Search This Blog

Follow by Email

Thursday, May 26, 2016

Is there a pill to make you live longer? The HOPE-3 trial and the hype that will surely follow ...

Today in the New England Journal of Medicine (NEJM) an article has appeared reporting the results of the Heart Outcomes Prevention Evaluation-3 (HOPE-3)trial. Exciting simplifications are sure to follow in the news. If you would like to stop reading now, the answer to the question posed in the title is probably no.

The question:

For many years researchers have discussed the possibility of a creating a pill that might contain several kinds of medication that would reduce people's risk of dying of cardiovascular disease. It is an attractive thought. Since cardiovascular disease is the major cause of death globally, reducing that risk has the potential to vastly reduce death and disability. Originally we thought that a suitable "polypill" might contain something to lower the blood pressure, something to lower cholesterol and something to reduce the risk of blood clots. Many studies have looked at the mortality benefits of various blood pressure pills, cholesterol lowering strategies and anti-clotting drugs and we have found that some of them help some people to some extent and some cause some people problems. Drat. So complex.

The history:

The original HOPE trial looked at a blood pressure pill (ramipril) which was added to whatever other medication a group of people at high risk of cardiovascular disease were taking. It lowered the risk of heart attack, stroke and death significantly enough that the study was stopped early and drugs in the ramipril class (Ace inhibitors) were aggressively prescribed, especially for patients with diabetes, probably to good effect.

The HOPE-2 trial looked at using vitamins to reduce homocysteine levels, which appeared to be a significant risk factor for heart attack and stroke. The studied doses of vitamin B6, B12 and folate did not reduce any of the outcomes they looked at. HOPE-2 was dashed.

In 2007 HOPE-3 began, looking at a combination of blood pressure medication and cholesterol medication, specifically to reduce death from heart attack or stroke, new heart failure, cardiac arrest or non-fatal heart attack or angina. The study was performed in academic medical centers all over the world, including the US, Europe, China and India, South Africa and Southeast Asia. Over 14,000 patients were enrolled, limited to men 55 and over and women 65 and over with intermediate cardiovascular risk (calculated at about 1% per year). Women 60-65 years of age were eligible if they had one of several specific risk factors, including glucose intolerance, smoking and high waist to hip ratios. These were people who had never had heart problems or strokes.

Drug company involvement:

Astra Zeneca helped fund this study along with the Canadian Institutes of Health Research. The cholesterol and blood pressure medication used in the study was on patent and sold by Astra Zeneca. The cholesterol pill was rosuvastatin (Crestor) and the blood pressure medication was candesartan/hydrochlorothiazide (Atacand/HCT). There were many drugs to choose from in the categories of effective blood pressure reducers and statin cholesterol medication, some of which would have been cheap and generic. The drugs they did choose are now generic but not yet cheap. I suspect Astra Zeneca hoped the study would have ended in time for them to reap financial benefits as their particular drug proved miraculously effective, or that the combination of the drugs could be made into a new formulation that would extend their market share.

So what happened? Getting to the point.

First there was a run-in phase. The 14,000+ patients were given the study drugs and about 2000 of them dropped out because they didn't tolerate them or didn't want to participate anymore. 12,000 patients were left. Of these about 3000 took placebo pills, 3000 took the blood pressure pills only, 3000 took the cholesterol pill only and 3000 took both blood pressure and cholesterol medication. After about 5.5 years, the patients taking blood pressure medication did not have a significantly lower risk of death, heart attack or stroke or the other cardiovascular outcomes. The patients taking cholesterol medication did have a reduced risk of death and cardiovascular outcomes. Taking both blood pressure medication and cholesterol medication was no better than just taking the cholesterol medication.

How big was the effect and how much does this cost?

The article quotes a hazard ratio of 0.71 for patients taking the combination medication, meaning that the medicated patient has 7 tenths the risk of a bad outcome of the unmedicated patient. The raw numbers are way more interesting. After about five and a half years of taking the combination of drugs or placebo, 163 patients in the active drug group had died and 178 of the placebo group had died, a difference of 15. At the present cost of these drugs, about $280 per month, it would cost about $3.7 million per life saved, which is pretty steep. Just taking the cholesterol medication, which is really the only one that had a positive effect, would cost about $2.6 million per life saved. If there were drugs on a $4 a month plan that did the same thing  we might get down to a very reasonable cost of $53,000, assuming no other associated costs like doctor visits or lab testing (which is not a fair assumption.) Death is not the only thing we care about, of course, but other differences, such as numbers of hospitalizations, were not much more impressive when I reviewed the supplemental data.

How about side effects?

To begin with, 2,000 of 14,000 patients did not like or tolerate the study drugs, so some of them probably had intolerable side effects. During the study, however, there weren't many important side effects except a bit more dizziness in the patients on blood pressure medication, which apparently didn't do them much good anyway. There was some muscle pain, but patients on placebo also had that. The dose of the cholesterol pill was very low, which might have explained the pretty awesome side effect profile. Also the patients most bothered by side effects probably dropped out in the run-in period. By the end of the study only about 70% of patients were still taking the prescribed drugs, so there may have been intolerance that was not reported. There was no excess of development of diabetes, perhaps because of the very low statin dose. Good studies have shown more development of diabetes in statin users, but doses are often significantly higher.

Other interesting findings from the supplemental data:

I was very pleased that the NEJM published a well written summary of data that didn't get discussed in the article. There were some really weird things that showed up.

  1. South Asians and to some extent Chinese subjects did not have near as much benefit from the study drugs as did subjects of European descent. 
  2. Stroke risk was actually higher for patients with borderline high blood pressure when they were treated with the blood pressure lowering drugs, though reduced for patients with higher and lower blood pressures. 
  3. Patients with the highest cholesterol levels (LDL greater than 141, mean of 166) did not benefit from treatment with cholesterol lowering drugs. That's really strange and counterintuitive.

Bottom line?

We have not discovered a pill yet that will make everyone live longer. Taking a statin drug which lowers cholesterol (who knows if that's the important effect) may lower your risk of cardiovascular disease or death, but you need to take it a long time to make a small difference and we aren't sure which one is most effective. Most people who take statin drugs (Lipitor, Crestor/rosuvastatin etc.) for primary prevention (meaning they haven't had heart attacks or strokes) will not benefit from them. Low doses of Crestor/rosuvastatin usually have a low risk of side effects. Treating normal or slightly high blood pressure with low dose candesartan/HCTZ doesn't reduce the risk of heart attacks, strokes or cardiovascular death. Navigating the pros and cons of all of these pills and potions is neither obvious nor simple and is best done with a doctor who you know and trust.

Thursday, May 12, 2016

Medical error--the third leading cause of death, or not nearly so simple?

Splashed over news feeds this last week is a reference to an article published last week in the BMJ (formerly known as the British Medical Journal) that states that "medical error [is] the third leading cause of death in the United States."

What is this article actually about?

The new article refers to research done over the last 2 decades, and most recently an article by NASA toxicologist and patient safety advocate John T. James PhD in the Journal of Patient Safety in 2013. Dr. James evaluated 4 previous studies and estimated that 400,000 people, approximately, die yearly in the US of conditions that were either caused by or, more often, made worse by medical error.

That article, too, was met by outrage and was widely quoted and misinterpreted. I think hardly anybody actually read the article. Headlines read "Deaths by Medical Mistakes Claim the Lives of 400,000 People Each Year."

No new data, but a new way to think about it:

As far as I can tell from reading this most recent article, in BMJ, there is no new data. There is just the realization that if this many people are dying of medical error, it should be showing up somehow as a major cause of death, just like heart disease or cancer. The reason that it is not is because the questions that are asked on death certificates do not include anything about whether medical error had occurred. The article, quite correctly, states that this should be documented. Documenting this is the first step toward making major changes that will keep patients safer.

So what did the original article actually say?

When Dr. James' article came out in 2013, I read it and wrote a blog commenting on the study, its methods and what medical error looks like from my experience. Dr. James kindly responded in a comment,
"Thank you for putting my JPS study in an appropriate perspective. Some folks like to sensationalize the results as if these patients die only because of a medical error, and that is not what I wrote or intended."
The blog from 2013 was actually pretty good and completely relevant to the recent news release. You may want to read it. It does put the data into perspective.

Patient safety--very very important!

I am glad that this recent article has grabbed international attention for patient safety. What we read in USA Today and the rest of the popular press is not an accurate depiction of the problem, but the problem remains very real. Medical error is contributing to the deaths of hundreds of thousands of patients yearly. We, as physicians and healthcare providers, are trusted with the lives of vulnerable people, and it is up to us to build processes that allow our good intentions to be translated into good care.

Sunday, May 1, 2016

Practical and commonsense research from Ontario's Physician's Services Incorporated Foundation: Let them drink apple juice!

Today in JAMA online (the Journal of the American Medical Association) a delightfully practical article was published regarding how to help little children recover from gastroenteritis (stomach flu.) The authors, Doctors Stephen Freedman, Andrew Willan, Kathy Boutis and Suzanne Schuh, compared the health of over 600 children aged 6 months to 5 years, when given a medical rehydration solution to drink versus diluted apple juice along with the fluids that the child preferred. These children were not chronically ill or severely dehydrated, but were sick enough to have been brought by their parents to the emergency department and were at risk for needing hospitalization or intravenous fluids.

In previous studies oral rehydration with electrolyte solutions such as Pedialyte, which contain over twice as much sodium and 6 times as much potassium as sports drinks, along with much less sugar, had performed well in preventing hospitalization or return to the emergency department. Theoretically, children would need electrolytes to replace the ones lost in diarrhea or from vomiting, and extra sugar would just draw more fluid into the intestines, worsening the diarrhea. In the children studied in this recent research, however, giving a teaspoon of dilute apple juice every 2-5 minutes and treating vomiting with a very effective anti-nausea medication, ondansetron, by mouth, worked at least as well, actually significantly better than the same routine using bottled electrolyte solutions. This was particularly true in the children who were over 2 years old and so had definite taste preferences. Children given dilute apple juice and then liquids that they preferred were less likely to be admitted to the hospital or need intravenous hydration than children who were provided oral electrolyte solutions.

There has been a magic associated with giving children Pedialyte or similar solutions for their mild intestinal ills, despite the fact that it is more expensive (1 liter costs about $9, though may be as cheap as $5 when bought in bulk) and doesn't taste very good. Children mostly don't mind it, but some really do, which can be a big factor when they are feeling queasy. This study presents evidence to allow parents to give their children pretty much any liquid that sounds good, making an extra trip to the pharmacy with a sick child unnecessary.

Now oral electrolyte solutions and apple juice have both been available from many manufacturers for many years. What drug company, then, would be interested in funding this lovely research? None, of course. Could it have been funded by a governmental agency? Yes, I guess so, though much of that grant money is very hard to get. This study, though, was funded by a private foundation with a very interesting story.

In 1947 physicians in Ontario, Canada, joined together to create a pre-paid health plan. The physicians pro-rated their fees to stay within budget, and eventually 8000 doctors participated. In 1969, the Physician's Services Incorporated health plan ceased to exist and was replaced by a state run entity, now the Ontario Health Insurance Plan. There was still money in their account when they stopped providing services (they must have been doing something right) and the physicians decided to use it to create a foundation to improve the health of the citizens of Ontario. The original investment of those physicians was $16.7 million, and in the years since inception over $167 million has been paid out in grants. They now have around $90 million from which they fund over $3 million in grants yearly.

If you are interested, read the 2015 annual report for the foundation. The projects they fund are wonderful. There are little grants, one for $5500 to a physician who wanted to study lower doses of a testosterone product for female to male transsexuals. There are larger grants as well. There is a project to grow mats of human cartilage from small numbers of cells to heal the damage of osteoarthritis, reducing the need for joint replacements. There is a fellowship for a researcher who will take scientific research and bring it to the bedside, this year a nutritional expert who works on creating healthy diets. The money granted to this person is to free them up to do work that will not provide them any likely financial benefit. There is a study to look at the side effects and impacts of those side effects on patients who take the (often not very effective) drugs for Alzheimer's disease. There is a study looking at whether giving children who receive antibiotics in the hospital a probiotic milk product (kefir, I'm thinking) reduces their risk of getting antibiotic associated diarrhea. The studies are all like that: practical, meaningful, unlikely to increase healthcare costs and likely to improve health. They are studies that are not often done in the US, where drug companies are our major funding source, expecting to make more money selling drugs than they spend on research.

The PHI foundation is much smaller than the Rockefeller, Howard Hughes or Bill and Melinda Gates foundations which privately fund research in the US. Even so, they helped Dr. Freedman et al publish a paper that will allow the next generation of babies and toddlers to drink dilute apple juice rather than electrolyte solutions and be more likely to stay out of hospitals as they recover from their tummy troubles. Their other grants are just as good. They are a model of medical research funding I have not seen before. They are private individuals, doctors even, who have worked together to do something great. They are not super rich philanthropists, drug companies or federal government agencies.  This organization is making a powerful and positive impact by funding and encouraging researchers who are wise and curious.