Modern Medicine’s Sleight of Hand
by Bill Sardi by Bill Sardi
Here are news headlines derived from published studies of four different types of drugs:
- Cholesterol-lowering statin drugs lower the risk for a heart attack by 30%.
- Tamoxifen, the drug used to treat breast cancer, cuts the number of cases of breast cancer by 49%.
- Bevacizumab (Avastin) plus chemotherapy had a 26 percent reduction in the risk of death (hazard ratio of 0.74).
- For males, finasteride (Proscar), a drug often used to treat prostate inflammation, has been shown to reduce risk for prostate cancer by about 25%.
A naïve reader would get the false impression that these drugs will result in 30 fewer heart attacks per 100 statin drug users, 49 fewer cases of breast cancer among 100 tamoxifen users, 26 fewer deaths for every 100 Bevacizumab users and 25 fewer cases of prostate cancer for every 100 users of finasteride.
In reality, these are only relative numbers. Buried deep in the data for these clinical studies is the absolute percentage in "hard numbers" of how many patients in 100 will actually benefit from the drugs. The absolute risk reduction is far less than what is quoted in the newspaper headlines or television reports.
A 25% relative risk reduction can be cited when the absolute occurrence of a disease among patients taking an inactive placebo is 2% versus 1.5% taking a drug. However, the absolute difference is only 0.5% (1/2 of one-percent).
For example, if the incidence of a disease is low, like heart attacks, that only occur among 2 to 4% of adults over a 5-year period, then the number of patients who would have to take a statin drug to prevent one heart attack would be 1000 to 2000.
Let’s say heart attacks occur among 3% of 1000 patients over 5 years — that would amount to 30 heart attacks. A 30% relative risk reduction of 30 would be 9 less heart attacks (21 total over 5 years). That’s two heart attacks that were prevented every year among 1000 statin drug users, or just 0.2% difference (2/10ths of one percent) between users and non-users of statin drugs. Some 1,825,000 statin pills would be consumed to prevent 9 heart attacks — but not 9 lives saved. Statin drugs in the widely-quoted meta-analysis (review of many studies where statistics are pooled together) did not reduce all-cause mortality rates. Statin drugs slightly prevented heart attacks, but not mortal heart attacks.
In the 4 statin drug trials analyzed to come up with the widely-used figure of a 30% reduction in heart attacks, here is what the hard numbers looked like: 67 of 7065 (0.948% — 9/10ths of one-percent) statin drug users died versus 102 of 7054 (1.446% — nearly 1 of one percent) who did not take the drug, for an absolute difference of 0.498% (1/2 of one percent, absolute difference). [British Medical Journal 323: 1—5, 2000]
But cardiologists will look directly into their patients’ eyes and tell them they may die of a heart attack if they don’t take the drug. Actually, there is little chance of benefiting from statin drugs, and more risk of experiencing side effects, like elevated liver enzymes and muscle degradation, which is a mortal risk in itself.
Another part of the shell game is that the cholesterol studies only involved high-risk individuals, but the cholesterol drugs are being prescribed to healthy middle-aged adults. So their chance of benefiting from the drugs is nil x nil.
Cancer drug deception
The same deception is true for the anti-cancer drug bevacizumab (Avastin). Only when used with another cancer drug did bevacizumab increase median survival for colon cancer patients from 15.6 months to 20.3 months, a 4.7-month difference over chemotherapy alone. On a relative basis, this was a 26 percent reduction in mortality. [New England Journal of Medicine 350: 2335—2342, 2004]
Tamoxifen shell game
The same shell game is used to promote Tamoxifen, a drug that many women refuse to take because of its side effects.
The National Cancer Institute (NCI) reports that 10 million American women are at high risk for breast cancer and maybe ought to see their doctors about taking tamoxifen. The NCI concedes only 2.4 million women are likely to benefit from taking the drug, and out of this group of 2.4 million the number of cases of breast cancer is likely to be reduced from 58,148 to 28,492. The number of cases of breast cancer appears to be cut in half (49 percent) when viewed as a relative percentage.
But in reality, only 2.4 percent of the 2.4 million women would be expected to develop breast cancer if they didn’t take tamoxifen and 1.2 percent if they did. So the difference in hard numbers is only 1.2 percent. [Journal National Cancer Institute 95: 526—32, 2003]
This is hardly a good way of selling Tamoxifen to American women, so the relative numbers are used instead. Then it’s called a blockbuster drug. It’s very misleading. Yes, admittedly when looking at hard numbers about 20,000 additional women would be spared from breast cancer. The problem is you have to get over 2 million women to take the drug to realize this benefit.
Another one of the hidden facts about tamoxifen is that half of the women with advanced estrogen-positive tumors immediately fail to respond to this drug and are removed from drug therapy. So the studies don’t always include these failures. [Drugs 61: 1721—33, 2001]
Current efforts to expand tamoxifen therapy as a preventive agent for breast cancer have begun. High-risk but otherwise healthy women derive some risk reduction if given tamoxifen. Of the 6600 women who took tamoxifen, there were 69 fewer tumors compared to 6000 other women who took a dummy pill. In other words, tamoxifen benefited only about 1 in 100 high-risk women as a preventive measure. Yet it was widely hailed as a breakthrough! [Associated Press Oct. 30, 1998] There has been resistance to the idea of giving healthy women a drug that runs the risk of serious side effects. The National Women’s Health Network asked: "Does this trial represent disease prevention, or disease substitution?" [Breast Cancer Action, Feb. 26, 1996]
Prostate cancer prevention is illusory
The same ruse is used to promote finasteride (Proscar) for prostate cancer prevention. Finasteride is reported to reduce risk for prostate cancer by about 25%. It is estimated that widespread use of finasteride by adult males would add 300,000 years of life over a 10-year period. [Urological Oncology 22: 362—68, 2004] However, it would cost about $3.4 billion per year and require every male in the U.S. age 55 to 80 years of age to take the drug, to achieve this aggregate health goal. [Journal National Cancer Institute 96: 1410—12, 2004]
For a hypothetical group of 1000 sixty-two year old men treated with finasteride, an increased survival of 140 life-years is forecasted, or a statistical increase in survival of about a month (0.14 years) per individual. [Journal Urology 175:934—8, 2006]
Here are the hard numbers for finasteride:
The landmark finasteride prevention study involved 9060 men. Among the men taking finasteride, 18.4% developed prostate tumors versus 24% in the placebo group, for a real difference of about 6 in 100 men. This means, over a 7-year period, for every 100 men taking the drug, 94 would experience no benefit while 6 would. While most men will develop cancerous cells in their prostate gland if they live long enough, few die from it.
In the finasteride study, only 10 men died of prostate cancer, five in each group taking finasteride or placebo. Men taking finasteride who were found to have prostate cancer were more likely to have a fast-growing kind, which would be expected to be more deadly. So finasteride may slightly reduce the risk of prostate cancer, but not dying of prostate cancer. Men taking finasteride had more erectile dysfunction and loss of interest in sex, but also had fewer urinary problems. [New England Journal of Medicine 349:215—224, 2003]
News reporters and health agencies all participate in the charade
A bigger problem here is that news reporters, many of them physicians or biology-trained, know the magnified numbers in these drug studies are hardly worth reporting. They are supposed to be "the fourth estate," the guardians of public welfare. News reporters these days appear to be allied with the pharmaceutical companies in reporting these sham numbers. When the public is informed of the absolute numbers, they are buried at the bottom of the news story.
Public health agencies are not a cut above the hype news reporters either. The National Institutes of Health uses these relative numbers to announce results from freshly-completed clinical studies. The Food & Drug Administration uses numbers like these to approve new drugs.
Even doctors are fooled by the relative numbers. A study of internists and general practitioners who prescribe cholesterol medications found they were less likely to prescribe these medications when the benefits were reported in absolute rather than relative percentages. [British Medical Journal 309: 761—64, 1994]