Proceedings of the Standing Senate Committee on
Social Affairs, Science and Technology
Issue 9 - Evidence - March 27, 2014
OTTAWA, Thursday, March 27, 2014
The Standing Senate Committee on Social Affairs, Science and Technology met this day at 10:30 a. m. to continue its study on prescription pharmaceuticals in Canada.
TOPIC: The nature of unintended consequences in the use of prescription pharmaceuticals.
Senator Kelvin Kenneth Ogilvie (Chair) in the chair.
[Translation]
The Chair: Welcome to the Standing Senate Committee on Social Affairs, Science and Technology.
[English]
I'm Kelvin Ogilvie, a senator from Nova Scotia and chair of the committee. I'll ask my colleagues to introduce themselves, starting on my right.
Senator Seidman: Judith Seidman, from Montreal, Quebec.
Senator Eaton: Nicky Eaton, Ontario.
Senator Seth: Asha Seth, Toronto, Ontario.
Senator Enverga: Tobias Enverga, from Ontario.
Senator Cordy: I'm Jane Cordy, a senator from Nova Scotia.
Senator Eggleton: Art Eggleton, senator from Toronto and deputy chair of this committee.
The Chair: Thank you, colleagues. We are continuing our discussion and study of prescription pharmaceuticals in Canada. This is part four of a four-phase study on prescription pharmaceuticals. In this section, we are dealing with the nature of unintended consequences.
Today, we have the good fortune to be hearing from researchers who address drug therapies. I will recognize each of them in turn, they will present, and then we will open up the session for questions. By agreement, Dr. David Juurlink, who is head of the Division of Clinical Pharmacology and Toxicology at Sunnybrook Health Sciences Centre and the Institute for Clinical Evaluative Sciences will begin.
Dr. David Juurlink, Head, Division of Clinical Pharmacology and Toxicology, Sunnybrook Health Sciences Centre, Institute for Clinical Evaluative Sciences: Thank you, and good morning. As Senator Ogilvie mentioned, I'm a physician at Sunnybrook Health Sciences Centre. I specialize in internal medicine and clinical pharmacology. In addition to my clinical work, I conduct research in the field of drug safety.
I've been asked to speak about the unintended consequences of prescription medication use. This is an extensive topic, as you know, but it's relevant to every Canadian because almost everyone either takes a prescription drug or will do so at some point in the future.
Canadians leave doctors' offices by the thousands every day with prescriptions in hand. They expect and hope that these drugs will do what the doctor intended — make them feel better or make them perhaps even live longer. Unfortunately, many of the prescriptions we issue do neither of these things. Every drug we prescribe carries some risk and, as a result, Canadians are seriously harmed and sometimes even killed by medications that are prescribed with the best of intentions. No one can tell you exactly how often this occurs but, as someone who sees this first-hand every day and who studies it, I can tell you that it happens a lot more often than patients and doctors appreciate.
When drugs are prescribed to large numbers of people, some harm is inevitable. That's just the nature of drug therapy. It's worth considering some of the reasons why drug-related harm happens so much more often than it might. Without such a discussion, there really is no prospect of improvement.
I want to be clear that the unintended consequences of prescription medications are not exclusively the fault of doctors, patients, regulators or drug companies, but each of these groups plays a role in the genesis of the problem and in improving it.
The issue of how drugs get to market is important. For a new drug to gain market approval, the manufacturer has to provide Health Canada with some evidence that the drug does what they say it does, something beneficial, and that it does so with minimal risk. This is usually done in the context of a clinical trial, a type of research where some patients get the drug and some don't. Well-done trials can give us a sense of what a drug can do.
These studies aren't perfect. They involve patients who are carefully selected and often very different from those in the real world. It's difficult for me to overstate the importance of that point. The trials are often small and of short duration. They often examine drug effects that sound important but, frankly, aren't — changes in cholesterol or blood pressure or blood sugar, rather than, say, reductions in heart attacks or strokes. When companies find unflattering results about their products, they are under no obligation to publish them. As a result, doctors and patients can be misled, and what we think we know about a drug's benefits and risks is often wrong.
Not only do we operate under those uncertainties, we write far more prescriptions than we should. There are a variety of reasons for that. First, there is, to an extent, a societal expectation that if you have a problem, there must be a drug to treat it. Many patients, for example, insist that they cannot sleep without taking a pill. My grandmother is that way. Others believe that a sore throat or a cough automatically warrants an antibiotic. Of course, not everyone feels that way, but many patients visit their physicians with a problem and the expectation of leaving with a prescription.
If we're a nation of pill takers, that's not exclusively the patient's fault. It's clearly partly the doctor's fault. As a group, doctors prescribe drugs far too freely. This sometimes happens because patients pressure us to do that, but more often than not it's because we have a desire to help people. That's why most of us went to medical school. Prescription medicines are one of the main tools at our disposal, but we don't always, I think, undertake a fair assessment or appraisal of the risks and benefits of a drug when we reach for the prescription pad.
The drug industry has a major role to play in this issue. They are in business not to make the world a better place but to make money, and they do that by selling their products. I've already mentioned that they don't always publish unflattering findings, but there is more to it than that. They deliver countless free samples to doctors' offices, which might sound like a good idea, but they're simply marketing devices and nothing more. They sponsor educational talks at fancy restaurants where experts speak about their products to other doctors and are paid handsomely to do so. The industry supports the development of treatment guidelines and patient advocacy groups, which in some ways is a good thing but is also in a sense a type of marketing.
These are some of the reasons why doctors write $30 billion worth of prescriptions every year in Canada, but this degree of prescribing carries enormous risks for the population. I'd like to give you one tangible example of the unintended consequences of prescription drug use. It bears on what I view as the most serious drug safety crisis facing Canadians today.
We have all experienced pain at some time in our lives, and that's a fact of life. Millions of Canadians live with pain daily from arthritis or fibromyalgia or spine disease and other conditions. This is a huge problem, and it can have a tremendous impact on one's quality of life. Patients don't want to be in pain, and doctors want to help. Prescription drugs are a common element of the treatment of chronic pain. The problem, however, is that for many patients these drugs don't work very well. What would you think if I told you that more than 10,000 Canadians have died over the past two decades as the result of prescription painkillers like OxyContin and Dilaudid? That sounds like a lot, and it is, but in fact no one at Health Canada can tell you exactly how many such deaths have occurred. When I say 10,000, I give you what I think is a conservative estimate. For every tombstone, there are dozens, probably hundreds, of people who have suffered from addiction and its consequences. This is an epidemic, and it started in the doctor's office.
I will explain briefly in general terms how this happened. Imagine a 45-year-old man, a mechanic, who goes to his doctor because of back pain. It's persistent and interferes with his work and quality of life. Of course the doctor wants to help. Depending on the cause, there might be some non-drug options, but perhaps they're expensive, like physiotherapy, or perhaps they're invasive, like surgery, or perhaps they're too hard to implement, like weight loss and exercise. In 1990, the doctor might have started treating the patient with acetaminophen or Tylenol. If that didn't work, he or she may have progressed to a drug like Motrin or Naprosyn or any of a dozen other anti-inflammatory drugs. These drugs help some people but can cause kidney problems and life-threatening bleeding from the stomach and ulcers. Every doctor knows this and by teaching and experience has come to respect the risk of these drugs.
In the mid to late 1990s things changed. Doctors began prescribing strong painkillers like OxyContin much more liberally. These drugs are called opioids and are chemically similar to heroin. If that sounds like hyperbole, believe me, it is not. Why did doctors start doing this? We did so in part because the other medications we have don't work well and because we are afraid of their side effects. We also embraced opioids because we were told to do so by experts, by key opinion leaders, that we should. These people were paid huge sums of money by drug makers to give lecture after lecture to front-line doctors. In those lectures they sometimes overstated the benefits of these drugs and minimized risks and suggested that these drugs were much less addictive than previously thought. They claimed doctors who weren't embracing opioids for their patients in pain were opiophobic and doing their patients a great disservice. Some of these experts even conveyed inaccurate information to thousands of medical, dental and pharmacy students and in at least one instance a textbook with incorrect, biased and dangerous information was produced by a drug company and given free of charge to thousands of medical and dental students in Canada. These people are now writing prescriptions.
The result of all this is a huge shift in how we treat pain. A few years ago a few of my colleagues and I published a paper in the Canadian Medical Association Journal showing that from 1991 to 2007 the prescribing of oxycodone, the active ingredient in OxyContin, rose by 850 per cent in Ontario. This is a drug more potent than morphine. Some see this as better pain treatment. I don't think that's true. I want you to consider this: These drugs were granted market authorization in Canada because in small clinical trials lasting a few months they reduced pain scores more than sugar pills or placebos. These were studies done in relatively healthy people without known risk factors for drug addiction, people who weren't taking sleeping pills or antidepressants or drinking alcohol to excess. That was the basis on which Health Canada allowed these drugs on the market and the basis on which experts told us to use them not just for a few weeks but for years, if needed, and at doses that would have been unthinkable in 1990 and would never have been permitted in the trials themselves. We now have millions of tablets of what amounts to pharmaceutical grade heroin being dispensed every year in Canada. Some patients get better, but many don't and they become dependent on these drugs and can't stop taking them without going into horrible opioid withdrawal. Their body gets used to the drug; the dose is escalated. Maybe they take an extra dose now and again or take it with alcohol or a sleeping pill and they just don't wake up. For others the drug's pleasurable effects take over their lives and they start to crush it and snort it or inject it. They borrow and steal drugs from family and friends or buy it on the street for $20 to $40 a tablet to pay for their habit. Heroin is cheap and it is making a comeback. That's exactly why that's happening.
In general terms, that is why more than 10,000 Canadians are now dead and why more lives have been ruined by these drugs than anyone can tell you. All of this sprang from well-intentioned prescribing that I think was misguided. That's just one glimpse into the unintended consequences of prescription medications in Canada but I think it's the most pressing threat we face today and there are valuable lessons to be learned from that story, many of which apply to other drugs and diseases.
With that I'll conclude and thank you very much for the opportunity to address you here today.
The Chair: Thank you, Dr. Juurlink.
I'll now turn to Dr. Abrahamowicz. Dr. Abrahamowicz is a James McGill Professor, Department of Epidemiology, Biostatistics and Occupational Health at McGill University. You are appearing as an individual. We're delighted to have you.
Michal Abrahamowicz, James McGill Professor, Department of Epidemiology, Biostatistics & Occupational Health, McGill University, as an individual: Thank you very much for the invitation. It is both an honour and a challenge to be here. I am a researcher, not a clinician. Actually, I am a biostatician, so I see these things from a more aggregate research aspect. My talk may be partly complementary to Dr. Juurlink's talk. I produced a monstrous presentation; I will just give you the main issues here and the information.
I will start with another dramatic example of the well-established unintended effects. This is probably the most established study in the history of mankind. It concerns the extremely popular drug marketed under the name of Vioxx — the pharmaceutical name is rofecoxib — which was extremely popular for treating different arthritis conditions, rheumatoid arthritis, and so on.
On September 30, 2004, the pharmaceutical giant Merck decided to withdraw this drug from the market. The drug was approved by FDA, as Dr. Juurlink described, five years earlier, in 1998. The reason Vioxx was withdrawn from the market was the result of a trial that was not intended as the primary goal to look at the cardiovascular safety; it was for a different purpose. However, as a by-product of analyzing the results of ongoing trials, people noticed a significant, more than doubling risk of serious, potentially life-threatening cardiovascular events, especially strokes and so-called myocardial infarctions, or heart attacks in common language. Subjects who were exposed to Vioxx more than doubled the risk of those events than those who were not. Because it was a randomized trial, it was sure that it must have been related to the drug.
What are the implications of this for society at large? Vioxx was aggressively marketed by Merck. By 2004, when it was withdrawn, it was estimated that the annual sales were in the order of US$2.5 billion worldwide. It was sold in more than 100 countries worldwide. It was estimated that about 80 million people worldwide have used Vioxx at some point in time during the five years it was on the market. By this time, it was estimated that there were more than 10 million prescriptions for Vioxx every month in the U.S. alone. It was a huge impact on the safety of the population.
The initial approved trial and some other studies indicated more than doubling and, specifically numerically speaking — I'm a biostatistician so numbers appeal to me — it was estimated that about 1.6 per cent of patients would have additional cardiovascular events if they used Vioxx as opposed to the situation where they were not using Vioxx. Specifically, there was a 3.5 per cent incidence in those who were taking the medication versus 1.9 per cent among those who did not. This implied that for every million long-term users of Vioxx — and I will explain why I say ``long-term'' — there would be about 16,000 or somewhere in the order of 10,000 to 20,000 additional serious cardiovascular events that could have been prevented if the patient did not use this medication.
Based on this, there were high-level publications by one of the top cardiologists, Dr. Topol, who claimed that probably several thousand and maybe even hundreds of thousands of those events could have been prevented worldwide if the cardiovascular risk associated with this drug was detected earlier. It was on the market for five years.
Interestingly, we can ask: What was the reason for such a delay? Why did it take five years? First of all, the initial trials that lead to approval, as Dr. Juurlink said, are mostly for the effectiveness or efficacy of the drug. These are short-term, with a limited sample size and a limited number of patients. For the specific trial we did to approve Vioxx, the average duration of follow-up was only nine months. On the other hand, in the approved and other studies, the cardiovascular risk associated with Vioxx started to increase after about one year of treatment. So the initial trial was too short to detect the full impact. In spite of this, when I looked at the original publication, interestingly they already had statistical significance about non-random increase in the myocardial infarction, not in strokes. However, this was dismissed because the numbers were small. It was 0.4 per cent in one group in those who used Vioxx. Remember, this was a small duration of the trial.
Interestingly, in the meantime when Vioxx was on the market, there were some large-scale epidemiological studies of the type Dr. Juurlink is performing, in the U.S., showing systemic increase of cardiovascular risk. These were ignored or dismissed based on the general claim that these were not randomized trials and therefore were not rigorous enough, even if these were state-of-the-art studies and the result was systematic.
In addition, basic science evidence was developed by people who work on the chemical and biological properties. They proved experimentally the specific mechanisms which I, as a statistician, cannot fully describe, but by which Vioxx and similar drugs increase blood pressure. This leads to an exaggerated response to plaque rupture, and this induces serious cardiovascular events.
All this was ignored for a long time, until finally it was impossible to ignore it and the company withdrew the medication. Of course, this is as serious an example as it could be. Some of our colleagues ended up on CNN news and so on; this was such a big story. To put this into perspective, of course, problems with drugs are not necessarily related to the biological unsafety of the drug. There may be other problems.
There was an interesting study which looked at more than 1,700 drug recalls in the U.S. between 2004 and 2011, so almost two recalls per week. Almost every two days there was a recall of a drug. Interestingly, only 5 per cent of these recalls were related to the inherent unintended effect of the drug if it was used as prescribed, which was the case with Vioxx. Most of the recalls were related to human errors in the administration, packaging, the way the drug was given, mixing up the dose, mislabelling, and so on. Still, this 5 per cent is very important, systemic evidence of unintended drug effects.
Further, to put things into perspective, not all drugs in this category are as serious and life-threatening as was the case with Vioxx. As Dr. Juurlink mentioned, almost all drugs have side effects. These may range from minute side effects, such as scratching or problems with sleep, to serious, to life-threatening.
Joel Lexchin, from the University of York and University of Toronto, took big pains to reconstruct the causes for 41 withdrawals of drugs from the Canadian market — now I'm speaking specifically of Canadian data — between 1963, when systematic assessment of the drugs started to be routine, to 2004.
In four of my slides I reproduced directly the reasons. We see that the reasons vary. Some of these often happen, like hepatotoxicity, but some are relatively mild reasons. It depends on the drug, of course. It depends on the mechanism, the study, appropriation, and so on. Specifically, WHO, the World Health Organization, classifies different types of adverse effects of drugs. There's some interpretation, but I would say they identify at least these five types of adverse effects.
One is an almost immediate effect, related directly to the drug being in the patient's plasma and people reacting to the drug. An example of this would be the well-established effect of different psychotropic medications on cognition, especially among elderly people, people who are using so-called benzodiazepines, a popular drug like Valium. These drugs affect cognition. People have a tendency to be a bit fuzzy and they don't notice the stairs or the sidewalk ends, and they fall and break legs or have serious injury. These are well-established, the immediate effect of being under the drug at this moment. There are other effects. Some effects are called idiosyncratic, meaning they are systematically observed in epidemiological studies, but the mechanism is not known, so the evidence is slightly weaker. There is also a cumulative effect related to chronic, long-term use of the drug. There is also a delayed effect, sometimes delayed by years. For instance, a mother may be using a drug during pregnancy and it has an impact on the infant after birth. In addition, different effects may be related in different ways to how the drug is used. It's not only taking the drug but the dose, the duration of the treatment, and so on.
I use an example of the Canadian study recently performed by our group. We looked at the association between oral glucocorticoids, a very popular therapy used to control rheumatoid arthritis to reduce the activity and the inflammation, versus the risk of serious infections. Using a large Quebec administrative database, we identified, among 16,000 elderly patients, 1,800 serious infections, so serious that it led to hospitalization. We could demonstrate, as was demonstrated before, that the use of glucocorticoids was statistically significantly associated with increased risk.
However, what I want to emphasize is that the risk depended largely on how the drug was used. People who used the drug for the last four weeks at a moderate dose may have a 10 per cent risk increase, which is probably minor. If they used the same dose for the last six months, it will be a 50 per cent increase. If they were on this dose for three years, they would have double the risk.
On the other hand, the same doubling of the risk may occur if the patient was on this drug for only four weeks but the dose was high, 40 milligrams. There's a dose/duration trade-off. If they were on the high dose for the last three months, we have a dramatic fivefold increase of risk of serious infection.
All this is to claim that I think while we need to evaluate the drugs, we also need to evaluate this in relation to how the drugs are used. All drugs are admitted to the market because of some evidence of their therapeutic benefits. We have to do a better job of evaluating the trade-off between the risks and benefits, and this may help us in optimizing the treatment and deciding that, for instance, we should avoid accumulation of the drug. Maybe there should be so-called drug holidays; people should not stay on the same therapy, especially on a high dose, for too long a period of time.
There are other methodical challenges, but I will not have time to dwell on them. Thank you very much, and sorry for my going too fast.
The Chair: Not at all.
Before I open the floor up to questions, I remind colleagues that we have studied the clinical trial system and we have made very significant recommendations in these areas with regard to many of the issues that have been raised today. I will follow your questions carefully, and those that we've already made significant recommendations on that parallel the points being made, I may urge a truncation of the detailed response. We want to get to new information that relates directly to unintended consequences, which of course is what they are addressing here today. Not everyone has been a part of the four-part study that we have been dealing with.
Senator Eggleton: Thank you very much, gentlemen, for your presentations and for your ongoing excellent work that both of you are noted for.
In responding to your presentations, I want to focus on three areas: physician prescribing behaviour; the roles and responsibilities of the pharmaceutical companies; and the regulator, the regulator being Health Canada.
In terms of physician prescribing behaviour, a number of things come into play here. We've talked about prescription drug misuse. We've talked about overuse. We've talked about addictions, and you've referred to these as well. Drug-to-drug interactions, I didn't hear, but that's an issue in all of this as well.
Much of this would lead to the suggestion that there needs to be a lot more physician education and ongoing training in terms of their prescribing activities. Do you have any particular suggestions as to how that might happen, and should it be harmonized across Canada?
Dr. Juurlink: That's an excellent question. I think that education is a critical element of ongoing maintenance of competency for physicians. I'm not sure that education on its own is going to change a great deal with how we prescribe drugs.
One of the problems is that physicians are — as you are, too — busy people. Some of them have 60 or 70 patients to see in their offices daily, they have paperwork to do and they have kids and a family to get home to, so they don't have as much time for education as you might think or as they would like. We are required to maintain a certain level of ongoing education. For many people, the easiest way to do that is to go to one of those talks I alluded to in my presentation. Go to a talk where an expert speaks about a topic. The experts who do that are generally well- intentioned, but they have biases, conscious or unconscious, and they are giving the talks repeatedly for the same company where they say what the company would like them to say.
I think education is important, but sometimes the way in which education is delivered is suboptimal. I think the best way for doctors to be educated about these issues is to attend conferences. Sometimes those, too, are sponsored by drug companies but in an unrestricted way so that the companies don't have any control over the content, the speakers or what is said.
Perhaps I have gone a bit off topic, but I think education is important. The extent to which it is going to improve physician prescribing — opioids is a good example. The reason we are in the predicament we are in now is because doctors were educated to do what we have done, and in hindsight, it was horribly wrong to do that. Unlearning is a very difficult thing to do. Learning is relatively easy, but modifying your practice and unlearning something you have been taught about and has been passed down from your mentors is hard to do. I don't dismiss the value of education entirely; I'm just skeptical as to what, in the grand scheme of things, it will accomplish.
Mr. Abrahamowicz: It is the same issue that for education, there is maybe more of a need for independent education, education provided by people like researchers who do not have vested interests. For example, our society, the international society for pharmacology, has a strong Canadian component that is dedicated to the study of the effects of medication. It is excellent. Sometimes we get small grants from it or other companies, but they have absolutely no impact on what is presented, how talks are selected or who are invited as speakers.
Dr. Juurlink: The issue of academic detailing, with drug representatives going to doctors' offices and imparting information that is patently biased, is one of the main ways companies get a good toehold and get their drugs on the market. It has been suggested by some people — this is a good idea — that maybe there should be independent academic detailing funded by people who do not have a vested interest in anything other than optimal prescribing. That is an expensive undertaking, but I think that would actually be helpful.
I think it would be better, instead of a drug representative who comes into physicians' offices with a big, fancy lunch, a bunch of samples and some selected publications, if someone who had no vested interest other than seeing the patients get the best therapy would come in and say, ``Listen, this is what the evidence is; this is what you should be doing.''
With the Therapeutics Initiative in British Columbia, there was no better institute on the planet than the TI in British Columbia. You can go on their website now and find all kinds of outstanding, easily digested information that is very useful to front-line physicians. As I understand it, the TI is struggling, at least in part, because of politics in British Columbia. I don't profess to know all of the details.
Senator Eggleton: I think you've pretty much answered my second question, which had to do with the pharmaceutical companies and the fact that sales representatives are there at many of the conferences reaching out, and education comes from those companies, although Rx&D says they've considerably restrained that in recent years in terms of the rules, but independent education or training is the message I hear from you.
What about the regulator? What should the regulator be doing in the case of this prescription issue? I don't know whether you've seen Bill C-17. We also had a report called First Do No Harm: Responding to Canada's Prescription Drug Crisis, which came out a while ago. Are these things part of the answer? What more do we need to do in terms of the regulator?
Dr. Juurlink: I think the regulator plays an important role in how drugs are used. I don't know the extent to which they have a role in educating physicians, pharmacists or nurses. One way in which they could improve what they do is by increasing transparency. Perhaps you have already talked about that in the course of your meetings. When a new drug comes on the market, most of what we know about it — unless you have the time, inclination and initiative to go to the medical literature and dissect it yourself, which very few clinicians do — most of what we know comes from the company, which is, as I said before, biased. It would be nice if Health Canada issued something more than just the summary basis of its decisions, which it does for a selected number of drugs that come on the market.
The process that informs regulatory decisions is inscrutable. I don't know what goes on at Health Canada — I can't tell you — but I am sure they have information that would be valuable to me as a front-line prescriber that I don't have access to. I suspect there are multiple reasons for that, such as trade secrets and whatnot, but I think the one thing Health Canada could do to improve the quality of prescribing by physicians would be to shed a little light on what exactly goes on behind their doors.
Mr. Abrahamowicz: The example of Vioxx demonstrates that there is this conundrum that on the one hand, when the drug is approved, safety cannot be fully evaluated because both the company and, to some extent, society are eager to start using the drug and they cannot wait long enough to get enough patients using it to see what the specific adverse effects are.
If there were some mechanism by which it's mandatory that after let's say one or two years of the post-marketing use of the drug, there would be an independent evaluation. We have a large database in Canada that would allow us to do so, although there is an issue of access to this; some provinces make it difficult to access. Then we may speed it up; rather than discover it after five years, we may discover it after one or two years. Serious side effects of a drug that is widely used may easily be discovered after two years.
At the moment, it's more haphazard. At one point, a physician or researcher may decide to look at the specific association of this drug with these events, maybe because of the anecdotal reports or some indication, something to do with a basic science study, such as in the case of Vioxx, where we suggested that there is a specific mechanism that is produced. There is a lot of potential for improving things through research.
I wanted to mention that we are part of a very useful initiative, which is the Drug Safety and Effectiveness Network of the Canadian Institutes of Health Research. It focuses much more at the moment on safety. We are getting inquiries from provincial or federal regulators, and we try to respond by evaluating specific associations. This would surely improve things in Canada, and it has already started. This is a very useful process.
The Drug Safety and Effectiveness Network also puts a lot of emphasis on the end users, so we hope to be able to educate physicians, although this is not my domain.
The Chair: We really appreciate the point you are making, Dr. Juurlink, when you refer to the issue of transparency. We have made significant recommendations and compared international practices in that area, so we agree entirely with your point and we have made those points. We recognize the importance of what you have said.
Senator Seidman: Dr. Juurlink, I'd like to ask something about your institute and the kind of research you do, which I think might help us a bit.
You have an extensive data repository composed of health service records for Ontario. I'm sure it's a very useful tool for population health research because it's not patient-specific, if I understand correctly, but it's linkable.
Dr. Juurlink: We have health data on individual patients. It is all anonymized — individuals, their prescription drug coverage, their doctors' visits, their hospitalizations, their vital statistics — and we can link all those, as Dr. Abrahamowicz does, too, in a thoughtful way to try to inform about the post-market safety of drugs once they are on formularies.
Senator Seidman: How do you prioritize and select your research topics?
Dr. Juurlink: That's a tough question.
First, I would like to put a plug in in for this kind of research, not just because I do it but I think it is important because when the drugs come to market, we have a sometimes woefully incomplete picture of how safe and effective they really are. And it is only after they have been used, like Vioxx was, and hundreds of thousands of millions of people, that you can realize the signals and can do something about it.
At any given time, I might have 15 or 20 different studies percolating in my head, and I prioritize the ones that are most feasible and bear on the largest number of people. This is why of late we've been studying opioids as much as we have.
Drugs for diabetes, drugs for heart disease and drugs that affect large numbers of people, we don't prioritize them only for that reason, but it's partly because we need data. If a drug is used by fifteen people, it's hard to study it, but if it's used by hundreds of thousands it's much easier to do.
I would make the point that probably the best place on earth to do this research was British Columbia until some years ago. Perhaps you know that they just can't do this sort of research anymore. The investigator-initiated work that I have the privilege of doing and that I think sometimes influences practice and changes what the doctor does in his or her office in North Bay or Shubenacadie or in Toronto — I like to think that sometimes it changes what the patient leaves the office with. I would love to have the sort of data they have in B.C., and even my colleagues in B.C. who do this sort of work are hamstrung by whatever is going on there now, and they can't answer the kind of questions Canadians need them to answer.
Senator Seidman: You started to get into the second part of my question, which has to do with the translation of the research into policy because, of course, that's the point. I know a lot of researchers say, ``We do the research, we're the scientists, and then the policy makers have to look at that and translate it into policy. That's not our job.'' I fully understand that, but, of course, there is little point in doing all this research unless it has some practical import and helps us better understand and deal with how we practice in the real world and how physicians practice in the real world.
I would like to move to your annual report in 2012-13.
Dr. Juurlink: My institution's report?
Senator Seidman: Yes. It included a study on OxyContin prescribing near the Ontario border with tamper-resistant drug formulation that was introduced in the U.S. I bring up that report because it had impact and a legislative change in Ontario. That's an interesting and important impact because it had a change to the Ontario Drug Benefit Program in preventing dispensing of generic OxyContin.
Could you explain how that was actually accomplished, how that actually happened? It's a good translation of research into policy and practice?
Dr. Juurlink: How the study happened or how the translation happened?
Senator Seidman: How the translation happened.
Dr. Juurlink: To summarize the study, a few years ago the manufacturer of OxyContin changed the formulation so you couldn't just crush it and snort it. It was now tamper resistant. It was still abusable; you just couldn't crush it and snort it.
That same phenomenon happened two years earlier in the United States. That was 2010, so in the period after that happened in the United States, we thought maybe people are coming to Canada and getting the old crushable abusable version of OxyContin, and that's exactly what was happening. Interestingly, I think we looked at six different border cities, and it was specifically in the Detroit-Windsor tunnel corridor where there was an enormous surge in the prescribing of OxyContin. It lasted for about a year, to the tune, if I recall correctly, of about a quarter of a million excess tablets, which is worth millions upon millions of dollars on the black market, which is where I assume most of those drugs went.
I think the reason that that research had an impact was it was done by our group, and I'm one of the principal investigators of this group, called the Ontario Drug Policy Research Network, and we are sponsored by the Ontario government. It's a collaboration between policy makers and researchers where we do questions; sometimes it is the questions they ask us, and sometimes they are questions we initiate, and that was a question we came up with ourselves. We thought they might be interested, and they said, ``Please run with that.'' So we ran with it and we gave it to them. I don't think on its own it was necessarily the reason why they chose not to put generic OxyContin on the Ontario formulary, but I think it was one of the bits of information. It was due to a close working relationship that my colleagues and I have with the minister and ADM in Ontario.
Senator Seidman: Communication was one of the things we discussed in this committee with other witnesses and communication amongst various stakeholders, health practitioners, pharmacists, physicians and even the industry and then with the patients, of course.
Based on your experience, and this example, which is quite a successful example, in my opinion, what recommendations might you make to us in forming some conclusions ourselves in this study of unintended consequences, because that's obviously what we're trying to do? Is there some way we can facilitate that translation from research to policy?
Dr. Juurlink: Well, that's also an excellent question, and I'm not sure I have a ready answer for it.
I think the crux of your point is that research that is done and isn't listened to might as well be thrown in the garbage, which I think is probably true and fair. So I think it is incumbent upon the researchers to try to get their message out, either through the media or the publications, but, frankly, people don't often read the publications. They read the Toronto Star, The Globe and Mail, and the National Post, et cetera, and we can use media, Twitter and whatnot to get our message out, and many of us are increasingly doing that.
How can the government facilitate translation? I'm not an expert in knowledge translation, so I don't know that I have a good answer for your question. The single best thing you can do is recognize the value of this sort of work, if I may, and be supportive of it. I have got to say it's awfully cheap to do this sort of work. We can do an important study for $20,000, but we need data to do that. Wouldn't it be nice if in Ontario I wasn't relegated to studying people over 65 or people on social assistance? I would love to study people who are 40 or 45 who have drug coverage, as they can do in B.C.
But in the absence of data and the absence of funding, not much will get done. I think it's a rather opaque answer to your question, but it's the best I've got on the spot.
Senator Seidman: No, but it's a good one. I appreciate that. Thank you.
Senator Eaton: I would like to talk about drug holidays, drug cocktails. Dr. Juurlink, you were giving us an example of the man with the backache, the mechanic who comes to you. Physicians, as a rule, are not given a great deal of education when they go through medical school on nutrition, exercise and sleep.
Dr. Juurlink: No.
Senator Eaton: None of those things have any impact. But has there been any thought, when you're dealing with chronic illness or opiates, to looking at cocktails, and some are not opiates? You can give a person a drug holiday or prescribe it: ``I'm giving you this drug now for two weeks, and after that you will have to stop for a while and you'll have to do this instead''? Is there ever any chat as to sort of alerting the patient? Drugs lose their effectiveness, do they not?
Dr. Juurlink: That's true. The issue of a drug holiday means interrupting drug therapy after a period of time. Sometimes it's appropriate and sometimes it's not. In the context of using opioids for chronic pain, it's actually somewhat dangerous because after a couple of weeks of therapy, the body has become accustomed to the drug, and if you take it away suddenly, the patient becomes very sick. They don't die from that, usually, but they become very sick.
Senator Eaton: Can you substitute after only a couple of weeks?
Dr. Juurlink: It depends on the patient. I have seen it happen in three days; it is really quite dramatic. Their body becomes tolerant to the drugs and becomes dependent. They don't realize it as dependence. They just realize it as, ``If I don't take my drug, I don't feel well.''
Senator Eaton: There is nothing you can't substitute with another therapy for a while?
Dr. Juurlink: Well, you could alternate from one opioid to another. It involves the same receptor. It is a little bit dangerous because the equivalence of dose varies from drug to drug and person to person. You are getting at the use of less toxic drugs to minimize the use of more toxic drugs; am I right?
Senator Eaton: Yes, or alternating them.
Dr. Juurlink: That is sometimes an option. It is a very patient- and disease-specific option. Doctors don't stop drugs as often as we should. We are very good at adding drugs to people's medicines, and I hope I can take your question on a tangent, but sometimes what we are doing is adding drugs to combat the side effects of another drug the patient is already on. Unless you step back for a minute and think about why you are adding the drug in the first place, you might miss that.
Drug holidays are an important thing. They apply to a small group of drugs. With opioids some people feel you can do rotations of opioids, but it's more about trying to keep the pain under control and less about minimizing the toxicity.
Senator Eaton: Do you ever talk to your patients about exercise nutrition, physiotherapy?
Dr. Juurlink: My patients are exclusively in the hospital. I don't have an outpatient practice or a clinic. I am not a family doctor. Patients are sick and coming to me with strokes and heart attacks and those sorts of things. Sometimes I do, I have to say, but more often than not I don't.
Diabetes is an excellent example where the patients need education about their diet from me, or even better from a dietitian, someone who knows more about diets than I do. We do try and identify the nondrug elements of therapy that are important to a patient's care. You asked about my personal practice. Most of the patients I look after have other more pressing problems.
Senator Eaton: Do you have prescribing guidelines to help reduce the incidence of drug toxicities in elderly patients? You were talking about elderly patients who are given another drug and end up falling. They are not told about the side effects, and they end up with a broken hip, and we all know what happens after that.
Dr. Juurlink: Welcome to my practice, yes. That's exactly what I see. To use that one example, most physicians who treat older patients know there are some drugs that you should not give to older people. They increase the risk of falls in older people. A great example is sleeping pills and sedatives. Drugs like Ativan and Xanax are used by hundreds of thousands if not millions of Canadians. We use them all the time. It is generally the case they are best avoided. We prescribe them because we want the patient to feel better. As I said in my opening remarks, we want the patient to feel as though they have had a good sleep, or if they are troubled by anxieties or stresses of life, it has become commonplace we give people these drugs. You have to step back and ask yourself, as a physician writing a prescription, what am I trying to do?
Senator Eaton: Do you say to the person, ``Listen, if I give you Ativan, you are going to fall down if you get up at night?''
Dr. Juurlink: I sure do. I see them after their fall and after they have broken their hip. When patients in the hospital ask me for a sleeping pill, I say no. I tell them that what happens first is that you will go to sleep a bit earlier. The quality of your sleep will be much poorer than you think. There is a good chance when the bells go off at 2 in the morning, or the patient in the next bed yells about something, you might get up and go to the bathroom and trip and fall, and I make it clear that this is not some innocuous pill that will help them sleep and it's that simple. I don't think that happens as often as it should.
Senator Eaton: Do you have anything to say about drug holidays? You used the expression.
Mr. Abrahamowicz: I don't see patients. I see databases and I see statistics and so on, but I do collaborate with people. This was mentioned to me a few times by my colleagues. When I decided to do some serious work in this field, I had to learn basics. We were concerned about detecting discriminative effects. Immediate effects are easier to detect. If you have a cumulative effect, it will not be detected early on because the drug has been on the market for only six months; so nobody could observe what will happen after two years of the therapy. We thought how to optimize this data, which is my obsession in my research between risk and benefits. Even with Vioxx, I spoke with a top rheumatology expert at one conference. I do a lot of research on rheumatology. Some of them admitted to us and to their colleagues that when Vioxx was withdrawn from the market, they actually stocked this for their patients because they felt some of the patients have no cardiovascular risk. These are young women. They still made profits from the effectiveness of this drug in controlling their primary disease.
We are trying to think whether the drug has an important cumulative effect. If from time to time you stop or gradually reduce the dose to avoid this withdrawal effect, which is also important, could we still keep most of the therapeutic benefits while reducing gradually the risk? This requires rather complex analysis, and we are developing methods and only getting a few results. I personally believe there is some potential here, but as Dr. Juurlink said, it may only apply to some drugs. The therapeutic effects last for some time. You don't have to be on the drug every day.
Senator Eaton: You can go up and down, in other words?
Mr. Abrahamowicz: Yes. We are only now on the cutting edge of developing methodologies. I wanted to back up and appeal to make better access to the data. It is a bit of a tragedy for us that some provinces — I will not mention which — make it extremely difficult for researchers to get access to totally analyzed data, there is no risk of breaching their confidentiality. Their computers are not even connected to the Internet, so there is total security of the data. In spite of this, we sometimes have to wait two and a half years to get access to the data. In the meantime, if you think about something like Vioxx, hundreds of thousands of patients may have serious side effects which could be prevented if we could do the research. It is not directed to your question, but it is an important message that the research community wants to bring to the table.
Senator Eaton: I have one last question. We've heard about antibiotic shortages coming up. We have heard about antibiotics being not as effective in some diseases anymore as others. Should antibiotics be prescribed with the care that opiates are — in other words, very carefully and gingerly — or not?
Dr. Juurlink: I think it's true that antibiotics are overprescribed, and the usual sequence of events is the patients in the office, they have what they think is something that warrants an antibiotic, it's a viral infection, and it absolutely does not warrant an antibiotic; and a good doctor will spend the time to tell the patient you don't need this. Sometimes, and for a variety of reasons, the doctors accede to the patient's request, and the patient goes off, and the rationale in the doctor's mind is, well, it is just one prescription; it's probably not going to hurt the patient. Someone else is paying for it, and I've got another patient to see in the waiting room. I don't know how often that happens. I think it happens less now than it did 15 or 20 years ago.
I do think we need to be more careful with the use of our antibiotics. In addition to the patient-specific risk, the respect of some life-threatening skin reaction or some fatal cardiac arrhythmia, and people do die of those things from the antibiotics, well-intentioned prescriptions.
When you do it to one patient, it is one thing. When you do it on millions of patients over the years, it breeds resistant organisms, and that is an entirely different threat. We should be more careful. I don't know that they — part of my initial reaction is should they be treated as gingerly as opioids? My initial reaction was no because of how inherently dangerous and addictive opioids are. But I do think care is warranted through every prescription, and the doctor, before he or she writes the prescription, should ask: Listen, with this drug and with the exposure of the patient to the risks of this drug, do I really think I'm going to make them feel better or live longer? Because if I don't, I shouldn't write the prescription.
Senator Seth: This is quite an interesting topic.
Dr. Juurlink, in your presentation you mentioned that there are five types of adverse events of the drugs. It could be immediate, cumulative, long-delayed, idiosyncratic or drug-drug interactions. Do you know which of these adverse events are most common in Canadian drugs?
Dr. Juurlink: I think you made that point. Did you want to field that one?
Mr. Abrahamowicz: Maybe you had better answer. I would say all of them, and some are easier to detect than others, but you would be better to answer.
Dr. Juurlink: It's certainly fair to say there are different types of adverse drug events. There are the direct toxicities, the allergic reactions, the interactions and so on as you've commented. No one on the planet can tell you which of those is most common. Part of the reason is because it's hard to recognize. If a patient on a blood thinner begins to bleed, you don't need to be a good doctor; you can be a high school student and realize it probably has something to do with the blood thinner the patient is taking. But if an 85-year-old patient comes to the hospital with confusion, which is a nightly occurrence at Sunnybrook, where I work, it's sometimes hard to know that that's a drug-related problem unless you actually go digging. It could be an infection; it could be a stroke; it could be any number of different things.
The reason why it's difficult for me or anybody to tell you which of those is most common is because they are often so difficult to recognize. When a child is born and they have misshapen limbs and the mother took a drug during pregnancy — I'm using the thalidomide example from the early 1960s — it's easy to make that link. When an 85-year- old or 80-year-old or 60-year-old has a sudden death from a heart attack, we don't think to blame the Vioxx the patient was taking for the last three years and had tolerated just fine until then. We say the person is at risk of heart disease by living in society. We don't think to make the link. It's sudden death in an 18-year-old that makes the eyebrows go up. That is just one example. You're touching on a really important topic. There's a great deal we don't know and, frankly, a great deal that isn't knowable about the extent to which drugs cause harm.
Senator Seth: I understand you have conducted research on mortality related to prescription opiates consumed even though they are given for chronic pain patients. You mentioned 10,000 people died with OxyContin and the painkiller medication is epidemic. Are opiates responsible for increased mortality? Given they are prescribed, should there be more guidelines for prescribing, a greater time-period interval, more patients visited and follow-up, et cetera? What would you suggest for this?
Dr. Juurlink: I think there are multiple aspects to your question. The first is: Is it the drugs themselves that are killing people? The answer is an unequivocal yes, it's the drugs, but they're doing it in different ways. It sometimes happens that a patient is put on an opioid by a well-meaning physician and they take their drug diligently and they don't take extra doses, and they don't go to multiple doctors and multiple pharmacies, but they have the misfortune of taking it with something else, like a sleeping pill or alcohol or both. That's a drug interaction of sorts and the patient doesn't wake up, as I alluded to earlier.
Sometimes these drugs kill people because the person transitions from a legitimate pain patient into a patient whose life is consumed by these drugs and they become addicted to them. We were taught it was less than 1 per cent of patients, which is patently untrue; it's a false claim. Those patients end up crushing and injecting and snorting and they die of drug overdoses and are labelled as drug addicts. But they didn't wake up one morning and say, ``I think I'd like to be a drug addict.'' They started with a prescription from their dentist or their doctor, and that's how they got where they are.
Sometimes they die because of misadventure. Teenagers sometimes have what they call ``Skittles parties'' where they go to their parents' medicine cabinets and they look at what's interesting in the cabinet that they can bring to a party, and they go into a bowl and everyone just takes something from the bowl. I know of at least one instance where a person died of doing exactly that. The person had never been on an opioid, took 80 milligrams grams of OxyContin and didn't wake up.
There are many ways in which these drugs can kill people, and it behooves us to be more careful in how we prescribe them. I'm not sure I answered your question in full, though.
Senator Seth: How can this be avoided? That's my question.
Dr. Juurlink: It's awfully hard to avoid because it has become an entrenched practice, the prescribing of these drugs.
I was a pharmacist in Nova Scotia from 1990 to 1995. It was the case when I was a pharmacy student in the late 1980s and a pharmacist in the early 1990s during medical school that when a patient came to the pharmacy and got a prescription for morphine, they had cancer. They almost always had cancer. Ten years after that, it wasn't cancer; it was back pain and whatnot.
I've talked to you about how this all unfolded. This is why this is an important epidemic. If 13 people were sickened tomorrow by E. coli in Brandon, Manitoba, it would be all over the news and there would be a big public outcry. Nobody would be arguing for more E. coli or that we should continue to sell tainted food or beef. In this epidemic, there are people who hold a very different opinion than I hold and they say that we should be using these drugs, but we should simply be teaching doctors to use these drugs more carefully. I agree that teaching doctors to use these drugs more carefully is a good idea, but the drugs are inherently toxic. It's not hyperbole to say that they are pharmaceutical grade heroin. I said it deliberately in my comments.
The idea that we should be giving huge swaths of the population heroin in pill form and that somehow that will make them feel better or live longer is a very tenuous one. It does make some people feel better. We just don't know who they are at the outset and in the process of finding those people, we get others addicted to the drug. I'm not sure if I went off topic a bit there.
Senator Enverga: That was a great presentation. My question is for you, Mr. Abrahamowicz. It's about your statistics. I've seen all those drugs that were taken off the shelves. After 2004, after 10 years, has there been any improvement? Have we made more recalls than ever?
Mr. Abrahamowicz: I'm developing methods that help make research in this field better. By developing these methods, I'm also applying these methods to specific studies, but I don't have the perspective Dr. Juurlink has. I would say from my personal observation we are doing a bit better, because actually the infamous Vioxx study did help raise concerns about this, which translated to more research, more funding for research in this field, and more acute awareness of this. Even the pharmaceutical industries are more serious, partly because they are forced to be and partly because they don't want to have a black spot on their name.
There is some better detection. We are still far away from the idea. The idea is impossible. In mathematics, we speak about something you want to reach but never reach fully. It's impossible because there are so many possible combinations, so many drugs and so many possible side effects that I'm sure there are some we will never know about and there will be some people killed by some adverse effect of some medication nobody has studied. The mechanism I mentioned about safety of specific drugs being transmitted via the Drug Safety and Effectiveness Network to researchers is an excellent mechanism because it speeds up the process. There's so-called rapid funding, so by the time the queries are received by the network, managed by the Canada Institutes of Health Research, within two or three weeks they send a query with some basic information to researchers. We respond within a month, or sometimes earlier. Several groups respond and say how we are able and how we can study this project. Then there's a quick, mini-grant, which is a much less painful and complex process as the standard research initiated grant. Within maybe four months after the query was initially sent, we start doing active research on the drug and hopefully a year later we have the response but it depends.
Again, data accessibility is a major issue. Tomorrow we have a four-hour meeting of my research team involving people on teleconferences from 14 universities across Canada. The main focus will be how we can get better data to answer different types of queries. People have clinical data, administrative data and so on. We want to go beyond that. Could we use U.K. date, could we use German data? Basically, the drugs probably act similarly on Germans as on Canadians. This is where there's progress. However, I don't have solid statistics on how many successful studies there have been. I would say there are more, there's better.
I remember after President Obama decided to put US$1 billion into research on post-marketing effectiveness and the safety of medication, I saw one review paper that said there were 600 new studies using U.S. population data initiated in a single year. There's a lot of research going on. Canada is more or less proportional to the U.S., and probably using better methods and better research standards so there is some improvement.
Senator Enverga: Are we making any progress at all?
Dr. Juurlink: I think we are. One example that comes to mind bears on the antibiotic question that was asked earlier is a paper that we did, which I'm actually quite proud of. In 2006 or so, there was an antibiotic that many hundreds of thousands of people in Canada took called Tequin, which was commonly used for chest and sinus infections.
A graduate student came to us one day and said, ``I think this drug is causing dangerously high blood sugars and in some patients dangerously low blood sugars.'' We reviewed the literature and it was clear she was on to something. It was clear that it was a potential big story, not from a news perspective but from a clinical, front-line perspective, because people were oblivious to it. We went from conceiving of this study to implementing it in about a week and a half.
I remember being up at 2 a.m. working on successive iterations of this manuscript and it was out the door about a week later. It was extraordinarily fast, and it was only possible because the data were there. We didn't have to apply to a granting agency and waste eight months waiting to be funded to do it. The manuscript was out and I sent it to Health Canada and to the FDA before we sent it to the journal. The journal published it quickly and two months after the publication the drug was taken off the worldwide market by the manufacturer.
I want to give you an anecdote. When this paper was under review I happened to run into a rep from that company and said: ``What do you make of this link between the blood sugar problems that I'm reading about?'', knowing full well that this drug was 16 times more likely to cause high blood sugar than comparable drugs. He said, ``It's reporting bias, and patients, when they get sick, their blood sugar goes up. Just ignore that stuff,'' and I said, ``Okay.'' That drug is no longer on the market, thankfully. That's a good example of the information that doctors get from drug reps.
Having access to the data is critical, as well as being able to do these studies quickly. That's a good example and I think really important.
Senator Enverga: The reason I was checking these records is because of hepatotoxicity, which affects the liver and sometimes the kidney. Everyone should know that our liver and kidneys are the ones that clean up all this bad stuff in our body. Should that be the focus of these drugs?
Dr. Juurlink: That's true. One of the main reasons for drugs coming off the market is because they cause inflammation or injury to the liver, and sometimes they cause irregular heartbeats and arrhythmias, which are very dangerous. You don't know this happens until the drugs have been in widespread use.
Let's say the drug has been studied for this antibiotic. I don't know how many people were in the preclinical studies but it might be 500 or 1,000 people. They're given a drug and the drug seems to work. When the side effect is rare, it's very hard to uncover in a study of 500 people or 1,000 people. It's not until the drug has enjoyed widespread use by thousands and hundreds of thousands of people that you can actually pick these signals up. We know to look for them and when the studies come out about a new drug we know to look very carefully and dissect, to the extent we can, where the possible signals of harm might be. In the Vioxx study, it was clear in the abstract of that study that there was an unexpected increased risk in cardiovascular events with this drug and the researchers in the company tried to explain it away with an a implausible explanation. People knew that this was something that had to be zeroed in on. I'm conscious of having gone off topic a little bit.
Senator Lang: I'm sitting in for Senator Stewart Olsen. I find your description of how drugs are prescribed, drug upon drug quite interesting, and then your description of Skittle parties, which I've never heard of before, quite alarming.
I want to go back to the example that's been raised since we started the proceedings, and that's the question of Vioxx and the injurious results that came of that drug over such a long period of time. Both of you are very familiar with how it came into being and how it was prescribed and then how it was found.
Looking back — and perhaps you could provide it to the committee here — if you were in charge today and knew what you do today 10 years ago, what would you say the government should do legislatively to help prevent that?
What should the company be further required to do in respect to ensuring that the drug they're selling is doing what they say it's doing and it's not having those effects? Is there a legislative requirement or policy that's further needed? Perhaps you could comment on that.
Mr. Abrahamowicz: I'm not very strong on what legislation should do, but from the research perspective the initial study in 2000, in the New England Journal of Medicine, led to the approval of Vioxx. As I pointed out, there was a significant increase, although based on small numbers; this was a study of average follow-up duration of only nine months. On average, people were on Vioxx for only nine months.
It was already a statistical signal and of course at this moment the company should be obliged — or somebody who is independent would be even better — to follow up immediately with a larger study of post-marketing or maybe even the approval should be delayed until the safety issue is evaluated based on the larger number of subjects. We both believe that part of the reason was that the trial was too small and lasted too short a time to have solid evidence.
Now I have a bit of a problem reconstructing the timing. There is a paper that I cite by Dr. FitzGerald. The paper was published in 2003, but he says that more or less parallel to Vioxx approval they already identified the mechanisms. I'm not sure whether there were some other papers or whether they were complement representations which indicated a specific mechanism, very well and logically defined, that could actually explain why there's an increase.
This is an additional and very strong kind of corroborating evidence to support findings from epidemiological studies. Apparently Vioxx, and to some extent FDA, as far as I understand from these secondary sources, were ignoring these messages, discarding them, saying the most rigorous design is a randomized trial. It's an experimental study and almost the only logical difference for increase of risk or decrease of risk is the drug itself because everything else is equal. A large number of patients received the drug and another compatible group of patients does not receive and we compare them.
Other reports are accumulated over the same time period based on population-based study, which we call observational study. This is what we both and many other Canadians are doing. We look at what happens in the population when there's more variation in the use of the drug. As I mentioned, there will be different durations, different doses and different types of patients. This creates some challenges for the researchers. One challenge is, is it the drug itself or the disease for which the drug is given that causes the observed adverse effect? We continue to make progress, and we do have good methods. These studies indicated increased cardiovascular effects were done according to state of the art, and it was difficult to explain why they were so easily dismissed.
As Dr. Topol in his very forceful post-withdrawal paper argues that in the same period when Merck was continuing to dismiss these reports and refused to do a trial specifically looking at the cardiovascular toxicity of Vioxx, they were investing up to $100 million into marketing the same drug. The trial would cost maybe 10, 20 per cent of this amount. For sure, look for it on the legislative side and probably a bit of intentional behaviour.
Dr. Juurlink: You asked me what I would do differently. If I could go back 15 years, I would do what one of the provisions of Bill C-17 allows. As I understand, one of the provisions empowers the minister to compel the manufacturer to conduct additional studies or to impose restrictions on a licence. The pivotal trial of Vioxx — and I don't have it at hand, but as I recall from the abstract, it showed a four or fivefold increase in the risk of cardiovascular events compared with others. It may have been a chance observation, but it should have been a clear signal. It's easy in hindsight. The thing to do would be, ``Listen, maybe this drug shouldn't be rushed into the Canadian marketplace. Maybe before you're allowed to sell this drug in Canada you should provide more evidence about the cardiovascular safety of this drug, because your own study raises that as a potential concern.''
Senator Lang: One further question, Dr. Juurlink: You mentioned in your opening comments the summary of drug evaluations by the federal government, by Health Canada, I believe. Could you elaborate further on the fact they're not made available to you and perhaps tell us why? More importantly, in comparison to the United States or other countries like Britain or Australia, is that made available?
The Chair: We'll come back to that, because we've dealt with the transparency issue and Health Canada on this. I appreciate, senator, you may not be aware of that. I'll come back to that issue with regard to a summary. We have already dealt with that in significant recommendations, if you don't mind.
Senator Lang: That's fine.
Senator Cordy: I'm wondering how much research is being done in Canada on unintended consequences.
Dr. Juurlink: I think a lot. It has been promoted substantially by the implementation of the Drug Safety and Effectiveness Network. Although many of the questions they answer are questions that are posed by Health Canada, and are sometimes the questions that are not of most interest — not to dismiss Health Canada's concerns. I've seen their questions sometimes, and sometimes they are questions the answers to which are already known or are clinically not that important or clearly not answerable. The most important questions are ones raised by people on the front lines, like my graduate student who raised the issue of blood sugar problems with that drug. It wasn't even on Health Canada's radar. They were happy to receive the paper when they got it, but they wouldn't have asked the question back then had they been given the opportunity. I think it happens a lot. I think it happens less than it might, in part because our most valuable data repository in B.C. is largely inaccessible. It has hampered my work substantially, including on new anticoagulants that are used excessively and with some risk to the people who take them. It happens a lot. It could happen more. We clearly have a bias in favour of it, but perhaps that's an unnecessarily long answer to your question.
The Chair: In addition, senator, we have already made recommendations about more opportunity for DSEN to listen to people on the ground to initiate studies, not just to take the direction in the situation that exists at the moment. We're all on the same page here, and it's an important question.
Senator Cordy: Do we know how many Canadians are addicted to a drug that was actually prescribed to them by their doctor? Do we have access to any of that information?
Dr. Juurlink: I don't think we do. Somebody somewhere knows exactly how many prescriptions are issued for these drugs. It's a controlled substance. The records are kept. I have no idea who keeps them or who has access to them. In theory, it should be very easy. It should be a summer student job to find out exactly how many prescriptions are dispensed. Knowing how many patients are addicted is a very different matter, because different people mean different things when they say the word ``addiction.'' I can tell you that the manufacturer of OxyContin, specifically Purdue Pharma, made false claims about the risks of its drug, patently false, and Health Canada knows that. Unlike in the U.S. where a lawsuit was brought and Purdue Pharma pleaded guilty to the felony of misbranding its drug and paid a $600 million-plus fine, nothing happened in Canada. The false statements that led to the large swathes of the population being addicted to these drugs were made and recognized; and nothing has happened, as far as I know, with regard to them.
Senator Cordy: Maybe OxyContin has received a lot of coverage in Nova Scotia particularly, but I'm not sure that patients understand how prevalent and easy it would be to become addicted, because Canadians have a stereotype about what a drug addict is, and it's not being addicted to a drug that you're being prescribed by your pharmacist.
Dr. Juurlink: That's exactly right. People have the perception that because it's written by a doctor and comes from a pharmacy and it's received the blessing of a committee somewhere that pays for it, that it must be safe and effective. I think doctors share that perception sometimes.
One of the seminal papers that was cited to say the risk of addiction with these drugs is less than 1 per cent consists of five sentences in a letter to the editor in the New England Journal of Medicine. It's not a study. That didn't stop it from being quoted hundreds and hundreds of times to support that assertion which, at the time, frankly, doctors should have known better. We should have known better. You can't give people heroin in a pill form and expect them not to become addicted. It's crazy to think that you can do that, and yet we swallowed that message hook, line and sinker, in part because we wanted to hear it. We were happy to hear that we no longer needed to rely on these old drugs that caused stomach ulcers and bleeding and kidney problems and heart attacks and that we had these other drugs and could do it safely.
Senator Cordy: New and improved, right? If a person dies from taking a drug that has been prescribed to them, and if they had used the drug according to the instructions given to them by the doctor and pharmaceutical companies, is this recorded anywhere as an adverse effect?
Dr. Juurlink: I think sometimes it is, but not always. An example is somebody who is on a blood thinner to prevent stroke from an irregular heartbeat. It's common, evidence-based practice. It happens that people, when on blood thinners, sometimes bleed, and sometimes those bleeds are fatal. It's incumbent upon the physician to report it to Health Canada. I don't know how often that's done, probably not as much as it could be, but when it's done, a record is kept at Health Canada. It's a different matter when someone dies of something not recognized. Let's say they went on Vioxx and had a heart attack and dropped dead, or they were on an antibiotic that caused an irregular heartbeat and they dropped dead. Unless it's recognized as an adverse drug event, it's not going to be reported. That's one of the big problems that we face. When someone comes to the hospital on an antibiotic and their blood sugar is through the roof, we say, ``Oh, you have diabetes and you're sick; that's why.'' We don't think sometimes it's related to the drug, and that's a difficult thing to do, especially when the adverse effect is a common clinical occurrence. We don't think to blame the drug sometimes.
Senator Cordy: Going back to Senator Seidman's comments about the challenges of bringing forward research, I think you stated quite well earlier in your presentation that the information given to doctors about a drug is, I would say, 90 to 100 per cent of the time given to them by the drug company at one of the better restaurants in the city, or a weekend golf tournament. I have friends who are doctors. That happens all the time. That's a much easier way to get your information than to sit down and read for the weekend about it. In the United States, they have a sunshine law where doctors will have to say they had a golf weekend with whatever drug company you wanted to name. How do we overcome that in Canada?
Dr. Juurlink: There are a couple of issues there. First, there is the issue of the sunshine law. There is a commentary in The Globe and Mail today on that. There is a move to bring some light on the relations between physicians and the pharmaceutical industry. I think it would be helpful to know who is being paid how much, by whom, and what for.
The idea that we should get our medical information from a drug rep is nonsense. They have a job to do and their job is often at odds with our obligations to our patients. They are not our friends. Although sometimes they become friends, that shouldn't happen. It's incumbent upon doctors to get their information but, as you rightly point out, the easiest thing is often to do what we do. Watching what you eat and exercising is harder than taking a pill. Getting information from someone who provides it to you willingly is a lot easier than going through the literature and doing your own research. Minimizing the extent to which doctors interact with pharma is a good idea. It is already much better than it was 15 years ago, but it all boils down to doctors being empowered and enforced, I think, to maintain their education — as they are to a certain extent — with as unbiased information as is possible.
Senator Nancy Ruth: In this whole business of knowledge synthesis and dispersions, you keep referring to provinces that you can't get data from or to B.C. which has problems now. Can you tell us about that? To put it in a positive way: What is it you want the provinces to do that would help your research?
Mr. Abrahamowicz: I don't usually negotiate directly, but I am sometimes involved with the team which has to negotiate. For instance, my province of Quebec has an excellent administrative database. Every prescription given to a patient who is above 65 or on social assistance is entered in the same administrative database called RAMQ. This is fantastic in terms of the numbers. It avoids all the problems we discussed about not enough patients and not enough follow-up. We have the full history of prescription taking. We don't know whether or not the patient took the drug, but we know what prescriptions we were given.
If we decide today that there is an important concern about a specific unintended effect of a specific drug and we start a new negotiation with RAMQ, the provincial bodies, especially le Commission d'accès à l'information that is concerned about patient confidentiality, and so on, has extremely complicated procedures and so many previous demands that are not processed that it would take up to two years before they would allow us to use the data. In the meantime, patients are taking the drugs and are possibly having side effects and no one can prove it.
I pointed this out to the people who manage safety and effectiveness, but now we have the federal-provincial dichotomy that says you, as part of the Canadian government, want us to respond to these queries while other parts of the government at the provincial level block us from access to the necessary data and we have to go through complex negotiation. Could this be resolved somehow? We understand patient confidentiality but it is only one part of the equation. If we ask the patient if they are more concerned about possibly dying because of a side effect of a drug that may be optional to reduce their pain — and maybe they will suffer a bit; maybe they will have an alternative medication — are you more concerned about that or more concerned that by some extremely unlikely process they would hook up with someone who has bad intentions and try to find out from the database that this patient is living in a small village in northern Quebec and might have this disease because they were given this medication? What are the cost benefits for society?
I am sorry about my emotions concerning this.
Dr. Juurlink: The issue is really one of access to data. One of the reasons why my colleagues and I have had good fortune with our studies is we work in a place where the infrastructure exists and has been in place for 20 years now with anonymized data that is treated carefully, and we can ask and answer important questions for a large population base.
Ideally, we would have data available on everyone in real time. That should be the case. My kid can go online with a video game and play with somebody in Australia in real time but a drug store like the Rexall across from my hotel, for example, has no knowledge about what was dispensed at the Shopper's Drug Mart down the street. It is crazy when you think about it. That data is all collected. Would it not be nice if we had a much better, and earlier, sense of drug- related concerns because more data were available for study, as was the case in B.C. until things went wrong.
Senator Nancy Ruth: You are not telling us what things went wrong but we'll talk later.
Dr. Juurlink: I don't know what things went wrong. That's the problem. No one does.
The Chair: I will go to the second round before I ask my questions.
Senator Enverga: Aspartame was removed from the market and then OxyNEO came. How did they allow the generic to come in if they knew it was bad?
Dr. Juurlink: OxyContin was removed by the manufacturer and replaced with the new formulation that was tamper resistant, ostensibly as a patient safety manoeuvre. Coincidentally, it happened a few months before the patent expired on the original product. The generic companies — and there were several — recognized a potential market here. To be fair, if everyone took generic OxyContin and if we had assurances that it would not be crushed as it clearly and would be taken as the doctor intended, people would have less to gripe about.
I was vocal in my opposition to the idea that this should be approved and I think a strong case can be made from a public safety perspective that it should not have been approved. On the other hand, a case can be made that it should have been approved because it is less expense than the brand name drug.
I don't know why it was approved. I think Minister Ambrose alluded to some changes coming down the pipeline a few months ago. I don't know what those are and I don't think anything has happened there.
Senator Enverga: Do you think the policy should be changed? Should it be automatic?
Dr. Juurlink: There are pros and cons of putting it on the market. I personally think the cons outweigh the pros. I have regular contact with a recovering drug addict who lost about 13 years of his life because of his addiction, which started with an experiment with some Percocet tablets. One of his most dramatic relapses was after the generic OxyContin came back and he couldn't resist the drug. It's out there; it's being used, and that reaffirms my original position that this should not have been allowed to come to market.
The Chair: To follow up on this issue, we've had this very question in a previous session or so. As you are responding, Dr. Juurlink, outside of the walls one does not necessarily know what regulatory authority would be available to prevent the authorization of a generic form in terms of the way the generic drug companies are set up and the fact that a drug has been approved and you're dealing, supposedly, with an identical chemical component.
However — and I don't want to pursue this any further at this point — I think there are some who would argue that the authorities do exist with regard to that particular example for two reasons: first, it's a controlled substance; and, second, there is a clear known additional adverse effect of that controlled substance in its generic form versus the so- called tamper-proof replacement. In terms of protecting the health of Canadians, certain authorities do exist at this point. We will pursue that in our discussions as a committee in terms of what we think we can recommend in that area.
By the way, we have a copy of many of your papers, as you would imagine, and your chart in particular, which shows the implementation of PharmaNet and the drop in the use of opioids. It is dramatic.
Dr. Juurlink: It is. We have a similar publication under way right now in Ontario.
The Chair: It is absolutely dramatic. I will use that to come to this whole issue of data, and I will approach it from two points of view: First, the general issue that we've been dealing throughout all of our studies, which deals with the clinical trial process through to the post-approval monitoring, through to, as you have very effectively pointed out, in addition to all that we have learned, the absence of good data collection in all levels of drug interactions across the country — and I'm talking broadly — even though there are clear examples of how effectively it can be done, and the importance of that; then I want to come to the Vioxx issue as a specific example of the complexity of the issue of benefit and harm.
As a committee, we are not only very sympathetic to your arguments with regard to the need for data, the need for access to that data, and then the interpretation and transmission of the information from that data into policy and effective protection of Canadians with regard to the pharmaceutical system, but we have already been making substantial recommendations in that area. But your testimony here today in this particular section further enhances the significance and importance of that.
Dr. Abrahamowicz, you clearly identified — as did you, Dr. Juurlink — the limited number of persons in an initial clinical trial. We know they are chosen carefully; and we know, and we have reported on the fact, that the elderly, children, women — particularly pregnant women — are not even included in those trials; and therefore the subsequent use of the drugs for them represents an off-label use immediately.
We've made significant recommendations in terms of the information that should go to physicians before they can even make a prescription off label. But in that study we also recognized the value of off-label prescribing. We understand the complexity, but it all comes back to information, the information available to the physicians prescribing and the collection of information from society once the drugs are approved. Because, as you have further illustrated, once a drug is approved and it gets into the general public, that is the ultimate clinical trial. The collection of data from the use of the drug in the broad spectrum of the population, with all the subsets of the population, is critical to ultimately giving the ultimate protection to our society.
So we, like you, have not understood why it is in this day and age, when Google can tell you what suit you are going to buy next week, as you have pointed out, why you can't know what the pharmacist down the street has prescribed to the same patient in the same day. We don't understand why we are not further along this line in Canada with regard to the collection of data, because it is absolutely essential to the protection of Canadians. As you have given clear examples to us today, the use of such data, suitably protected, can bring great and immediate benefit to Canadians when it's in that use.
I'm not going to ask for more debate on this from you today, but I want to ask you the following, because you are experts in the use of data; you've had access to data sets that you have brought benefit to society on, or given illustrations of where more benefit could occur; and you've also seen what happens when data sets that are available suddenly don't become available to be used in the way that they have been used.
I will ask you both, on our behalf, after you leave here, to think about, if you were making specific recommendations to us, how you think we could make recommendations on, first of all, the importance of getting that data, the ability to collect, methods and examples of how that data can be collected to protect patients, and any other aspect of that that you can think of, after you leave here, and give us that in a truncated form.
This data, in all our studies, is the essence of the issue. We know we have spent billions, through the development of a so-called information highway in this area, and we have very little to show for it as a national system. Yet, examples clearly exist of how it can be done.
We also know that in the United States there are very large data sets that occur through the insurance plans that deal with prescriptions. Great sets of data can be collected and individual patient confidentiality can be protected.
If you would address that point clearly; we're capable of understanding it very well, so if you could give it to us in a focused form, it would be tremendously helpful to us.
I would like to come back to the example that Dr. Abrahamowicz used on the Vioxx situation. This, to me — and I'm speaking personally now — represents a tragedy on two levels. I'll digress for a moment to repeat to you: We have made what we consider to be very significant recommendations in our earlier reports with regard to the necessity of transparency in the data collected during clinical trials. We've compared international situations and we have made what we think are very significant recommendations. We agree with you that transparency is a key to all of that.
Now, coming to Vioxx and the two tragedies: First, the great tragedy of the number of people who died unnecessarily, apparently as a result of the way the drug was introduced and used; and on the other hand, somewhat lesser but still, in my view, a tragedy in terms of benefit to health, namely, the loss of the drug to the enormous number of people for whom these drugs are the ultimate in terms of dealing with certain types of pain with minimal side effects.
This brings us to where we are today and beginning to move forward on the issue of personal health and the reality of genetic markers. What we have in the situation of Vioxx is a subset of a subset of the population for whom the drug is lethal. We have the subset with cardiovascular indications, and within that set there is a very clear set that if they take this drug, they are going to die, under certain dosages and certain times and so on. As a result of that, we had the tragedy of many of them apparently dying and, on the other hand, those people for whom this pain medication can bring enormous benefit to them.
To look a little down the road in this, I want to ask you if you can see a way, with the lesson of Vioxx that is there, that we can look down the road and avoid the subset tragedy, and learn what that will be in advance of perhaps approval, and at the same time give under the label the very clear directions as to how that drug can be legitimately prescribed. Are we there yet? Are we going to get there within the foreseeable future?
Dr. Juurlink: That's quite a question. I think you made an excellent point; the release of a new drug really is a mass experiment on the population. The whole crux of your question is the optimizing of prescribing so drugs will be given to people who will benefit and not to people who will be harmed. There are a good number of people who bemoaned the disappearance of Vioxx because it made their lives much better, notwithstanding the risks they were being exposed to.
With the idea of personalized medicine, I think one day we'll be able to get a little cheek swab or a blood test that says this is a drug that will work for you or this is a drug that you must not take. In fact, I do this from time to time in my own practice; there are drugs that I know will cause grave harm to a patient, so I will order a genetic test before prescribing it. There are very few of them, but there are a couple of examples. So it is now commonplace — in fact, the standard of care — to do this particular test before you put a patient on a drug.
I think it might be the case that one day we have the ability to do that in broader practice, and I think drug companies know that and they are interested in studying drugs in that way. When drug companies run these trials, they are clearly interested in excluding people who might be at risk, and that can sometimes be determined genetically. They are interested in having people who are likely to respond to the drugs in the trial because that will make their trials cheaper, faster and shorter in duration. I think it's great, and I think that one day we'll be able to do that through harnessing the power of genetics.
It will only be useful to the population if we apply those same testing attributes when we unleash the drug. My worry with the personalized medicine issue or the issue of pre-market trials based on genetics is that we will show that drugs work great in some people and they are very safe when you focus in on their genetic makeup, yet when you market the drug, you do not apply that same testing protocol. That may be in 20 or 30 years; I don't know how far away it is. It is very drug-dependent.
Mr. Abrahamowicz: I share the same general concern, maybe two aspects. I also believe there is some promise in personalized medicine and pharmacogenomics. On the other hand, the subset is a very challenging issue. Unless there is a basic science study that explains which subset this drug or the opposite should have, it would be a particularly bad response with safety. It is a bit of what we call a fishing expedition. Even in a relatively large study, if you don't know which subset to focus on, to variably identify that this will work for elderly men but not for elderly women or this will work for people with this kind of co-morbidity but not another kind of co-morbidity, it is extremely difficult. It basically requires any study of so-called interaction or effect modification. The effect will depend on some characteristics of the subjects and requires at least four times larger number of subjects, even if they are men versus women with very well-distributed characteristics. If we are talking about a small subset, it becomes even trickier.
One way out with some of the drugs like Vioxx is international collaboration. With my McGill collaborators, we started an initiative and we are now running analyses in the U.K., U.S., Canada and France. For confidentiality reasons, the government will not allow us to put the data together. We use statistical methods to combine the results or look at the consistency of the results. This helps a bit, but it is a long process. That does not mean we should not start. We should keep trying, without being overly optimistic.
There is another issue I wanted to briefly bring up. In all fairness, the results from a proof trial actually showed increased cardiovascular risk even in people with no previous evidence of cardiovascular disease. On one hand, the subset can work on two different levels. First, sometimes the drug may have unintended consequences only for that subset. Second, it may increase risks for everybody, but if your baseline risk is very low, then twofold increase of risk means you are moving from 1 over 1,000 to 2 over 1,000 and then there is unfortunately a need to somehow balance the benefits.
My favourite example from one of the great philosophers is that we always make decisions about probabilities and seriousness from the different risks. For instance, it may seems like nobody will actually risk death for small gain, but people cross the street, therefore risking being killed by a car, because something costs 10 cents less on the other side.
We need to quantify this risk, but there will always be some judgment necessary. Is it warranted to put a patient at a slightly higher risk of these events in order to prevent major pain and increase the quality of life?
The Chair: Thank you both very much. We look forward to your advice as to how we can recommend an absolutely failsafe and effective data collection system that will bring great benefit to the health of Canadians.
I want to thank my colleagues for their questions. It has been, as usual, a very interesting session. Again, we thank you for taking the time out of your incredibly busy careers to appear before us.
With that, I declare the meeting adjourned.
(The committee adjourned.)