The Right Stuff

By January W. Payne
Washington Post Staff Writer
Tuesday, August 16, 2005

A new study affirms the feasibility of using the Internet to conduct gold-standard medical research. The authors say the findings are the first to be based on a randomized, placebo-controlled, double-blind clinical trial conducted entirely via the Internet.

The eight-week trial looked at the effects of two herbal remedies, kava and valerian, that are used widely to treat anxiety and insomnia. The supplements' effects were not remarkable, according to the study; neither provided more insomnia or anxiety relief than placebo.

While the results echo earlier findings, the researchers' use of the Internet for everything -- from recruitment to patient consent to data collection -- makes the study unique. Previous Internet-based trials used the Web for most, but not all, steps of the research process.

The all-electronic method offers several advantages, the authors said.

"You can roll out a study much quicker, which allows you to get results much quicker," said Bradly Jacobs, lead study author and assistant clinical professor of medicine at the University of California, San Francisco. Results "can be made available to the public much quicker," he added.

Jacobs is an owner of an IT company whose products were used in the trial. The results appear in the July edition of the journal Medicine, which is published by Philadelphia-based Lippincott Williams & Wilkins. Drawbacks and limitations of pure online research include the inability to test therapies that require someone to treat or observe participants as well as safety concerns about testing pharmaceutical products. Jacobs said hybrid studies that mix Internet and conventional methods can be used for research where a physician's presence is needed.

Experts familiar with Internet-based research are enthusiastic about its efficiencies.

"The minute you finish [an Internet-based] trial, . . . the data entry is done," said Tim McAlindon, a rheumatologist and director of the Center for Internet-Based Epidemiologic Research, which focuses on developing and validating Internet research, at Tufts-New England Medical Center.

He said online research costs a quarter to a half the price of a traditional trial.

A 2003 online clinical trial investigating the effects of glucosamine on osteoarthritis of the knee cost about $914 per participant, compared with $1,925 for a similar hospital-based trial, according to the computer-based study, for which McAlindon served as lead researcher. (The online trial used regular mail to collect hard copies of the required consent forms.)

Finding participants for Internet-based trials is relatively easy. In the kava-valerian trial, Jacobs and his team created a Web site for the study, advertised on two women's health Web sites and used an e-mail list from Alternative Medicine magazine to draw more than 1,500 registrants.

Because using the Internet requires no face-to-face contact, the researchers took extra steps to confirm participants' identities. The volunteers' name, birth date, address and other information were verified by a private company, Jacobs said, a process that narrowed the pool down to 1,241 people.

The field was then trimmed to 391 people whose scores on anxiety and insomnia assessments met the study's guidelines. Of those, 135 people were randomized to receive placebo, 121 received kava and 135 were put in the valerian group. Those participants received either the supplements or placebo by mail but did not know which group they were in.

In the 2003 glucosamine trial -- funded by the Arthritis Foundation and the National Library of Medicine -- patients received glucosamine or placebo by mail and filled out online questionnaires that measured their symptoms at least every two weeks, said McAlindon. Both trials included security measures -- such as encryption, passwords and a firewall -- to protect personal data transmitted over the Internet.

The speed at which participants signed up for the Internet-based trials surprised the researchers. Jacobs said recruiting took only six weeks for his study, compared with the nine months to two years that he would have expected using conventional methods.

In traditional clinical trials, "you sweat blood trying to get patients in," said McAlindon. But in the glucosamine trial, he said, "we just sat there while these e-mails bounced into our in box."

Internet trials are, at least for now, best used for studying over-the-counter medications and supplements, experts said. Obstacles exist in testing stronger drugs over the Internet, including making sure participants are monitored closely for side effects.

Still, Internet-based trials provide benefits that can aid researchers and speed the release of findings.

"The goal here," Jacobs said, "is really to increase access" to trials and data "for the U.S. population." ยท

© 2005 The Washington Post Company