Better CR sites succeed through faster initiation

Competitive pay and experience also matter

It's a question all sponsors and clinical research sites ponder: Why do some studies and some sites finish significantly faster than others?

A new study provides some answers.

"Good sites start enrolling quickly," says Harold E. Glass, PhD, MSc, professor, health policy, of the University of the Sciences of Philadelphia in Philadelphia, PA. Glass also is the president of TTC, a data base company in Philadelphia.

It's almost that simple.

"The amount of time it takes them to sign a contract is shorter, so they're in the field much faster," Glass says. "What happens is not only do they enroll on a weekly rate, but they actually enroll well and have more time to enroll because they're out there early."

Sites that begin enrolling earlier than others have more time to enroll and apparently use this time to improve enrollment, he adds.

"The study ends at the same time for everyone, so the only way to have more time is to enroll more quickly," Glass says.

The site performance study compared research data from 262 phase III studies that involved 2,047 sites in the United States and Europe.1

Experience counts

Besides having a faster initiation speed, faster sites also have more experienced investigators, Glass says.

Researchers looked at the amount of studies that investigators had done for a particular sponsor in the previous three years by calculating the number of investigational new drug forms completed and submitted to the FDA, Glass says.

"We looked up how many times someone's name appeared in [form] 1572 records," he explains. "So sites that did more 1572s on record and did more work for pharmaceutical companies did a better job."

While this would appear to be an obvious finding, it apparently runs counter to conventional wisdom.

"You hear things like, 'Just because someone has done a lot of studies doesn't mean they'll be good in the future,'" Glass says. "Maybe someone will fool lots of people, and no one realizes they're actually terrible, but my guess is they're not bad; they're being selected because they did a good job."

While the better-performing investigators do receive more repeat business, it's also true that the larger sponsor companies are less likely to capitalize on this experience, Glass notes.

"The larger the sponsor company the less likely it is they'll use an investigator for the second time," Glass says. "I suspect it's because they're big companies and there's turnover in people and they lose track of them."

The best strategy would be to mix up the mix of new and experienced investigators, Glass suggests.

"You want new people now and then, but experience counts," he says.

Competitive contracts work best

The study also found that there is no relationship between how fast a site completes its work and how much it's paid, although how it is paid does matter.

"We have a model that determines what is the market rate in terms of clinical grants for the particular work in a country, looking at it in relative amounts," Glass says. "So being paid in the 99th percentile was top payment for comparable levels of work."

When researchers looked at this one measure, they found that how much a site is paid had almost no impact on completion speed, he says.

"If anything, and this is bizarre, it's negative: The more you pay the worse performance you get," Glass says.

"What does matter is how they are paid," Glass adds. "If you have competitive enrollment, then when you get more subjects in, you get paid more, and that makes a big difference in trial speed."

What worked best were competitive contracts.

Sites that were paid competitively, meaning they were paid for performing against certain milestones and if they didn't enroll subjects they were dropped, were faster enrollers, Glass says.

The study also found that sites that received upfront, nonrefundable start-up payments generally completed a higher number of patients.1

"Cash flow is important, so you want to make those milestones so you'll get paid," he adds.

Collect performance data

The study's findings have several implications for sponsors and clinical research sites.

First, CR sites should collect their own performance data, and sponsors should develop their own list of top-performing sites, Glass suggests.

"They should develop a list from the FDA of investigators by experience level," he says. "Then they should see what information they can find out about sites."

There already is considerable information available in national databases, but sponsors also have access to valuable site performance information in their own databases, he adds.

"There are lots of ways to construct investigator databases, especially if you find that sites that perform well start early in the process," Glass says.

Sites need to keep their own data because large sponsors have not yet embraced the idea of keeping site performance data and using it in the site selection process, he notes.

"People think these gigantic organizations are well-oiled machines, but they're made up of hardworking people who struggle from day to day," Glass says.

And they make mistakes, including losing or not maintaining their own site performance records.

Also, sponsors need to end contracts with sites that do not enroll or do not enroll quickly, Glass says.

"Sponsors hang on to sites way too long," Glass adds. "It's hard to let them go."

Manage the contract process

Secondly, sponsors and CR sites should work on signing contracts faster and negotiate for faster payments for performance, Glass suggests.

It's important to speed up the contract process, he says.

"What irritates sites more than anything else is having a problem with cash flow," Glass says. "They need to work out something with sponsor companies to maximize their cash flow."

CR sites can start by being realistic about the financial side of research.

"That's where they get into trouble is by not believing that big companies can take so long to pay," Glass says. "But it does take a long time."

On the other hand, the more successful a CR site is, the more power it has in the marketplace. These sites should be able to negotiate a payment schedule that will serve both sides' interests.

"When sites first start research they have no idea what a hassle it is wait for payments, and they get caught off guard by this," Glass notes.

"What sites have to realize is there is no rule about how much they have to be paid, but it has to be reasonable for the work they're asked to do," Glass says. "There's a lot of variation in fair market rates, so I tell sponsor companies that it's only too much if they won't pay it, or it's not enough if they won't accept it."


  1. Glass HE, DiFrancesco JJ. Understanding site performance differences in multinational phase III clinical trials. Int J Pharm Med 2007;21:279-286.