DRUG BUST Alan Cassels
How do you calculate your net worth, the value of your company or your chequing account balance? You don’t have to be an accountant to understand that these amounts are determined by looking at both columns in the ledger – your assets and liabilities and your credit and debit. In Double-Entry accounting, your assets are listed on one side; your liabilities or debts on the other. Subtract one from the other and you arrive at an idea of your net worth or the value of your company or the amount remaining in your chequing account. Simple, right?
In fact, this calculus can be helpful in determining the value of prescription drugs. For one thing, drug effects are largely described through numbers. Drugs are put through years of study, often involving thousands of patients and hundreds of staff carrying out detailed and meticulous note taking to gather data on the health effects – both positive and negative – of these treatments.
It is a truism in medicine that therapies designed to help you could also hurt you. A statin drug to lower your cholesterol has the potential to lengthen your life or shorten it. You would only ever take this drug if you understood both sides of the ledger and if the likely benefit of swallowing the pill every day exceeded the harm, given your individual health circumstances.
Applying this accounting analogy to the world of prescription drugs can often reveal some shocking information. For starters, the so-called “benefit to harm ratio” is never clear-cut. Often, the numbers on the minus side of the balance sheet – in terms of numbers of people in a clinical trial who have suffered a “Severe Adverse Event” (such as being hospitalized, an extended hospital stay or dying) – are often hard to come by, misleading or simply absent.
In financial terms, how could you ever know the value of your company or the size of your chequing account unless you also identified your liabilities and the size of your debt? The further you dig into this, the more you find that what you believe about the value of your drug may be biased, incomplete or simply wrong.
The lack of harm data in published research is of growing interest to many researchers, including Dr. John Ioannidis, a Greek physician and researcher who has published extensively in this regard. One of his studies, published nearly 10 years ago in the Journal of the American Medical Association, examined the quality of safety data from large, randomized trials (the ones of the highest quality) and found that “generalizable data on drug safety reporting are sparse.” In other words, we’re not getting the full picture of the liabilities of many of the most widely studied drugs.
Looking at research across seven different medical areas, Ioannidis found that these minuses – the adverse events and toxicities from the drugs – were adequately defined in less than 40 percent of studies. When people drop out of a study because of the toxicity of the treatment, we want to know why they dropped out. He found that “only 46 percent of trials stated the frequency of specific reasons for discontinuation of study treatment due to toxicity.” Let me put this another way: less than half the studies revealed the percentage of study participants who couldn’t even handle the treatment long enough to remain in the study.
Another of Ioannidis’ articles focusing on this issue was published in the free, online medical journal Public Library of Science (www.plos.org/journals/). Entitled Why Most Published Research Findings Are False, the article described, mathematically, why much of what is published in the medical world is simply not ‘true’ in terms of a generally agreed upon definition of ‘true.’
One article about SSRI antidepressants – drugs like Paxil, Prozac or Zoloft – asks is this an “evidence myth constructed from a thousand randomized trials?” When you get huge numbers of trials and sift them by clever, selective reporting of results, you arrive at a pile of dung smelling like a rose garden. This method of reporting almost always makes drugs appear much more effective or safer than they actually are.
The balance sheet approach to calculating the net health effects of drugs has attracted the attention of some local champions as well. A few years ago, I attended a presentation by Dr. Jim Wright at UBC, in which the major studies of statins (used to lower cholesterol) were discussed. The findings revealed that the trials did not adequately take into account the full range of “Serious Adverse Events” due to statins. Even though SAEs must be reported and documented in clinical trials, when Wright and his colleagues looked into it, most published trials of statins did not report the total number of people with at least one SAE. So even if the drug shows benefits for preventing heart attacks, for example, we don’t know how many additional people would have been hospitalized or stayed in a hospital longer because of the drug.
To get the full picture of the net effects of statins, Wright and his colleagues have spent considerable energy writing to the authors of studies, politely asking them to provide their data – the complete set of SAE data from specific studies. Sometimes they were ignored and occasionally they received explanations or additional data. However, at no time did they ever get the specific SAE data they were requesting. My interpretation of this is simple: there are many weaselly ways to hide the minus side of the drug ledger.
So what to conclude about researchers running a drug trial who can’t or won’t give you the complete picture of serious adverse events? Do we assume they are dishonest and hiding the bad effects they find in the clinical studies? I think it is safest to assume no news isn’t good news. If you aren’t getting the full data on a drug’s potential serious adverse events, it is because those data are likely not flattering. Without those data, it will be easier for the marketers to create a rosier picture of the net worth of the drug. To me, the “trust us, we’re experts” response to queries about full safety data does not deliver warm fuzzy feelings of confidence.
And what can we say about the statins – drugs like Crestor, Lipitor or Zocor – that make up the most widely consumed class of prescription drugs on the planet? What do we really know about their net health benefits?
Dr. Wrights’ research, which is helping paint a clearer, overall picture of the value of statins, suggests that for about 80 percent of the people currently prescribed statins, the benefit of taking the drug does not exceed the harm. For the other 20 percent, – men who have had a previous heart attack, for example – the benefit may outweigh the harm by only a marginal amount, probably helping only about one in 20 of these higher risk patients. And there is good evidence that women and the elderly do not receive any overall health effects from statins. In other words, the most widely sold drug class in the world has both good and bad effects and when you include both of those entries in the ledger, for most people, the net effect is next to zero.
The implications of this are huge. Through public drug plans, we spend hundreds of millions of dollars every year on this drug class. Our private insurers and our own bank accounts contribute hundreds of millions more to drugs that provide no overall net health benefits for the majority of people who take them. Patients who are actually harmed by the drug would naturally add billions more in costs to our health system.
Bad news is slowly starting to trickle in on the statins. A small study in June, reported in the Canadian Medical Association Journal, suggested that statins are potentially even more harmful than we expected. This study found that, in some patients, statins have a measurable toxic effect on muscles (muscle weakening, being one of the most common adverse effects related to statins). This effect is measured by performing muscle biopsies and evidence of muscle related damage was found in about 60 percent of the individuals taking statins who complain of sore muscles. More importantly, the study also found evidence of muscle damage in about five percent of patients who had no symptoms whatsoever.
Taking other drugs may also exacerbate the risks. As the authors note, the “risk of statin-associated myotoxic adverse effects is enhanced by concomitant use of some medications.” And, well, most people are taking other medications in addition to their statin.
If you have sore muscles and pain from your statin drug, you can just stop taking it, right? Well, one of the problems is that the muscle damage is not readily reversible and it can persist in patients who stop the drug for four months or even longer.
I find it especially troubling when I constantly hear the experts spouting the company line that statins are “safe,” “well-tolerated” and “highly effective,” given the incomplete reporting of safety data from clinical trials. We would never enter into a financial arrangement where the full transaction costs aren’t made clear so why do we engage in transactions that involve taking prescription drugs without knowing what the downside – the costs, debits, losses – might be?
It doesn’t have to be this way. Luckily, the Canadian government is embarking this fall on a countrywide strategy to study drug safety and effectiveness. It’s a strategy long in the making and it comes with huge expectations. It will also likely be years before anything fruitful comes of it.
In the meantime, you must ask yourself one simple question: “Do I know what is on both sides of the ledger before I enter into a drug transaction?” If not, ask more questions. And don’t be satisfied with weaselly answers.
Alan Cassels is a drug policy researcher at the University of Victoria and author of The ABCs of Disease Mongering, an Epidemic in 26 Letters. He has never had his cholesterol level tested and knowing what he knows about statins, he never will. firstname.lastname@example.org