Selling Outcomes

In the halls of big drug companies like Pfizer and AstraZeneca walks a new breed of manager that didn’t exist just two years ago. With puzzling titles like “vice president of real-world data,” they are charged with gathering information on patient outcomes—on how well their products work once a patient leaves the doctor’s office or the pharmacy.

Although monitoring the effectiveness of a product in the market may seem like standard practice for any industry, pharmaceuticals is not just any industry. Its business model has developed around a scientific process of establishing a drug’s safety and efficacy in a series of clinical trials, after which a drug is approved by the Food & Drug Administration and sold. Attention has traditionally turned, at that point, back to the laboratory where work is under way on the product that will pick up the slack when the newly marketed one goes off patent and is muscled out by a generic.

The hiring of a vice president of real-world evidence at a major drug company may or may not signal a big change to this model. But evidence suggests that a shift to a more patient-focused way of doing science and business is under way.

The idea that drugmakers might be paid for patient outcomes rather than pills, vaccines, and injectable drugs may seem drastic. Surely, the skeptics will say, pharmaceutical companies will get paid for selling drugs for as long as there are pharmaceutical companies.

But nothing is that simple in health care anymore. Drugmakers, regulators, and insurance companies (payers, in industry lingo) are adapting to a shift in the economics of health care. That move is away from payment on the basis of the volume of individual services and products rendered and toward payment determined by improvements in an individual patient’s health. The push for value-based health care, which predates the 2010 Patient Protection & Affordable Care Act but is also a key tenet of this law, involves risk sharing on the part of health care providers, payers, and drug companies.

As a new health care finance system evolves, drugmakers are facing increased requirements from regulators and payers to provide detailed information on the benefit of their products compared with alternative therapies, including generic drugs. Pharmaceutical firms have launched major efforts to get information on how their products are used and how they are working for patients in the “real world.”

There is no shortage of real-world data, which is part of the problem. The rise of electronic medical records and the information technology (IT) and data banks necessary to identify, corral, and analyze information have helped make sense of it. But figuring out the complicated puzzle of just how drugs are changing the health and lives of patients in the real world is a daunting task.

G. Steven Burrill, CEO of the investment firm Burrill & Co., says the ongoing shift addresses gross inefficiencies that arose during the so-called blockbuster drug era of the 1990s, when drug companies launched one billion-dollar-selling drug after another.

“Fifty-five percent of drugs used in America don’t work for the patients they are prescribed for,” Burrill says. “That’s because we have lived in a one-size-fits-all world. Clinical trials are done on the smallest possible population for FDA approval on the basis of safety and efficacy, after which the drug is thrown over the wall into a health care system where there is incredible patient variability.”

Drugs fail, he says, because they don’t work in the same way for every patient. One example is Vioxx, a broad-spectrum pain relief drug that Merck & Co. took off the market in 2004 after adverse cardiac events started to emerge. “The system hasn’t been wired to deal with the diversity,” Burrill says.

At the same time, payers are increasingly challenging drugmakers on the cost of their products relative to the benefit they deliver, Burrill says. “Drug companies have come to realize that they need to create value in their product and then capture that value.” But the means of determining value are unclear, Burrill says, because of the variability in what payers and regulators around the world regard as the value of a medicine.

Burrill and others point to a concerted effort in which drug companies, payers, and regulators are working, sometimes in partnership, to harness the data available in clinics, doctors’ offices, and patients’ homes—anywhere electronic data are being generated on patient health. “We are clearly moving into the era of big data and data analytics, and everyone is trying to figure out what data are meaningful and how to use them,” he says.

Information on the effectiveness of a drug compared with its alternatives is becoming more important throughout the health care enterprise, agrees Robert Temple, deputy director of clinical science at FDA. “What everybody is interested in now is comparative data: Which antidepressant works better?” he says. “But those are not easy studies to do, because when the effect of something is not huge, it is hard to show the difference between two drugs. It’s hard enough to show the difference between drug and placebo.”

Even placebos become an issue, Temple says, because the use of a placebo and other standard practices in controlled clinical trials is not feasable in real-world outcome studies. What is clear, however, is that when undertaking such studies, patient-reported outcomes are of more value than physician-reported outcomes.

Physician-reported data are less reliable, Temple says, as they are influenced by the skills of the physician in communicating a patient outcome. Patient-reported outcomes do a much better job at getting the full picture of a drug’s impact on day-to-day life, Temple explains.

Marc L. Berger, vice president of real-world data and analytics at Pfizer, agrees that drugmakers need to continue studying the safety and efficacy of their drugs beyond traditional clinical trials. He says methods for doing so are opening up, energized by technology and access to data.

“Science moves forward when the opportunity is there,” Berger says. “Our insights are getting stronger and the availability of data is allowing us to utilize those scientific insights.”

Berger, who came to Pfizer from UnitedHealth Group last year to head a new group studying patient outcomes, notes that Pfizer recently formed a partnership with Humedica, a health care analytics company, to mine Humedica’s database and glean insights into the health impact of therapies.

Sachin H. Jain, chief medical information and innovation officer at Merck, also acknowledges a focus on outcomes in the pharmaceutical industry tied to a heightened effort to deliver personalized care. Although the industry has always been concerned with real-world outcomes, Jain says, access to data and advances in health care IT have catalyzed a surge in activity.

“The increased focus on patient outcomes across the industry is driven by our ability to measure it,” he says. Electronic medical records have been a boon, but only lately have drug companies and academic research institutions begun the arduous task of amalgamating and standardizing the data from a vast field of sources.

Like other drug companies, Merck has acted to access data through partnerships, including one with Regenstrief Institute, a nonprofit medical informatics and health care research organization affiliated with Indiana University School of Medicine. The partners are studying patient health outcomes in the area of osteoporosis with data culled from Regenstrief’s information on 13 million patients.

“We are involved in a number of conversations with potential partners,” Jain says. The idea is to garner meaningful information to advance what has long been the company’s mission of improving patient health. That mission is not changing, he says, but with better information on patient outcomes, the company’s business model will change.

AstraZeneca is also working with partners to access patient data, according to Greg Rossi, vice president of payer and real-world evidence. In addition, the company employs a group of epidemiologists, bioinformaticians, and other health outcome researchers who focus on determining real-world evidence of its drugs’ benefit to patients. “It’s an increasing capability that I think the industry needs to have and we have to develop,” he says.

The challenge, according to Rossi, is getting at data outside the controlled environment of clinical trials. Drug companies have to shift their focus from gathering information on people with a particular disease, the kind that is derived in the clinic, to gathering information about individual patients. “If I can measure in close to real time, not at a population level but at a patient level,” he says, “I can start thinking about how I can link performance incentives to payments based on the outcomes actually being achieved.”

This is not a new idea, Rossi points out. “Disease management has been talked about since the 1990s, but it’s very difficult to do,” he says. “Now we are starting to really be able to integrate electronic medical records, which enable us to think much more clearly about quality indicators and the performance of the health care system.”

Rossi says the information gleaned from actual medical records will go toward answering questions that AstraZeneca is already fielding from insurance companies about why their products are worth what the company charges for them.

AstraZeneca is in the middle of a four-year contract with HealthCore, the health outcomes research subsidiary of WellPoint, a big managed-health-care firm that keeps a huge database culled from the 44 million patients it covers. The partners have conducted studies on a number of AstraZeneca drugs that have enabled the firm to put together a patient outcome narrative. “We are able to use that information in a very successful way with payers as well as with our internal R&D organizations to better design our own critical programs,” Rossi says.

Timothy Wright, the global head of development at Novartis, says his firm is also channeling outcome data into drug development research. He expects that outcomes will increasingly influence the entire science and business life cycle of a drug.

“Typically in clinical trials, we would launch a drug and try to get as many people to take it as possible based on its safety and efficacy profile,” he says. The assumption followed that a drug would have a patent-protected life span during which Novartis could explore options for follow-up drugs.

“What’s different now,” Wright says, “is that we’re looking at our portfolio, figuring out how we can find something of transformational value—something that can change the outcome in a significant way such that the drug will become a new standard of care, a new drug of choice for physicians to offer patients.”

Whereas this used to be a postmarketing exercise, “we are now bringing this thinking and some of these end points into our development program,” Wright says. “Soon we will not just be launching and selling the pill, but selling it with the outcome data that show the value of it to the patients, physicians, and the payers.”

Many of the health care data analysis firms that have partnered with pharmaceutical companies have been collecting outcomes data for decades. Like the drugmakers, they cite the rise of electronic medical records in the health care industry as important in advancing patient outcome research. They also cite payer pressure for outcome data as a catalyst for change in the drug industry.

When drug executives speak of selling outcomes rather than pills, “they are speaking the language of the payer,” says Jon Duke, an assistant professor at Indiana University School of Medicine and a research scientist at Regenstrief. “Health care reform writ large is about value-based care. Yes, they are trying to sell products, but the way to sell in the future is through outcomes. That is the new reality.”

Regenstrief began collecting clinical data in the 1970s to support a health care information exchange. By the 1980s, the organization was gathering data from a network of Indiana hospitals. It currently manages a data bank of 4 billion data points from 13 million patients, including lab studies, radiology studies, prescribed medications, test reports, insurance claims, and other routine medical data.

A venture that started as an academic and public health system information resource, Regenstrief now has booming collaborations with industrial partners. In addition to the Merck partnership, Regens­trief has worked with GlaxoSmithKline and its neighbor in Indiana, Eli Lilly & Co. “It’s a great time to do what we do,” Duke says. “Someone described it as selling umbrellas in a rainstorm.”

Marcus Wilson, president of HealthCore, also sees a growth market because drug companies lack the resources to add real-world studies onto their current clinical trial organizations. “If they studied every population, it would cost $20 billion to get a drug to market instead of $10 billion,” Wilson says. “It would take 20 years.”

Wilson points to a complex, shifting health care landscape in which the drug companies are only one element undergoing change. Physicians are assuming greater risk as reimbursement for care becomes tied to improving the health of the patient through a regime in which prescription drugs are only one facet.

HealthCore was launched in the early 1990s and acquired by WellPoint in 2003. Wilson says the group operates “behind a fire wall,” making it possible to partner with drug companies whose drugs are reimbursed through WellPoint. Wilson characterizes the partnership with AstraZeneca as an effort to “build an effective environment to evaluate patient care, no matter who is asking the questions. We both acknowledge this is likely going to be a 10-year project.” That being the case, the partners have begun negotiations to extend their initial four-year agreement, which expires next year.

Humedica, founded in 2008 by former pharmaceutical executives to conduct drug effectiveness studies, also sees a big rise in outcomes research. “Now, 40% of our revenue comes from outcome studies as opposed to drug effectiveness studies,” says Steve Davis, the firm’s vice president of life sciences. “We didn’t even go after that business. It came to us.” Humedica is currently doing work for 20 drug companies, he says.

The Boston-based company, which claims to have the world’s most comprehensive clinical database, says it has access to information from health care provider organizations treating approximately 30 million patients in more than 38 states.

Drug firms and their partners agree that gaining access to such large stores of data on patient outcomes will lead to near-tectonic shifts in health care. But the change may be more radical for pharmaceutical companies than anyone can anticipate.

“Right now, we are still in the business of selling products—the pill or injections,” Novartis’ Wright acknowledges. “But in the future, if we were to go down the road of actually delivering specific outcomes, there is a bit of a scenario where pharmas become like insurance companies,” making risk-based determinations of outcome criteria and appropriate payments. “It’s a matter of whether the system evolves in this way. If it does, it will require a different set of skills.”

Whether drug companies will, like insurance companies now, make decisions on which patients will be reimbursed for drugs, “is a good question for a glass of wine,” HealthCore’s Wilson says. “I don’t think patients will be denied drugs.” But the question remains: How well do the drugs work? “All decisions today are based on suboptimal information,” Wilson says. “None of the decisionmakers have all the information needed, and the biggest price is paid by patients.”