In the short life of the internet so far, electronic data capture (EDC) came of age at an inopportune time. The top EDC solutions were born in the age of beeping modems and underpowered PCs.
As luck would have it, the EDC leaders were all built in the technological Pleistocene, before companies like 37Signals (project management), Picnik (photo editing) and Netflix (streaming video) designed clean, fast, network-based web applications that ran not on malware-clogged desktop computers or ancient mainframes but in the cloud. Long before elegant wireless tablet computers from Apple, the DNA of the EDC pioneers had been built to operate at a slower pace.
So the research side of the pharmaceutical world has 10-year old applications running on today's internet. It's not a great match. People working for the industry have one experience of the internet as consumers—and quite another at the office.
Cmed, with headquarters in Horsham, England, is trying to rectify the EDC time warp. Comprised of both a contract research organization (CRO) and a technology division, Cmed has ties to the University of Oxford and built its EDC system a few years after the leading EDC firms. Cmed has assembled a roster of top executives from some of the biggest CROs in the industry, as readers can see on this page.
In the autumn, Cmed won an EDC project for a malaria trial in South America. The sponsor: GSK. That suggests that the large EDC companies may not have the entire market sewn up yet.
James Haughwout, Cmed's VP of commercialization, is blunt about the limitations of well-established EDC suppliers: "The real world systems are slow and crash. People are asked to get off [the systems] so repairs can be made. A lot of those systems are not giving the people the experience that Google or Gmail are giving."
Haughwout previously worked at Amgen, as well as at technology leaders like AOL and Netscape. He says one key to the performance of the Cmed system is its slightly later arrival, allowing the company to start with a blank sheet of paper.
Cmed says it uses the Oracle Berkeley database, which is more distributed, allowing better performance and quicker page turns. "We can tap multiple servers in multiple places to do a database lock," Haughwout says. "It is a true distributed architecture. You can make changes without crossing out a whole bunch of pages that have been generated."
Faster Page Turns
Cmed is trying to give ordinary users of its system the same experience they have when browsing the web at home, where performance is rarely an issue. At the office, of course, clinical trial sites are not always happy with the rapidity of EDC systems, especially when trials are starting up or winding down. With Cmed's system, even those peak periods offer no latency or spinning hourglasses. "You're not all hitting one database, one server," Haughwout says. "We are sharing the data in a set of resources."
In effect, Haughwout says, some older EDC solutions have built-in areas of traffic congestion: a single program, a single database that cannot easily handle hundreds of simultaneous users.
Haughwout goes on to say that there is often a gap between the real-world speed of an EDC solution and what clinical trial executives choosing such systems see in a carefully controlled demonstration. Especially in hospitals, he says, sharing copper phone lines or network access can result in frustrating experiences. "It is not as fast as the software engineers think it is," he says.
In some cases, he says, the complexity of modern trials leads EDC suppliers to suggest work-arounds to their own systems. These may solve one problem only to create another, harming the trial's overall productivity. For example, suppose a sponsor has requested all manner of automatic edit checks from a system that cannot actually support dozens of users doing that. d9A2t49mkex
A technology vendor with a slow platform, watching its system grind to a halt, may suggest that the automatic edit checks be turned off. Why? To boost the performance of the rest of the system. For Haughwout, that's a compromise that his customers should not have to make.
Nor do his customers have to pick a time (4:00 a.m., anyone?) to shut down an EDC system for an amendment to the system. The Cmed system can push out updates in real time, to selected geographic areas, with no outages or offline time for anyone. "You don't have to tell the physician to get off the system," Haughwout notes. In an era of global projects, that will support more continuous operations. "There is no such thing as 4:00 a.m. when you are in that many time zones," he says.
By design, Cmed's technology is intended not only to work in crowded hospitals with an abundance of firewalls and IT security provisions, but also in the developing world: places like Africa, where modern telecommunications infrastructure is nonexistent.
In a demonstration for a reporter, Cmed's Osie Gaines III and Karis Lovett displayed an EDC interface that was cleaner and faster than many of its peers. The Cmed system does use icons, but fewer of them, creating a less visually noisy design.
The company also showed off a solution specifically designed for challenging IT environments where poor connectivity or IT security measures interfere with the work of the clinical trial research associate or monitor. The Cmed solution amounts to a paired wireless base station and a smaller device (a laptop or tablet) that feed data to a central server in another location. The local IT environment, in short, becomes a non-factor. There is a news release about that solution here.
As Lovett and Gaines entered a few data points and quickly loaded multiple screens of information, it was clear that Cmed has succeeded in making the workflows of clinical trials a bit smoother and more fluid. Source document verification (SDV), they say, can be handled with a single click, much as one might buy a book on Amazon. "It is built to be always on and always fast, regardless of where you are," says Haughwout.d9A2t49mkex