Strategic Solutions, Inc.

Definition of insanity: doing the same thing over and over again and expecting different results.
—Albert Einstein

I had refrained from completing this last part of our series on health care until the presidential candidates became known and their positions on health care elaborated. As it turns out, they are not as radical as I thought they might be.

Most western countries have a mix of socialized medicine and private health care, although the mix varies considerably. For example, the UK, Germany, France, and Sweden have mostly socialized systems with some private health care plans on the side; in the USA it is mostly private health care until one reaches the age of 65, when individuals become eligible for the government-run Medicare.

Although in the United States there seems an aversion to government-run systems, most socialized health care runs remarkably well. There are the usual gripes about waiting lists, particularly for elective operations, but most studies show that people get what they want 90% of the time within 6 months. So, most of the horror stories are myths propagated by folks who have a vested interest in maintaining the status quo.

Senator McCain offers tax credits ($5,000 for a family; $2,500 for individuals) for existing health plans, plus a lot of legislation aimed at ensuring that the uninsurable have access to health plans. Much of what he has to say makes good sense, but it’s not enough to make a large impact.

Senator Obama would offer some of what McCain has proposed, but says additionally “Individuals and families who do not qualify for Medicaid or SCHIP but still need financial assistance will receive an income-related federal subsidy to buy into the new public plan or purchase a private health care plan.” I’m guessing here that you will have to try and get Medicaid first, but we’ll see. He proposes lots of federally based plans and will make coverage of children mandatory.

Neither of these gentlemen are really providing radical approaches to health care. A lot of what they propose are steps in the right direction; however, they don’t go to the root of the problem. Any country that has a decent health care system has to provide a lot of money to support it, and in the end, resources are always finite, which means making tough decisions. In my opinion, we need to start from scratch.

Everyone needs affordable health care according to his or her ability to pay. That essentially means that rich folks subsidize poor folks. A socialized medical system means we finance it from wage earnings, and if you can’t earn for any reason, the government will pay for it. In our case that means paying a lot more from our wages into some kind of national health fund—like Medicare, but applicable to all. After that, if you still want private health insurance, it will be up to you. I very much doubt that this will happen in the USA because private health care has too much vested in its operations. But of one thing I’m sure: if we keep going the way we are the percentage of uninsured will keep going up and the cost to the insured will skyrocket. Eventually only millionaires will be able to afford health care and by that time the system will have collapsed. It’s your choice people; we can start really making a difference or we can keep struggling the way we are and pass the buck to our children and grandchildren. I know which one my money’s on.

In Part II of this series, we examined the role of cost-effectiveness in medicine. In this part, we take a look at infrastructure, primarily electronic Health records (EHR), and incentives—pay for performance (P4P). In the USA, thus far, EHRs have made little difference, with a few exceptions, notably the Veterans Administration and a small number of institutions (Frieden & Mostashari, 2008; Kupersmith et al 2007; Asch et al 2004; Linder et al 2007; Chaudhry et al, 2006; Shekelle et al, 2006). Since EHRs have been touted as the best thing since the discovery of penicillin, why are we not making much progress? The first thing we need to discuss is what an EHR is, since not all EHRs are created equal. 

The Institute of Medicine defines an EHR as a system that includes: 

(1) longitudinal collection of electronic health information for and about persons, where health information is defined as information pertaining to the health of an individual or health care provided to an individual; (2) immediate electronic access to person- and population-level information by authorized, and only authorized, users; (3) provision of knowledge and decision-support that enhance the quality, safety, and efficiency of patient care; and (4) support of efficient processes for health care delivery. Critical building blocks of an EHR system are the electronic health records (EHR) maintained by providers…and by individuals (also called personal health records) (NIH Committee on Data Standards for Patient Safety. Key Capabilities of an Electronic Health Record System: Washington, DC: National Academies Press 2003). 

Originally the idea of an EHR was to replace paper records that were becoming too voluminous for systems to handle, and horribly inefficient to search. But, as the concept evolved and was put into practice, it became evident that EHRs could be much more. Some of these degrees of sophistication are reflected in the levels of EHRs, which imply also the degree of integration with other systems, and capability. 

It is estimated that of all the physicians in small practices, only 5-10% have EHR systems. Yet, these physicians are probably the group that could benefit most. So, why has implementation taken such a long time? For one thing, EHR systems are not cheap. For example, it could cost $20 billion over the next 5 years just to equip health care services in this country with EHRs. While this is not chump change, it is only a fraction of the $50 billion originally envisaged by the Bush administration for the Iraq war, which will cost the U.S. at least $2 trillion when all said and done. The second obstacle is training–training all these physicians to use the new systems and change their operations to take advantage of what can be achieved. That is a lot of investment in time and resources, and does not take into account the resistance of changing long-held ways of doing business. In addition, the use of EHRs has to be better integrated into P4P programs. 

A basic P4P program would reward physicians for following clinical practice guidelines based on the information they gather from patient EHRs. So far, though, particularly for pilot P4P programs that are CMS-based (Centers for Medicare and Medicaid) the results are lamentable. Lets imagine, however, what is possible. 

An obese female patient aged 34 comes into a physician office complaining of a multitude of symptoms that the physician recognized as incipient type II diabetes. Because her health service facility has a well-designed P4P program it recognizes that an ounce of prevention is worth a pound of cure. Not only does our physician properly diagnose our patient’s disease, but taking into account all the clinically available information on the EHR system, prescribes several prevention programs for our patient to lose weight and change her lifestyle, based upon EBM, and which is paid for by the healthcare facility. Tracking programs within the EHR system also encourage the physician to make follow-up appointments to see if the patient is really following the medical advice that has been dispensed. The physician gets paid a bonus for the patient when she avoids all the horrible comorbidities that come with type II diabetes, and the patient gets better. A win-win situation. The physician does not spend 10 minutes prescribing a bunch of expensive pills, either, since the EBM data available in the HER system tells her this isn’t worth it.  Fiction? Perhaps for now. But if we’re ever going to get a handle on health care costs and improve outcomes, this has to happen. 

In Part I of this series, we examined the role of evidence-based medicine. In this part, we take a look at the cost-effectiveness of medicine. 

What’s Wrong with Our Health-care System? Part I 

In the year 2000, the World Health Organization ranked the United States 38th in regard to overall health care (World Health Report 2000). Whiners predictably said it was because the WHO does not like private health insurance. Yes, that’s part of the problem. Today an estimated 50 million individuals in the United States don’t have any kind of health insurance, and that rate is growing at 2 million per year, with health premiums and costs rising at 7% per year. But that doesn’t count the 15-30 million persons who are underinsured. That problem has been stated another way: 29% of people who have health insurance are not covered for catastrophic events, drugs, or even going to see a physician when they are sick.  Health costs have risen so much that they would be at the top of list for bankruptcy filings were subprime mortgages not also in crisis. At this rate within 10-20 years, the system is likely to collapse because the few people left who still might be able to afford insurance then will be in effect subsidizing the rest of us who can’t; moreover the rest of us will be so sick that by the time we arrive in the hospital we will need 10 times more work than had we been able to afford a physician visit in the first place. 

How did things get so bad? There are 4 pieces to this puzzle: evidence-based medicine; cost-effectiveness studies; infrastructure (electronic health records, etc); and health care funding. In part I of this series, we start by looking at the contributions of evidence-based medicine (EBM). 

What is EBM? Crudely, it defines what treatments and screenings work in medicine, and how well they work. More scientifically, it attempts to “rate the evidence” concerning a given procedure or treatment in series of scales using a hierarchy of different types of clinical trials or studies. The “gold standard” has always been the randomized controlled trial (RCT). In its simplest form, in this type of study, one group of patients is randomly allocated to the control drug, treatment, or procedure—whatever that may be, while the other group is allocated top to the experimental drug, treatment, or procedure. In the best design, neither the patient nor the physicians involved in the study know which group they are in. This type of design is called “double blind.” The results are then clinically and statistically analyzed by other persons not involved in the trial itself. 

Many factors also impact how well the results of the RCT are rated. For example, small numbers (typically < 20 in either group) make it harder to detect a difference in effects between the 2 groups, what we call the “effect size.” When patients or physicians know which group they are in or treating, this lack of masking or blinding can affect the outcome through a process that is termed bias. Finally, if the trial populations are not representative of more general populations, then the results of the RCT will not be applicable to more general populations. Until fairly recently, this issue was largely ignored, but is increasingly being raised as a problem. For example, if the exclusion criteria include a long list of diseases or clinical problems--hat are termed “comorbidities”--then you can imagine we will not know whether those individuals will benefit from the trial, regardless of the results. The quality of RCTs reported in the medical literature has always varied considerably. To make up for small-number RCTs, the results are often pooled in a study called a meta-analysis, which is conducted (hopefully) by investigators with no conflicts of interest. But, as many people know, the results of many of these trials have often sparked intense debate. Why? 

As we pointed out in a previous posting on statins, millions of dollars are often spent on trials with little chance of success. It’s their money you might say, so why should we be concerned? That may be so, but you’re going to pay for the results one way or another if you end up with a prescription for the drug that was tested. And those drugs can be very expensive. I’m not saying that all drug trials are like that; many are worth it, but we need to be far more selective in what trials are funded. For now, we won’t discuss problems with how RCTs are rated, but concentrate on the application of the results. Let’s use a simple treatment as an example. If you are diagnosed with a venous ulcer, the most effective treatment we know is to apply compression using a multilayer bandage. There are some contraindications to compression, primarily whether you have peripheral arterial disease, and usually a quick screening is done to know if you have that problem, too. (The test is called the ABI, and it is a ratio of the blood pressure measured at the ankle to the blood pressure at the wrist–technically the brachial artery.) So if you are diagnosed with a venous ulcer, and go visit your doctor, you would assume that 100% of the time you would be immediately prescribed compression treatment if the ABI is okay, right? Unfortunately, that’s not always the case. These were a few of the conclusions of a Canadian study carried out on venous ulcers: 

“During 1 month, 107 physicians reported having 226 patients with leg ulcers; only a few patients had had ultrasound assessment. Few physicians (16%) were confident about managing leg ulcers; 61% reported not knowing enough about wound-care products. More than 50% were unaware that compression is effective treatment for venous ulcers. Problems reported were lack of evidence-based clinical practice guidelines for leg ulcer care (82%); absence of evidence-based protocols in home-care agencies (72%); lack of access to wound-care products (69%) and wound-care centres (66%); and poor communication among health care workers (60%).” This and many other studies show that our overworked physicians don’t always know what to do. So what is the answer? 

Evidence-based clinical practice guidelines (CPGs) were developed precisely to help educate physicians and medical specialists. They started emerging in the mid nineties and have been growing more prolific every day, as one can see on the Agency for Healthcare Research and Quality (AHRQ) web pages. The problem is that not all health-care personnel are aware of them, although this number is generally rising. The other issue is that many treatments and practices have not been tested with RCTs or other kinds of clinical trials, and until they are, these treatments will remain unverified at best.  Medicare is getting in on the act too, and has a number of P4P (pay for performance) programs that are aimed at health-care providers. Essentially these programs, if they are all implemented nationwide, will force providers to comply with CPGs or lose payments; health-care providers that do well will be rewarded for their effort. How much money Medicare’s approach will save in the long run is unknown, and there are a lot of wrinkles to iron out, but it’s a good start even if it’s not perfect, because patients will get better more scientifically based treatment. The application of evidence-based medicine does promise to save lives, and help ensure that patients get the best treatment, but the major factor missing is how cost-effective treatments are. This is the subject of Part II. 

The Great Statin Debacle 
February 3, 2008 

With the publication recently in BusinessWeek exploding the story about Vytorin, it is time to examine the role of statins, cholesterol, and heart disease. The results of the new drug trial, termed the ENHANCE study, which tested a combination of Zetia (ezetimibe), manufactured by Schering Plough and Merck, and Zocor (simivastatin), manufactured by Merck, were scheduled to be published in early 2007, but this was delayed. A press release on January 14, 2008 indicated that there was no benefit to be derived from the combination. Indeed, the randomized controlled trial, which randomly allocated either Vytorin (10/80 mg) or Zocor (80 mg) to 720 patients that had heterozygous familial hypercholesterolemia–a rare condition present in about 0.2% of the population–and measured the intima-media thickness (arterial wall thickness) at 3 sites in the carotid arteries of patients found that wall buildup increased by 0.0111 mm for the Vytorin-treated group, nearly twice of that of the Zocor-treated group (0.0058 mm). These results were not statistically significant, meaning that the probability of the difference happened by chance. (In statistics, we often use a probability of .05 as the minimum evidence that two results in a trial did not occur by chance, but were due to some effect.) The full results of the enhance study will be presented at the American College of Cardiology meeting in March, but probably won’t be published in a medical journal until the end of the year. Meanwhile, more trials of Vytorin are planned. I guess the drug companies are happy to spend your money on worthless trials, because they have invested so much time, money, and effort in them. So, are statins a complete waste of time and money? This depends in part on whether you believe the cholesterol “hoax.” Why do I say hoax? All the statins were developed on the principle that high cholesterol—specifically, low density, also know as “bad” cholesterol—is one of the major causes of heart disease. If you believe this, then we might as well believe the tooth fairy is real. Start by reading Anthony Colpo’s article about the myth of bad cholesterol published in Journal of American Physicians & Surgeons. If you want more evidence, go read the article by Krumholz et al. published in JAMA in 1994. (And if that doesn’t satisfy you, write me an email and I’ll send you a bunch more references.)  But, let’s pretend for the sake of argument that lowering cholesterol has some merit, and your physician, who has been liberally supplied with both samples and literature, checks your cholesterol and finds it’s high. You take a statin and deal with the side effects, the muscle pains, and so forth, and the increased risk of cancer (your doctor surely knew about that, and told you, didn’t he/she?) and wait for the drug to work. A few months later, your cholesterol comes down. Problem over, right? Wrong! Statins do lower cholesterol, but except in 1% of cases, they won’t stop you from getting a heart attack. How do we know that? Because of something called the NNT. The NNT, or number needed to treat is a parameter that is not often quoted in many drug trials, although it should be. It is basically the mean or average number of patients that have to be treated to in order to have 1 additional event happen. So, if you treated patients with an antibiotic for some infection, and the event was “cure,” and the NNT was 2, then that would mean that for every 2 patients treated, 1 would be cured of the infection. That’s a pretty good number. What’s a typical number for a statin? It’s about a 100 in regard to the mortality of coronary heart disease (see Table 2, statins). So for every 100 patients, we might expect 1 person to benefit. One! And this number does not take into account the number of people in that group of 100 who might be harmed by the drug, which is the number needed to harm.  Statins do actually have some interesting anti-inflammatory properties and have even been proposed as treatments or prophylactics in avian influenza (H5N1). But that’s another story.  So, what’s my advice? Unless you already have severe coronary heart disease, and actually might be a candidate for statin treatment, ignore the cholesterol mantra and stay away from statins. If your physician insists that your cholesterol is high and needs statin treatment, ask him or her to read some of the literature we have cited. See if your doctor understands about the NNT parameter. In terms of heading off cardiovascular disease, you are far better off leading a healthy lifestyle (no stress, now!) and eating a balanced diet free from processed foods. Consider taking supplements of unadulterated omega-6 and omega-3 cold-pressed oils in about a 2.5 or 3:1 ratio.    

Marissa J. Carter, PhD