Optimizing Strategies to Maximize Patient Recruitment Yield

Recruitment of participants presents a significant and persistent bottle neck in the research process. Failure to reach participant enrollment targets may delay or halt study progression, wasting limited resources. Insufficient rates of participant accrual are the leading cause for termination among clinical trials with results posted on ClinicalTrials.gov. Additionally, studies must enroll representative samples to ensure generalizability and comply with NIH guidelines. Investigators often cite recruitment difficulties - including lack of interest, difficulty reaching specific populations, and limited resources - to explain failure to reach enrollment goals. However, these recruitment difficulties may not actually be due to a lack of willingness to participate among eligible patients. A national survey conducted by Harris Interactive in 2001 found that 85% of patients were either unaware or unsure that participation in a clinical trial was an option at the time of diagnosis. 75% of these patients said they would have been willing to enroll had they known it was possible. Innovation in recruitment is sorely needed to make recruitment more efficient and more representative.

In recent years, researchers have begun to leverage patient portals and electronic health records (EHRs) for recruitment. This method allows for engagement of broad populations within a health system in a cost-effective way. Emerging evidence of EHR recruitment shows promise. The researchers of this study have examined the use of EHR recruitment to identify and contact eligible patients via MyChart direct-to-patient messaging across a range studies within Emory Healthcare. Having completed a pilot program within the Emory system, this service is now available to all Emory investigators for whom MyChart recruitment is appropriate.

While the feasibility of MyChart messaging has been documented, little is known about best practices for how recruitment messages should be designed. There are, however, good reasons to believe that how messages are designed and framed may impact potential participants' responses to them.

This project is a platform trial that will allow for serial testing of various aspects of portal-based messaging. Consistent with platform trial design, the researchers will use a common set of outcome measures across various studies, and the portal-based messaging strategy allows for easy randomization (generally at the patient level) of patients to specific messaging approaches.

In the context of broader use of MyChart messaging, the researchers have designed this study to examine the impact of message design on patients' responses. Specifically, this study will examine the impact of including positively-valenced language that describes the value of the study and the impact of inclusion of information regarding any payments that participants will receive in the study. Both of these factors are ones that are grounded in behavioral economics and may be meaningful to potential participants.

The first proposed trial will involve randomly assigning patients who are identified as potentially eligible for studies using the MyChart recruitment mechanism to receive a recruitment message using 1 of 4 templates: 1) the current standard template; 2) a template with inclusion of positively-valenced language describing the study's value; 3) a template containing information about payment to participants; and 4) a template with both positively-valenced language and payment information. For all studies using the MyChart mechanism, all 4 messages will be created, and assignment will be done based on the last two digits of the patient's medical record number (MRN).

The primary endpoint is the rate at which individuals respond that they are interested in participating in the study to which they are invited. Additionally, this study will examine the non-response rate and the rate of individuals responding that they are not interested in the study. It will also track the rate at which individuals exposed to a MyChart recruitment message change their permission settings for contact about research opportunities. Study teams whose study invitations are used in this trial will be contacted about the rate of enrollment among patients who responded as "interested" to the message so that the researchers can estimate enrollment rate within each intervention group. Finally, select patients will participate in short interviews to understand their impressions of the message they were sent and their attitudes towards research more generally.