By Dr. Bradley Eilerman and Len Testa
Part 1: Complexity
This is the first in a series of articles on a framework we’ve developed for recommending medications for treating patients with type 2 diabetes. We begin by defining the scope of the problem that physicians face when making these choices. Future articles will explain how to incorporate patient cost, side-effect trade-offs and other decisions that need to be made. For our last article, we will introduce our decision-support software, which automates much of this decision-making.
A thousand times today, in offices all over America, hospital patients will be diagnosed with type 2 diabetes. When that happens, a healthcare provider has to make a treatment decision of enormous complexity, often with partial information to go on, and in the span of just a few minutes.
Consider that in this typical office visit, there are at least 60 different single drug treatment options that the healthcare provider can prescribe to bring that patient’s blood glucose to goal. These options include at least 50 branded medicines approved by the FDA since 1992 and generics such as metformin. Most healthcare providers will also recommend diet and exercise as therapies, too.
A healthcare provider will typically prescribe 1 to 5 of these treatments, based on the patient’s specific condition. There are almost 6 million different combinations of “1 to 5” medicines when choosing from a list of 60. (If you’re interested in the math, it’s an “n choose k” problem where n=60 and k=1, 2, 3, 4, and 5)
Many of those combinations don’t make sense, of course, such as those that use a rapid-acting insulin without a basal insulin, or those with a medicine that is contraindicated by a patient comorbidity. But eliminating every obviously invalid option can still leave a healthcare provider with anywhere from a few thousand to over a million possible therapies to consider.
But choosing the right medication from millions of options is just one of the healthcare provider’s challenges. Another problem appears when the healthcare provider has to consider the patient’s out-of-pocket cost for these treatments. After all, a patient won’t take medicines they can’t afford.
There are more than 40,000 individual health insurance plans offered in the United States, each with their own drug coverage and prices. So, in addition to sorting through millions of therapy combinations, a healthcare provider has to know how much each medicine costs for the patient’s insurance plan. And if the patient’s prescription coverage comes from Medicare Part D, the healthcare provider has to consider that plan’s coverage gap, commonly known as the “donut hole,” in the context of the patient’s budget and treatment.
No other industry asks its professionals to solve problems this big without technology to help. And yet we expect healthcare provider to do it, usually in the middle of a 15-minute office visit. Why? More importantly, what can we do to help?
One way to help healthcare providers deal with this complexity is the use of professional treatment guidelines. Among the most popular is the American Association of Clinical Endocrinologists/America College of Endocrinology Comprehensive Type 2 Diabetes Management Algorithm (“the AACE algorithm”). First published in 2013, the AACE algorithm is a 10-page guide with flowchart-based treatment options. These options include eleven classes of anti-diabetic medications, plus diet, exercise, and surgery options.
The AACE algorithm enables healthcare providers to customize therapies that include more than just lower blood glucose levels. Medicines that show benefits such as weight loss, cardiovascular advantages, or reduced risk of hypoglycemia are favored more than medicines without those properties. In addition, the AACE algorithm also seems to favor aggressive use of combination therapies for patients, potentially bringing them to goal faster than step therapies.
However, we believe the AACE algorithm has limitations that affect its ability to identify promising treatment options. For example, the AACE algorithm doesn’t ask the patient how much they can afford per month for their medication. But patients won’t take medicines they can’t afford.
Even if it asked the patient for a monthly budget, the algorithm has no direct way to incorporate this information into its decision, because it has no mechanism to use the details of the patient’s insurance plan.
We also believe it’s difficult to use the AACE algorithm when discussing a set of treatment options along a continuum of cost, side-effect, and efficacy with the patient. For example, we believe current guidelines are not easily used to answer patient-centered questions such as:
- Is an additional 0.3 reduction in blood glucose levels worth $30 per month to you?
- Is an extra $40 per month in medication cost acceptable to you for a medicine that may lead to weight loss?
All of this complex decision-making is pushed back on to the healthcare provider to handle. It’s not realistic to expect consistently good outcomes in these situations. A different approach is needed.
In the remainder of this series, we’ll describe a software program we’ve written to help physicians treat patients with type 2 diabetes. It evaluates all possible combinations of medications against a standard set of rules, patient insurance plan, and budget to offer a set of therapies that both the patient and physician agree best fits the patient’s goals.
Our software program was awarded 1st place from the American Association of Clinical Endocrinologists annual conference and poster competition.
Next Week: Part 2 — “Cost Vs. Outcomes”
Bradley Eilerman MD MHI obtained his medical degree from the University of Kentucky in 2001. He completed his internship at Vanderbilt University Medical Center in 2001 and his residency in Internal Medicine/Pediatrics at the University of Cincinnati in 2005. Dr. Eilerman went on to complete a fellowship in Diabetes, Endocrinology and Metabolism in 2008. He finished his Master’s in Health Informatics in 2016. Dr. Ellerman works as the lead physician, director of clinical research, and clinical endocrinologist for the St. Elizabeth Physician’s group where he cares for a patient base of approximately 2,500. He is licensed to practice medicine in the state of Kentucky.
Len Testa is a computer scientist at GlucosePATH.com, focusing on the combinatorial optimization engine and architecture. Len’s other work includes optimization software that has helped millions of people minimize their waits in line at Disney World. (He’s happy to talk about that, too.) That said, it’s widely acknowledged that Len is the “pretty face” of a much more talented group of people, and that “pretty face” is charitable, at best.