Why lowering drug costs can start with machine learning and better patient interaction.
While it appears there is very little most Americans can agree on when it comes to healthcare these days, one exception is that prescription drug therapies cost too much.
Whether the plan is government-sponsored or commercial, in a blue or red state, aimed at Boomers or millennials, or any other permutation, the feeling is drug prices are already high and threatening to go even higher-unless someone does something about it. By which people mean the government, despite all the risk and disruption legislation and regulation can bring.
Whether at the federal or state level, politicians on both sides of the aisle have been talking a great deal lately about finding ways to lower the cost of prescription therapies. But so far that’s all it’s been-talk. In the meantime, prescription drug prices continue to rise, and members must cover large and increasing out-of-pocket costs for the medications they need. On the member side, these high costs can lead to difficult choices-especially for Medicare and Medicaid members.
Some members will look to stretch their supply of medications by taking smaller doses or taking them every other day instead of every day. Others may forego them entirely when faced with the choice of paying for medications or purchasing other necessities such as food or rent. Still others may sacrifice their own medications so their children can get the care they need. No matter which choice they make, however, their health outcomes are negatively affected, reducing their quality of life while increasing their health plans’ long-term costs.
The good news for health plans (and their members) is they don’t have to risk a political solution to reduce the patient’s out-of-pocket cost for drugs. Instead, they can use a combination of machine-learning (ML)-based predictive analytics and hypertargeted communications to inform members about lower cost (but equally effective) alternative therapies-both generics and therapeutic equivalents-and motivate them to either make the change themselves or speak with their physicians about their options. Here’s a look at how that might work.
Determining the target audience
For this hypothetical example we will use a multi-state health plan with 250,000 covered lives including Medicare, Medicaid, and commercial members. Using pre-configured drug-switching rules, the analytics will look for members taking non-preferred products who have not previously tried (and failed) the lower-cost alternatives. The targets include members taking 200 name brand drugs which could be converted to lower cost generics, and 20 classes of high-cost drugs that could be converted to lower cost therapeutic equivalents.
These analytics yield a total cohort of 2,083 members who are presumed to be unaware that there are lower-cost alternatives to the medications they’re currently using.
Related: The Ripple Effects Of High-Priced Drugs
Getting the word out
Once the targets have been confirmed, the payer begins its efforts to inform those members about alternatives. Rather than simply sending out generic messages, however, they use the analytics to search the members’ history, demographic and sociographic profiles, etc. to give them information they can use to create hyper-targeted, personalized messages that address each member’s specific needs and concerns.
The most obvious example is a message informing a member that he/she is currently paying $X per month for a particular medication, resulting in a cost of $XX per year if taken as-prescribed. But by switching to the preferred alternative, the member can reduce that cost by $Y for a total out-of-pocket expense of $YY.
Payers can also use patterns in paid claims to detect when members are attempting to “stretch” a prescription across additional days, a strategy that can render the medication ineffective. A message can be created saying, “we understand how expensive it can be to take medication X each day, and that it can be tempting to try to make it last longer by not taking it as-prescribed. But to produce the best health results it is important for you to take your medication as prescribed. To address both needs, we can recommend this generic or alternative therapy that will deliver the same results at a substantially reduced cost, helping you gain the full benefits of your medications while fitting your budget better. Talk to your doctor about it.”
By using analytics to determine what challenges members may be facing, and then hyper-targeting the messages to show how moving to a generic or therapeutic equivalent will help members achieve their best health outcomes while addressing their financial concerns, payers can help drive the behaviors they desire more effectively.
How much can a health plan expect to save through these efforts? Based on our clients’ experience, the analytics show that of the 196 members targeted to switch from a branded drug to the generic equivalent, 45 will be converted-a rate of 23%. That yields an estimated savings per conversion of $3,392.73, or $152,673 per year.
The 1,943 candidates for conversion to a generic therapeutic equivalent (i.e., switch from one generic to a different but therapeutically equivalent generic) will typically yield a 5.5% conversion rate. Those 107 members will deliver a savings per conversion of $1,467.89 for total estimated savings of $157,064. Finally, the 907 candidates for shifting from the non-preferred brand to a branded equivalent are expected to convert at a rate of 7.4%. These 67 members will generate savings of $667.83 per conversion for a total estimated savings of $44,823.
The total net result is a blended conversion rate of 7.2%, which generates $354,560 in annual savings and a per member per year (PMPY) cost reduction of $1.42. All while making the cost of medications more affordable for their members.
Taking the lead
There is always risk that federal and state governments will act to help rein in the high cost of prescription medications, given the current political climate. That doesn’t mean payers are stuck doing nothing, however. By applying the right ML-based and other analytics to their data and using the results to send hyper-targeted messages to members, payers can take the lead in bringing skyrocketing drug prices back under control in a way that benefits the members-and their own bottom lines.
Robert Oscar, RPh, is chief executive officer, RxEOB, Richmond, Virginia. RxEOB is a leader in the health informatics and technology industry, serving millions of health plan memnbers.