Investigator Advances "Desktop" Theory of Medicine

Computers, biomarkers, risk assessments and other new tools are redefining how we measure disease and the ways medical professionals diagnose and treat patients. A Robert Wood Johnson Foundation Investigator makes recommendations for adapting to this brave, new world.

    • March 10, 2011

Not so long ago—a mere 20 years—assessing symptoms at a patient’s bedside and carefully reviewing a patient’s complaints was the key to understanding a disease. Yet increasingly sophisticated computational technology, genetic research, carefully targeted drugs and other factors have permanently changed this model of medicine, according to Jason Karlawish, M.D., who won a Robert Wood Johnson Foundation (RWJF) Investigator Award in Health Policy Research in 2009.

“The change starts with a very different way of conceptualizing disease as risk,” Karlawish said. “Simple diagnostic labels no longer define disease because what we see as disease is now a series of risk factors assessed using certain—sometimes arbitrary—cutpoints along a dimension.” An expert in aging and bioethics who primarily studies dementia, Karlawish added, “in the case of Alzheimer’s disease, for example, I am witnessing how diagnosis means knowing a patient’s downstream risk of developing disabling cognitive impairment, rather than simply working with a diagnosis of Alzheimer’s disease.”

To further analyze this new approach to medical practice, Karlawish’s Investigator research project—The Making and Unmaking of Alzheimer’s Disease— asked the question: “How have concepts of risk transformed the way we think about health and aging?” To begin to answer that question, Karlawish— an associate professor and the director of the Alzheimer's disease Center’s Education, Recruitment and Retention Core at the University of Pennsylvania Department of Medicine—has been examining the changing definitions of Alzheimer’s disease.

His recent study, “Preservation of the Capacity to Appoint a Proxy Decision Maker: Implications for Dementia Research,” published in the February 2011 issue of the Archives of General Psychiatry, showed that patients who have received an Alzheimer’s diagnosis may have very different levels of cognitive function. The work assessed the ability of Alzheimer’s patients to decide whether they wanted to participate in a randomized clinical trial for a new Alzheimer’s drug or appoint a research surrogate to make a decision for them.

Karlawish and his team of five psychiatrists found that 37.7 percent (40 of 106) of those deemed incapable of consenting to a drug trial, and 54.8 percent (86 of 157) of those deemed incapable of consenting to a neurosurgical drug trial, were found capable of appointing a research proxy. Only seven of 186 (3.8 percent) were deemed capable of consenting by all five psychiatrists.

Karlawish’s work has important implications for better understanding the ethical issues related to conducting research on Alzheimer’s disease patients. “What’s happening in Alzheimer’s disease is an uncoupling of the pathology from the clinical expression of the disease,” Karlawish explained. “Researchers are starting to put forward data showing how brain scans such as amyloid imaging will indicate whether an individual will suffer cognitive problems down the road. These measures of the risk of cognitive decline are beginning to become Alzheimer’s disease.”

Karlawish asserted that this transformation is incomplete because a critical actor is missing, namely a drug that can have an impact on a patient’s progress toward cognitive decline. Drawing on the histories of cardiovascular disease and diabetes, Karlawish argued that drugs are key to transforming a disease from a bedside to a desktop diagnosis. Once a drug is created that can effectively treat a set of risk factors, those risk factors may be redefined as a disease. “In the model of desktop medicine, drugs don’t simply treat patients, they transform a measure of risk into a disease.”

The Desktop Doc

“I picked the word ‘desktop’ to describe my theory because just like the phrase ‘bedside,’ it captures where the patient and physician meet,” Karlawish said. “When I was in medical school [the 80s], we were told if you want to learn medicine, go to the bedside. Now, this education starts at the computer on the desktop, in the outpatient office or the hospital work station,” he added, referring to the plethora of data that now measures a patient’s health, and to the computer’s capacity to use that data to generate a risk profile.

Beyond Alzheimer’s disease, Karlawish suggested other examples of diseases defined at the desktop, such as osteoporosis. “This is now diagnosed not by symptoms presented at an exam, but by measuring the degree of bone thinning and then using that measure—as well as information on gender, age, smoking and other characteristics—to calculate the patient’s 10-year risk of a major fracture. Based on the risk, a person is treated for osteoporosis.”

This desktop approach, Karlawish theorized, “is boundless, constrained primarily by what defines risk and whether the cost of treating that risk is worth it. The next step is how physicians should train and practice desktop medicine, now and in the future.”

First, Karlawish said, we should ask: What type of student should choose medical school and how should they be trained? “The criteria for diagnosing osteoporosis make a clear point. Doctors now need to learn statistics and economics. Pre-medical education needs to address these proficiencies. In addition, the desktop model demands accurate, electronic medical records, and this will become even more important as an increasing number of physicians work in medical homes. In this environment, the database is the key to caring for a group of patients based on their risk factors,” he said.

While Karlawish sees most of this trend toward desktop medicine as good news for medical professionals and patients, there is one cause for concern that he is addressing—in part—in his new research. “A lot of the information hospitals and physicians will need to practice desktop medicine is increasingly patented or owned by private interests, potentially driving up costs and shaping the data to suit their needs. I think we need to rethink some of the public/private distinctions that guide who owns genetic tests, risk calculators and the biomarkers used to determine risk,” Karlawish advised.

The broad availability of medical information also “weakens the physician’s authority to some degree, but it may help patients direct what they want.” This is a plus, Karlawish said, “as long as patients do not misdirect their care.”

Research Underway–Understanding Biomarkers

At the center of the desktop revolution is the biomarker—a way of combining biology and risk to determine the possibility of illness developing in an individual. “I’m conducting new research, that may be published in 2011, that is a history of the biomarker in medicine,” Karlawish said. “The concept—which is only about 10 years old—has taken hold in desktop medicine. I am looking at when it developed and why, as well as how it relates to my work and experience with Alzheimer’s disease. There are several biomarkers under development for Alzheimer’s disease that could be helpful, but they don’t exist yet.”

“I think we need more precision in defining biomarkers as they will give us our best chance to intervene—especially in Alzheimer’s disease—with effective treatments in the future. But I will also look at the ethical and policy issues related to biomarker-based diagnoses and the associated costs,” he added.

As he considers the extensive reach of the research he is now conducting—information that will have an impact far beyond the needs of Alzheimer’s disease patients—Karlawish credits RWJF for “giving me the ability to do the enormous amount of reading, thinking and research needed to develop this work. These projects would not have happened without that support and the input I’ve received from the community of scholars I’ve met as an Investigator,” he said. “In addition to being a geriatrician, I’ve always been interested in exploring concepts of aging and disease, so this is now the direction of my work.”

The Robert Wood Johnson Foundation’s Investigator Awards in Health Policy Research provide funding to highly qualified individuals undertaking broad studies of the most challenging health, health care and health policy issues facing America. Grants are awarded to Investigators from a wide range of disciplines. Their work reflects thinking that is creative and conceptual and crosses disciplinary boundaries in search of knowledge and innovative solutions to critical health problems and policy issues.

Most Requested