in my garden.
posted by bitteroldman at 9:26 AM on September 10, 2009 [1 favorite]
But I also wonder what is going on in the minds of the prescribers - I'm sure some of them get a cut of the profits, but I would think that others are saying to themselves "better safe than sorry" - it's better that more suffer the extra cost and side effects of these drugs instead of one person being diagnosed with this disease.
Our family doctor orders rounds and rounds of tests at the slightest cough. It's likely paranoia on the MD's part. But some of my family members appreciate it - "You see, our doctor orders rounds and rounds of tests, and you see what they found out? If only Johnny had our doctor, then maybe they would have caught his disease sooner..."
I admit, it's a difficult ethical issue for me. How would I feel saying "No, you don't have osteopenia" to my patients, only to find out that I've made a mistake and a few of them actually have?
By prescribing a drug, I am doing the most that I can (and at the same time, have washed my hands of all responsibility - if you get the disease now, it's the drug's fault, not mine).
It's easy for me to judge, but then again, the only thing my mistakes kill are the plants in my garden.
posted by bitteroldman at 9:26 AM on September 10, 2009 [1 favorite]
Fun fact: at the currently ongoing Fosamax trial of one particular patient vs. Merck, the judge ordered a "cooling off period" in jury deliberations because they had gotten too acrimonious, with occasional screaming easily audible outside the room. One juror made an "unsubstantiated" complaint that someone else threw a chair at her.
posted by rkent at 9:34 AM on September 10, 2009 [2 favorites]
posted by rkent at 9:34 AM on September 10, 2009 [2 favorites]
BUTT REFORMIN HEALTHKARE IS SOCIALZMMNTtt!!
Obviously as this exemplifies, your health and well being should absolutely be dictated by capitalism. NOT. A kindly reminder derail to contact your congressional representative(s) if you are in the US and ask them 'can I haz reform nao plz?'
posted by cavalier at 9:58 AM on September 10, 2009
Obviously as this exemplifies, your health and well being should absolutely be dictated by capitalism. NOT. A kindly reminder derail to contact your congressional representative(s) if you are in the US and ask them 'can I haz reform nao plz?'
posted by cavalier at 9:58 AM on September 10, 2009
So you are telling me that "social anxiety", "occasional sleeplessness" and "erectile dysfunction" may not require a lifetime of expensive medications?
I am shocked. SHOCKED!
posted by briank at 10:01 AM on September 10, 2009 [1 favorite]
I am shocked. SHOCKED!
posted by briank at 10:01 AM on September 10, 2009 [1 favorite]
I think the real problem here is that Boniva is a ridiculous name for a drug.
posted by chinston at 10:05 AM on September 10, 2009 [10 favorites]
posted by chinston at 10:05 AM on September 10, 2009 [10 favorites]
Don't forget about social dysfunction, occasional anxiety, and erectile sleeplessness!
posted by jamstigator at 10:06 AM on September 10, 2009 [2 favorites]
posted by jamstigator at 10:06 AM on September 10, 2009 [2 favorites]
I was diagnosed with osteopenia when I was about 30, and my doctor essentially said "this is worrisome, but there's no really safe way to treat it in pre-menopausal women, so keep lifting weights and taking calcium supplements and we'll talk about it again after you hit menopause." Was that unusual, or are we mostly talking about older women here?
My sense was that my doctor thought my bone issues were worrisome because of my age, not because of the extent of the bone loss. She said that if I'd been 60, it would have been normal.
posted by craichead at 10:11 AM on September 10, 2009
My sense was that my doctor thought my bone issues were worrisome because of my age, not because of the extent of the bone loss. She said that if I'd been 60, it would have been normal.
posted by craichead at 10:11 AM on September 10, 2009
Our family doctor orders rounds and rounds of tests at the slightest cough. It's likely paranoia on the MD's part. But some of my family members appreciate it - "You see, our doctor orders rounds and rounds of tests, and you see what they found out? If only Johnny had our doctor, then maybe they would have caught his disease sooner..."
This sort of thing is a major reason why our health care is so expensive compared to the rest of the world. A similar thought process is at work in the "executive physical", which was recently criticized in NEJM (unfortunately the complete article is only available to subscribers). The problem is, when you do unneeded tests, you are likely to find a spurious abnormal result, which then leads to more test but ultimately ends up being of no concern. Sometimes the results are worse, however. I remember a case many years ago where a patient who was scheduled for a routine operation (something like a hernia repair or gallbladder surgery) and got a preop chest X-ray (which was pretty routine at the time, although even back then I didn't see the utility of it). The x-ray showed a mass in one of his lungs; it was probably benign but given his smoking history and so on they decided to biopsy it. Well in the middle of the operation the surgeon was struggling to get the mass off the subclavian artery and the artery tore. The patient bled to death in a matter of minutes. Similar stories abound, although the patient usually survives.
As far as the article in the FPP, what is also troubling is that as mentioned briefly in the article, many of those women will suffer side effects of the drug. In aggregate, the cost of treating those may outweigh the cost of treating the one vertebral fracture that might be prevented. There is also the consideration of additional pain and suffering caused by those side effects. How do such marginally effective drugs get approved? Among other things, the people who do studies that would normally require large numbers of people (as mentioned in alsm's link) can often scale them down by looking at surrogate endpoints. So for these drugs they have been shown to improve the surrogate endpoint of bone density and it is assumed that means they also improve the clinical endpoint of osteopenic fractures. This is also discussed at length in Dr. John Abramson's book Overdosed America (blog link, Amazon link), which I have recommended before.
posted by TedW at 10:12 AM on September 10, 2009 [6 favorites]
This sort of thing is a major reason why our health care is so expensive compared to the rest of the world. A similar thought process is at work in the "executive physical", which was recently criticized in NEJM (unfortunately the complete article is only available to subscribers). The problem is, when you do unneeded tests, you are likely to find a spurious abnormal result, which then leads to more test but ultimately ends up being of no concern. Sometimes the results are worse, however. I remember a case many years ago where a patient who was scheduled for a routine operation (something like a hernia repair or gallbladder surgery) and got a preop chest X-ray (which was pretty routine at the time, although even back then I didn't see the utility of it). The x-ray showed a mass in one of his lungs; it was probably benign but given his smoking history and so on they decided to biopsy it. Well in the middle of the operation the surgeon was struggling to get the mass off the subclavian artery and the artery tore. The patient bled to death in a matter of minutes. Similar stories abound, although the patient usually survives.
As far as the article in the FPP, what is also troubling is that as mentioned briefly in the article, many of those women will suffer side effects of the drug. In aggregate, the cost of treating those may outweigh the cost of treating the one vertebral fracture that might be prevented. There is also the consideration of additional pain and suffering caused by those side effects. How do such marginally effective drugs get approved? Among other things, the people who do studies that would normally require large numbers of people (as mentioned in alsm's link) can often scale them down by looking at surrogate endpoints. So for these drugs they have been shown to improve the surrogate endpoint of bone density and it is assumed that means they also improve the clinical endpoint of osteopenic fractures. This is also discussed at length in Dr. John Abramson's book Overdosed America (blog link, Amazon link), which I have recommended before.
posted by TedW at 10:12 AM on September 10, 2009 [6 favorites]
You forgot to add a "Metafilter:" in front of that, jamstigator.
bitteroldman makes a good point, in the current environment of medical malpractice lawsuits, not prescribing something that could prevent an adverse medical condition in one out of 270 patients may be too high of a risk for the prescribing doctor depending on how many patients they treat.
posted by porpoise at 10:13 AM on September 10, 2009
bitteroldman makes a good point, in the current environment of medical malpractice lawsuits, not prescribing something that could prevent an adverse medical condition in one out of 270 patients may be too high of a risk for the prescribing doctor depending on how many patients they treat.
posted by porpoise at 10:13 AM on September 10, 2009
Based on my reading of the study, it seems that its exclusive focus on vertebral fractures (to the exclusion of other disabling and expensive fractures) and its framing of ostopenia as merely "risk for being at risk" when it is in fact a particular stage of bone weakening has a distorting effect on its conclusions.
I hope my experience might shed some light on this issue.
I was diagnosed with ostopenia (in some of my vertebrae) and osteoporosis (in other vertabrae and my hip) four years ago, when I was 34. Despite taking calcium pills compulsively and diligently lifting weights since I was a teenager, two years ago I had a fracture from a minor accident. It required three days in the hospital and multiple follow-up visits to an orthopedic surgeon.
My doctor had recommended that I go on Boniva, one of the drugs mentioned in the article, but my insurance company refused to cover it. At $60 a month, they had apparently made a calculation that it was not worth it to include it in their "formulary." My hospital stay and doctor's bills cost them over $10,000. This seemed to me to be a clear case of statistical shortsightedness and indicative of the insurance industry's emphasize on expensive post-facto interventions as opposed to preventative medicine. I don't know if taking Boniva when I was first diagnosed would have prevented the fracture -- as with all preventative treatments, it's impossible to determine their precise affect (theoretical or actual) on outcomes -- but I would be interested to see data to see if it might have. That data isn't available here.
I know my exception doesn't prove a rule, but this study seems to have a skewed cost-benefit analysis. Vertebral fractures aren't the only fracture women with weakened bones risk, and other fractures can be formidably -- if not equally -- expensive and disabling. I think some other systematic studies are needed before buying in to the conclusions drawn here.
posted by foxy_hedgehog at 10:18 AM on September 10, 2009 [4 favorites]
I hope my experience might shed some light on this issue.
I was diagnosed with ostopenia (in some of my vertebrae) and osteoporosis (in other vertabrae and my hip) four years ago, when I was 34. Despite taking calcium pills compulsively and diligently lifting weights since I was a teenager, two years ago I had a fracture from a minor accident. It required three days in the hospital and multiple follow-up visits to an orthopedic surgeon.
My doctor had recommended that I go on Boniva, one of the drugs mentioned in the article, but my insurance company refused to cover it. At $60 a month, they had apparently made a calculation that it was not worth it to include it in their "formulary." My hospital stay and doctor's bills cost them over $10,000. This seemed to me to be a clear case of statistical shortsightedness and indicative of the insurance industry's emphasize on expensive post-facto interventions as opposed to preventative medicine. I don't know if taking Boniva when I was first diagnosed would have prevented the fracture -- as with all preventative treatments, it's impossible to determine their precise affect (theoretical or actual) on outcomes -- but I would be interested to see data to see if it might have. That data isn't available here.
I know my exception doesn't prove a rule, but this study seems to have a skewed cost-benefit analysis. Vertebral fractures aren't the only fracture women with weakened bones risk, and other fractures can be formidably -- if not equally -- expensive and disabling. I think some other systematic studies are needed before buying in to the conclusions drawn here.
posted by foxy_hedgehog at 10:18 AM on September 10, 2009 [4 favorites]
I was diagnosed with osteopenia when I was about 30, and my doctor essentially said "this is worrisome, but there's no really safe way to treat it in pre-menopausal women, so keep lifting weights and taking calcium supplements and we'll talk about it again after you hit menopause." Was that unusual, or are we mostly talking about older women here?
it depends on how severe the osteopenia is and if bi-yearly bone density tests show any further degeneration. Part of the problem is that most doctors/endocrinologists aren't accustomed to treating pre-menopausal patients with osteoporosis, and don't really know what to do with them. If you are interested in a referral to a physician in your area who specializes in pre-menopausal osteoporosis, send me a MeMail and I can make some calls for you.
posted by foxy_hedgehog at 10:21 AM on September 10, 2009
it depends on how severe the osteopenia is and if bi-yearly bone density tests show any further degeneration. Part of the problem is that most doctors/endocrinologists aren't accustomed to treating pre-menopausal patients with osteoporosis, and don't really know what to do with them. If you are interested in a referral to a physician in your area who specializes in pre-menopausal osteoporosis, send me a MeMail and I can make some calls for you.
posted by foxy_hedgehog at 10:21 AM on September 10, 2009
First, the 270 figure is the upper bound of the confidence interval. The 95% interval is 104 to 270.
Fosamax (alendronate sodium) is available as a generic, which costs about $20/month. That appears to be cheaper than Actonel or Boniva, which cost about $40 and $80/month, respectively.
$20 * 12 * 3 * 270 is $194,400. That's a lot of money to prevent one fracture. How much does a vertebral fracture cost? This 2004 study estimated the cost at £2613 ($4,356).
That seems pretty lopsided. But patients with a vertebral fracture are at a 4-5 fold risk of subsequent fracture. 26% will refracture within 1 year of the original fracture. So the argument is that by preventing the first fracture, the drug is really preventing the patient from entering a vicious downward spiral of fracture, atrophy, refracture, repeat.
Of course, that doesn't necessarily mean it's a good therapy for most (or even many) people, just that maybe it's not quite so obviously outlandish as it may appear at first.
Another example of incredibly overprescribed drugs with a high number needed to treat is the statins. For example, Lipitor has an NNT of 250. That article has a good example of how the pharmaceutical companies spin things in their marketing ("Lipitor reduces the risk of heart attack by 36%...in patients with multiple risk factors for heart disease.") to sound much better than they are. When things are explained in terms of NNT and costs, things become more clear:
"What if you put 250 people in a room and told them they would each pay $1,000 a year for a drug they would have to take every day, that many would get diarrhea and muscle pain, and that 249 would have no benefit? And that they could do just as well by exercising? How many would take that?"
posted by jedicus at 10:27 AM on September 10, 2009 [5 favorites]
Fosamax (alendronate sodium) is available as a generic, which costs about $20/month. That appears to be cheaper than Actonel or Boniva, which cost about $40 and $80/month, respectively.
$20 * 12 * 3 * 270 is $194,400. That's a lot of money to prevent one fracture. How much does a vertebral fracture cost? This 2004 study estimated the cost at £2613 ($4,356).
That seems pretty lopsided. But patients with a vertebral fracture are at a 4-5 fold risk of subsequent fracture. 26% will refracture within 1 year of the original fracture. So the argument is that by preventing the first fracture, the drug is really preventing the patient from entering a vicious downward spiral of fracture, atrophy, refracture, repeat.
Of course, that doesn't necessarily mean it's a good therapy for most (or even many) people, just that maybe it's not quite so obviously outlandish as it may appear at first.
Another example of incredibly overprescribed drugs with a high number needed to treat is the statins. For example, Lipitor has an NNT of 250. That article has a good example of how the pharmaceutical companies spin things in their marketing ("Lipitor reduces the risk of heart attack by 36%...in patients with multiple risk factors for heart disease.") to sound much better than they are. When things are explained in terms of NNT and costs, things become more clear:
"What if you put 250 people in a room and told them they would each pay $1,000 a year for a drug they would have to take every day, that many would get diarrhea and muscle pain, and that 249 would have no benefit? And that they could do just as well by exercising? How many would take that?"
posted by jedicus at 10:27 AM on September 10, 2009 [5 favorites]
Many years ago, I heard of a small town in Arizona with the highest incidence of diabetes in the USA. Something in the water? Ethniciity? Heat? None of those. The local doctor diagnosed his patients with diabetes if they were in the high end of the normal range. That's how he saw diabetest. An awful lot of people took insulin who didn't need it at all.
posted by Carol Anne at 10:30 AM on September 10, 2009 [1 favorite]
posted by Carol Anne at 10:30 AM on September 10, 2009 [1 favorite]
@foxy_hedgehog: the study actually agrees with you on that point. It says:
Aside from the tendency to emphasise the relative over the absolute risk reduction, the authors of three of the four reanalyses focus exclusively on vertebral fractures, rather than long bone and hip fractures, which are more relevant to patients.9 10 12 In addition, two reanalyses use morphometric rather than clinical vertebral fractures as their outcome of interest.11 12 Two thirds of vertebral fractures are subclinical or asymptomatic and may not affect quality of life. As a consequence showing that drugs reduce vertebral fractures may not be as important to patients as it seems.
The authors are criticizing the reanalyses because they're citing only the benefits with respect to vertebral fractures. But even if you accept that there is a benefit, the "number needed to treat" for these drugs is still alarmingly high.
posted by storybored at 10:31 AM on September 10, 2009
Aside from the tendency to emphasise the relative over the absolute risk reduction, the authors of three of the four reanalyses focus exclusively on vertebral fractures, rather than long bone and hip fractures, which are more relevant to patients.9 10 12 In addition, two reanalyses use morphometric rather than clinical vertebral fractures as their outcome of interest.11 12 Two thirds of vertebral fractures are subclinical or asymptomatic and may not affect quality of life. As a consequence showing that drugs reduce vertebral fractures may not be as important to patients as it seems.
The authors are criticizing the reanalyses because they're citing only the benefits with respect to vertebral fractures. But even if you accept that there is a benefit, the "number needed to treat" for these drugs is still alarmingly high.
posted by storybored at 10:31 AM on September 10, 2009
This article seems a trifle disingenuous. The main goal of treating osteopenia is not the short-term reduction of fracture risk over three to five years. The main goal is the long-term prevention or delay of osteoporosis. Given that the lifetime risk of an osteoporotic fracture in women is 40%, and that a hip fracture is frequently the event that ends a senior's independent life or even their life altogether, this is kind of important.
posted by timeo danaos at 10:34 AM on September 10, 2009 [1 favorite]
posted by timeo danaos at 10:34 AM on September 10, 2009 [1 favorite]
"...bone density fails to capture a poorly understood factor known as bone quality, which declines with age, researchers say.
'A 50-year-old woman with a bone density that is clearly in the osteoporotic range has a much lower risk of having a fracture in the next year or the next five years than an 80-year-old woman with the same bone density,'' Dr. McGowan said. ''It's definitely something about the quality of the bone changing.''"
posted by storybored at 10:40 AM on September 10, 2009
'A 50-year-old woman with a bone density that is clearly in the osteoporotic range has a much lower risk of having a fracture in the next year or the next five years than an 80-year-old woman with the same bone density,'' Dr. McGowan said. ''It's definitely something about the quality of the bone changing.''"
posted by storybored at 10:40 AM on September 10, 2009
The problem is that the success of excess testing is very visceral. A disease is discovered and a persons life (or quality of life) is saved. On the other hand, it's failure is hidden. Some tests are done, they come up negative. You get a little piece of mind and healtcare costs go up just a tiny bit for everyone.
Some tests have negative effects as well, for example a mamogram uses radiation, and actually increases breast cancer risk. But it's more likely to discover a tumor then cause one, for women over 40. But if a 30 year old gets a mamogram and discovers cancer, her life is saved, whereas someone who get's breast cancer later in live would have had lots of other risk factors and would never think to blame the mamogram she got as a younger woman.
---
As far as the costs of the tests go, I'm going to go against the conventional wisdom and claim it's probably bullshit. The tests probably cost a fortune for the same reason all other kinds of medical stuff is expensive.
If you walk into a Perl Vision eyeglasses cost hundreds of dollars, but you usually don't pay that price yourself because it's covered by insurance. But this post about $8? How could glasses possibly really be worth that much money? It's just a few ounces little plastic, metal and glass. Maybe the expensive ones are higher quality, but for $8 you could replace them every month and still save a ton of money.
But because most people don't pay for their own glasses, it's a non-issue for them.
I think that the costs of these tests are expensive for the same reason name-brand drugs are so expensive. There's no motivation to develop a cheaper test. In fact the incentive is to make them as expensive as possible. And because of the patent system it would be difficult for competitors to create cheaper tests.
Take Amniocentesis for example, a test for fetal abnormalities. It's a procedure that actually carries some risk for a baby, and it looks like it $1000-1500 (I actually had no idea how much it costs until just now, I'm kind of shocked, actually!)
But look at this: The British NHS is developing a cheaper, non-invasive blood test that might be able to replace Amniocentesis. That's something they have every motivation to do, because they pay for the care, rather then profiting off of it.
If there was a financial incentive for people to figure out how to make this routine testing cheaper, I'm sure there that costs could go down a lot. A lot of the lab work could be automated, etc. That would be a much better alternative then denying people these tests, which give them piece of mind even if they're not the best idea in the average case.
Actually, I think that kind of incentive system could be grafted onto our current system. Simply provide a cash bounty to researchers who come up with cost saving advances, and stop letting people abuse the patent system to squeeze money out of sick people (cash bounties could also be used to incentivize development of new important tests, rather then gambling with the patent system)
posted by delmoi at 10:42 AM on September 10, 2009 [5 favorites]
Some tests have negative effects as well, for example a mamogram uses radiation, and actually increases breast cancer risk. But it's more likely to discover a tumor then cause one, for women over 40. But if a 30 year old gets a mamogram and discovers cancer, her life is saved, whereas someone who get's breast cancer later in live would have had lots of other risk factors and would never think to blame the mamogram she got as a younger woman.
---
As far as the costs of the tests go, I'm going to go against the conventional wisdom and claim it's probably bullshit. The tests probably cost a fortune for the same reason all other kinds of medical stuff is expensive.
If you walk into a Perl Vision eyeglasses cost hundreds of dollars, but you usually don't pay that price yourself because it's covered by insurance. But this post about $8? How could glasses possibly really be worth that much money? It's just a few ounces little plastic, metal and glass. Maybe the expensive ones are higher quality, but for $8 you could replace them every month and still save a ton of money.
But because most people don't pay for their own glasses, it's a non-issue for them.
I think that the costs of these tests are expensive for the same reason name-brand drugs are so expensive. There's no motivation to develop a cheaper test. In fact the incentive is to make them as expensive as possible. And because of the patent system it would be difficult for competitors to create cheaper tests.
Take Amniocentesis for example, a test for fetal abnormalities. It's a procedure that actually carries some risk for a baby, and it looks like it $1000-1500 (I actually had no idea how much it costs until just now, I'm kind of shocked, actually!)
But look at this: The British NHS is developing a cheaper, non-invasive blood test that might be able to replace Amniocentesis. That's something they have every motivation to do, because they pay for the care, rather then profiting off of it.
If there was a financial incentive for people to figure out how to make this routine testing cheaper, I'm sure there that costs could go down a lot. A lot of the lab work could be automated, etc. That would be a much better alternative then denying people these tests, which give them piece of mind even if they're not the best idea in the average case.
Actually, I think that kind of incentive system could be grafted onto our current system. Simply provide a cash bounty to researchers who come up with cost saving advances, and stop letting people abuse the patent system to squeeze money out of sick people (cash bounties could also be used to incentivize development of new important tests, rather then gambling with the patent system)
posted by delmoi at 10:42 AM on September 10, 2009 [5 favorites]
You know, I tend to be quite conservative when it comes to doling out prescriptions but the article you're sighting, itself, could be construed as playing with the statistics to paint a bleak picture. Indeed, when the absolute risk of a clinical event is relatively low, describing the impact of a drug in terms of relative risk instead of absolute risk difference leads to more apparently impressive numbers. That's Biostatistics 101.
However, 270 is the statistical upper estimate of their calculated number needed to treat. The point estimate is actually 133 with a 95% confidence interval of 104-270. Focusing on the end of the confidence interval that pushes your case might be described as far from a dispassionate analysis of the data.
Moreover, let me put this in a bit of different context. Let us assume that the point estimate for the number needed to treat is accurate. Now, realizing that that NNT is based on studies of women with a mean age of about 65, and that the studies were conducted over about a 3 year period with an absolute risk of clinical (ie symptomatic) vertebral fracture of about 1% in the control groups over that time period, what does that suggest about the potential NNT over the next 15 years of life?
Let's do a bit of naughty extrapolation. Assume that the baseline risk remains stable. It probably doesn't. In fact, it increases with age, but let us assume that it remains stable as assuming otherwise will make the NNT appear even more impressive. Over 15 years, the absolute risk of a vertebral fracture (back of the envelope), would then approach 5%. If the impact on bone mineral density attributable to bisphosphonates is preserved over that time period, the NNT over 15 years to prevent a clinically significant vertebral fracture would be about 27. Now consider the alternative combined outcome of clinical vertebral fractures or hip fractures, and your baseline absolute risk in osteopenic women might go up even further, yielding an even more dramatic absolute risk difference (and even smaller NNT) with therapy over a more prolonged clinically relevant period. What if my back-of-the-envelope guess was that over their lifetime, bisphosphonate therapy in osteopenic women in their 60s might yield an NNT of 5-10 to prevent clinically significant fractures?
Playing with statistics sure is fun. While I agree that perhaps in published studies, documenting estimates of the NNT and absolute risk difference in addition to relative risks might lead to a more balanced interpretation of the data by the uninformed reader, the numbers are already there in black and white. All you need is a cocktail napkin and a pen. The real systems-level problem is that many practitioners don't have anything approaching a basic understanding of biostatistics, nor do they have a whole lot of time to evaluate the literature themselves. Depending on drug reps for this purpose (continuing medical education) or even taking authors' conclusions at face value is without a doubt, a crummy way to practice medicine.
posted by drpynchon at 10:45 AM on September 10, 2009 [5 favorites]
However, 270 is the statistical upper estimate of their calculated number needed to treat. The point estimate is actually 133 with a 95% confidence interval of 104-270. Focusing on the end of the confidence interval that pushes your case might be described as far from a dispassionate analysis of the data.
Moreover, let me put this in a bit of different context. Let us assume that the point estimate for the number needed to treat is accurate. Now, realizing that that NNT is based on studies of women with a mean age of about 65, and that the studies were conducted over about a 3 year period with an absolute risk of clinical (ie symptomatic) vertebral fracture of about 1% in the control groups over that time period, what does that suggest about the potential NNT over the next 15 years of life?
Let's do a bit of naughty extrapolation. Assume that the baseline risk remains stable. It probably doesn't. In fact, it increases with age, but let us assume that it remains stable as assuming otherwise will make the NNT appear even more impressive. Over 15 years, the absolute risk of a vertebral fracture (back of the envelope), would then approach 5%. If the impact on bone mineral density attributable to bisphosphonates is preserved over that time period, the NNT over 15 years to prevent a clinically significant vertebral fracture would be about 27. Now consider the alternative combined outcome of clinical vertebral fractures or hip fractures, and your baseline absolute risk in osteopenic women might go up even further, yielding an even more dramatic absolute risk difference (and even smaller NNT) with therapy over a more prolonged clinically relevant period. What if my back-of-the-envelope guess was that over their lifetime, bisphosphonate therapy in osteopenic women in their 60s might yield an NNT of 5-10 to prevent clinically significant fractures?
Playing with statistics sure is fun. While I agree that perhaps in published studies, documenting estimates of the NNT and absolute risk difference in addition to relative risks might lead to a more balanced interpretation of the data by the uninformed reader, the numbers are already there in black and white. All you need is a cocktail napkin and a pen. The real systems-level problem is that many practitioners don't have anything approaching a basic understanding of biostatistics, nor do they have a whole lot of time to evaluate the literature themselves. Depending on drug reps for this purpose (continuing medical education) or even taking authors' conclusions at face value is without a doubt, a crummy way to practice medicine.
posted by drpynchon at 10:45 AM on September 10, 2009 [5 favorites]
Holy crap, alms, The Last Psychiatrist is a [good] link.
posted by infinitewindow at 10:54 AM on September 10, 2009
posted by infinitewindow at 10:54 AM on September 10, 2009
It's amazing how differences in the way that trial results are reported can really affect perceptions about risk vs. reward for different medical procedures, particularly for "preventative care" (which everyone knows is always a good thing, right?). Usually, results are reported that explain the percentage drop in risk, without ever providing info about how large those risks are in the first place. Shifting to an explanation about how many people will need to be treated for how long in order to prevent one occurrance provides that context, sometimes in surprising ways.
Other recent examples that come to mind:
*The effectiveness of prescribing cholesterol-lowering drugs to middle-aged adults who don't have high cholesterol: reduces heart attack and stroke by 50%! Of course, because the number of these people who would have a cardiac event is already so low, that works out to treating 120 people for 2 years in order to avoid one heart attack, stroke, or death. At an average cost of about $3.65 per day for Crestor (and some nasty nasty side effects), it's definitely worth doing!
*Middle-aged men who haven't ever been diagnosed with prostate cancer can take Propecia (yes, the baldness stuff) and lower their risk of getting prostate cancer by 25%! Wowee! Of course, 71 men would have to take it for 7 years each to prevent one case of cancer. At $3 a day plus the risk of impotence or incontinence, a small price to pay, right?
*For women between ages 50 and 70, you have to give an annual mammogram to 2,000 women for 10 years in order to prevent one death from breast cancer. As a bonus, if you're not lucky enough to be the one-in-2,000 whose life is saved, maybe you'll be one of the 10-in-2,000 who gets a false positive and gets to undergo biopsies and other unnecessary treatment. Sign me up!
I know that the vast majority of people involved in these sorts of trials are good people who are working hard to save lives; I know the vast majority of doctors that end up prescribing this stuff are only trying to do right by their patients. It's just boggling to me that the end result is this culture where we're by most measures doing more and more screening and treatment of people who aren't sick, at enormous cost to our health care system. And it's not even the wasted money that is the worst part; as someone who has gone through the whole "we're pretty sure you have cancer, let's do a biopsy; well, the biopsy's not clear but let's do surgery just to be safe" thing, I'm well aware that it's not costless in terms of people's *actual health.* Throwing a zillion 50 year olds on Crestor doesn't just tick up medical cost inflation by another point, thus making insurance just a bit more unaffordable, but also results in some real negative side effects for a lot of people. I'm sometimes convinced that our entire medical system is systematically blind to the "costs" side of the cost vs. benefit equation.
Anyway. Clearly an issue a bit near-and-dear to my heart. Thanks for yet another (depressing) example of why the U.S. really needs to be throwing more money at comparative effectiveness research rather than just more money to NIH.
posted by iminurmefi at 11:03 AM on September 10, 2009 [1 favorite]
Other recent examples that come to mind:
*The effectiveness of prescribing cholesterol-lowering drugs to middle-aged adults who don't have high cholesterol: reduces heart attack and stroke by 50%! Of course, because the number of these people who would have a cardiac event is already so low, that works out to treating 120 people for 2 years in order to avoid one heart attack, stroke, or death. At an average cost of about $3.65 per day for Crestor (and some nasty nasty side effects), it's definitely worth doing!
*Middle-aged men who haven't ever been diagnosed with prostate cancer can take Propecia (yes, the baldness stuff) and lower their risk of getting prostate cancer by 25%! Wowee! Of course, 71 men would have to take it for 7 years each to prevent one case of cancer. At $3 a day plus the risk of impotence or incontinence, a small price to pay, right?
*For women between ages 50 and 70, you have to give an annual mammogram to 2,000 women for 10 years in order to prevent one death from breast cancer. As a bonus, if you're not lucky enough to be the one-in-2,000 whose life is saved, maybe you'll be one of the 10-in-2,000 who gets a false positive and gets to undergo biopsies and other unnecessary treatment. Sign me up!
I know that the vast majority of people involved in these sorts of trials are good people who are working hard to save lives; I know the vast majority of doctors that end up prescribing this stuff are only trying to do right by their patients. It's just boggling to me that the end result is this culture where we're by most measures doing more and more screening and treatment of people who aren't sick, at enormous cost to our health care system. And it's not even the wasted money that is the worst part; as someone who has gone through the whole "we're pretty sure you have cancer, let's do a biopsy; well, the biopsy's not clear but let's do surgery just to be safe" thing, I'm well aware that it's not costless in terms of people's *actual health.* Throwing a zillion 50 year olds on Crestor doesn't just tick up medical cost inflation by another point, thus making insurance just a bit more unaffordable, but also results in some real negative side effects for a lot of people. I'm sometimes convinced that our entire medical system is systematically blind to the "costs" side of the cost vs. benefit equation.
Anyway. Clearly an issue a bit near-and-dear to my heart. Thanks for yet another (depressing) example of why the U.S. really needs to be throwing more money at comparative effectiveness research rather than just more money to NIH.
posted by iminurmefi at 11:03 AM on September 10, 2009 [1 favorite]
As others have noted, what's up with the deceitfully hyperbolic 270 number? Is storybored going to come back and correct her lies damned lies statistics?
posted by anotherpanacea at 11:33 AM on September 10, 2009
posted by anotherpanacea at 11:33 AM on September 10, 2009
But now many health public health experts say it's a case of disease-mongering.
I Guess| Someone made
----------------------------
a boner| YEEEEEEEAAAAAAAAAAHHHHHHHHHH
posted by Inspector.Gadget at 11:35 AM on September 10, 2009
I Guess| Someone made
----------------------------
a boner| YEEEEEEEAAAAAAAAAAHHHHHHHHHH
posted by Inspector.Gadget at 11:35 AM on September 10, 2009
So, I know virtually noting about the disease, so I have to ask: Is vertebral fracture the only issue with osteopenia, or are they just cherry picking?
I mean, people who are careless with table saws seldom die, but given the number of woodworkers who can't count to ten on their fingers, you'd have to be some kind of sociopath to suggest that it's OK to be careless with a table saw. Similarly, I saw a paper once that showed that the lifetime medical expenses of smokers were, on average, below those of non smokers. The lifetime medical expenses of children eaten by dingos is lower still.
And to follow iminurmefi's logic - doctors are becomming less likely to prescribe antibiotics for just anything that comes down the pike. As we approach cold and flu season bring this up and see how many people absolutely howl about about not being allowed to make a hefty co-pay for a drug that won't do them any good at all.
As many of you know, I am a pawn of the big pharma. Just saying.
posted by Kid Charlemagne at 11:43 AM on September 10, 2009
I mean, people who are careless with table saws seldom die, but given the number of woodworkers who can't count to ten on their fingers, you'd have to be some kind of sociopath to suggest that it's OK to be careless with a table saw. Similarly, I saw a paper once that showed that the lifetime medical expenses of smokers were, on average, below those of non smokers. The lifetime medical expenses of children eaten by dingos is lower still.
And to follow iminurmefi's logic - doctors are becomming less likely to prescribe antibiotics for just anything that comes down the pike. As we approach cold and flu season bring this up and see how many people absolutely howl about about not being allowed to make a hefty co-pay for a drug that won't do them any good at all.
As many of you know, I am a pawn of the big pharma. Just saying.
posted by Kid Charlemagne at 11:43 AM on September 10, 2009
If you walk into a Perl Vision
...then the salesdroid greets you with a hearty
posted by ROU_Xenophobe at 11:45 AM on September 10, 2009 [5 favorites]
...then the salesdroid greets you with a hearty
@P=split//,".URRUU\c8R";@d=split//,"\nrekcah xinU / lreP rehtona tsuJ";sub p{
@p{"r$p","u$p"}=(P,P);pipe"r$p","u$p";++$p;($q*=2)+=$f=!fork;map{$P=$P[$f^ord
($p{$_})&6];$p{$_}=/ ^$P/ix?$P:close$_}keys%p}p;p;p;p;p;map{$p{$_}=~/^[P.]/&&
close$_}%p;wait until$?;map{/^r/&&<>}%p;$_=$d[$q];sleep rand(2)if/\S/;printposted by ROU_Xenophobe at 11:45 AM on September 10, 2009 [5 favorites]
As others have noted, what's up with the deceitfully hyperbolic 270 number?
Did someone call :). The FPP clearly states "Up to 270". "Up to" as in "up to and including". It ain't hyperbolic, it's meta-bolic.
posted by storybored at 11:50 AM on September 10, 2009
Did someone call :). The FPP clearly states "Up to 270". "Up to" as in "up to and including". It ain't hyperbolic, it's meta-bolic.
posted by storybored at 11:50 AM on September 10, 2009
chinston: "I think the real problem here is that Boniva is a ridiculous name for a drug."
Well, it would be a fine name if it were an erectile dysfunction drug.
posted by Kadin2048 at 12:02 PM on September 10, 2009
Well, it would be a fine name if it were an erectile dysfunction drug.
posted by Kadin2048 at 12:02 PM on September 10, 2009
I'm 33 and just had a bone scan that indicated I'm on the cusp of osteopenia, myself.
My sense was that my doctor thought my bone issues were worrisome because of my age, not because of the extent of the bone loss. She said that if I'd been 60, it would have been normal.
My limited understanding is that bone density follows a curve with the maximum density reached in your late 20s/early 30s. So if you're already osteopenic at that point, it doesn't bode well for when you're older.
That said, I too am just following a regimen of calcium supplements and exercise. Not sure you can do much more, but it can help to flatten the decline curve.
posted by misskaz at 12:45 PM on September 10, 2009
My sense was that my doctor thought my bone issues were worrisome because of my age, not because of the extent of the bone loss. She said that if I'd been 60, it would have been normal.
My limited understanding is that bone density follows a curve with the maximum density reached in your late 20s/early 30s. So if you're already osteopenic at that point, it doesn't bode well for when you're older.
That said, I too am just following a regimen of calcium supplements and exercise. Not sure you can do much more, but it can help to flatten the decline curve.
posted by misskaz at 12:45 PM on September 10, 2009
But I also wonder what is going on in the minds of the prescribers - I'm sure some of them get a cut of the profits
Where does this idea come from? Are there really doctors taking money in exchange for prescribing drugs? Every doctor I've ever met gets violently angry at the suggestion.
posted by Pope Guilty at 12:46 PM on September 10, 2009
Where does this idea come from? Are there really doctors taking money in exchange for prescribing drugs? Every doctor I've ever met gets violently angry at the suggestion.
posted by Pope Guilty at 12:46 PM on September 10, 2009
Are there really doctors taking money in exchange for prescribing drugs?
Yes.
posted by jedicus at 12:54 PM on September 10, 2009
Yes.
posted by jedicus at 12:54 PM on September 10, 2009
That's horrifying. I thought that was flat-out illegal.
posted by Pope Guilty at 1:09 PM on September 10, 2009
posted by Pope Guilty at 1:09 PM on September 10, 2009
I think the real problem here is that Boniva is a ridiculous name for a drug.
Gives me a little shiver every time Sally Field checks her calendar to see if it's time for her monthly Boniva.
You like me! You really like me!
posted by palliser at 1:10 PM on September 10, 2009
Gives me a little shiver every time Sally Field checks her calendar to see if it's time for her monthly Boniva.
You like me! You really like me!
posted by palliser at 1:10 PM on September 10, 2009
Are there really doctors taking money in exchange for prescribing drugs?
While outright bribery like that is very rare (and probably illegal), many physicians are paid to be on speakers bureaus sponsored by drug companies, given help writing papers, and so forth. All of those activities are supposed to be free of commercial bias, but someone who was just paid $500 to give a talk about, say, statins, you are certainly going to think favorably about the company that paid you.
posted by TedW at 1:15 PM on September 10, 2009
While outright bribery like that is very rare (and probably illegal), many physicians are paid to be on speakers bureaus sponsored by drug companies, given help writing papers, and so forth. All of those activities are supposed to be free of commercial bias, but someone who was just paid $500 to give a talk about, say, statins, you are certainly going to think favorably about the company that paid you.
posted by TedW at 1:15 PM on September 10, 2009
The trouble with using drugs like Fosamax to treat osteopenia instead of osteoporosis is that we do not have data indicating that the drugs build strong bone that will not fracture ten years from now. Longterm use might build weak bones that are worse than you would get with no drug treatment at all.
posted by Ery at 2:50 PM on September 10, 2009
posted by Ery at 2:50 PM on September 10, 2009
What about vibrations for increasing bone mass?
posted by [insert clever name here] at 3:20 PM on September 10, 2009
posted by [insert clever name here] at 3:20 PM on September 10, 2009
My grandmother broke her hip and died not long thereafter. My significant other has osteoporosis and takes Boniva for it. It's a scary thing. Whenever she's going out, I tell her she should let me encase her in bubblewrap. Mostly though, I try not to think about it or else I'd be constantly worrying and that'd cost me years off my life without doing any good. But if she died, that'd cost me years off my life too, as losing a spouse (or equivalent) is tremendously stressful and bad for your health; I can't really imagine life without her now (we've been together 20 years as of this past spring). She's by far the best thing to ever happen in my life, weak bones or not.
Me, I don't get ads in my mail about bone mass. I get ads about BONER mass. Heh.
posted by jamstigator at 5:16 PM on September 10, 2009
Me, I don't get ads in my mail about bone mass. I get ads about BONER mass. Heh.
posted by jamstigator at 5:16 PM on September 10, 2009
I love NNT (number needed to treat). It's the perfect hammer with which to smash people who, through intentional deceit or otherwise, try to inflate pharmaceutical or surgical effects with relative reductions.
"Oh, well see, levocrapadine reduces symptoms of eructation of 58%"
"But it was really just an absolute change in incidence of 3.2% to 1.34%."
"Yeah, 58% reduction!"
"You need to treat 54 people to get a positive effect."
"Er...."
"And twenty of them will go blind as a side-effect!"
posted by adoarns at 5:26 PM on September 10, 2009 [1 favorite]
"Oh, well see, levocrapadine reduces symptoms of eructation of 58%"
"But it was really just an absolute change in incidence of 3.2% to 1.34%."
"Yeah, 58% reduction!"
"You need to treat 54 people to get a positive effect."
"Er...."
"And twenty of them will go blind as a side-effect!"
posted by adoarns at 5:26 PM on September 10, 2009 [1 favorite]
CBC radio's Ideas did a good two-part show (available on podcast) about this sort of thing, called You are Pre-Diseased.
posted by zadcat at 6:59 PM on September 10, 2009 [1 favorite]
posted by zadcat at 6:59 PM on September 10, 2009 [1 favorite]
Zadcat, that's a terrific show. It covers prostate and breast cancer screening, with some interesting facts on screening efficacy.
Dr. Gilbert Welch: I think the generic problem is somewhat like the "check engine" lights on your car. Do you have check engines lights? My first car was a '75 Ford Fairlane. There were only two things monitored: my oil pressure and my engine temperature. I now drive a Volvo that is 10 years old, but it is checking about 25 different engine functions. And sometimes a check engine light comes on, and you’re really glad to know, and it leads to something you want to do something about. Sometimes the check engine light is just a nuisance, and it just keeps flashing on and off and the mechanic can’t fix it. And some of the audience might have this experience where they went to get it fixed and it made matters worse. And if you had that experience, you’ve had some of the experience of overdiagnsosis and that’s what I’m worried about. We’re putting more and more check engines lights on the human body. We have to ask ourselves if that is really the best way to get to a healthy society. We’re constantly monitoring for things to be wrong. Is that really the best way to achieve health?
posted by storybored at 8:07 PM on September 10, 2009 [1 favorite]
Dr. Gilbert Welch: I think the generic problem is somewhat like the "check engine" lights on your car. Do you have check engines lights? My first car was a '75 Ford Fairlane. There were only two things monitored: my oil pressure and my engine temperature. I now drive a Volvo that is 10 years old, but it is checking about 25 different engine functions. And sometimes a check engine light comes on, and you’re really glad to know, and it leads to something you want to do something about. Sometimes the check engine light is just a nuisance, and it just keeps flashing on and off and the mechanic can’t fix it. And some of the audience might have this experience where they went to get it fixed and it made matters worse. And if you had that experience, you’ve had some of the experience of overdiagnsosis and that’s what I’m worried about. We’re putting more and more check engines lights on the human body. We have to ask ourselves if that is really the best way to get to a healthy society. We’re constantly monitoring for things to be wrong. Is that really the best way to achieve health?
posted by storybored at 8:07 PM on September 10, 2009 [1 favorite]
The show references an interesting book: Know Your Chances - Understanding Health Statistics.
posted by storybored at 8:13 PM on September 10, 2009
posted by storybored at 8:13 PM on September 10, 2009
« Older "Folk music for people who don't like folk music." | La Pura Vida Newer »
This thread has been archived and is closed to new comments
posted by alms at 9:15 AM on September 10, 2009 [3 favorites]