Predictive Analytics Case Study: Ian Duncan
Actuaries are risk averse by nature. In order to keep up with the transformative nature of today's healthcare environment SOA Fellow Ian Duncan encourages actuaries to become entrepreneurs and hands-on business leaders.
The Affordable Care Act represents career-expanding opportunities for actuaries and a chance to have a more direct impact on the delivery of healthcare.
"You have to come out of the back room and take off the green eyeshade and take a role in managing the company," said Duncan, whose actuarial training has propelled him on a 30-year journey through consulting, pharmaceuticals, entrepreneurship and academia. Duncan, FSA, MAAA, FCIA, FCA, FIA, is a professor at the University of California, Santa Barbara.
Duncan has maintained a special focus on using predictive analytics to help healthcare companies provide higher quality care without sacrificing their bottom lines.
His recent work has focused on helping Accountable Care Organizations (ACOs) and hospice providers identify the most financially taxing end-of-life treatment regimens through early identification of cases likely to result in long-term hospitalization or over medicalization. Duncan and his team of actuaries and palliative care clinicians then establish interventions to reverse the trend.
Duncan acknowledges that it is a morbid and difficult topic, especially for the average physician who is trying to keep people alive. Only an actuary could get excited talking about this sensitive end-of-life model, Duncan said.
The first ACO he shopped the model to wasn't buying it. But Duncan eventually met a businessman who runs a hospice and is used to dealing with patients and families who have palliative care needs.
"What it tells me is you can develop a wonderful predictive model, but if no one is going to use it or is interested in it, well that's it. The model and business have to work together," Duncan said.
"What I've tried to get people to understand is one model doesn't fit all. You need to have an understanding of what your business problem is. You need a solution. And then you can develop a model that is specific to that need. An off-the-shelf model doesn't work in this business anymore."
When Duncan was the VP for Clinical Outcomes & Analytics at Walgreens the company constantly tested hypotheses and piloted programs to see what would work. One theory Walgreens wanted to test was whether it could identify pharmacy users who might benefit from some sort of intervention from pharmacists. A long-standing problem for pharmacy companies and manufacturers is patients not sticking to their meds.
While pharmacies can't measure how many pills customers take, because they can't observe users, they can measure how many prescriptions individuals fill. Nationally, people only fill about half their prescriptions, Duncan said. So Walgreens recognized that the general health of people with chronic diseases would be significantly improved if pharmacists could help move the needle.
The company's original hypothesis was that if the pharmacist got to know the patient and offered medical advice that would affect the outcome. But Walgreens found that the most successful interventions (in terms of volume of patients and return on investment) were often automated - when customers got a call reminding them it was time to refill their prescription. One reason it was successful was because it was "basically costless," Duncan said. "And even if it just improved adherence slightly, the return on that program was more significant than programs that involved human intervention."
As actuaries continue to strengthen their position in the field of healthcare predictive modeling it is important to remember there is not a single answer to every question. It's a tough lesson to learn, Duncan said, especially after years of actuarial education and training that emphasize there is a "right" answer.
If actuaries can find comfort in the unknown and risk taking, Duncan believes they will learn more than they would sitting at a desk cranking out reserves, they will also add value to healthcare companies and the lives of the patients they serve.