Recently I've been reading Peter Singer's The Most Good You Can Do. Singer is a philosopher and bioethicist who belongs firmly to the utilitarian tradition dating back to Jeremy Bentham. "The most good you can do" is Singer's modern-day twist on the Benthamite "greatest good for the greatest number."
To Singer, the "greatest good" is human life, more particularly life lived without fear, illness, or starvation. Singer argues that we should be "effective altruists" in pursuit of this greater good. By "effective altruist" he explicitly means performing cost-benefit analyses that assign our charitable giving to those causes that demonstrably save the most lives or do the most to reduce human suffering.
Once one accepts this basic premise, certain things follow. One should support, with one's benevolence, those causes that intentionally support these goals. It is not just, says Singer, that we should prioritize anti-blindness campaigns that target childhood trachoma over art museums. Within charitable causes targeting illness, a childhood vaccination program in Sub-Saharan Africa should be prioritized over flying two children who are conjoint twins from a poor country to a rich country for expensive first-world surgery.
Singer explicitly uses the example of just such a case, performed at Stanford's Lucille Packard Children's Hospital. As I bike by Packard every morning on my way to clinic, this example is particularly close to home. It is not that it is bad to spend your money this way, per se, but for the cost of shipping the conjoint twins across the ocean, performing modern surgery, using an ICU, and all the other paraphernalia of modern medicine, one could have saved far more lives. It was not, in short "the most good you can do."
Effective altruists, says Singer, make choices, and those choices should be both economically rational and global in nature. Those who judge charitable organizations, like Charity Navigator, should not judge based on what percentage of the money they raise goes to the charitable goal, but rather based on what the goal actually is, and how effectively the money is used in pursuit of that goal. In contrast to Charity Navigator, which dislikes money spent on organizational infrastructure, Singer favors infrastructure that allows one to do "the most good you can do."
Article in The Lancet
More or less when I was finishing Singer's book, I came across a fascinating article in The Lancet on the Global Burden of Disease Study 2013 (GBD 2013). GBD 2013, funded by the Gates Foundation, attempted to bring together all available epidemiologic data to "enable comparisons of health loss over time and across causes, age-sex groups, and countries," to quote the paper (Lancet 22 August 2015;386:743-800).
The tools used by the investigators were DALYs (Disability-Adjusted Life Years) and HALE (Healthy Life Expectancy). To jump to the paper's punch line, between 1990 and 2013 life expectancy at birth rose by 6.2 years, from 65.3 years in 1990 to 71.5 years in 2013. DALY's decreased, HALE increased, and the world is a remarkably healthier place than it was just a generation ago.
But that's not the whole story. Again, to quote the paper, "For communicable, maternal, neonatal, and nutritional disorders, global DALY numbers, crude rates, and age-standardised rates have all declined between 1990 and 2013, whereas for non-communicable diseases, global DALYs have been increasing."
To put this in context, when one eliminates or significantly reduces easily preventable causes of death (infectious diarrhea) people grow old and get old people's diseases. If you die of a preventable childhood illness through lack of vaccination, you don't smoke, eat too much fat, and develop congestive heart failure five decades later.
What is astonishing about this paper is the suggestion that most of the world is now making the transition, accomplished in high-income countries over 50 years ago, to chronic disease. This transition has not come about by chance, but rather by the concerted efforts of governments, the World Health Organization, and private foundations (the Gates Foundation and a host of others).
The Gates Foundation, in spending its billions of Microsoft dollars, explicitly chose a Singerian "most good you can do" approach, focusing on infectious diseases in low- and middle-income countries, with stunning success. Their success, and the larger success celebrated by GDB 2013, is largely through the application of old, cheap technologies rather than through the development of new technologies (though the Gates Foundation has been a great supporter of vaccine research).
In addition to public health measures, the rising tide of economic success on a global level, the result of much-maligned globalization, must surely be playing some part in this story.
An Old Tension
As a physician I look at The Lancet article and say "hurrah for public health." To increase the average life span by 6.2 years in a 23-year period on a global basis is an astonishing success. Think of it: for every four years that passed, average lifespan improved by more than a year. That must be unprecedented in human history, outside of a few more limited settings, but certainly has never previously occurred on a global basis. It provides stunning support for the "effective altruism" approach.
The Gates Foundation, as I have mentioned, intentionally avoided the cancer problem as being too messy and intractable. Its support of HPV and Hepatitis B vaccines will certainly have important cancer prevention effects, but those will roll out over the decades.
As a doctor who treats cancer for a living, what is one to make of these results? Well, the reduced DALYs for acute diseases are partially offset by the increased DALYs for more chronic disease, as one would expect. Oncologists are not going out of business, nor are cardiologists or neurologists, in the Singerian utopia. We are still mortal, we still age.
On a global basis, we are in the midst of an explosion in cancer cases in Low and Middle Income Countries (LMICs), partly the consequence of people living longer and partly the consequence of picking up bad Western habits (Camels and McDonald's).
So public health measures are all well and good, but what am I supposed to do with the 37-year-old triple-negative breast cancer patient sitting in my clinic whose liver metastases just blew through front-line chemotherapy? Or the same patient, in an office in Bangladesh or Tanzania?
Singer's book is focused on charitable giving, not patient care. I have no doubt that he would consider the appropriate choice, at both the individual and health care system levels, as really no choice at all: "the most good you can do" would re-direct resources to where they could lengthen life the most, and that place is not third-line chemotherapy. The moral calculus is all on the side of vaccinating children or providing mosquito netting to their parents.
This is not a new argument, of course. The tension between public health measures (which treat populations in a cost-effective manner) and individual health care (which focuses, sometimes expensively and often for too long, on the patient in front of you) is a real one. In the United States and many other countries it is hard-wired into the payment systems and into the bureaucracies that manage health care.
There are deep historical reasons why the CDC and Medicare are dueling organizations, rather than living under one roof, but that is not my focus here. My ethical responsibility, one embodied in the Oath of Hippocrates I swore many years ago, is to the patient sitting in front of me in clinic. I am supposed to keep that patient out of harm's way, not some other doctor's patient half a world away. It is also, under the laws of the state of California, my legal responsibility.
There are many diversions in this argument. ASCO has recently and quite explicitly moved in the direction of including cost-benefit analysis in clinical decision-making, in its "Conceptual Framework to Assess the Value of Cancer Treatment Options."
Similarly, the movement towards limiting heroic measures, use of hospice care, and advanced care planning all represent attempts to avoid the unnecessary and even punitive experiences shared by dying cancer patients.
But these are arguably not Singer-like "most good you can do" prescriptions. Rather, they all live comfortably within a context Hippocrates would have understood, all focused on the patient sitting in front of you: don't subject your patient to drugs that will bankrupt his family and usually won't work. Don't subject your patient to diagnostic tests that have little value and some real risk. Make sure that your patient has the opportunity for a good death, surrounded by those she loves, rather than a miserable wasting away, intubated and in pain in an ICU.
What is the Most Good We Can Do?
But there is, in addition to the doc who cares for Mrs. Smith, another part of me, a part that resonates with "the most good you can do" argument. And at a global level, in cancer, we have a pretty good idea of what constitutes "the most good you can do."
If I speak just of breast cancer (and I can speak intelligently of little else) it is pretty clear that fairly simple local-regional control measures would save more lives than any other intervention.
I say "fairly simple," though there is nothing simple about it: an operating room presumes the steady flow of electricity, of teams of nurses and surgeons and anesthesiologists, of pathologists to read the slides, and radiation oncologists with not inexpensive machinery. Many of these, taken for granted in high-income countries, are simply unavailable to most in low-income countries.
After surgery, one could make the case for flooding the world with tamoxifen, and perhaps-perhaps-1980s-style off-patent chemotherapy, presupposing one could test for estrogen receptor. Tamoxifen has probably saved more lives on a global basis than any other cancer drug.
But past that point things get very expensive and the gains progressively smaller. HER2-targeted adjuvant therapy almost makes the cut from an efficacy standpoint, but is totally out of reach for the vast majority of the world's HER2-positive patients from a financial standpoint. And forget mTOR inhibition and CDK 4/6 inhibition and second-generation HER2-targeting, whose expenses and unproven benefits in the adjuvant setting all would fail "the most good you can do" sniff test.
Cost-Benefit Analysis
Recently some of my Stanford colleagues published a cost-benefit analysis in the Journal of Clinical Oncology (Durkee BY et al. J Clin Oncol 33, 2015, epub ahead of print Sep 8 2015) looking at the addition of pertuzumab to a standard taxanes+trastuzumab combination.
This three-drug combination is a big deal in metastatic HER2-positive breast cancer, bringing median survival from 40 months up to about 56 months. We seldom see survival improving by more than a year in the metastatic setting, in breast cancer or any other disease. But the price paid for the improved survival comes in at an astonishing $713,219 per quality-adjusted life year.
Things change, of course: drugs go off patent and get cheaper, and countries that were previously impoverished undergo economic takeoff and create robust health care systems, its citizens now sufficiently well off to obtain that which was previously denied. This is certainly happening in many places.
Advocacy can play a role as well. Singer, in discussing the British charity Oxfam, notes that some of its most effective work has involved advocacy, specifically influencing governments in low- and middle-income countries to "do the right thing" on behalf of its citizens.
Drug Patents
The obvious advocacy issue in oncology involves drug patents. If a drug is highly effective at saving lives, but is out of reach for large proportions of a population, then the moral calculus favors eliminating or significantly limiting patent rights, as occurred in Sub-Saharan Africa with HAART for AIDS/HIV.
Such advocacy requires an "on the ground" understanding of how the health care system works, as well as a sophisticated and nuanced understanding of the benefits of therapeutic interventions. Revoking the patent for an mTOR inhibitor in Zimbabwe would not pass the sniff test, for many reasons.
But perhaps for trastuzumab? Maybe, maybe not: I would love to see a cost-benefit analysis on this. Administering a year of Herceptin involves more than just the cost of the drug. It require effective and somewhat complex infrastructure.
"The most good you can do" is a concept that many of us will not be totally comfortable with. I'm certainly not, my Hippocratic tradition still warring with my cost-benefit public health instincts. But certainly we can start a dialogue on what it would involve, both at home and abroad.