A long-standing debate in the life sciences has been the role of nature versus nurture in determining individual characteristics. For example, how much of an individual’s height is determined by their genes and how much by their nutrition – both in childhood and prenatally?

In the last few decades advances in our understanding of genetics has shifted this dichotomy to describe it in terms of genetics versus environmental factors, and expanded our appreciation for the role nature/genetics play in causing all manner of human diseases. For example, it was discovered that genetically determined slow serotonin transporters in the brain can predispose individuals to developing depression. However, the more we learn about genetics, similarly exciting discoveries are being made into how environmental factors influence the activity of genes, and thus affect an individual’s health.

Thus - like many things with multifactorial causes - the pendulum swings one way and then the other, i.e. between nature/genetics and nurture/environment.

But what has also become clear is that genetics and environmental influences are not  separable - they interact in important ways, and both can cause health problems and be routes to solutions too.

What recently caught my eye in the nature versus nurture debate were two studies about smoking cessation that came out in the last couple of weeks. The first, by Nicholas Christakis published in the New England Journal of Medicine (and reported in the May 22nd New York Times) describes the importance of social factors for people trying to quit smoking. This study demonstrates how an individual’s social environment plays a dramatic role in their ability to stop smoking, and how people can help each other stop smoking as a collective activity.

On the nature side of the equation, NIH researcher George Uhl led a study that shows how genetic markers can predict which of two smoking cessation treatments will be more likely to work for an individual. From these two studies it is also possible to postulate that individuals may have genetic predispositions that effect the ability of social networks to help them to stop smoking – or possibly to start smoking in the first place. (Didn’t we call this peer-pressure in junior high?)

I find both these studies very interesting because for years smoking has often been described in terms of individual choice (or weakness), and as such something that is completely separate from genetic or environmental factors. These studies refute that popular myth, and more generally illustrate the fundamental point that most illnesses and healthcare problems are complex, and that the solutions to these problems need to be similarly sophisticated. This also reminds me why I cringe a bit when someone says to fix a healthcare problem, “all we need to do is…….”

What are your favorite examples of interacting genetic and environmental factors, or where simple solutions have gone awry when used for complex problems?

While watching the Red Sox-Milwaukee baseball game on TV last Saturday, the announcers were discussing Julio Lugo, the Red Sox shortstop who had missed 5 games with a concussion. One of them noted that the team had done baseline neuroimaging testing on all the players so if they had a concussion during the season they would have a baseline to compare to their post-concussion tests. I assume that they also did non-imaging neurological testing, such as reflexes and memory, etc., and this was an extension of all the other evaluations the players went through.

The reason I found this interesting is that it highlights the importance of having an understanding of both a patient’s baseline and their goals. For example, the ability of someone to play the piano after a hand injury is healed depends upon whether or not they could play the piano before the injury. (Sorry – that’s from a very old joke.)

Obviously after a professional baseball player has a concussion, the important factors include reflexes and visual focus since hitting a 90+mph fastball and catching a rocketed ground ball requires precise physical abilities. But for the rest of us - and overall healthcare quality - baseline functioning and goals are important because they are different for every patient, and should be a crucial component of the conversation between a patient and their clinical team. Only when the clinician understands what the patient’s abilities were before the illness or injury, can they work with the patient to develop a treatment plan that they hope will get the patient to their goals.

This may all sound silly or simplistic, but too often I have heard stories about patients who complain that their physician has given them the treatment plan without considering their desires or perspectives - as in “this is what where going to do.” Unfortunately, physicians sometimes only see one reasonable treatment course. But patients are as different as baseball parks – and I mean globally from sandlots to Yankee stadium – and their perspectives, concerns and treatment goals really are just as diverse.

I’m ready for blasts from physicians and others on this one, so let me know what you think.

At a policy related forum for medical residents at the Massachusetts General Hospital last week, one of the questions was about end of life care. The other panelists and I answered from a clinical perspective - and I also put a plug in for the Kenneth B. Schwartz Center which does educational programs for clinicians about strengthening patient-caregiver relationships and communications.

However, during the informal discussion after the panel, a number of the residents raised questions about how much of our healthcare spending goes for treating people at the end of life. In talking with them about this, I recalled three things having looked at this issue several years ago: First, there is a lot of public misunderstanding and misrepresentation of information about this topic. For example, percentages of Medicare spending turning into percentage of total healthcare dollars, or spending in the last year of life being reported as spending in the last month of life. Second, the percentage of healthcare spending in the last 6 months of life has not changed in decades. And third, people agree that if a patient’s life expectancy was clear, then it would be easy to reduce spending at the end of life. This last point was one of the reasons for hospice care, which attempts to both improve clinical care and reduce costs. Hospice has been a big clinical success since - unlike hospitals and general home care services - it is designed to fit the needs of the patient and their family.

So what about costs at the end of life? I did a quick (and admittedly non-comprehensive) search for more information, and found that:

• As of the late 1990s, care in the last year of life consumed about 27% of Medicare’s spending and this percentage hadn’t changed in 20 years. These costs represent 10-12% of total healthcare costs, and 30-40% of this spending occur in the last month of life.
• There is great variability of costs incurred during the last 6 months of life across the country – with up to a 250% difference. But this might also reflect the overall variability in healthcare costs by location.
• A 2007 study found that hospice use reduced Medicare’s costs in the last year of life by 23%, but a 2004 study found that savings from hospice were only with younger patients with cancer, and that overall, hospice enrollment cost Medicare more money.
• A great 2002 article reviews the literature and the issues about end of life care – including how Medicare financial incentives may contribute to more intensive care and higher costs.

So what are my conclusions? Any focus on reducing spending on end of life care as the solution for our rising health care costs is misplaced. These costs are significant, but they are only part of the problem since their growth reflects the overall rise in healthcare spending. And Buntin and Huskamp’s conclusion in their 2002 article, that we have “critical gaps in our knowledge about how to design a better end-of-life care system for Medicare beneficiaries,” is right on target.

What are your thoughts about end of life care, hospices, and educational organizations for clinicians like the Schwartz Center?

Are people who participate in clinical trials patients or subjects? This may seem like a minor rhetorical difference, but I believe it has tremendous implications for health and biomedical research policy. Let me explain why -

Clinical trials are experiments to discover new knowledge. Their intent is to see if a new way of treating a specific disease or condition is better than, the same as, or worse than, another option – either a placebo or an established treatment. Therefore, when people agree to participate in a clinical trial they are participating in this experiment, and their fundamental goal for participating should be to help future patients by expanding biomedical and/or clinical knowledge. As part of their participation, they may receive some benefit – IF the experimental therapy does prove to be beneficial. Therefore, they are not patients; they are subjects within the clinical trial.

I often see phrases like “clinical trial patients” or “patients in clinical trials,” and I understand what they are talking about, since for many patients with serious conditions that lack good treatment options, clinical trials offer some hope. However, I think that the term muddles the distinction, and can lead to conflicting responsibilities for the clinician-investigators conducting the trials and the patient’s primary physicians. This is one reason that I believe that patients who elect to participate in clinical trials should maintain a relationship with their primary care physician during the clinical trial – even if the researcher in charge of the clinical trial is responsible for all their clinical care. This also means that the clinical investigator and the primary care physicians for the subjects in the clinical trial need to have regular communications - both before and during the person’s participation in the clinical trial.

When the person’s primary care physician is also the clinical investigator in the clinical trial, these lines are easily blurred, and this can lead to problems – something that all clinician-investigators struggle with. This is why clinical trials have protocols that are pre-approved by at least one Institutional Review Board (IRB) that is charged with protecting the rights and welfare of subjects in the clinical trial. (IRBs are also involved with monitoring the clinical trials and modifying the protocols as needed during the course of the trial.)

I was reminded about this issue by an article in the May 1^st issue of American Family Physician that discusses when and how it is OK to use a placebo in clinical practice, i.e. for patients rather than clinical trial subjects. The article lays out a bright line for physicians - “do not lie” - as well as referencing an older article* that lists five conditions that should be met before giving a placebo to a patient:

1. There is a well-established, durable physician-patient relationship
2. There is a concrete diagnosis that does not mandate or support the use of other “active” interventions
3. The patient specifically requests that the physician provide some form of intervention
4. The use of such agents is a consideration of last resort
5. The use of such agents does not substitute for, or interfere with, diagnostic and therapeutic diligence

The distinction here is between using a placebo clinically, and using it in a clinical trial. For someone to become a subject in a clinical trial, they need to first give their informed consent so they are aware of the risks involved with their participation. Having a physician give such informed consent in clinical practice when using a placebo might actually undermine the placebo effect, and thus remove any benefit the patient might receive from the placebo. The AFP article specifically counsels that if a patient asks about a recommended treatment that is a placebo, that an appropriate response from their physicians should be something like - this may not directly affect your condition, “but it may turn on other mechanisms that might be important for your health and make you feel better.”

Any additional thoughts on this issue?

* Bok, S., “The ethics of giving placebos,” Sci. Am. 1974; 231 (5): 17-23

The Institute of Medicine put out a report yesterday titled “Retooling for an Aging America: Building the Health Care Workforce.” The report discusses how the aging of the baby-boom generation will create greater needs for health care providers (of all types) who are trained in caring for the elderly with chronic conditions. The report’s recommendations fall into three categories: training, system transformation and financing. Like many reports about health system improvement, their recommendations all make sense - particularly within the context of the three categories. However, like many IOM reports, the writing by Committee process is a bit evident in that, (at least from the Executive Summary), it doesn’t seem to describe a complete plan, nor does it prioritize any of its recommendations - either in terms of funding or which actions should be done first.

In addition, while the report recognizes that the elderly in the coming decades will be healthier than those of 20 or 30 years ago, it doesn’t seem to fully address how this will change the healthcare services needed by the future elderly.

It seems to me, that one of the major challenges facing the healthcare system of the future is how to better manage chronic conditions - regardless of the patient’s age. Thus, rather than retrain clinicians (or train more caregivers) in geriatrics, there needs to be more across the board efforts in chronic care management and coordination among all levels of caregivers. This would benefit the growing elderly population - many, but not all of whom will have multiple chronic conditions - as well as the non-elderly with chronic diseases like diabetes, and the many neuromuscular degenerative diseases like MS or rheumatoid arthritis. This type of system-wide transformation seems like a better use of resources than segmental/specialized retraining and recruitment.

What are your thoughts?

Medical information can change how clinicians treat patients, how patients care for themselves, and how healthcare payers promote or prevent the use of treatments and diagnostic tests. However, this information can act as either a broad sword or a scalpel, and produce good or bad outcomes.

A recent report from a Canadian new service about an article from the Canadian Medical Association Journal describing the outcomes from warning about the use of anti-depressants in children brings this issue down from a general concept to being very specific. This news report stated:

Two years after Health Canada warned about prescribing anti-depressants to children, the number of children and teens who died by suicide increased 25 per cent after years of steady decline, major new Canadian research shows.

And the increased suicide rate coincided with a 10-per-cent decrease in the rate of visits to doctors for the treatment of depression in children.

For the study, researchers tracked what happened in Manitoba before and after Health Canada warned in 2004 that newer antidepressants may be associated with an increased risk of “suicide-related” events in patients under 18.

They found the warning was followed by an overall 14-per-cent drop in antidepressant use among children and adolescents, fewer visits to doctors for depression, and - among eight- to 17-year-olds - increased rates of completed suicide.

More than 90 per cent of the children and teens who killed themselves were not taking antidepressants when they died.

Published Tuesday in the journal of the Canadian Medical Association Journal, the study is the first to document “such a wide range of unintended health consequences” from a major drug warning, the authors say.

Lead author Dr. Laurence Katz, a child and adolescent psychiatrist in Winnipeg, warns the increased risk of suicide could be a “random fluctuation.”

“We can’t say the warning, or the change in antidepressant use or the physician office visits caused changes in suicide rates,” says Katz.

The suicide rate among children and teens was also still relatively small, from 0.04 for every 1,000 children and adolescents before the warning, to 0.15 per 1,000 after.

But Katz worries the widely publicized drug warnings have led to more cases of untreated depression, and an impact “beyond what was intended.” The drop in doctors visits for depression suggests that some vulnerable children are getting no treatment, including psychotherapy, at all. He says his hunch is that families were afraid to go to the doctor for fear their child would be put on medication.

“But that’s not the only treatment for depression. Not going to the doctor deprives you of all forms of treatment.”

If anything, researchers expected office visits to go up after the warning was issued because physicians were urged to increase the monitoring of patients for potential adverse reactions.

Katz, an associate professor of psychiatry at the University of Manitoba, says the drug warnings and media response may have “generated a lot of fear.”

“Understandably parents who kept bringing their children, their teenagers in for troubles with depression were already struggling, and fearful (and) often appropriately cautious about whether their child or teenager should be put on a medication.”

Katz believes the findings could be applied to any Canadian jurisdiction. Other studies coming out of the U.S. are showing similar results. [Emphasis added]

The antidepressant warning involved drugs known as SSRIs, or selective serotonin re-uptake inhibitors, a class that includes Prozac, Paxil and Zoloft, as well as serotonin noradrenaline re-uptake inhibitors (SNRIs), which include Effexor. The drugs have not been approved in Canada for children, but doctors have prescribed them “off-label,” which they are legally permitted to do, to tens of thousands of toddlers, children and teens for depression, social phobia, anxiety and obsessive-compulsive disorders.

In 2003 the U.K. banned antidepressants for children. The only exception was Prozac. Studies have shown the drug is safe and effective in children.

A year later, Health Canada warned that people taking the newer-generation antidepressants may experience behaviour or emotional changes that may put them “at increased risk of self-harm or harm to others.”

Katz says he didn’t have a problem with the warnings themselves. But he says some people leaped to the assumption “that these medications lead people to kill themselves.”

[The report also notes that, “For young adults, there was no significant change in the rate of completed suicide.”]

Obviously, the outcomes found in this study are very worrisome, but it also a too dramatic example of the principle of unintended consequences.

It also reminds me of how a news story in the early 1990s about adverse reactions with the second medicine to treat AIDS.  This news report caused many AIDS patients to stop taking the medicine, and given that there was only one other medicine to treat AIDS, this certainly wasn’t a good thing for their long-term survival

Although correlations don’t prove causations, I think this study definitely underscores the importance of healthcare regulators - and their media colleagues - carefully considering how they present new health information and notices to the public - for both good findings or dire warnings. With all the proposals to empower patients to make their own decisions through consumer directed insurance plan, and to give people more health information, there should also be much more research into how people respond to health related information delivered in different forms from various sources.

Like many complicated problems in the US healthcare system, setting a goal can be easy, but achieving it can be hard. In recent years there has been a number of proposals for increasing the use of primary care clinicians to help patients with chronic diseases (like diabetes) manage their care and avoid long-term complications - and to presumably lower long-term healthcare spending. For example, the “Patient Centered Medical Home” proposal supported by a dozen organizations, (including companies and family practice and pediatrics associations), doesn’t specifically use the term “primary care,” but it gets to the same result - heightened relationships and communications between patients and a particular clinician or clinic. This .

Increasing the use of primary care clinicians is a great concept, but as has been seen in Massachusetts, increasing relationships between primary care clinicians and patients often runs into a wall when you improve insurance coverage. A recent New York Times article reported on what people here in Massachusetts have been talking about for some time: Before the start of the Massachusetts insurance expansion program there were parts of the state where it was hard to find primary care doctors taking new patients. Now that the program has expanded insurance coverage to about 340,000 more people, it has gotten really hard - particularly in Western Massachusetts. (A friend of mine moved to the most Western part of the state last summer, and almost had to go 90 minutes to Hartford, CT to find a primary care doctor.)

The options for increasing the number of primary care clinicians have both logistical and cultural challenges: Training more physicians takes many years, and as the Wall Street Journal on-line reported, until a few years ago, it had been 30 years since a new medical school granting the MD degree had opened in the US. (Several new osteopathic medical schools have opened.) Another option is retraining specialized physicians to practice primary care. This not only takes time for the retraining, but almost invariably, the potential income from practicing primary care will be less than practicing in a specialty area - which is why physicians aren’t lining up to go back into primary care.

The cultural barriers to increasing the supply of primary care clinicians involve biases that vary in different parts of the country towards osteopathic physicians (ODs DOs) and foreign born physicians, as well as against physician assistants (PAs), and nurse practitioners (NPs). Too many patients - and particularly around places like Boston - want to see an MD that trained at an Ivy league medical school and hospital. (Having graduated from an Ivy league medical school, I can tell you that’s a credential which alone is certainly not a perfect measure of a clinician’s quality.)

Another interesting proposal has been raised in the United Kingdom, where according to the Financial Times, the UK’s National Health Services wants to start having pharmacists provide some primary care services - “pharmacists prescribing drugs and administering tests for minor ailments will free up time for family doctors.”

Of course what this all comes down to is money. The UK pharmacists welcome the proposal, as long as it is backed up with funding to pay them for their increased work. And with the battle over Medicare payments to physicians looming with the 0.5% increase in the Sustainable Growth Rate set to expire at the end of June, (and then revert to the default formula’s 10.1% reduction), it would be nice if Congress (or Medicare itself) could weigh into this policy problem and adjust how they pay clinicians to increase incentives for primary care services. However, my understanding is that Medicare still pays according to the type of service (e.g. intermediate office visit), without regard to the type of physician (e.g. primary care or specialist).

So my conclusion is that while money seems to be a driving force behind the primary care clinician shortage, there needs to be some more good thinking about how to redirect money to increase the supply and availability of primary care, as well as how to use other forms of influence to address some of the cultural and logistics problems keeping the functional supply of primary care clinicians smaller than it could be.

Any ideas? Criticisms? Rants? Raves?

Information technology tools has been touted for years as a cornerstone for improving the quality of healthcare and reduce spending. While, clearly this has not been achieved, many e-health initiatives are being launched, and it is unclear how successful or efficient these will be. I am often concerned about the effect e-health systems have on the health care quality when my own physician spends so much time looking into and typing on his laptop. But to avoid discussing the challenges of e-health based upon my n of 1, below are 4 perspectives that are more expansive and analytical:

The Commonwealth Fund recently released a report about 27 state governments’ e-health activities in 2007. One remarkable thing about this report is that while there was consensus “that health information technology (HIT) can significantly improve health care delivery and quality and reduce its costs,” states are pursuing a broad range of HIT initiatives:

While Health Information Exchanges (HIEs - which enable the sharing of medical information among clinicians in an area/region), are the highest priority for most states, after that, the consensus starts to break down.

What worries me, is not the lack of activity, but the lack of consistency in implementing any of these systems. I don’t think it would be wise for everyplace to use the same medical records system, etc., but too often the “lets put in an e-health system” decision is supported with a lot of money but without enough up-front time and planning to make the new system actually provide value to patients, clinicians or the overall health system.

The Commonwealth report also has information about the barriers to accomplishing e-health priorities:

• Funding (26 states)
• Stakeholder Impacts, Buy-in (20)
• Establishing Business Case, Sustainability(13)
• Lack of Standards (12)
• Privacy/Security Concerns (11)
• Legal Issues, Technical Resources & Governance Issues (4 each)

The lessons learned by the states were equally diverse:

• Collaboration & Engagement (24 states)
• Time Required (10)
• Communication & Education (8)
• Unknown/Too Early & Funding (7 each)
• Governance, Leadership & Consumer Focus/Involvement (5 each)

Clearly Funding and Collaboration are important. (These are two aspects of health system reform that I am working on in my book project from the perspective of how social and economic forces drive the structure and operations of the US healthcare system.) The importance of the other factors in the Commonwealth report likely depend upon the culture and capabilities of the local health system and healthcare professionals.

The UK National Health Services’ “Choose and Book” system was discussed in the special e-government section in the February 16th issue of The Economist. This system is intended to enable patients and physicians to choose and make appointments for specific services. A couple of bits from this story capture its problems:

• >50% of physicians surveyed by the British Medical Association said that their experience with “Choose and Book” was negative or very negative.
• The doctor interviewed for the piece said that only about one in ten of the appointments she makes with the system turns out fine. And it is slow - consuming up to 2 minutes of a 10 minute appointment - making her a “highly paid and expensively trained doctor who is acting, in effect, as a secretary.”

The City of New York’s is launching a new initiative with $60 million to “equip doctors with computer software that can track patients’ medical records in order to provide better preventive care.” This initiative combines electronic medical records and a local health information exchange. Hopefully, the 2-year prep-time they have spent will enable this initiative to be launched smoothly. And it appears that the NYC health commissioner is setting realistic expectations, since his perspective according to the New York Times is that “medical practices will most likely see a decrease in productivity for about six months as doctors’ offices become familiar with the technology and transfer data from paper records or other systems.”

The Department of Veterans’ Affairs in September 2004 scrapped a multi-million dollar computer system when its installation at the VA hospital in Bay Pines, FL - one of the busiest in the VA system - was deemed a failure. This Core Financial and Logistics System was intended to integrate hospital’s clinical operations with its inventory, budgeting and planning programs. However, it either didn’t work as designed or the VA employees couldn’t understand how to use it - as some of the system’s proponents claimed. In either case the result was the same: Productivity dropped as the number of surgeries that could be scheduled declined, and supplies could not be found. This led to Congressional hearings and Federal criminal investigations about how the tens of millions of dollars spent on this system had been used.

The Bottom Line is that while you can spend a lot of money putting in electronic health systems, getting value from them in terms of more efficient care delivery, better prevention or more efficient financial operations depends a LOT on how the systems are designed and how user friendly they are for both patients and healthcare professionals.

Does anyone else out there have interesting experiences or perspectives on electronic health systems?

A very interesting article in yesterday’s New York Times discusses how the National Health Service (NHS) in England is clamping down on patients using both their own money and the NHS services to get treatments for the same condition at the same time. The article primarily discusses the case of a woman with breast cancer where the NHS wouldn’t pay for Avastin, and told her if she paid for it herself, she would have to pay for all her medical treatments for breast cancer.

The article also discusses the complexity and apparent confusion within the medical community and the NHS about how this policy is supposed to be implemented. But it doesn’t delve into how this policy would undermine patient-physician communications, quality of care, and the integrity of the medical record. What would happen if a patient knowing that paying on their own for treatments not covered by the NHS could cause the NHS to deny them coverage for this condition. Would they then not tell their physician about this “other” treatment they were receiving? Certainly this could affect the quality of their care and possibly lead to drug-drug interactions. And then if the physician found out about it, how would this undermine their trust in what else the patient told them? And what if the patient decided it was important to tell their physician that they were getting this other treatment, but asked that it not be recorded in their medical record? Could the physician be liable if something happened because of this - possibly because of treatments ordered by another clinician who didn’t know about the “other” treatment not included in the medical record?

This situation reminds me a little of 10+ years ago when patients were reluctant to talk to their clinicians about the complementary treatments - such as acupuncture or herbals - they were be taking. Now it it encouraged for physicians to ask, and patients to tell ,their clinicians about these non-prescription treatments because of the known (and possibly unknown) interactions that can occur with prescription treatments.

It all just makes my head shake and shoulders shudder. But at least I don’t have to wonder why the NHS is adopting this policy - it’s all about money. I guess that they have decided that if a patient can afford to pay for something expensive that the NHS hasn’t approved for them, then they can afford to pay for all of their care for that illness.

And now for the irony. Towards the end of the New York Times article it notes that “…Mrs. Hirst was told early this month that her cancer had spread and that her condition had deteriorated so much that she could have the Avastin after all — paid for by the health service. In other words, a system that forbade her to buy the medicine earlier was now saying that she was so sick she could have it at public expense.”

Any thoughts?