[ed. note: Michael Jorrin, who I call Doc Gumshoe, is a longtime medical writer (not a doctor) who writes for us about medicine and health a couple times a month. He has agreed to our trading and disclosure restrictions, but does not generally write directly about investment ideas. His ideas, thoughts and words are his own, and you can see all his past pieces here.]
Yesterday, the check-engine light in my Jeep came on. This, annoyingly, happens from time to time even when the vehicle is running just fine. As we all know, when the check-engine light starts blinking on and off when you’re driving, it’s a sign that something is really wrong and we shouldn’t obliviously keep going until disaster strikes, but even if it’s not blinking, we should get it checked out at “our earliest convenience,” which, for me, was this afternoon.
I drove to the establishment of the auto mechanics who have been taking care of our vehicles for about twenty years, and the owner plugged the little computer in, switched off the check-engine light, and told me that he would attend to it if came back on. This has happened quite a few times before, and the cause can be as trivial as the gas cap not being screwed on tightly. Or as non-trivial as a leak in the fuel line.
As I drove home, it struck me that having a reliable long-time auto mechanic who knows you and knows your car is something like having a long-time primary care physician who knows you and knows your state of health. And perhaps most important, “and cares about you.”
Here’s an example.
I had the same excellent primary care physician for 22 years, and, in addition to relying on his care for such ailments as afflicted me from time to time, I always had my annual complete physical with him. The routine was well established. I would pop into his office about a week before the physical and permit his nurse to draw blood for the complete blood count (CBC). Then I would return for the physical, and he would poke and press and peer according to the usual ritual. After the physical part of the physical, I would put my clothes back on and go into his office where he would review the results.
About 17 years ago, here’s what happened. First, he went over the results of the examination and the blood work – all excellent, he was glad to report. Except! One little detail in my CBC was disturbing, that being my PSA results. (PSA is prostate-specific antigen, in case you forgot – an early indicator of the possibility that a cancerous tumor is growing in the prostate gland.)
The PSA assay had been part of my regular CBC for nearly ten years at that point, and it had reliably been down in the “no worries” range. But then it had started to go up, quite slowly. And this particular year it had moved from the “no worries” range into the “worries” range, although perhaps only mild worries. It had gone from about 2.2 to 2.8 to 3.5 to 4.2 in a time span of just over three years.
My excellent doctor looked me in the eye and said, “Michael, I know you know everything. But I want you to go to a urologist and have this checked out. I’m not letting you out of my office unless you promise me you’ll do it.”
We discussed that a bit. I said that I had thought that a PSA of 10 or higher was the marker that predicted prostate cancer. He nodded in semi-agreement. “That’s right. But you don’t want to wait until you definitely have prostate cancer. The uptick in your PSA has been accelerating. If that had been your first PSA test in life, and it had been 4.2, I would have said, ‘let’s give it a year and see where it goes.’ PSA levels can fluctuate. But yours have been going up – not fluctuating.”
We talked some more. He told me that the needle biopsy that the urologist would perform was quick and essentially painless – a shot of local anaesthetic would deal with it. I might have a bit of bleeding, but that was about it. He recommended a urologist he knew very well – in fact, he was the urologist’s primary care physician.
So I agreed to go.
Indeed, the needle biopsy was nothing much. But a few days later I received the dreaded telephone call. A cancerous tumor was growing in my prostate gland, and I had to come in to the urologist’s office to discuss options.
Before the office visit, I did a bit of sleuthing on the subject (as is my wont!) and got some data on the outcomes of the available options, which included open prostatectomy, robotic prostatectomy, radioactive seed implantation, and external radiation beam treatment. I did not much like what I learned. The least invasive was the external beam procedure, but if that did not totally eliminate the cancerous mass, surgical follow-up was problematic because the area around the prostate gland would be so messed up that it would be difficult to tell healthy tissue from tissue invaded by cancer cells. According to a Mayo Clinic paper (which by that time was about seven years old) the best long-term outcomes were in patients who had had open prostatectomy.
(I should add here that when the prostate cancer is confined to the prostate gland itself, treatment is generally highly successful. But if cancer cells have escaped the prostate gland – metastasized – the outcomes are much more doubtful and treatment is much, much more difficult. And, incidentally, since PSA testing has declined, based on the dubious recommendations of the US Preventive Services Task Force, the incidence of metastatic prostate cancer has increased, almost certainly because in many untested men, the cancer is not detected until it has progressed to metastasis.)
Open prostatectomy, in fact, is exactly what the urologist recommended. (That urologist, whom I have now been seeing annually for 16 years, is now chief of urology at a major New York City hospital). The surgery took place in April 2002. I spent three nights in the hospital, came home feeling pretty spry, and have had a PSA in the undetectable range ever since. So I pronounce the whole experience a success.
Regarding the importance of having a primary care physician, consider this: prior to having settled on that particular doctor (and having stuck with him until he retired), I used to have annual complete physicals at a large New York City medical center that specialized in preventive care. No doubt, the examination was quite thorough, but never did anyone at the clinic communicate with me in person – no real feedback, no advice. I would get a very complete written report on the results of all the tests. Presumably, I was supposed to share that with my primary care physician. But I didn’t really have a primary care physician, so those results fell into a vacuum.
What prompted me launch into this topic was a piece in the New Yorker (01/21/07) by Atul Gawande, entitled “Tell Me Where It Hurts.” Atul Gawande is a surgeon, practicing at Brigham and Women’s Hospital in Boston. He is on the faculty in the Department of Health Policy and Management at the Harvard School of Public Health, and a professor of surgery at the Harvard Medical School. He is a regular contributor to the New Yorker and has written several books, including The Checklist Manifesto, which argues for checklists in hospitals, particularly relating to the essentials in a surgical procedure, and the very large impact checklists have on reducing errors in those procedures.
The piece by Gawande is mostly about incremental medicine, which essentially means persistently searching for and trying a succession of interventions to address stubborn medical problems. He described the case of a 57-year-old man who had been suffering from unimaginably severe migraine headaches for 40 years and had gone from one to another to yet another form of treatment without success, until he landed at a headache center in Massachusetts where a physician patiently worked with him, trying a range of different options, until, after four years, one particular regimen started to work. When Gawande met this man, he had not had a severe migraine for a year.
Gawande contrasted the spectacular, and highly-paid work by brilliant surgeons and interventional cardiologists with the more routine, and much less well-paid work done by primary care physicians (PCPs). He was discussing this with an associate when he “made the mistake of saying that (he, Gawande) had more opportunities to make a clear difference in people’s lives.” His associate, an internist, was having none of that and pointed him to studies showing that people living in states that have higher ratios of primary-care physicians have lower rates of general mortality, infant mortality, and mortality from specific conditions such as cardiac disease and strokes; also that individuals whose chief health-care providers are PCPs have lower five-year mortality rates than the general population overall.
A bit of sleuthing for the substantiation of the assertions by that associate of Gawande’s led me to a hefty paper by Barbara Starfield, Leiyou Shi, and James Macinko entitled “Contribution of Primary Care to Health Systems and Health,” published in Milbank Quarterly (2005;83:457-502; PMID 16202000).
The Starfield magnum opus
"reveal" emails? If not,
just click here...
Barbara Starfield (1932 – 2011), a physician and health services researcher, was University Distinguished Professor and professor of health policy and pediatrics at Johns Hopkins University. She turns out to have been among the most fervent and persuasive advocates for the essential nature of primary care in health care. She was the author of six books and 56 journal papers on the subject. Based on that, one could conclude that her views on the subject are perhaps somewhat biased, but we should look at the data she presented.
The paper starts out by attempting a definition of primary care, following the lead of both the World Health Organization and WONCA, the World Organization of Family Doctors. (The acronym WONCA is derived from the first five initials of the World Organization of National Colleges, Academies and Academic Associations of General Practitioners/Family Physicians. I can see why they needed an acronym.)
The four main features of primary care, according to WHO and WONCA, are:
- The primary care provider is the patient’s first contact access for each medical need.
- The care provided is focused on the person, and not primarily the disease; e.g., the provider is treating Irene who is experiencing shortness of breath, rather than the provider is treating a case of asthma (or it might be congestive heart failure).
- The primary care physician can provide comprehensive care for a wide range of health needs.
- If the patient’s health needs require that care be provided by a physician other than the primary care physician, the primary care physician is involved in the coordination of care.
The Starfield paper reviews a colossal amount of evidence about the effects of primary care on health. The more than six hundred references provide three types of evidence on those interrelations.
- Studies that demonstrate that health is better in areas with more PCPs.
- Studies that demonstrate that individuals who receive health care from PCPs are in better health.
- Studies that demonstrate that the specific characteristics of primary care, as defined above, are associated with better health.
Health outcomes in areas with more versus fewer PCPs
Doc Gumshoe’s eyebrows emitted signals of mild skepticism when he looked at those types of evidence. After all, aren’t “areas with more PCPs” likely to be considerably more affluent than those with fewer PCPs? And aren’t “individuals who receive health care from PCPs” also considerably more affluent than those whose first response to an illness is to look for some kind of over-the-counter medications at the corner grocery store? But let’s not be in a hurry to consign the whole matter to the dust-bin on the grounds that better health is chiefly associated with a healthier bank account. Starfield does state that the studies were carefully controlled for various population characteristic, such as income, education, and racial characteristics.
The outcomes analyzed in these studies included total and cause-specific mortality, low birth weight, and self-reported health. (Low birth weight is a very common indicator of poor health in the mother.)
Providers were considered to be PCPs if they were family and general practitioners, general internists, and general pediatricians. (I would add that being a specialist does not rule out also being a primary care provider. The crucial difference is that the PCP knows the patient and is thoroughly aware of all aspects of the patient’s health, focusing on the patient’s total well-being rather than on a single specific medical condition. For some women, for example, a gynecologist may be their “medical home.”)
Here are some findings:
Starting in the 1990s, studies of health outcomes in states in the US with higher ratios of PCPs to population found that people in those states had lower rates of all causes of mortality, including heart disease, cancer or stroke, and infant mortality as well as low birth weight, even after controlling for sociodemographic measures such as percentages of elderly, urban, and minority populations; education; income, unemployment, pollution, and lifestyle factors such as seatbelt use, obesity, and smoking.
For example, a study examining stroke mortality in 549 individuals over a period of 11 years found that the availability of primary care essentially wiped out the impact of income inequality on stroke mortality, even after controlling for education levels, unemployment, racial and ethnic composition, and urban/rural ratio (P < 0.0001).
Analyses examining the positive influence of an adequate supply of PCPs by comparing the ratio of physicians to patients in US counties also showed that all-cause mortality was 2% lower, cancer mortality was 3% lower, and heart-disease mortality was 4% lower as well as cancer and heart disease mortality were lower in counties where the supply of PCPs was greater.
In Florida counties, each 1 per 10,000 population increase in family physician was associated with a decrease in mortality from cervical cancer of 0.65 persons per 100,000 population. The math works out to suggest that a one-third increase in family physicians was associated with a 20% decrease in cervical cancer deaths.
A study in England found that each additional general practitioner per 10,000 population was associated with a 6% decrease in the standardized ratio for mortality.
Summarizing the many studies analyzing those links suggest that in the US, increasing the availability of PCPs by 1 per 10,000 population, a 12.6% increase, would result in an improvement in a number of health outcomes and prevent as many as 127,617 deaths per year.
So far we have been looking at the relationship between the supply of PCPs and health status on a population-wide basis. But what about evidence about health outcomes in individuals who mostly initially consult PCPs versus those who consult specialists?
Starfield refers to a survey in the US which showed that respondents who reported having a PCP as their usual source of care had lower subsequent five-year mortality rates. This held true after being controlled for initial differences in health status, demographic characteristics, health insurance, health perceptions, reported diagnoses, and smoking status. A likely reason for this is that individuals who initially consult PCPs are likely to do so at a much earlier stage in whatever medical condition is afflicting them, while those who rely on specialists are likely to wait longer, when the condition or disease has progressed and is thus more difficult to treat.
She also calls attention to the results achieved in Cuba from the national emphasis on providing community health workers, which was instituted by the Castro regime in the second part of the last century. Regardless of the many seriously negative results of that regime, the national health care system has been a remarkable success, such that both infant mortality and life expectancy in Cuba are now superior to those health indices in most of the rest of Latin America, and nearly on a par with those in the US.
A very recent specific example
This one came in over the transom from Canada. A study in the Canadian Medical Association Journal (Nakhla M. et al, CMAJ 2018:190;E416-E421; PMID 29320236) reported that 26.9% of young people between the ages of 1 and 17 with type 1 diabetes (T1DM) presented with diabetic ketoacidosis (DKA). DKA is a condition that develops as blood sugar mounts; it is sometimes fatal, but easily preventable. But among those who had been under treatment of a family physician or a pediatrician, these young patients were much less likely to have developed DKA – 31% less likely if their caregiver was a family physician and 38% lower if the caregiver was a pediatrician.
One of the most alarming findings from the analysis was that over 40% of the total cohort reportedly made zero visits to a usual care provider during the 2 years leading up to the diabetes diagnosis.
The study authors suggested the following reasons for this effect:
- The benefits of “continuous primary care,” where the healthcare provider is aware of the child’s baseline health and would therefore be more attuned to recognizing new symptoms;
- Prevention was achieved by parents knowing whom to call in case of illness and learning about general health through regular physician visits;
- A strong underlying relationship between the physician and patient having a beneficial effect on the family’s overall health-seeking behaviors, such as when to seek non-emergency care.
A scenario that is perhaps relevant
A NY Times Magazine piece by Siddhartha Mukherjee (whom you may know as the author of that massive tome about cancer, The Emperor of All Maladies) was focused on what he called “the epidemic of cost.” His concerns with the costs of health care are certainly legitimate, but the example he cites leaves a lot of room for skepticism.
His hypothetical case study is a 60-yar old man who has survived a heart attack. He is a non-smoker, but overweight and diabetic. He has undergone angioplasty and the placement of a stent to keep his coronary arteries open. The physician now must decide between two drugs to reduce the chances of a second MI. One is Plavix (clopidogrel), a very well-established blood thinner to prevent clot formation. The other is Brilinta (ticagrelor), a much newer blood thinner from AstraZeneca.
Brilinta has been demonstrated to be more effective than Plavix – but not a great deal more effective. The clinical trial cited in the Brilinta prescribing information (PI), which compared 9,333 post-MI subjects on Brilinta with 9,291 similar subjects on Plavix (or generic clopidogrel) reported a difference, favoring Brilinta, of 1.9% in all-cause death between the two cohorts. The Brilinta cohort experienced fewer cardiovascular deaths (2.9% vs. 4.0%) and non-fatal MIs (5.8% vs. 6.9%) than the Plavix cohort. However, Plavix/clopidogrel came out ahead with regard to strokes – 1.4% of subjects taking Brilinta experienced strokes, versus 1.1% of those taking Plavix/clopidogrel.
That nearly 2% difference in the all-cause death rate is significant, without question. But what lands Mukherjee in a quandary is the difference in cost. Plavix/clopidogrel, formerly a prized property of Bristol-Myers Squibb, having gone off patent in 2012, is now generic and pretty cheap – 25 cents per pill. Brilinta might cost up to about $6.50 per pill – 26 times more, says Mukherjee.
Mukherjee says the following: “Should the doctor prescribe the best possible medicine, assuming that the man has private health insurance that will pay the bulk of the costs? Or should she…” (Mukherjee is being politically correct here, I assume) “…try to conserve health care costs by prescribing the cheaper medicine that is nearly as good? And consider this: If the cost to you was the same – you have maxed out your copay and will end up with the same out-of-pocket expenditure – would you agree to take the slightly inferior drug to benefit the system as a whole? You’ve just had a heart attack, for God’s sake. You pay thousands of dollars for health insurance. Is it fair to ask you to bear the slightly increased risk to enable some broader social goal?”
I am not particularly dumfounded by Dr Mukherjee’s rant. From my own perspective, I would absolutely want the somewhat more effective drug. I am in good health and would like to shoot for a few more decades of entertaining the Gumshoe tribe with my musings, even if I had to shell out $6.50 per pill out of my own pocket. But that’s not the point.
The point, in the context of my rant (and not Mukherjee’s!) is that the decision –whether to opt for the more expensive pill or stick with the cheaper one – is best made through conversation between the patient and the patient’s primary care physician (whether he or she – mine at present is a she).
Mukherjee suggests that perhaps “deep biological markers” might be discovered that would identify the “few men and women who will benefit substantially from Brilinta over Plavix.” He compares that with defining the subset of individuals who would not benefit from knee replacement, perhaps by means of some algorithm that the harried orthopedic surgeon would access via his/her iPhone.
To this I say, “Spare us from algorithms.” Please, let important decisions regarding our health and, indeed, our survival, be made by human beings who know us and have concern and empathy for our health and well-being.
To be fair to Mukherjee, his piece in the NY Times mag discussed the costs of health care in the US, and what might be able to be done about it, in a much broader context. I don’t disagree with a lot of what he had to say, only with the particular suggestion to shun the more expensive newer drugs.
Here are a few more specific points:
- The study referred to by Mukherjee (not specifically cited in his article) evaluated those outcomes at the 12 month marker. (Wallentin L. “Ticagrelor vs. clopidogrel in patients with acute coronary syndromes.” N Engl J Med 2009;361:1045-1057. PMID 19717846) If the Brilinta cohort had 1.9% lower mortality at 12 months, what would that difference have amounted to after 10 or 20 years? Quite a lot higher, I should think – maybe 20% or more.
- Regarding costs, Plavix/clopidogrel was very expensive when it first came on the market back in the 1990s. It was only when it went generic that the price dropped so markedly. There are now eight generic clopidogrels on the market.
- The $6.50 per pill price quoted by Mukherjee is what a non-insured person would pay if he or she just went to the drugstore with a prescription. Health insurers make deals, and deals for individual patients are available through AstraZeneca. And the price of Brilinta will come down over time.
The PCP as our defense against “one-size-fits-all health care”
By which I mean strict adherence to guidelines. For example, whereas the chief associations of cardiologists have promoted guidelines whereby the target for high blood pressure treatment is a systolic (the higher number) BP of 130 mmHg or less, the American Academy of Family Physicians have stuck with the principle of individual care, and kept the target for hypertension treatment in persons 60 years of age or older at systolic BP of 150 mmHg. In diverging from the pronouncements of the AAC and the AHA, the AAFP reasserted the commitment of family physicians to treating the patient and not the medical condition. No doubt some patients – depending on a number of health conditions – would clearly benefit from having their BP brought to the lower target. But others would not, especially taking into account the inconvenient fact that to achieve those lower target levels in most patients, three separate drugs are required, each with its own profile of physiologic effects, including side effects.
Guidelines and algorithms are certainly of value to the harried physician. But they are poor substitutes for contact with and concern for the patient. Regarding the Brilinta vs. Plavix decision, which to Mukherjee is representative of the underlying problem with health-care costs in the US, there are some patients who would do fine on Plavix and others who would benefit from Brilinta, but an off-the-shelf algorithm is not the best way to determine which is which.
A few more general conclusions
Starfield, as well as a number of other investigators in this area, emphasize the distinction between specialists and generalists. The specialists, whether cardiologists or oncologists or orthopedists or infectious disease physicians, are more concerned with specific disease-related measures and adherence to guidelines for those diseases. The primary care or family physicians, on the other hand, are more targeted to multiple aspects of health, which could be termed “generic” health. Primary care provides superior care by focusing on the condition in the context of the patient’s other health problems or concerns.
PCPs also focus considerably more attention to prevention. For example, in states with higher ratios of PCPs to general population, smoking rates are lower, obesity rates are lower, and seatbelt use is higher. A study in 60 communities in the US found that good primary care is associated with higher percentages of smoking cessation and influenza immunization. And primary care is also associated with earlier detection of breast cancer, colon cancer, cervical cancer, and melanoma. Most mammograms are ordered by PCPs, and a study also found that each 0.1% increase in the availability of PCPs was associated with a 4% increase in the likelihood that cancers would be detected in an early (rather than late) stage.
So, in addition to improving the general health of patients through better comprehensive treatment, primary care improves preventive measures and leads to earlier detection of cancers – as it did in my case!
* * * * * * * *
Something else I was going to get to in this installment is the question of choosing the best health-care practitioner, especially those specialists (such as surgeons) with whom we don’t have much of a chance to interact prior to those high-stakes moments when we’re about to go under the knife. An excellent book (given to me by a young man well known to Travis), entitled Unaccountable, by Marty Makary MD, a surgeon on the faculty at Johns Hopkins, provides much perspective on this nervous-making question. I will review his thoughts in an upcoming epistle. Best to all, Michael Jorrin (aka Doc Gumshoe)