Decision support, predictive modeling may speed clinical research

The clinical research community has been looking for new and better ways to conduct expensive and time-consuming clinical trials--using clinical trial management systems, recruiting patients online and using electronic health record data to conduct research more efficiently, for example.

A new software program takes a different approach--it uses decision support technology and predictive modeling for its "virtual clinical trial" offering.

Users access Archimedes' modeling algorithms online and can run multiple scenarios based on different kinds of patient populations, specified health conditions and various treatment regimens. It predicts outcomes by using algorithms derived from mathematical modeling of large public databases.

Archimedes, founded by evidenced-based-medicine expert David Eddy, M.D., is partnering with clinical research organization Quintiles on the project.

According to Eddy, this approach can be used in comparative effectiveness studies that contrast different ways of treating a particular condition. To perform the same kind of studies using traditional clinical trials would be more expensive and would take longer, he says in an online video.

Researchers are trying a variety of "virtual" models in order to accelerate research and cut costs.

Some organizations are using data from electronic health records in practice-based trials. The Affordable Care Act provided funding for comparative effectiveness research and set up the Patient-Center Outcomes Research Institute (PCORI) to supervise such studies.

Other companies are looking to the Web, recruiting patients and conducting research studies online. Mayo Clinic, for example, has experimented with recruiting research participants using social media.

But drug giant Pfizer "stumbled" over its efforts to complete a virtual trial, FierceBiotechIT reported earlier this month.

After launching the trial last year, the company found that its online recruitment effort fell short, stalling the trial's progress, InPharm reported. According to the article, Miguel Orri, senior director of clinical science at Pfizer, said that the study "didn't recruit" because patients felt uneasy about sharing their health info online and struggled with a tricky sign-up process.

The company says it has fixed some of the bugs in the study and is expanding the trial to Europe, according to FierceBiotechIT.

But decision-support and predictive modeling such as Archimedes' ARCHeS might be able to go where online efforts cannot.

The tool is based on a familiar technology--Archimedes' IndiGO decision support tool, which is in use at Kaiser Permanente and Fairview Health Services in Minneapolis, according to a recent InformationWeek article. Besides identifying at-risk patients, IndiGo can also show physicians the best treatments for preventing a heart attack, stroke or other adverse health event in a particular patient.

Kaiser subsequently studied IndiGO in its Hawaiian division. A soon-to-be-published paper on that study will show that patient outcomes improved as a result of using IndiGO.

The Food and Drug Administration has said it will help other companies that want to test new ways to conduct research, including virtual trial models. Janet Woodcock, director of the FDA's Center for Drug Evaluation and Research, commended Pfizer on its pilot and encouraged other manufacturers considering novel ideas to advance clinical trials to talk to the Agency about trial design and oversight, according to FierceBiotechIT.

To learn more:
- read the announcement
- see the YouTube video
- here's the FierceBiotechIT article
- here's the InPharm story
- check out the InformationWeek article