Sound bites that hurt (Part II)
As I wrote in my last blog, this paradoxical result of patients developing diabetes taking a statin is not news to us. Published in Lancet in 2010 (Lancet2010; DOI:10.1016/S0140-6736(09)61965-6) is a meta-analysis of this issue.
A meta-analysis is not a study, but it can provide important information and hypothesis-generating work. Randomized clinical trials can then be fashioned to prove or disprove the hypothesis. The more alike the trials included in the analysis, the more valuable it becomes.
This meta-analysis analyzed 13 trials each with at least 1,000 patients who were divided into two arms that were identical and followed for at least a year. 91,410 patients were included in the analysis. During the four-year followup, 4,278 patients or 4.6% developed diabetes. 2,226 patients were treated with statins and 2,052 with placebo. The difference was 9%. In English, I have to treat 255 patients with a statin for four years for one of them to develop diabetes. Those 255 patients had 5.4 deaths or myocardial infarctions avoided in those four years.
Death, myocardial infarction reduction or diabetes? You decide. Now let me mention that thiazide diuretics, beta blockers and the beloved niacin all increase the risk of diabetes in patients taking them to some extent.
Let me close this portion of the blog with the following: patients who have had a cardiovascular event need statins regardless of what they do. Patients who have moderate or high risk should be on statins and carefully followed to determine their glucose status. Everyone needs to eat better and exercise more.
If you are at low risk, a study such as calcium scoring will determine your cholesterol burden, and patients should be placed on statin when appropriate.
That is the information beyond the sound bite.
Now, let’s address the sound bite contention that Pradaxa causes myocardial infarctions. This discussion is also very complex, and when reduced to a sound bite, does not benefit anyone.
Some facts: When Pradaxa at 150 mg twice a day is compared to warfarin, embolic events are 1.11%/yr vs. 1.69%/yr. Hemorrhagic stroke is 0.10%/yr vs 0.38%/yr. Ischemic stroke is 0.92/yr vs 1.20%/yr.
Major bleeding is 3.11%/yr vs 3.36%/yr. Death is 3.64/yr vs 4.13%/yr. The reduction in mortality is approximately 10%. This data is what caused the FDA to approve the drug.
Added to this are warfarin’s very real problems. I will not dwell on all of them here, but they include — and are not by any means limited to — only 55% of eligible patients who receive the drug for various reasons, and this number increases as patients get older. When warfarin is used, it is often ineffective. During the first year, warfarin is therapeutic 58% of the time. The lack of adequate therapy is often due to patient non-compliance with dose and/or food or drug interactions. Warfarin is, however, inexpensive.
Does Pradaxa cause more myocardial infarctions?
Sound bites that hurt (Part 1)
I knew it was going to be a long day when I started to get messages that statins cause diabetes.
There is an adage in TV journalism that “if it bleeds, it leads.” The more grotesque something can be made the better. Partial truths in medicine do not count. This is not Presidential politics. Patients for some reason often make health decisions based upon what they hear on the news, or worse, what their friends tell them.
I blame doctors for most of this. I believe that people would rather hear it from medical professionals, but not many of us will give them the time to explain complex topics simply. Some doctors can’t, some doctors won’t. As I have said before, it’s the major reason that I spend the time to write my blog.
I certainly have things to write about at this time. I will start by explaining why statins don’t cause diabetes, and then I will move on to why Pradaxa does not cause heart attacks. Finally, I will touch on Paula Dean and why as Soupy Sales used to say “people who live in glass houses shouldn’t throw stones.”
First, it is important to understand that randomized clinical trials test only the hypothesis that is being tested. In the case of statins, the hypothesis is that if you take them, you have less of a risk for an ischemic event than if you do not. The risk reduction increase with increasing doses of statin that decrease the level of LDL cholesterol.
This topic has been widely discussed and understood for some time among those of us who follow these sort of issues. The report which engendered the current rash of news stories is possibly the worst of the studies to quote. The study was published in the Arch Intern Med 2012 (DOI: 10.1001/archintermed.2011.625). It reports results from the WHI or Women’s Health Initiative. This study encompasses 153,840 postmenopausal women 50-79 years old. Data was collected at baseline and year three, and the data is from 2005.
Only 7% of the women in this study were taking statins. Of those taking statins, at baseline 71% developed an “increased risk” of diabetes. After adjustment for “confounding variables” the risk was 48%. (My quotation marks)
First of all, the absolute numbers are not provided. What is meant by that is the following example: Plavix was approved because it had a reduction of stroke risk by 20%. What is not understood is that the risk of stroke in the control group was 1.8%. The risk in the Plavix group was 1.3%. This is a 20% reduction but you need to treat 200 patients to get one benefit. The largest use of Plavix, stent protection, has nothing to do with this study. It was the study that was used by the FDA to approve Plavix and then we used Plavix however we wanted.
What was the real risk of developing diabetes? What was the characteristics of this group? What was it about this 7% that made their doctors put them on a statin in the first place? Were they more likely to develop diabetes anyway? Did their weight increase over the time of the study? Was there any offsetting benefit by the use of statin? None of these questions are answered only that “statins cause diabetes.”
Some more pertinent facts next.
The deadly conversation (Part 2)
In my last blog, I began writing about how the discussion of a death with the loved ones can engender a myocardial infarction. The fact that it happens is not much of a surprise. The results of this study, however, are quite a surprise, and that is why it was published in a prestigious journal such as Circulation.
Of the 1,985 participants who participated in the study cited, 270 reported the death of someone within the past six months. 19 patients reported the death within 24 hours of the news. Seven patients in 24-48 hours before, five patients in 48-72 hours and 21 in four to seven days.
The study reports that the chance of having an acute myocardial infarction within 24 hours was 21 times what your risk would be if you were not told of the event. This risk of a myocardial infarction declines each day but lasts for at least one month. Age, sex, frequency of physical activity or history of coronary artery disease did not adjust the risk or explain it.
To put this in perspective, let’s look at some other causes of myocardial infarction. Anger causes a 2.4 fold increase in events, marked anxiety a 1.6 fold increase, an exacerbation of depression a 4.3 fold increase. Extreme physical activity that is unusual for a patient, like shoveling snow, has also been shown to be a cause.
What causes this marked increase in myocardial infarctions? The short answer is that we don’t know. One of the authors of this paper is currently conducting a randomized clinical trial as to whether aspirin and or low dose beta blockers might help.
The number may be even greater. That’s because since the time frame that this data was collected, 1989 to 1994, our definitions of myocardial infarction has changed. We did not have the sensitive biomarkers that we now have. Troponins, the exquisitely sensitive markers of cardiac cell damage, were not available. The line between unstable angina and myocardial infarction is razor thin now. Many more patients who experienced chest pain or “collapse” after being told of the death of someone may, with today’s technology, qualify in this type of analysis.
One answer, at least in women, may be the “Broken Heart Syndrome” or as it is known, Takotsubo syndrome. Perhaps the best term for it is stress cardiomyopathy. Since I blogged about this several years ago, I have seen four more cases. Always women, although we don’t know why. One after an argument with her husband; one who opened the mail and found deportation papers; one with tremendous financial stress and one after being told that her son died in an accident.
One notable problem with the study cited was that the patients in the study were not catheterized. In today’s practice, almost everyone with the symptoms and presentation undergoes cardiac catheterization. The four women mentioned above all were and all had normal coronary arteries. The extensive damage seen at cath almost always reverses, and patients are usually fine after six months…at least their hearts.
Cardiac medication will not fill the void left by someone’s death. It seems we are all at risk. To those of you who know what I am blogging about, you understand the difficulty. To those who do not, you will at some time in your life. It is the human experience.
There is a saying “what doesn’t kill us makes us stronger”. Maybe it’s true.
The deadly conversation (Part 1)
We will all die. As they say (and you all know who they are), the only things certain are death and taxes. I can say without hesitation that the hardest situation anyone finds themselves in is when that individual has to tell someone about a death. I can tell those of you who have not gone through it that there is no right way and no good way. It always hurts everyone involved. To those of you who have not had the experience, I wish you the best.
This is not an unusual circumstance for physicians, and the closer you are to the front lines, the harder it is. Some deaths are expected, but the finality is still a shock and a burden to the loved ones. Even when expected, death hurts. If the death is sudden and unexpected, it is even worse. Closure for the family never occurs, and the trauma often lingers for years. Cardiac arrest that arrives in the hospital and does not survive, large myocardial infarctions that spiral down to death, procedural death — they are all painful to speak about and watch.
The first person I ever had to tell that someone had died was when I was 30 and deep in my medical training. I was still at the point where I was not responsible for the discussions as I was not the senior member of the team. It was my great grandmother who was 96 at the time. I was elected to tell her that her last friend had died. She still had her immediate family, but her last connection to her independent life had died. She looked at me and said, “It’s not good to be the last.” I will never forget the look on her face.
It seems that this conversation is even tougher than I thought. It seems that this conversation can actually kill you, particularly if you are at high cardiac risk.
Published online on January 9, 2012 in Circulation is an article titled “Risk of Acute Myocardial Infarction after the Death of a Significant Person in One’s Life: The Determinants of MI onset Study.”
As always, some background is needed. A study was conducted between 1989 and 1994 to determine what the potential triggers are to a myocardial infarction. This study included 1,985 patients of which 1,318 were men and 590 were women. All patients were interviewed about one week or less after their event. During the interview, one of the questions posed was, “During the past year, did you hear the news of the death of a friend, relative or someone who was very significant in your life?” If they responded yes, the timeframe of the death and the relationship status were determined. The participants were also asked to rate the significance of the death in terms of slightly, moderately or extremely meaningful.
These participants were in a case-crossover study. A case-crossover study means that each patient is his or her own control. Using the demographics and risk factors, an estimate of what the death rate should be was calculated for that patient’s situation. This can then be compared to what actually occurs.
The answer is sobering.
What would you do?
Medical research is broken down into three basic types. The first is pure research, which occurs in a lab and starts with testing in small and then larger animals. The second is first in man studies where the substance is given to human volunteers to test toxicity. Sometimes these are healthy subjects, and sometimes because the substance is unique, these are very sick patients who volunteer. The third are randomized clinical trials that are sometimes small and sometimes massive. The small trials are at the beginning to test safety, and the massive ones are at the end to test whether the drug provides a benefit: mortality, morbidity, lower blood pressure, longer eye lashes, whatever is being proposed as a benefit.
This blog post focuses on a unique and promising treatment for myocardial infarction that has been discovered by researchers in England and supported by grants from the British Heart Foundation. It represents true bench research and is an example of great science.
Stem cells have generated a great deal of excitement but not much is available to show for it. This is to some extent the result of politics being involved with science, which is never a good combination. As I have blogged about before, many countries are far ahead of the United States in this field. There is a considerable amount of basic science which is not being done.
What if your natural cells could be primed to repair heart damage if it occurs? The heart has cells known as “progenitor cells,” which formed our hearts in the first place. These cells lie dormant in our epicardium. These scientists had an idea about awakening them from sleep and allowing them to heal a heart that has been damaged.
This work was reported in the journal Nature 2011;DOI:10.1038/nature10188. They used mice for the work. Here is the rub. I know, like Gilda Radner said, “it’s always something.” The something is that you have to take the medicine before you have the heart attack. If you were at risk, would you take a medicine that might help you recover from what you have but didn’t ask for? Patients take aspirin to reduce death from a heart attack by 20%, they take statins to reduce the risk of a heart attack; would you take a pill every day so that if today was the day your heart was damaged, it would heal itself?
These researchers found that by giving mice the peptide thymosin B4, the mice who were then given heart damage and another booster dose of thymosin B4 sustained a 25% improvement in ejection fraction after myocardial infarction, compared with the control group that received placebo.
That’s an increase, which we presently cannot match with medication. Further, these mice were not given that medicine, so it is possible that the benefit of everything we presently have available would be substantial.
Mice are not human, and often what works in mice does not work in a human, or if it works, the effect is less. Further, now that these scientists know the pathway exists, they are working on finding drugs which turn these cells on in a more effective way. This work is presently beginning to be done in humans.
The journey has begun…the future is promising.
The fourth night is free Part II
Often, a study is done for one thing, and it turns out to be provocative for another reason. The Assessment of Pexelizumab in Acute Myocardial Infarction was done to prove that the drug pexelizumab, which is an antioxidant, would decrease damage in patients who sustained a myocardial infarction. I was one of the investigators in this study. It did not achieve success, but as is often the case, the data served as the source of another study.
As I have discussed in other blogs, the use of data in this way is known as Post Hoc. A Post Hoc study can not be used to define an answer, but it can be used to define a problem. These studies often form the basis for further work that is randomized and controlled.
Published in JAMA 2012; 307:66-74 this past week and titled, “International variation in and factors associated with hospital readmission after myocardial infarction,” it reviews the country with the worst readmission rate.
And the answer is…you guessed it, the United States. In the process of this study, 60% of the US patients were discharged in three days, and 14.5% were readmitted within 30 days. This contrasts to 54% rate of six days hospitalizations with a 9.9% readmission rate. Germany had the longest length of stay at 9 days. Importantly, this has nothing to do with mortality rates, which were not different within this study.
Clearly patients are returning for some reason. Further, and more compelling, is the proof once again of the law of unintended consequences. Medicare now holds the length of stay for a myocardial infarction at three days, and you get “dinged” if it goes beyond. If this “quality measure” is not correct, what other “quality measures” are incorrect? Maybe we should rename them? Maybe we should change what we measure?
The study showed that patients with multivessel disease, a baseline heart rate elevation and being in the United States had the greatest likelihood of being readmitted. Within the United States, a pattern of difference cannot be found. We all do it poorly it seems.
Clearly, we are not saving money with the approach we presently have. Being readmitted to a hospital is a hassle for everyone involved. Multiple tests are repeated, and all the paperwork has to be redone. Often, other people are involved in the patients’ care, and at times, a hospital is involved, which is a real problem. Repeat cardiac cath is often necessary to look at the site of angioplasty to see what its status is. If added up, I would venture that we spend more sending patients home early than if we waited another day or so.
Hospitals are left with the problem of a DRG payment and length of stay that is set in stone and will definitely not change just upon the basis of this study. More studies will surely come. In the mean time, if you have a myocardial infarction, don’t be surprised if your doctors asks you to spend the fourth night. Even if it’s not free.
The 4th night is free (Part I)
I’m sure that you have at one point received a promotion in the mail stating that if you stay at a certain hotel for three nights, the fourth night is free. Does what works for hotels work for the United States health care system?
Since the beginning of care standards for myocardial infarctions, much has changed. In the 1960’s, the standard was six weeks of bed rest and oxygen tents. This six weeks was spent in the hospital. All patients were anti coagulated with warfarin, and the use of anti platelet drugs were minimal. Coronary artery bypass surgery was not yet defined, and morphine and oxygen was about all we had to offer.
In the 1970’s, as bypass became more accepted, we as cardiologists struggled to find the proper timing for bypass surgery. Should it be immediate, or should six weeks go by? (Everything in medicine is six weeks if you hadn’t already figured that out.)
In the 1980’s, it became clear to cardiologists that we needed to do something about a heart attack instead of just watch it. As I have blogged about before, we didn’t even know that heart attacks were caused by clotting of the artery until the seminal work of DeWood in Seattle. This knowledge that “it’s the clot, stupid” led to the formation of the TIMI group which stands for Thrombolysis in Myocardial Infarction. This group was led by Dr. Eugene Braunwald of Harvard until he stepped down last year.
I, and many others, spent the 1980’s and early 1990’s researching various drugs and methods to stop heart attacks. By the mid 1980’s, it became clear to cardiology that the best way to stop heart attacks was to perform angioplasty to the infarct related artery. This became even more successful in the 1990’s as stents became available leading to, at last, a new standard of care.
At the present time, many hospitals provide immediate angioplasty for the treatment of myocardial infarction. This includes hospitals that do not have immediate coronary artery bypass surgery available. If no angioplasty is available, then a transport to a hospital that does is warranted. A byproduct of this treatment is that patients often have very limited damage to their heart muscle. Every heart attack has some damage, but prompt presentation to the hospital and the paradigm of 90 minute door to ballon times have significantly benefited patients. “Time is muscle” and the sooner patients present to the hospital and the artery is opened, the better patients do.
All this has led to patients who often feel quite good the morning after, and they want to go home. What should we as cardiologists do? What should hospitals do? As we cardiologists often do, we did a study, and it was determined that the new length of stay for a myocardial infarction was to be three days.
The question is…did we get it right?
Another bright idea bites the dust
Well, it’s a new year, and we start off with another cautionary tale about medicine by cookbook gone awry. Often, we in medicine have an admirable goal and then devise a proposed solution. This is where it gets dicey. The solution is often put in place before the research is done that proves or disproves the solution. This is not done purposely but we doctors do this over and over again. This is where cardiology’s use of randomized clinical trials excels at attempting to keep us on the right path.
The latest victim is the use of low molecular weight heparin in the treatment of acutely ill medical patients to prevent deep-vein thrombosis and pulmonary embolism.
Some background. Pulmonary embolism is a catastrophic and often fatal problem. A review of 6,833 autopsies showed that 81% of fatal pulmonary embolism occurred in acutely ill medical patients (J Clin Pathol 2004;57:1254-1257). Roughly 10-20% of medical patients and 40-60% of orthopedic patients are at risk for deep-vein thrombosis.
Does this risk warrant exposure of every single patient that enters the hospital to blood thinners to prevent deep-vein thrombosis? Presently at Holy Cross Hospital, and most other hospitals in the United States, the default provision is that everyone gets low molecular weight heparin injections unless the doctor “opts out” by an order. This is not assigned by perceived risk, it is just done by an order sheet that every patient gets on admission.
I have spent considerable time reversing these orders as many of the patients I have are not candidates for this treatment. Further, it is annoying to patients and causes at times large bruises when not administered correctly.
Please do not misunderstand me. I am in favor of this treatment for many patients, but it should be individualized by the physician and not by fiat. This, by the way, costs a considerable amount of money but less now that generics are available.
Where did we go wrong? As published in the NEJM (N Engl J Med 2011; 365:2463-2472) on December 29, 2011, it turns out that the rate of death from any cause did not differ between the group that received the drug from the group that wore graduated compression stockings.
This study did not take place in the United States. It took place at 193 sites in China 24.9%, India 48.8%, Korea 4.6%, Malaysia 3.5%, the Philippines 7.0%, Mexico 4.8% and Tunisia 6.4%. It enrolled 8,392 patients from January 2008 to September 2010. (Remember what I have said about research being done in other countries particularly Asia and Europe.)
The study enrolled acutely ill medical patients. One group received injections and the stockings and the other group received the stockings and placebo injections. At 30 days, 4.9% of the patients who received drug died and 4.8% of the patients died in the placebo group. The most common cause of death was pulmonary failure in both groups. One patient in each group died from a pulmonary embolus — the reason for the whole issue in the first place.
It is clear that there is a group of patients that benefit from this treatment. What is also clear is that the simple use of stockings could make all of our lives easier. It is time to rethink our one size fits all approach and use the skills we were taught in school to assign what should be given and not.
2011
As 2011 draws to a close, I thought we could review some of the highlights and low lights that the year has brought us.
But first, I need to alert my readers to an important albeit discouraging article written by Tara Parker-Pope in the New York Times yesterday. Ms. Pope is the editor of their Wellness Blog and writes an article entitled “The Fat Trap.” She has been overweight her entire life, and the article recounts her struggles and the research, which increasingly shows that we may be doomed to stay the weight our bodies want and not the weight we want.
I have blogged about this extensively on November 1st and 3rd. It’s good to know that at least I’m topical. This was the topic of several other blogs throughout the year.
In February (1st and 3rd), I discussed a common misconception among both patients and doctors. That is that you need to have a severe blockage in your heart artery to matter.
Throughout the year, I discussed the exciting yet troubled debut of drugs which compete with Coumadin for the prevention of embolic stroke in atrial fibrillation (01/06, 02/08 , 04/21, 08/02, 10/06, 12/09, 12/13).
In April, I discussed what will become the focus of this year. Whether what we do is “appropriate or not” (04/07, 04/12, 04/14). Starting January 1st, the federal government will review 100% of the admissions to the hospital for Medicare recipients who receive angioplasty with or without a stent. I warned about this in this blog (12/06 ).
Another topic I blogged about extensively is why we are losing our competitive edge to Europe and Asia (04/19).
In May, I took on the challenge of high triglycerides (05/03, 05/05) and then reviewed the death of niacin (05/31, 06/02, 06/07).
I have spent considerable time this year discussing the changes wrought by the Affordable Care Act (05/24, 05/26, 06/14, 06/16, 06/21, 07/05).
I am dismayed at the attitude of many of my patients who I believe, in general, reflect the attitude of the population. If people understand this massive change and understand that this is just the beginning of a process that must occur, they would be enthusiastic about it. No one to date has stepped up and given a cohesive narrative about these necessary changes. That is I believe one of the great failures of this year.
The new aortic valve procedure, TAVI, is now approved and stay tuned as we should have some exciting news in the near future. This is the beginning of a new paradigm in cardiology, and in the future, much of the valve work maybe done in this way (01/20, 07/12, 10/27).
In August, I reviewed the enormous obstacles we face with generic drugs (08/25, 08/30, 09/13, 09/15).
In October, I reviewed the Nobel Prize and its idiosyncrasies (10/13, 10/18). I also continued my assault on our congress (09/22, 09/29, 10/04).
In November, I learned something as I reviewed the origin of the BMI and its misuse (11/08, 11/10).
December reviewed what the year has brought in the way of scorecards for cardiology (12/02, 12/22).
It has been a long year for many people. I look forward to a better year in 2012, and I want to thank you all for taking this journey with me.
Happy New Year.
Remember what I said about statistics?
One of the reasons that I write this blog is to point out the difference between the medical press and the lay press. We have before us an excellent example.
Published in JAMA 2011; 306:2588-2593 is an article entitled “Association between chlorthalidone treatment of systolic hypertension and long-term survival.” This study is a 20-year follow-up of one of the original studies that represents the basis of our current understanding of blood pressure control and treatment, which is that you need to treat blood pressure because it prolongs life. But does it? This article is an answer to that question.
The original study enrolled 4,736 elderly patients with a mean age of 72 years old, between 1985 and 1988. The treatment was with atenolol, a beta blocker, and chlorthalidone, a diuretic. The placebo group got nothing, but there was a crossover if the placebo group participants pressure went too high. This study would be unethical today since most physicians are convinced that this condition needs to be treated. At the study endpoint, there was an important difference in the incidence of stroke and major cardiovascular events.
I have blogged before that the ACCORD study showed that strict control of both diabetes and hypertension independently do not provide better care in diabetic patients. We now have an understanding of what happens to a large group of patients who were treated for hypertension that was primarily systolic.
Hypertension comes in three flavors. You can have an isolated increase in the top or systolic number, an isolated increase in the bottom or diastolic number or an increase in both the systolic and diastolic numbers. The last type is perhaps the most important. Systolic hypertension tends to occur in women, and we at this point, 20 years later are no better in treating or diagnosing it.
We now understand that the systolic number maybe an artifact in this patient population which tends to be petite women. This is because the blood pressure measurement becomes difficult when the brachial artery becomes non compressible by the blood pressure cuff because of calcification. In English, if we simultaneously stick a needle in the artery and measure the pressure and at the same time measure it with a blood pressure cuff, the two numbers would be quite different. The patient may not really have a high systolic blood pressure number.
This discrepancy between the actual and perceived values of blood pressure leads to over treatment in this group, and the symptoms of dizziness and fatigue as the actual blood pressure may be quite low on treatment.
What did this published article show? It was heralded as “blood pressure treatment prevents death.” What the article actually showed is that we all die of something. If it’s not a cardiovascular problem, it’s something else. The treatment group had a death rate of 59.9% and the control group had a death rate of 60.5%. The actual numbers were a gain of 105 days for the treatment group for all cause mortality and 158 days for cardiovascular mortality. Each month of treatment resulted in one extra day in life expectancy.
It should be pointed out that after the study, all patients were treated in an ongoing manner. The effect here is from just the time that the treatment difference occurred. The five years that one group was treated and one not.
As Artie Johnson would say, “Very interesting.” The take-home message is that treatment of this condition does seem to matter when it comes to strokes and cardiovascular events, but in the long run, pick your poison.
For whom the bell tolls, it tolls for us.
About the Institute
Categories
Archives
Follow Us
