Ain’t What It’s Cracked Up To Be
Many of the medications we’re prescribed have a reasonable probability of not working for the problem for which they’re being prescribed.
According to a senior physician executive with GlaxoSmithKline…
“The vast majority of drugs – more than 90% – only work in 30 or 50 percent of people…
“Drugs out there on the market work, but they don’t work in everybody.”
The most toxic category – oncology, or chemotherapy drugs – has the lowest response rate (see table below).
How many cancer patients are aware that the pain and suffering – physical and financial – that they’ll endure on chemo will likely be for nothing in three out of four cases?
Medications Often Benefit Only a
Minority of Patients
The bottom line for our health and our healthcare is that many medications are effective in only a minority of patients – generally in the 25-60% range of effectiveness.
The following table lists some drug response rates by therapeutic area:
Response Rates for
Select Therapeutic Areas
In some cases – such as depression – these results are barely better than placebo (inactive sugar pill).
According to Best Choices From the People’s Pharmacy:
“Placebos generally produce improvement in 30 to 50% of depressed patients, so the drugs are just a bit better than the dummy pills.”
Effectiveness Varies With Stage of Disease
All these shortcomings – which are discussed in more detail in Our Healthcare Sucks – have to be weighed against the probability of benefit from the medication(s) prescribed for your condition.
Those with early-stage disease symptoms are less likely to actually realize a drug’s purported benefits.
This relative expected benefit is calculated for individual drugs using something called the Number-Needed-to-Treat, or NNT (see the video below for a discussion of NNT – but remember that, despite the speaker’s faith in them. randomly controlled trials can also be manipulated – and click here for NNT examples for specific conditions).
More About NNT
NNT is basically an effort to get a better handle on which drugs work best for which conditions. Medical experts who specialize in evidence-based medicine have analyzed the clinical literature and developed a methodology based on the clinical evidence.
This allows them to calculate what the expected probabilities of a drug’s effectiveness will be in a typical patient (which is subject, of course, to the caveats about every patient being uniquely affected by all medications).
The resulting methodology is called the “Number Needed to Treat”, or “NNT”, for a particular drug. This represents the number of patients needing to take a given medication for a single patient to realize the drug’s expected benefit.
A low NNT number is good – it means fewer patients need to take the drug for one to benefit, giving each patient taking the drug a higher probability of realizing benefit by taking the drug.
A high number is the opposite – more patients required for a single patient to benefit reduces a given patient’s odds of receiving any benefit while the risks of side effects, complications, and expense remain the same (though perhaps greater or lesser in any given patient).
The NNT considered effective differs depending on whether the drug is being prescribed as treatment or as a preventive measure.
Many physicians regard an NNT of 10 or less as representing an effective therapy and an NNT of 20 or less as effective for prevention.
This is because you’d expect to have to apply a drug as a preventive measure to a larger population of people without symptoms to realize any benefit than you would for those with symptoms and in current need of treatment. This is discussed further below.
First, however, let’s look at what baselines exist for judging a medication’s effectiveness.
Medications vs. Surgery
According to the evidence-based specialists at Bandolier, the NNT for bypass surgery (for left main artery blockage) is 6. This means for every 6 patients undergoing this highly-invasive procedure, only one will benefit by avoiding death within 2 years. I suspect most such patients think their odds for success are much better than 5-to-1.
The NNT for a carotid endarterectomy to prevent stroke or death over 2 years is 9, meaning 8 of 9 such patients receive no benefit.
While these aren’t very impressive odds for such high-risk interventional procedures, they do serve as baselines for comparison with NNT numbers for more commonly prescribed medications (click here for a table of NNTs for other cardiac interventions).
While none of those medications listed manage to get below the 10 NNT threshold suggested for medication treatment effectiveness, treatment with these medications was far more effective (lower NNT numbers) for severe disease than for mild disease.
An NNT of 11 for statin medications in those who’d had a prior cardiovascular (CV) event is obviously more effective than an NNT of 35 for those who had not, meaning that statin drugs are more effective in preventing a second heart attack or stroke than preventing a first such event.
For the former, one in eleven patients can be expected to benefit, while 10 of 11 won’t; in the latter group, only 1 in 35 can be expected to benefit.
For those who’ve never had a heart attack or stroke, this translates into 97% (34 of 35) receiving no benefit in the 4 year timeframe measured while incurring all the risks of side effects, complications, and expense for that 4 year period.
The real vs. perceived benefits of statins is discussed in other posts and in the book. For now, it’s clear they work better to prevent the recurrence of an “adverse event” than they do to prevent an initial event.
For blood pressure treatment, 1 in 15 patients with severe high blood pressure can be expected to benefit from drug treatment over a one year period, while 14 of 15 likely will receive none of the described benefit during that timeframe.
While this is probably far less effective than most people realize, it’s far better than results for those with mild hypertension (slightly elevated blood pressure), where 699 out of 700 are likely to receive no benefit.
Combined with information below about lack of benefit in driving blood pressure below certain levels, a prudent person with only moderately elevated blood pressure might consider non-drug approaches to blood pressure reduction – which remains an important goal for longevity and avoidance of a broad range of diseases, not just cardiovascular disease.
Breast Cancer, Too
In Know Your Chances: Understanding Health Statistics, the physician-authors cite breast cancer treatment where even those who benefit may suffer off-setting negative effects that negate a drug’s benefit.
According to their report, while the drug tamoxifen (Nolvadex) works for some high-risk patients to prevent a first occurrence of breast cancer, it may also cause more life-threatening complications from other causes than the invasive breast cancers it reduces.
Even this may overstate the benefit.
There was no difference in death from breast cancer or from all causes between those women taking tamoxifen and those taking a placebo.
And although there were 16 fewer cases of invasive breast cancer per thousand women taking tamoxifen, there were 21 more cases per thousand of invasive uterine cancer and blood clots.
And despite early assurances from the National Cancer Institute (NCI) that another drug, Raloxifene (Evista), offered comparable benefits with fewer side effects than tamoxifen, subsequent analysis found there were no meaningful differences in the increased rates of dangerous blood clots and aggressive uterine cancers with either drug.
Better Options For Most People
The point of these disappointing examples is that most medications aren’t as effective as the public is led to believe.
For many people, this could make the analysis of their often significant risks vs. such limited benefit tilt more in favor of drug-free approaches to disease prevention, in particular.
At a minimum, it should give pause to those inclined to listen to drug company propaganda – whether on TV or in the doctor’s office – extolling the virtues of their pricey medications to prevent disease when cheaper, safer, and better lifestyle measures are available (see “Better Than Drugs“).
A Bohemian Ending
To close on a lighter note, take a look at this parody of our polypharmacy problem – to the tune of Queen’s Bohemian Rhapsody…
Multiple medication use ‘not always hazardous,’ say researchers Improving the Quality of Life for the Frail and Elderly In praise of the NNT Drugs for Fibromyalgia: How Good Are They?