The process of choosing a disease treatment option can often overwhelm a patient already dealing with the shock of a diagnosis. Health economist and biologist Christopher Jones, Ph.D., University of Vermont assistant professor of surgery and director of the Global Health Economics Unit based in the Center for Clinical and Translational Science, is obsessed with efficiency. Together with a collaborative group of UVM colleagues, he has invented an algorithm that gives patients a helping hand in determining which treatment is best for their personal circumstances.

Jones discussed this work at a Center for Research on Vermont seminar, titled “Made In Vermont: The Use of Vermont Medical Data to Improve Health Care Decision-Making,” on April 15.

The research is ongoing. A recent paper by Jones and colleagues, including vascular surgeon Andrew Stanley, M.D., associate professor of surgery, in the Journal of Health Economics and Outcomes Research examines the use of their algorithm – a point-of-care cost analysis (POCCA), which will be housed in a much larger library of algorithms called ForMyOdds.com, a Vermont-based spin-off company he created – in determining the best treatment for an unruptured aortic aneurysm. He and his coauthors on the paper note the importance of weighing the benefit of the intervention against the expense.

“ForMyOdds is all about using data to be proactive – to personalize care to patients’ unique circumstances,” says Jones. “Euclid brought us ratios around 280 B.C., but we can now use a similar approach to tell a patient – to a very granular level of detail – their best options in light of cost per unit of clinical benefit. This is based on the costs and benefits others have experienced who have walked in their shoes.”

In the Journal of Health Economics and Outcomes Research study, which reports on the development of a predictive model, the researchers used Fletcher Allen Health Care-specific patient and cost data from the Vascular Study Group of Northern New England. Under the guidance of Stanley, the group retrospectively analyzed data from 389 elective aortic abdominal aneurysm repairs (230 endovascular repairs and 159 open surgical repairs) that took place between 2003 and 2011. According to the study authors, this group prospectively collects a wide range of important information on preoperative, intraoperative and postoperative elements of patient care, such as gender, age, BMI, disease status and history, and medications.

Jones and colleagues examined which clinical characteristics contributed to the most (and least) expensive treatment course, assuming the clinical outcomes of the two repair modalities were both good. They compared demographic and preoperative characteristics of patients in relation to overall costs. Their analysis found an association between high cost and patients with preoperative risk factors that had not previously been reported, at least in their combination – age, hospital transfer status, and history of prior cardiac bypass surgery and chronic obstructive pulmonary disorder (COPD).

The research team concluded that “certain risk factors at the individual patient level are predictive of higher health care costs for these procedures. Under such circumstances, it is our expectation that such algorithms may be used to select the most cost-efficient treatment.”

In a grant from the UVM Department of Surgery, Jones has worked with UVM Professors Christopher Koliba Ph.D., and Asim Zia, Ph.D., to develop a system dynamics tool for estimating when patients should undergo dialysis versus transplantation.

Jones is also working with surgical oncologist Ted James, M.D., and emergency physician Peter Weimersheimer, M.D., on a mobile incentives tool to better engage patients towards meeting goals that are good for them, outside the hospital. That same team is building an algorithm that identifies not only the most efficient incentive for a given patient to meet certain clinical goals, but monitors how that algorithm can change with behavior, genetics, medicines, treatments, and over time.

But will insurers share this vision of offering data-driven decision tools and mobile incentives?

“Some are already starting to see that use of these tools will lead to healthier patients with more purchasing power, who in turn become the beneficiaries of more affordable and better coverage,” says Jones.

As one of several investigators funded through the Vermont Center on Behavior and Health’s Center of Biomedical Research Excellence award led by Professor of Psychiatry Stephen Higgins, Ph.D., a major effort over the next five years will be incentivizing patients to adopt healthier behaviors. Using the ForMyOdds platform may be a major step in the wider direction of making incentives fun and rewarding to the end user. 

As Jones points out, our millions of years of evolution placed enormous selective pressures on our genes and the ways in which we behave. “Most of the things we find to be fun are practice in a low stakes environment for what our genes have given us as preparations for action in a high stakes, or life threatening environment. That is why we like sports (vs. war), high calorie foods with sugar (vs. starvation), and money (vs. inability to support kin). Thus we may find that incentives work best when offered in fun environments rather than elaborate hospital settings. “We may want to think about Vermont-specific prizes like vacations, coffee, teddy bears and maple syrup,” Jones says.

(Christopher Jones, Ph.D., contributed to this article.)