Can value-based deals work if patients don't take the drug?
This feature is part of a series focused on drug adherence. To view other posts in the series, check out the spotlight page.
In the summer of 2015, two new cholesterol-lowering medicines launched to high expectations, demonstrating impressive efficacy in high-risk patients. The drugs, Repatha (evolocumab) from Amgen and Praluent (alirocumab) from Sanofi and Regeneron, were the first in a class of treatments known as PCSK9 inhibitors.
Both promised to be potent new options for patients whose cholesterol levels weren’t sufficiently controlled by statins, the standard of care for more than a decade.
But at a list price of more than $14,000 per year, Repatha and Praluent quickly ran into roadblocks with payers. Sales sputtered, and the presence of cheap generic versions of statins like Crestor limited the use of pricier options to the subset of patients at greatest risk.
Fast forward to earlier this month. Amgen, with data showing a cardioprotective benefit to Repatha in hand, has been aggressively engaging with payers and demonstrating it’s willing to experiment to prove the drug’s value.
Under an agreement with the New England-based insurer Harvard Pilgrim, for example, Amgen pledged to fully refund the cost of Repatha if a patient suffered a heart attack or stroke while taking the drug.
It is the latest example of what’s known as a value-based price contract, a relatively novel approach to linking cost to how well a drug actually works. While still not widespread in practice, payers like Cigna, Harvard Pilgrim and Express Scripts have begun to experiment more readily with these types of agreements, working with a small number of drugmakers willing to try new methods to secure favorable coverage.
For these deals to work, however, both sides have to get right one of the most persistent challenges in drug marketing: ensuring patients correctly adhere to their prescription regimens.
Necessary for enforcement
Value-based agreements can come in a number of different shapes, but all are essentially a form of risk sharing. Typically under these deals, a drugmaker will agree to a higher rebate or discount (i.e. lower net price) if a drug fails to work as advertised. If a drug performs better than expected, however, the rebate or discount given to the payer would be less, and therefore the drug would have a higher net price.
Payments are linked to an array of different health metrics — such as blood sugar for diabetes drugs or LDL-C levels for cholesterol drugs — or directly to patient outcomes, such as a heart attack or hospitalization.
But most value-based deals also feature adherence measures.
In some, it is explicit. A February agreement between Eli Lilly and Harvard Pilgrim, for example, ties the cost of Lilly’s osteoporosis drug Forteo (teriparatide) to improvements in adherence compared to a baseline.
In others, adherence is a caveat which underpins an overall focus on patient outcomes. Amgen’s deal with Harvard Pilgrim for Repatha includes a stipulation that patients must have taken the medication correctly for at least six months before a cardiac event for the drugmaker to refund the cost.
Some agreements also require adherence rates be maintained higher than a certain level for the outcomes-based component to take effect.
"There are adherence rates in many of these agreements that are necessary for them to be enforced," said Michael Sherman, chief medical officer at Harvard Pilgrim, in an interview.
Both payers and pharmaceutical companies already have an existing interest in ensuring patient adherence to prescription treatments.
Insurers don’t want to pay for preventable complications that can arise when patients don’t take their drugs. Low adherence rates across a health plan can push up costs for care substantially. For drugmakers, on the other hand, unfilled prescriptions represent lost revenues. Boosting patient adherence by 10% may be easier and less costly than finding 10% more customers.
Value-based deals can be a framework to reinforce those existing incentives and align them more directly with drug cost.
"I think it does create additional incentives for us and additional benefit for many of the things that we would do anyway, like trying to drive adherence," explained Sherman.
Over time, value-based deals could prompt payers to build out adherence programs like automatic prescription drug refills, or email/text reminders, said Sudip Parikh, a managing director for DIA Americas.
And wider use of such payment schemes could further motivate drugmakers to develop more patient-friendly (and competitive) treatment options, such as replacing injectors with oral formulations or packing multiple doses into one.
For the purposes of a value-based deal, tracking adherence is relatively straightforward in theory. Insurers use claims data to look at prescription refills, which acts as a proxy measure for gauging whether a patient is taking a drug on time.
In Lilly's deal with Harvard Pilgrim, for example, Lilly will reduce the cost of the drug if "meaningful improvements" in adherence are realized in Harvard Pilgrim's patients, compared to a baseline level. Harvard Pilgrim monitors refill rates to measure persistence and says it will work with both its pharmacy network and Lilly to boost those levels.
Calculations such as medication possession ratio, which compares the quantity of drug a patient holds to the number of days of therapy needed, are also commonly used as a metric.
But as value-based deals move out of pilot phase and include more patients, the challenge of monitoring grows more complex, especially if patient outcomes are involved.
Prescription refills can be tracked through pharmacy claims data, usually readily accessible to a payer. Outcomes like a heart attack or other health metrics, however, require medical claims data or patient records, which might not be well-integrated with pharmacy data.
Claims databases are also built to support financial transactions, so they may not give a full clinical picture for tracking patient outcomes.
Deals tying adherence to payment, though, have motivated some payers to develop more innovative approaches to identify which patients are most likely to be non-adherent.
In late 2014, Express Scripts inked a deal to make AbbVie’s then-newly approved hepatitis C drug Viekira Pak its exclusive formulary option for patients with the most common form of the liver disease. In addition to securing significant discounts on the pricey drug, the pharmacy benefit manager financially guaranteed patient adherence.
This was particularly notable since Viekira Pak's regimen initially involved taking multiple pills twice a day — a significantly greater burden than Gilead’s rival Sovaldi (sofosbuvir) that only required patients to take one pill per day.
"Adherence is a great value. We put a guarantee around it. It made us innovate."
"Adherence is a great value," said Steve Miller, chief medical officer of Express Scripts, to BioPharma Dive. "We put a guarantee around it. It made us innovate. We developed predictive models around adherence for hepatitis C."
Predictive modeling can help payers (or drugmakers, for that matter) better allocate resources to only those patients who most need support.
"If you are non-adherent because you can't afford to make your co-pay, sending you a pill box to remind you to take your medicine isn't very effective," Miller explained. "Likewise, if you are just scatter-brained and disorganized, lowering your co-pay doesn't make you more adherent."
In the end, Express Scripts was able to achieve a 92% adherence rate for patients receiving the drug through its Accredo Pharmacy using predictive modeling, and therefore paid out few of the guarantees.
Value for value?
Getting the adherence aspect of value-based pricing contracts right is just the first step, though. It still remains an open question if value-based deals can actually deliver the kind of savings and value they promise.
Against a backdrop of a broader debate on drug pricing, it is tempting to see value-based payments as a neat solution. But experts say these arrangements are more of a complement than a panacea, a way to broaden industry discussion on drug value and better understand real world efficacy.
"They are only part of the tool kit."
CMO, Harvard Pilgrim
"They are only one part of the tool kit, only one piece of a broader solution for managing pharmaceutical expenses," said Harvard Pilgrims’ Sherman.
Sky-high drug prices can also limit the impact of any value-based payments. "If the price is too high at the outset, then even if an arrangement looks good and makes sense as a way to demonstrate how a product is working, it may not be the value demonstration that it is cracked up to be," explained Dan Ollendorf, chief scientific officer at the Institute for Clinical and Economic Review (ICER).
Many of these deals have been for drugs in competitive therapeutic classes, such as Amgen’s PCSK9 inhibitor Repatha and Novartis’ Entresto (sacubitril/valsartan). Slower than expected sales and uptake may force a drugmaker’s hand and make value-based deals an attractive bet. In other areas with more limited choices, however, pharmaceutical developers have less incentive to experiment as payers are more likely to cover.
"Value-based arrangement is a good complement to other things that could be done to demonstrate and enhance product value," Ollendorf said. "But it is not a replacement."
Follow Ned Pagliarulo on Twitter