People, patients, polypills, primary care and POEMs: Making your health better in the real world

A recent article in the New England Journal of Medicine, ”Polypill Strategy in Secondary Cardiovascular Prevention”, also covered by the New York Times, demonstrated that people who had previous cardiovascular events, (ie., myocardial infarction -- MI, heart attack, stroke or urgent need for bypass for impending MI), had fewer recurrent heart attacks, including fatal ones, strokes, and urgent needs for bypass if they were treated with one pill a day (“polypill”) that combined their recommended medications than they did if they took multiple pills multiple times a day. The conclusion (from the Abstract) summarizes it as:

Treatment with a polypill containing aspirin, ramipril, and atorvastatin within 6 months after myocardial infarction resulted in a significantly lower risk of major adverse cardiovascular events than usual care.

This is good news, if not entirely surprising. The effectiveness of medications in preventing recurrent disease is a combination of both how effective the medication is if “taken as directed” and how difficult it is for a person (or, from a doctor-centered perspective, how “compliant” a “patient” is) to take the medication as directed. For an example of what this means, consider birth control, where “theoretical effectiveness” (how effective a method of pregnancy prevention would be if "used as directed" 100% of the time) is contrasted to “use effectiveness” (how likely a person who is ostensibly using a particular method of birth control is to get pregnant). Obviously, what matters to a person who does not want to get pregnant is the latter, and this is greatly impacted by how easy it is to not use it. Condoms and diaphragms have to be used every time and oral contraceptives have to be taken every day, which can contribute to a lower use effectiveness than forms of contraception that do not require this, such as IUDs and implants, known as LARC, long-acting reversible contraception.

So back to prevention of heart attack: Who is surprised that taking one pill a day results in better compliance than taking more pills more often? Hands? No one? Maybe cardiologists, since the alternative was “usual care”, which must have been “multiple pills multiple times a day”. Any sentient person would know, without needing to be a doctor, scientist, statistician, or epidemiologist, that taking one pill a day is not only easier but much more likely to happen than taking multiple pills, and especially taking pills multiple times. Remember, the medication only works if you take it, and the harder you make it to take, the less likely people will and the less likely it is to work. Even drug companies know this; you probably have noticed that when a drug is about to lose its patent one of the first “new drugs” that the company comes out with is a long-acting version of it that you have to take only once a day! I cannot prove that they would have been able to release this in the first place but instead held it in reserve just for this reason, but I would not be surprised. One pill once a day is also likely to be cheaper (except while under patent) and this is a big issue for people who have difficulty affording their medications (ie., most of us, but especially those with lower incomes. In the US, of course).

Though it seems obvious that one pill a day is more likely to be taken, this research study is not without importance. For starters, it showed that it worked to take one pill combining 3 drugs once a day. After all, if it didn’t help prevent disease, it wouldn’t be desirable. That it worked better than “usual care” is almost certainly related to it being one pill once a day, and how difficult it is to remember to take pills more often. [Even those of you who do not have chronic disease: have you ever been given antibiotics to take 4 times a day for 10 days? How often have you actually taken all 40 within the 10 days, even after you’ve started feeling better sooner?] We may presume that taking the multiple pills multiple times a day would have worked as well if people had done it, but because they don’t, it didn’t work as well. The contribution of the study is to show that the one pill once a day works, and because it is one pill once a day, works better.

Another important thing that the study did  is to look at meaningful outcomes: death, repeat heart attack or stroke, need for bypass. This also seems like a “duh”, if you are a regular person, but it has been common in the literature to not measure these but often “intervening variables” such as changes in cholesterol level (or blood sugar, or whatever you are studying) because this is easier to do and requires less follow-up. But having lower cholesterol, for example, only matters to the extent that you are less likely to have a heart attack or stroke!

When this study first came out, I commented on FB and Twitter that

‘Family physicians and other primary care clinicians know this better than subspecialists, since they are focused on "what works best for the disease?" and we on "what is going to work for this person?"!’

Of course, this is, probably, not completely true for all family physicians and all subspecialists, but it is certainly true that it is more likely for primary care clinicians to think about and be aware of how a treatment affects the whole person. Subspecialists are usually concerned with treating one condition, “their” disease”, and at how the treatment they prescribe ameliorates that, while primary care clinicians are looking at how it affects the person’s life. One simple example is drug interactions and “side effects” (which are actually just effects, but not the effects we want). Primary care clinicians care for a person with all the conditions that they have and have to not only see how, for example, their heart disease medicine works for the heart disease but if it is bad for another disease they have, or if its side effects (or difficult regimen, multiple pills multiple times a day) means that they don’t take it.

In research, family physicians and other primary care clinicians have looked for Patient-Oriented Evidence (POE) as opposed to Disease-Oriented Evidence (DOE). DOE looks at whether a treatment, usually in an experimental setting (which often has many differences from real life, such as free medicine and people to remind you to take it!) makes a disease better, while POE looks at whether it makes the person’s life, as a whole, better. This is important, especially if you are a person. (Or probably if you are an animal!) Indeed, family physicians have taken this a step farther to Patient-Oriented Evidence that Matters (POEMs), with sections reviewing recent research that does featured in several family medicine journals. POEMs is, in addition to being a cute acronym, has meaning; not all evidence, disease or patient oriented, actually matters. For example, the study cited above: showing that using a polypill decreases your risk of cardiovascular events and improves your life matters, while simply showing a change in a lab value might well not.

It is really good that there are treatments for diseases, whether common (like heart disease) or rare (like, ironically, I just discovered a rare blood condition also called POEMS!) that can make you better. It is also really good that there are subspecialists who know about them and can make recommendations for treatment (especially for the rarer ones) and who keep up on the literature. But it is also important that there is someone keeping an eye on the person, the patient, with all their diseases and medicines and treatment regimens and side effects, and, oh yeah, the stressors of their everyday life with money, and family, and work (or not having work) and how in heck, in this country, they are going to pay for their treatment. These are the real components of real life, sometimes called the “social determinants of health”, that are poorly addressed by US healthcare.

Polypills are good, as are POEMs. And so are primary care clinicians, especially when their employers allow them sufficient time and encouragement to actually provide comprehensive care for their patients.

Patient-centered research: answering the questions that matter to people

“Large investments are too often made in studies that provide poor-quality evidence,^1​ are overtly biased,² are not applicable to most patients,^3​ or yield results that do not address the real concerns of individuals facing clinical decisions.^4,5”  [1] (Patient Centered Outcomes Research Center, JAMA, April 18, 2012)

Indeed. This is not only “too often”, it is in fact characteristic of most research, whether sponsored by the National Institutes of Health (NIH), foundations, or private companies. The reasons for this are that it is easier to do and it relatively well funded (thus the “large investments”).

Many years ago, researchers, particularly those providing primary care, distinguished between “disease-oriented evidence” (DOEs) and “patient-oriented evidence” (POEs, or later, adding “that matters”, POEMs). The first is about showing that a treatment improves a disease. The second is looking at what treatments improve the life of a person. There is a big difference. A simple example might be in looking a regimen for diabetes that minimizes the complications from high blood sugar by keeping the average blood sugar much lower than had previously been the goal (called “tight control”). Disease-oriented evidence might show that using frequent insulin injections to keep the blood sugar in the low-normal range reduces the long-term negative effects of diabetes. A patient-oriented approach, however, looks at the overall impact on the person, not just the disease. Does the patient find it difficult to administer more frequent injections of insulin and check their blood sugar? Do they spend a significant amount of time with blood sugar that is in fact too low (after all, reducing the average increases the probability that sometimes it will be too low), and feel fuzzy-headed and unable to live the lives they wish to? Or feel dizzy? Or that a certain percent actually pass out from low blood sugar? Maybe break their hips and end up in the hospital or even dying from its complications? Even though their diabetes is in “good” control? “The disease was controlled but the patient died from a complication of treatment” is not a desirable patient-oriented outcome!

The motivation for private companies, most often pharmaceutical companies, to fund disease-oriented research is fairly obvious. Their sole agenda is to make profit, so they are interested in supporting research that shows that their drugs are effective for treating certain conditions. They have a number of advantages in this arena:

·         Showing “effectiveness” requires, by FDA criteria, only to show that a new drug or other treatment is more effective than placebo, not than the currently available treatments;

·         Since they are paying for it, they can suppress the publication of results that do not show their drugs in a good light;

·         They have to show only that it modifies the disease, not that it is the best choice for any individual patient (thus it is Disease-oriented, rather than Patient-oriented);

·         In lieu of patient-centered research, they have huge marketing budgets (far in excess of their research budgets) to advertise their products to both providers and directly to patients once they have been approved by the FDA.

Why NIH would mostly fund this sort of research is more complicated. Part of it is that an entire industry has been built around doing disease-oriented research, largely in biological science laboratories, but also in doing clinical trials in people. Thousands of academics and the institutions at which they work are dependent upon such funding to maintain and advance their careers and institutions. The review committees that make recommendations to approve or disapprove funding are “peer” committees, made up of people who do, largely, the same kind of work. This is good because they can understand and evaluate the science involved (a really bad thing would be for a group of politically-appointed ideologues to make the decisions, and this sometimes has occurred), but they are limited by their understanding of how research is done and what its goals should be. This relates to a second challenge – patient-oriented research is more complicated, more difficult to do, and leads to less “clear” outcomes. Of course, it is more relevant to patients and their providers making decisions about their care, but it is harder to fit into a rigid research model in which all but one variable is tightly controlled. This can lead to research that is done because it is possible to do it, rather than because it answers the questions that we have (see “Defining Streetlight Research”, February 26, 2009). Most traditional NIH researchers have no objection to patient-oriented research, but in an era of limited funding availability might be quite concerned if funding it decreased the amount available for the kind of work that they do.

Trying to address this, the Affordable Care Act of 2010 established the Patient-Centered Outcomes Research Institute (PCORI) to apply the rigorous standards of Clinical Effectiveness Research (CER) to treatment of patients, not diseases. One clear need here is that people, particularly older people, often have more than one disease. The “right” treatment for a patient’s cancer needs to take into account its effect on his/her diabetes – or heart disease, hypertension, arthritis, depression, alcoholism, glaucoma, poverty, and yes, maybe, another cancer – in any or all combination. Because we care for the person, who may have any or all of these conditions, primary care doctors are likely to be more sensitive to these interactions than are physicians caring for only one of them. It is for this reason that we have advocated for a National Center for Primary Care to look at the comprehensive care of patients differently than the existing disease-oriented Centers (e.g, the National Cancer Institute, the National Center for Heart, Lung and Blood Disorders, the National Institute of Mental Health) have.

The article in the April 18, 2012 issue of JAMA from which the opening quotation is drawn, “Methodological Standards and Patient-Centeredness in Comparative Effectiveness Research”, includes the following case as an example:

“A 78-Year-Old Man with Heart Failure, Diabetes, and Renal Failure. Mr B was admitted to his local hospital for the fourth time this year. Mr B experienced symptoms of shortness of breath and weight gain that led him to contact his physician, who recommended another hospitalization. He lives with his wife in a 2-story house located 30 miles from the hospital. He has recently been unable to climb the stairs to his bedroom. Mr B requires outpatient hemodialysis 3 times per week, but missed his last dialysis treatment because he felt too tired to go. This is his second hospital admission after missing a dialysis appointment. After each admission, Mr B received standard discharge instructions on how to care for himself at home. Mr B's highest priority is his independence, including the ability to drive and to remain in his own home.”

To make appropriate decisions about Mr B’s care requires not only considering the interplay of his several chronic diseases, but the circumstance of, and his own preferences and goals for his life. This cannot be done by employing solely the findings of disease-oriented research.

PCORI is not a National Institute for Primary Care, but it may help to achieve some of the same goals. In the same issue of JAMA, Joe Selby (the director of PCORI) and colleagues list the both the statutory criteria for PCOR-funded research:[2]

1. Effect on the health of individuals and populations
2. Probability of improvability through research
3. Inclusiveness of different populations
4. Current gaps in knowledge/variation in care
5. Effect on health care system performance
6. Potential to influence decision making
7. Patient-centeredness
8. Rigorous research methods
9. Efficient use of research resources

and their proposed priorities for national research:

1. Assessment of options for prevention, diagnosis, and treatment
2. Improving health care systems
3. Dissemination and communications research
4. Addressing disparities
5. Accelerating patient-centered outcomes research and methodology

None of these are disease specific, although unquestionably much of the research that is funded by PCORI will look at people with specific diseases. These criteria, however, recognize the need for treatment plans to take into account all aspects of a person, and by extension the family and community of which they are a part. This includes addressing the disparities that exist between groups, especially poor and minority groups; that is to say, addressing issues of social determinants of health and social injustice.

Maybe we will finally get funding for studies that provide us will more useful information for caring for actual people.

---

[1] Methodology Committee of the Patient Center Outcomes Research Institute, “MethodologicalStandards and Patient-Centeredness in Comparative Effectiveness Research”, JAMA, 2012;307(15):1636-1640. doi: 10.1001/jama.2012.466.

[2] Selby JV, Beal AC, Frank L, “The Patient-Centered Outcomes Research Institute (PCORI) National Priorities for Research and Initial Research Agenda”, JAMA 2012;307(15):2583-4. Doi: 10.1001/jama.2012.500