Decisions, Decisions: Michael Rosenberg on Adaptive Research

MG_eCliniquaLOGOBy Deborah Borfitz

September 06, 2007 - Knowing adaptive trials require fast decision-making by study sponsors and research teams, Michael Rosenberg, MD, MPH, built a clinical research organization 18 years ago to help them do it. The “tactical component” of adaptive trials, such as continuously tracking enrollment and better matching monitors to need, isn’t a new idea, he says. But, like sample size reassessment and dose finding on the strategic side, it remains the approach less taken because of false hopes placed on Web-based electronic data capture (EDC).

Fast, clean data and performance metrics are the key criteria for doing adaptive trials of any type, says Rosenberg, founder and CEO of Health Decisions (Chapel Hill, NC, and Oxford, UK). That also describes the two major disadvantages of Web-based EDC. “Someone needs to enter data by hand, and that inevitably leads to delays. The interval between when data is generated and when it’s entered can take a couple of days and, more often than not, a couple of weeks.”

Moreover, none of these systems measure performance, Rosenberg continues. “They miss the most important management component of a study, which has the most profound implications for study timelines and budgets.” Nearly a decade ago, as reported in Forbes, Health Decisions shaved 1.6 years off a five-year timeline and $32 million off a $100 million budget for a large, six-nation Alzheimer’s study using performance metrics to guide managerial decisions.

That was even before better data input devices, such as Health Decisions’ optical smartPen, came along, says Rosenberg. The SmartPen records strokes as case report forms are filled out and then docks in a computer station, transmitting data from the device to a central, Web-based database within seconds.

Desktop “widgets” have also been introduced, delivering automatic updates to sponsors on key study parameters. The technology works hand-in-hand with the SmartPen. “They can track patients in the consent process, in screening, enrolled, discontinued early, and those who have screen failed – or any other progress measure,” says Rosenberg. Project managers, study monitors, and company executives all have different widgets to provide the specific type of timely, clean data they need for rapid decision making.

In the world of adaptive research, the easiest way to turn incoming data into useful information is to use integrated electronic clinical development management systems where the focus is on entire study programs rather than individual trials, says Rosenberg. “The Achilles’ heel in the industry is the interval between studies.”

The Web is the tie that binds, allowing information to be collected from anywhere in the world. “Our job is to digest all that incoming information and make it actionable by different people in different roles.”

Establishing Benchmarks

The combination of the strategic and tactical components of the adaptive approach is “no less important than the introduction of the assembly line was for the auto industry,” says Michael Rosenberg, founder and CEO of Health Decisions (Chapel Hill, NC, and Oxford, UK). All trials today, at the very least, should incorporate some elements of adaptive trials – “especially on the tactical side. If you don’t consider these elements, you’re leaving something on the table.”

The inefficiencies of conventional study management techniques are mind-boggling, says Rosenberg. It generally takes a week to access key study progress metrics like site enrollment statistics. With technology used by Health Decisions, “we keep track of what’s going on every single minute at every single site and make this [information] available in real time to all study personnel, 24/7.”

There is still much confusion in the marketplace about constitutes an “adaptive” trial, says Rosenberg. In a broad sense, it is “any study that takes advantage of what is learned during the course of a study.” Health Decisions has run over 300 such trials and currently has about 30 active projects with big pharmaceutical firms and biotechnology companies of all sizes.

Throughout its history, Health Decisions has made it onto the Inc. 500 list of fastest-growing private companies three times, says Rosenberg. Between the most recent fiscal years, revenues have grown 50 percent and demand for the company’s adaptive services “skyrocketed.”

An example of how tactical adaptive management helps studies is demonstrated by Health Decisions’ record-setting rapid enrollment in four different therapeutic areas: cardiology, oncology, Alzheimer’s, and reproductive health. “For the Alzheimer’s study, we recruited 1,073 patients in seven months,” says Rosenberg. On a pivotal metastatic breast cancer study, Health Decisions’ record-breaking enrollment took three to six months off the expected timeline. “On 82 percent of our studies, we enroll on or ahead of time.” Industry-wide, only about 15 percent of studies have on-time recruitment.

Notably, the company was the first to get FDA approval to use sample size reassessment. “That component alone cut nine months off a large study,” says Rosenberg. “Overall, [the sponsor] saved a year in getting the product to market and [according to The New York Times] enjoyed an extra $366 million in income.”

Health Decisions’ Smart Monitoring process uses an electronic tool that tells users when there is enough work to send monitors into the field. The SmartPen is the data source for some studies, eliminating the need for monitors to do source data verification. “That can knock 20 percent off the cost of a study right off the bat,” says Rosenberg.

For companies, one of the biggest challenges for putting the adaptive approach into practice is the process change involved, says Rosenberg. “People need to make more decisions earlier than they used to. They can’t sit in the Bahamas for a year once a study starts. There are dose finding studies where they have to watch data every day…and make decisions [such as cutting a dose arm] on the fly. That’s how you cut timelines down.”

The upfront, per-patient cost of the adaptive approach is no more expensive than traditional methods, says Rosenberg. The savings are on the back end. On dose finding trials, for example, Health Decisions can shorten study duration by four months and shave 25 percent off study budgets. “Sponsors see that as a huge competitive advantage.”

The hard part is the number of unknowns, such as the duration of a trial. “Sponsors sometimes get overly optimistic about how long it will take to recruit patients,” says Rosenberg. “The risk is great so the payoff is great, too. To minimize risk, we make decision-making the best it can be.”

Promised FDA guidance on adaptive trials has yet to materialize, but Rosenberg doesn’t think it’s smart for companies to wait for the agency to tell them what to do. Sponsors can negotiate “on an individual basis” with the FDA and the European Agency for the Evaluation of Medicinal Products.

Rosenberg is currently penning a book on the practice of adaptive clinical trials, which is due out in the spring. The book will deal with design elements and address a lot of the “nuts and bolts” of how to put the adaptive approach into practice, such as data system requirements and how to keep the study team blinded.

Michael Rosenberg, M.D., MPH, is President and CEO of Health Decisions Inc., Chapel Hill, N.C., a full-service, globally capable CRO that leverages technology to simplify and improve processes for rapid, effective clinical evaluation of pharmaceutical products. Web site:

Share Button