Enrollment doubles with electronic reminder tool
Subject referrals had 10-fold increase
A new study suggests that clinical trial sites could significantly improve their subject recruitment with the use of an electronic medical record tool’s physician reminder.1
The study found that during a diabetes clinical trial there was a 10-fold increase in referrals after the tool was used. In the 12 months prior to using the recruitment tool, there were five referring physicians who averaged 5.7 referrals per month. In the 12 months of using the tool, there were 42 referring physicians and 59.5 referrals per month.
The diabetes trial’s actual enrollment doubled, with an enrollment rate of 2.9 per month in the 12 months before the tool was used, and a rate of six per month when the tool was used.
"The findings were remarkable," says Peter Embi, MD, MS, an assistant professor of medicine at the University of Cincinnati. Embi is the principal investigator on the study, which was conducted at The Cleveland Clinic.
While others have attempted to use emergency health records to remind clinicians of trials, no studies have been published showing such a significant benefit in the number of physicians making referrals and their recruitment rates, Embi adds.
The diabetes study was receiving subject referrals from five endocrinologists prior to the use of the electronic health record alert tool.
After the tool was used, there were an additional 36 general internists and sixth endocrinologist who made subject referrals, Embi says.
"There was an eight-fold increase in the number of physicians participating," Embi adds.
The study’s success shows that many physicians would make referrals to clinical trials if there were reminders and a simple process for doing so.
Some of the obstacles to physician referrals include a lack of time, unfamiliarity with the trial, and forgetfulness, Embi says.
"In my practice, I noticed I had missed a lot of potentially eligible patients for trials we had ongoing, and it was just because it wasn’t the first thing on my mind," Embi says.
This revelation gave Embi the idea of creating a tool that would remind physicians about clinical research while they were meeting with patients.
Part of the tool’s design also addresses the other obstacles, by having the tool do the initial subject screening and making the whole process very fast.
Since The Cleveland Clinic, which is where Embi practiced at the time, already had an electronic health record, he looked into adapting the tool to produce clinical trial alerts.
"The tool can be configured to alert for anything," Embi says. "Basically what you do is have it designate certain parameters."
Investigators decided to study the tool’s new use as a clinical trial reminder during an ongoing diabetes study. So the tool was configured to alert physicians if their patients were likely to be eligible for the trial. The criteria selected to trigger the alert were the patient’s diabetes diagnosis code, having a glycosylated hemoglobin level of greater than 7.4%, and being over age 40. "If those criteria were met, we had the alert come up during the patient’s encounter with the physician," Embi says.
The system was also adjusted to give the physician an opportunity to click a button that would send a message directly to the study coordinator through the electronic health record, Embi says.If the physician felt it was appropriate, and the patient agreed to be contacted by the study coordinator, then the doctor would click on that button, he explains.
Clicking on the button would document within the system that the conversation took place between the patient and physician and that the patient had given permission for the study coordinator to review his or her medical records. This made it compliant with HIPAA privacy requirements, Embi notes.
Also, when the physician clicked on the button, the system would send information about the clinical trial to an information sheet that would be handed to the patient as he or she left the clinic, Embi adds.
The written information would remind patients that they had given permission for the trial coordinator to contact them and that they might be called within the next couple of weeks. It also listed a phone number for them to call if they had any questions, and it informed them that they could change their mind about the trial at any time they chose, Embi says.
The Cleveland Clinic’s electronic health record was implemented throughout the institution and at all clinical sites, Embi says. "So while you’re in the room with the patient, you’re actually logged into the computer, and you’re doing your documentation and looking up labs for that patient."
The clinical trial alert system took less than a minute of doctors’ time, Embi notes.
Investigators spent about an hour or two adjusting the electronic health record to include the clinical trial alert system, Embi says.
"The hard part was figuring out how we were going to do it," Embi says. "We didn’t want to create something from scratch because that’s what has been done in the past, and that tends to make it difficult to make the system portable."
Investigators had to decide which tools to combine, which criteria to use, and to which physicians to send messages. They ultimately decided to send the messages only to endocrinologists and internists, Embi says.
They informed clinic administrators of the new alert system and informed doctors that they would activate the system and would study the physicians’ responses. In all, 48% of the physicians targeted responded to the alert notice, Embi says.
"At follow-up, even those who didn’t participate said it wasn’t intrusive and it wasn’t much hassle, and they wouldn’t mind seeing it in the future," Embi notes.
Although there was no cost analysis performed, anecdotal evidence suggests it was an efficient way to recruit subjects.
"From the perspective of the trial coordinator it was great," Embi says. "Her impression was that it was well worth it."
The clinical trial coordinator found the tool to be very effective when compared to other subject recruitment approaches, including some initiatives the site already was using, Embi notes.
Although there were more medical charts for the trial coordinator to review after the referrals increased 10-fold, it was fairly easy for her to determine which subjects would be suitable candidates without having to physically screen each patient referred, Embi says. "It was a little more efficient than it appeared," Embi says. "And it was very effective because it improved enrollments and didn’t involve that much more time on the part of the trial coordinator."
The next step is to adapt the same alert tool to another health care system and another type of electronic health record. Since the initial tool was built into a vendor-based electronic health record, which is similar to many others used in the U.S., researchers are optimistic the tool will work in many different products, Embi says.
"We got this to work in one health care system and with one electronic health record, and the question is, Can we get it to work elsewhere?’" Embi says. "We think we can, but we need to demonstrate this."
- Embi PJ, et al. Effect of a clinical trial alert system on physician participation in trial recruitment. Arch Intern Med. 2005;165:2272-2277.