Early Market Testing Can Benefit Health Care

Before seeking FDA approval, drug and medical device companies could avoid costly mistakes by seeing if their ideas will fly with patients

By Stephen Wunker and Joe Sinfield

Health-care companies often think experimentation is for the lab bench. Drug discovery and development requires remorseless trial-and-error—and only about 1 in every 10,000 compounds created will eventually make it to market. But once a molecule emerges from that process, experimentation tends to cease.

With tight regulatory supervision and a product pipeline that typically yields only a small handful of new launches each year, pharmaceutical companies invest heavily in the few products that do emerge. A typical drug launch can easily cost tens of millions of dollars or more. With patent expiration dates dictating the product life cycle literally to the day, firms want to maximize sales while they can.

Blowing the Opportunity

This situation is perilous. Pfizer (PFE), the world's largest drugmaker, offers a telling illustration. Beginning in 1996, it invested heavily in creating an inhaled form of insulin to treat diabetes. On its face, such a product seems an excellent idea. The market for injected insulin is several billions of dollars per year, and people do not enjoy sticking themselves with needles.

In 2006, Pfizer received Food & Drug Administration approval to market the inhaled insulin product Exubera, which it had licensed from biopharmaceutical company Nektar. Pfizer invested approximately $2.8 billion to get to that milestone and predicted annual sales of $2 billion. Then came the big commercial launch, with slick commercials and armies of sales reps. Total sales? $12 million a year. In 2007 the company pulled the plug on this extraordinarily expensive failure.

What went wrong? Diabetic patients were concerned with the ability to "feel normal" again. However, the Exubera product was a flashlight-sized tube that had to be used as if it were a large pipe. Dosing insulin into the tube required measuring in unfamiliar increments. Patients had to have a special lung function test before being allowed to use the product.

Further, as one endocrinologist put it, "I can teach someone how to use an insulin pen in five minutes, but it would take nearly an hour to teach a patient to use inhaled insulin." Finally, many doctors continued to worry about the fact that 90% of the inhaled insulin did not make it into the bloodstream and might have unknown effects in the lungs.

None of these concerns required an actual product launch to validate. Experimentation could have helped Pfizer avoid this expensive failure. Experiments could have included giving patients a mock-up of an Exubera tube to use and then asking them how they felt about the experience. Pfizer could have done research into the impact of unfamiliar dosing, lung function tests, training time, and safety concerns—research that would have cost less than even Exubera's eventual sales.

Levers for Experimentation

There's a big lesson to be learned from the Exubera debacle. Before spending serious money, pharmaceutical and medical device firms would do well to undertake a disciplined process of evaluating their major assumptions and risks, and how much they could experiment. Here are some areas ripe for such experimentation:

Which outcomes are critical to prove in a clinical trial? Does a weight-loss drug need to establish that it reduces incidence of diabetes, or that it reduces weight by a certain percentage, or that it reduces weight when combined with a specific diet? The answers depend upon which stakeholders will most affect the product's adoption—regulators, health insurance companies, or patients.

Business Exchange

Track and share business topics across the Web.