Merck offered an unusual proposition to consumers and insurers in 1998 as part of an effort to regain market share for its statin Zocor, which had been losing ground to then Warner Lambert’s Lipitor. The drugmaker’s “Get to Goal” guarantee offered refunds to any takers who failed to reach target cholesterol levels set by their doctor within six months of using Zocor and adjusting their diet.

The campaign, while focused more on consumers than payers, is viewed as the earliest example of the types of outcomes-based pricing agreements that are becoming more popular as payers seek to control costs and get value for their money.

As the world of healthcare moves away from fee-for-service payments to paying for outcomes, its seems only natural that drugs, particularly high-priced ones that work in select patient populations, would face efforts to create the same type of alignment between payment and results. At the heart of these efforts lie simple realities for both payers and drugmakers. Payers want to control costs. Drugmakers want to assure access. To do so, they are willing to share both risk and reward with payers.

Driving the expansion of pay-for-performance or outcomes-based pricing agreements are a number of factors, according to Josh Carlson, assistant professor in the Department of Pharmacy at the University of Washington, who studies the issue. These include the limited amount of evidence about new drugs when they first come to market (an issue intensified by accelerated drug approvals), increasing cost pressure from payers, the use of so-called reference pricing that pins what some countries pay for a drug to the price paid in specific reference countries, and the rise of health technology assessment programs and other policies intended to control the use of new technologies.

The performance-based pricing also serves a simpler purpose for drugmakers. It allows them to provide discounts that may be necessary to establish acceptable value in one market without affecting the price for a drug in other markets around the world as a number of payers peg the price they will pay for a drug to what price a specific country may negotiate with the drugmaker.

“These agreements are still playing out. There’s a lot still going on. They are very attractive, but they are hard to implement, so there’s a tension there and that’s not fully resolved,” says Carlson. “Whether they have real long-term viability is yet to be seen. There’s a lot of interest in them. They are expanding out from initial players to different health systems to more developing countries.”

In 2009, Cigna announced an agreement that it described as unique within the industry. The insurer crafted an agreement with Merck for its oral diabetes drugs Januvia and Janumet to be used in a plan with diet and exercise to control blood sugar levels in adults with type 2 diabetes. Under the agreement, Merck offered rebates for its diabetes drugs to Cigna and its patients if they reached agreed upon adherence levels and lowered their blood sugar. If patients taking the drugs lower their blood sugar levels further, they received additional rebates. The plan provided lower costs to patients and Cigna while increasing sales for Merck by improving adherence and raising their status on the formulary.

At the end of 2010 when Cigna first reported on the success of the program it said the 165,000 members using the Merck diabetes drugs lowered blood sugar levels on average by 5 percent. These patients also increased blood sugar lab testing by 4.5 percent during the same period. Drug adherence improved 87 percent.

“One of our strategies is moving from volume to value and really making sure it’s not just a focus on the discount on a medication or a procedure,” says Thomas Stambaugh, vice president of specialty pharmacy for the global health insurer Cigna. “Fair price and a value-based price is an important element, but what we’re focused on is the actual outcome and this is a very first step in that.”

Cigna, following the success of the agreement with Merck, established an outcomes-based agreement in 2011 with EMD Serono for its multiple sclerosis drug Rebif. It was the first outcomes-based contract in the United States for a specialty drug, according to Cigna. The intent was to reduce hospitalizations and emergency room visits by improving adherence and reducing flare-ups and related disability. Hospitalizations on average last for more than seven days and cost between $9,700 and $11,500. One third of Cigna’s cost for its patients with MS is tied to relapses. In the first year, Cigna’s Stambaugh says relapse rates for Cigna’s MS patients dropped to 1.5 percent from 2.4 percent.

But while Cigna plans on expanding such agreements—it soon expects to announce an outcomes-based plan for a rheumatoid arthritis drug—others are rethinking their approach.

The U.K.’s National Institute for Health and Care Excellence or NICE, which provides guidance to the U.K.’s National Health Service on the cost-effectiveness of medical treatments, has been a pioneer in forging outcomes-based agreements, but found them difficult to administer because of the need to gather clinical data to enforce them.

“Payers never liked these complicated schemes because you have to agree on endpoints, on costs, on paybacks,” says Ad Rietveld, director of RJW & Partners, a consultancy focused on global pricing and market access. “Most people said, ‘Why don’t you just lower price?’”

In the United Kingdom, that’s been the trend. These agreements, known as patient access schemes, initially tied payment to outcomes, but more recently have centered on straight discounts as a more expedited way of reaching the level of value NICE needs to recommend a drug for use in the United Kingdom.

“‘Patient access scheme’ is a polite way of describing a discount, which is what it is,” Andrew Dillon, chief executive of NICE told The Burrill Report. “Responder-based schemes are challenging because you have to put into place big data collection exercises to collect individual patient data.”

In addition to the challenge posed by data collection for outcomes-based agreements, discount agreements also pose a challenge for NICE. Because some countries tie what they pay for a drug to the price paid by what NICE negotiates, drugmakers that agree to discounts often do so on the basis that the amount of the discount is kept confidential. This lack of transparency creates a problem for any company with a competitor drug that needs to demonstrate it is more cost-effective than what’s currently available because it can’t determine the cost of what’s currently available.

“Health Technology Assessments without transparency are going to be a headache going forward,” says Rietveld.

But NICE has not entirely abandoned outcomes for discounts. In 2010, it entered an agreement with AstraZeneca for the non-small cell lung cancer drug Iressa that provides a means of linking payments to outcomes with a structure that features the simplicity of discounts without the need for collecting clinical data. The arrangement mirrors similar agreements used in Italy and Spain.

Under the plan, AstraZeneca provides the drug to patients at a one-time cost of $17,600 (£12,200) at the third month of use. Patients treated for less than three months pay nothing.

The thinking, according to the University of Washington’s Carlson, is that if the payer makes a single large payment after three months, it limits the payer’s cost while the continued use of the drug past the third month provides a proxy for clinical efficacy without the need for data. Cases where the drug is not working would not make it to the payment point because a doctor would discontinue use in non-responders by then. In those cases, the drug would cost the system nothing.

NICE has used similar agreements with UCB for its rheumatoid arthritis drug Cimzia (which the company provides to the National Health Service for free for the first 12 weeks of treatment), and with Zeltia for its cancer drug Yondelis, which is free for the first five cycles of treatment for patients with soft-tissue sarcoma.

Because the value of a drug can be difficult to determine until it is used by patients in a real world setting, pricing arrangements are being crafted to take into account the need for evidence-based pricing models. In the United States, the Centers for Medicare & Medicaid Services introduced its Coverage with Evidence Development policy in 2005 as a way to provide conditional payment for drugs, devices and diagnostics where clinical data needed to be gathered to demonstrate their value.

That same type of approach is being used elsewhere in some cases. GlaxoSmithKline was able to win a positive recommendation from NICE for its kidney cancer drug Votrient by agreeing to conduct further clinical testing to establish the value of the drug. In the absence of head-to-head data against the competing drug Sutent, GSK crafted a two-part plan. It offered a 12.5 percent discount for patients immediately. A second part of the plan, which is confidential, provides a future rebate should the drug perform adequately in a comparison trial with Sutent.

Such agreements continue to be works in progress. Their success will depend not just on how easy their terms are to enforce, but how well they align the incentives of payers and drugmakers, as well as patients and providers.

“It’s important when you are working with a pharmaceutical manufacturer that they feel everyone that is involved in care has an aligned incentive,” says Cigna’s Stambaugh. “As soon as one of those stakeholders has a misaligned incentive, that’s where things fail.”