A blueprint for high-volume, high-quality lung cancer screening that is detecting cancer earlier—and helping to save lives
Are you dreaming of a world where mass cancer screenings are discarded as wasteful and personal genetics and ‘big data’ inform individually targeted therapies? Wake up — that world is fast approaching.
It may seem paradoxical for the dean of the Jefferson School of Population Health to say this, but perhaps we do not need population data so much as better data at the individual level.
It is true that today we have FDA-mandated testing on populations, but we also have a widespread lack of an evidentiary basis for treatments for patients with hypertension, congestive heart failure, hyperlipidemia, and other serious disorders. Based on extensive data from thousands of people worldwide, clinicians can only take a guess at what patients’ therapies should be. This is incredibly expensive, not targeted, and very wasteful — and it probably leads to inappropriate utilization.
So there is a paradox in population health. How can we reconcile the challenge of personalizing medicine — paying for costly drugs such as Gleevec and Herceptin, along with other therapies that we do not know about yet — with the challenge of caring for populations, and with health care reform?
Let’s start by defining the term “population health.” One of the current challenges for both integrated delivery systems, such as the Geisinger Health System, and academic medical centers, such as the Jefferson School of Population Health, is that we are accountable to the public for our degree of success. If we ignore the complex population and social accountability issues in health care, we ignore them at our peril.
That is a big part of what population health is all about. Dr. Marc Williams has spoken of the concept of going “from volume to value.” Traditionally, at our university hospital, growth has been all about “more” — more people, more admissions, and more procedures. But as many of us realize, we are heading toward a very different world, a world that asks, “Why are we doing that test in the first place? How do we achieve greater value — better outcomes — for the money that we are spending?”
Instead of paying for “performance,” we should be asking about the results of the tasks and procedures that we are performing. This is a heavy cultural lift for many health care providers.
Another aspect of the cultural challenge has to do with what the media describe as “medical guesswork.” From heart surgery to prostate care, medicine often does not have a clear roadmap to follow. From an evidentiary standpoint, grade A, randomized-control-trial efforts inform decisions made at the bedside only about 20 percent of the time. As clinicians know, the other 80 percent of decisions represent the art of practicing medicine.
Now that the genome has been decoded, we may soon be able to achieve greater value at an individual level. Part of the challenge we face is getting to that place. Thus, the name of our institution — the Jefferson School of Population Health — sounded pretty corny four years ago. Now it seems like good thinking, because we are moving toward thinking about achieving value for a population.
The most widely accepted definition of population medicine includes three components:
What are the health outcomes and how are they distributed in the population? Here we look at typical measures of morbidity and mortality, with quality of life (QOL) being critically important.
What are the determinants that influence this distribution? At first glance, medical care would appear to be the cornerstone, but it is only about 15 percent of the QOL in any society — the actual “laying on of hands.” The other 85 percent of outcomes distribution consists of socioeconomic status, genetics, environment, social aspects, and, of course, “choosing your parents wisely.”
What are the policies and interventions that affect these determinants? Here we look at social, environmental, and individual factors.
Two classic studies 40 years apart in the New England Journal of Medicine — one in 1961 and the other in 2001 — answered essentially the same question. Of 1,000 people in the population, only one of them is admitted to a hospital each month. So where should we focus our efforts — on the 999 or on the one? The way we are currently organized, structured, and paid, almost everything is focused on that one hospital admission. Can we improve the health of the 999 if we figure out a way to make more precise care decisions for them?
Almost every sector of our economy has shown improvements in labor productivity during the last 20 years — but not health care. That might seem surprising, considering all of our scientific advances and the fact that 17 percent of the gross domestic product goes toward health care. But improving productivity often requires redesigning the delivery model. For health care, that means having more primary care physicians and fewer specialists, and engaging in higher-value activities.
Improving productivity requires delivery model redesign, different provider quantity and mix, and engaging in a much higher value set of activities. This is essential for controlling health care costs.
Source: Kocher & Sahni, Rethinking Health Care Labor, N Engl J Med. 2011;365:1370–1372
Dr. Eric Topol has written an excellent book, The Creative Destruction of Medicine: How the Digital Revolution Will Create Better Health Care. Dr. Topol’s thesis is that a new type of medicine will emerge from what he calls the “super convergence” of genomics, mobile connectivity, social networking, and new information systems. A rapid series of developments — from the invention of the cell phone in 1975 through personal computers, the rise of the Internet, the sequencing of the human genome, and the proliferation of social networks — has combined to create what Dr. Topol calls “the great inflection of medicine,” which will accelerate technology and give us a whole new type of medical practice within the next five years. Today, we are at the edge of this important breakthrough.
Source: Jefferson School of Population Health
Dr. Topol’s argument raises a few questions. Will these trends lead — as a front-page article in the Wall Street Journal has suggested — to gene sequencing becoming part of routine physician visits? As a general internist, I look forward to the day when a patient can present with high blood pressure and we can take a buccal smear, put it in the sequencer, and know within 15 minutes which antihypertensive medication the patient needs. That is an example of individualized medicine.
The private sector is betting billions of dollars that our ability to do that buccal smear and to select that drug in the office is less than a decade away. Analysts believe that it is going to be here by 2020 at the latest.
A major lobbying organization called the Personalized Medicine Coalition is fighting in Washington to get the Gleevecs and the Herceptins covered by insurance companies and to make sure they are still covered under the Affordable Care Act.
Dr. Topol speaks of the “democratization of information” — meaning that every person could have access to his or her genome. Some day this information could be put on everyone’s smart phone, and they would be able to visit their physician and have their genomic history at the touch of a finger. The physician would be able to analyze and use this information in a way that was never possible because of the power of cloud servers. Dr. Topol calls this “digitizing humans.”
Imagine a world where we won’t need mass screenings for breast cancer — where the notion of performing screening mammographies on a mass scale is seen as being wasteful and as leading to unnecessary additional tests and biopsies. According to Dr. Topol, this is possible because of our ability to dissect, decode, and define individual breast tissue granularity at the molecular level “from womb to tomb.”
The punch line in Dr. Topol’s book is that we will be able to deliver much more specific information on an individual-patient basis, empower patients and providers with this information, reduce waste, save money, and achieve better outcomes in the end. That sounds like an enticing and compelling future. But how does it connect to personalized medicine?
Clinical trials actually represent population data. What we need is real-world data at the level of the individual patient, and the only way we are going to get that is by eliminating mass screenings and treatments that promote waste. This is already happening to some extent.
If you visit Web sites that are promoting crowdsourcing, such as Patients Like Me, you will find that patients with certain diseases are sharing a variety of information — even about clinical trials that they are involved in.
Dr. Topol’s conclusion is that “the median of the bell curve” that we get from clinical trials is really not the message. The message is that individual patients have a much more detailed idea of their genetic makeup. Therefore, we won’t have to worry about false-positive mammograms and about all the women we send for biopsies. We will be able to think about which patient will benefit from Plavix, and we will also be able to know a priori who should receive tPA (tissue plasminogen activator) versus who should be treated with streptokinase.
Widespread PSA testing for prostate cancer is another clinical example of the cultural challenge we will face in the transition from widespread population-based screening to a future of individualized, genetically based screening and therapy.
Patients are being encouraged to oppose U.S. government action to eliminate prostate cancer screening for all men — when the U.S. Preventive Services Task Force has said that there is no evidence to support widespread PSA testing.
I think the message is that certain cultural hurdles need to be overcome in the period between a mass-screening, population-based paradigm and the individual “digitized” patient.
There is a full-page ad in AARP magazine that reads: “The power to fight advanced prostate cancer is already in you. Turn it on.” It is an ad for the powerful new drug Provenge, which was recently covered by Medicare. This drug costs $33,000 per dose, and patients need three doses. It is indicated only for men with metastatic prostate cancer that is resistant to all other therapies. According to FDA data, the drug prolongs life, on average, by only four to six months.
Without genetic testing of prostate cancer cells, a lot of money is going to be spent on Provenge until we can figure out who should be treated with this drug.
Are we ready to enter this world of the “creative destruction” of medicine? Are we ready to practice individualized care? Those are exciting questions from a primary care physician’s perspective.
Primary Care 2025, a report supported by the Kresge Foundation and developed by the Institute for Alternative Futures, looks at what medical care might be like in the near future. The report envisions patients taking charge of their own health care.
They will choose self-care and buy health-related products. They will have a higher demand for primary care providers. They will be armed with their own genetic information, and will know whether they need a PSA test or a mammogram.
Humana has said, in effect: “Wait a minute — we do not want to pay for Provenge. We do not want to pay for unnecessary tests. That does not make any sense from an economic perspective. We are going to try to keep patients away from centers that use expensive drugs and perform expensive tests. And we are going to do that by personalizing their care.”
Humana hopes to provide incentives that will connect people with their primary care physicians so that they will be compliant with diet and exercise and will avoid developing diabetes or end-stage organ damage. To that end, Humana has purchased several companies, such as Life Safe, Hummingbird, and Silver Sneakers, and is building Humana health centers around the country. Companies such as Humana are at the edge of practicing personalized medicine. They have the economic incentive to do so.
Similarly, a peer-reviewed journal called Games for Health is trying to figure out what kinds of “fun” tools and technology we need to engage patients better in their own health.
Some statistics are relevant in this regard. We know that global telecommunication devices will outnumber the world’s population by the end of 2013, and already more people across the globe own a phone than own a toothbrush. By 2020, at least 160 million Americans will be monitored and treated remotely for chronic conditions. According to a survey conducted two years ago, 88 percent of physicians would like their patients to track or monitor their health at home.
Using a Web-enabled cell phone and a special software program, U.S. Preventive Medicine has developed a health and fitness app called Macaw after the African parrot that has a long life span. The company feels that this is an effective channel to deliver behavioral messaging and interventions to large, targeted groups at a lower cost. The program is delivered as a stand-alone application on a person’s Web-enabled phone using an integrated wireless device. By means of this device, a health plan or even an employer can keep track of whether a patient is compliant with his or her medications or recommended exercise.
One health plan, such as Humana, and one app, such as Macaw, probably will not change the world. But it seems that while not everyone knows his or her genome, almost everyone has a Web-enabled cell phone.
If we could figure out how to individualize chronic therapies — determining who is at high risk for diabetes, for example — the implications of these tools from a population health perspective are striking.
Seventy-nine million Americans have prediabetes. Twenty-four million of these people go on to develop diabetes, seven million are undiagnosed, 17 million are diagnosed, and four million are not treated. Thirteen million Americans with prediabetes are treated, but 7.8 million of these are not treated successfully.
Sources: NIH, CDC
It turns out that of the 79 million people potentially at risk, only 5 million have their disease under control, and 18 million have diabetes that is not controlled. Of course, it is the 18 million uncontrolled diabetics who end up in hospitals with myocardial infarction, blindness, amputations, renal failure, and all the other end-organ damage that results from diabetes.
How do we connect the dots on this notion of risk? We can begin by figuring out who will progress to uncontrolled diabetes. We could probably do that with a deeper understanding of people’s genetic makeup. After we find the individuals at risk for progression, we could do a better job of organizing and managing their care with individualized tools, such as web-enabled phones. With respect to chronic diseases, some of our biggest opportunities to contain costs — or at least slow the rate of growth — lie in our ability to encourage behavioral changes.
A recent issue of the Harvard Business Review discussed the concept of “big data.” Without big data, we could not have decoded the human genome. According to the article, we are going to be at a whole new level of big data in another five years.
Imagine all those web-enabled telephones and everyone connected with information about taking their medicine, doing their exercise, getting sufficient REM sleep, and keeping track of their daily weight. From this behavioral “big data,” we are going to learn a lot about how to influence patients at the individual level.
But again we have a paradox. Population health at the “big data” level will involve learning all of that input for thousands, perhaps millions, of people with their web-enabled cell phones tracking their own health care. From that, better information will emerge.
It may go something like this: A patient with diabetes presents at my primary care office, and I say, “Michael, based on ‘big data’ that we have on patients such as yourself, here is a customized prescription for your weight-loss program that we know will work. And based on ‘big data’ from a thousand diabetic patients in our practice who have been giving us information at your stage of disease, here is something we know will work for you.”
So I am impressed by the idea of harnessing “big data” at the individual patient level. And I am not the only one who is thinking about this. Humana and many other insurance companies — but no hospitals, medical schools, or research centers — are collecting data from millions of Medicare enrollees around the country. This effort is being driven by the for-profit health care sector — by organizations such as Walgreens, Healthways, WellPoint, Aetna, Cigna, and Johnson & Johnson. The private sector, for better or for worse, has begun to move toward harnessing population-based “big data” and organizing this information at the individual patient level — and, in a way, delivering individualized care. Clearly, personalized, genetically based care is the next step.
Not too long from now, medicine is going to reconcile population-based care — clinical trials involving many thousands of patients worldwide — with individualized care based on decoding the human genome. However, the political journey from population-based medicine to individualized care is going to require a major cultural change. This change has several components.
First, we must develop a better evidentiary basis for what clinicians do every day at the bedside. Another component is better electronic systems that will provide us with all of the data we require on all of the patients that we treat.
In other words, we need better patient data in the form of a registry. Finally, we are going to have to change the educational paradigm as well. We need to change the model of treating one patient at a time, addressing one problem at a time, and guessing what would be the best drug and the best therapy based on limited clinical experience. And that change is coming.
A blueprint for high-volume, high-quality lung cancer screening that is detecting cancer earlier—and helping to save lives
Multiple Sclerosis: New Perspectives on the Patient Journey–2019 Update
Summary of an Actuarial Analysis and Report