Background
Participant recruitment is a major challenge in many research studies involving human subjects. Recruitment involves a number of activities, including identifying eligible participants, adequately explaining the study to the potential participants, recruiting an adequate sample based on study goals and design, obtaining informed consent and maintaining ethical standards, and retaining participants until study completion. Findings from several studies suggest that recruitment often takes longer than anticipated, projects incur higher costs than expected, and scientists routinely overestimate the number of participants available for enrollment in their studies. Even in studies that succeed in recruiting large numbers of patients, participation rates arelow; only 3-20% of the eligible participant pool chooses to participate. A survey of research studies found that 34% recruited less than 75% of their planned sample, and this reduction in the sample size leads to reductions in the statistical power of the study. Despite the widely acknowledged fact that recruitment is a challenge, details about recruitment efforts and outcomes are rarely published, resulting in limited knowledge about why recruitment rates vary between studies and what the most successful approaches are for recruiting study participants.
NRC Activities
The AHRQ National Resource Center (NRC) has held a series of one-on-one discussions with AHRQ health information technology (IT) grantees funded under the Ambulatory Safety and Quality (ASQ) Program in order to understand how their projects are progressing and discuss the challenges they are facing. Several grantees, especially those in the Enabling Patient-Centered Care Through Health IT (PCC) initiative that are using IT to support communication and/or the exchange of health information with patients, indicated that they had challenges recruiting patient participants into their projects. It became apparent that grantees might benefit from talking more directly with one another about how to effectively and efficiently recruit participants. To facilitate that discussion, the NRC team summarized the recruitment approaches used across projects into background information that was shared with grantees during an open forum focused on this topic.
The open forum took place on July 8, 2009, with 15 participants across 9 projects. Its purpose was to provide an opportunity for information sharing and discussion of the specific ways in which grantees conduct recruitment. The attendees represented projects that targeted either individuals with a specific condition and/or within a specific demographic population, or the general adult population. Some projects focused on particular chronic conditions such as diabetes, hypertension, cancer, chest pain, and obstructive sleep apnea, while others were focused more generally on medication management or prevention.
Approaches and Challenges with Recruitment
Through the course of the NRC activities noted above, the team collected information on grantees' approaches to and challenges with recruitment from a variety of sources, including grant applications, grantees' quarterly progress reports, one-on-one discussions with grantees, and the open forum.
Recruitment approaches
- Grantees are employing a variety of strategies to make potential participants aware of their projects, including:
- Sending letters or postcards (often from the patient's provider);
- Distributing or displaying flyers, pamphlets, brochures, and/or posters;
- Approaching patients one-on-one in their primary care provider's office or local community clinic;
- Conducting or participating in health fairs and/or other community outreach activities;
- Placing phone calls, emails, or sending messages via patient portals or listservs; and,
- Advertising in university newsletters, local newspapers, radio, or television stations.
Several grantees mentioned that they have found it necessary to use more than one recruiting approach - one (or more) to inform potential participants about the project and additional strategies for followup. In addition, most do not recruit differently depending on factors such as patient demographics, type of care setting, or type of health IT application.
Recruitment challenges
The overarching recruitment challenge noted during both the one-on-one calls and the open forum was identifying participants that meet project-specific inclusion criteria. In addition, grantees noted difficulties in communicating the complicated information required during the consent process and in overcoming general resistance to participation in research studies, while still encouraging patients' interest. Grantees also mentioned language barriers as a challenge, in that their rate of enrollment is slowed when the project does not have enough staff members who can speak the primary language of their potential participants. Finally, grantees noted participants' limited access to, discomfort, and/or inexperience with computers as a challenge in recruiting for health IT research studies.
Noteworthy Recruitment Experiences and Guidelines
During the open forum, grantees shared their projects' recruitment approaches and provided each other with suggestions based on their experiences and project commonalities. Grantees did not comment on the differences between the effectiveness and efficacy of particular methods. Rather, they remarked on their particular chosen recruitment methods and discussed the differences between overarching categories of methods. The following key themes summarize the information provided during the open forum.
Grantees aim to minimize the burden on participants in terms of their monetary costs incurred, time spent, and initiative required to join the project. Most grantees provide incentives for those who join, compensating participants for their time. These incentives are often monetary compensation, but may also be gift cards from local businesses. Grantees also provide medical devices such as continuous positive airway pressure machines, glucometers, and blood pressure meters, which participants are often able to keep after the research studies have finished. Grantees make an effort to minimize patients' time necessary to participate by reducing the steps required through the screening, consenting, and enrollment processes. They also mentioned organizing recruitment around patients' pre-scheduled primary care visits and traveling to patient locations instead of requiring them to travel. Finally, some grantees consider passive approaches, such as general advertising, to be less effective because the onus is on the patient to take the initiative and contact the project, resulting in lower response rates. The grantees prefer to have the project personnel actively contact potential participants and meet with them individually.
Grantees develop partnerships to collaborate with community organizations or individuals with relevant expertise. Grantees spoke of the benefit they found in developing relationships with other organizations to meet their recruitment goals. These partners either are other researchers in the same field or organizations that fill a gap in the project team's expertise. For example, a grantee working with patients with a particular condition found it helpful to identify other researchers interested in that condition in order to share ideas and refer eligible patients. Another example is a grantee that, after developing the first set of recruitment materials, consulted with a marketing specialist who used her knowledge to review and provide guidance on how to improve the materials. Grantees also found value in reaching outside the health care community to churches and other community organizations that serve their population of interest. A grantee focusing on African-Americans with hypertension developed relationships with local churches to assist in recruitment. While grantees noted that they could not rely on other groups to replace their recruitment efforts, the relationships they developed enabled them to expand the pool of potential participants.
When trying to recruit particular patient populations based on clinical criteria, grantees use "indirect" recruitment methods. The projects that target particular sub-populations found it more efficient to first analyze clinical patient data to determine if individual patients meet the project's inclusion criteria, and then to contact only those eligible for participation. These projects rely less on "direct" recruitment methods, such as health fairs, where project personnel interact one-on-one with potential participants. The grantees noted that, even though these approaches are effective ways to recruit participants, they are not efficient because of the level of effort required of the project personnel to screen those eligible for the project. In these cases, several grantees noted the value of identifying a person the grantee trusts to introduce the project to the participants, such as their primary care provider or church leader. This "person of trust" builds a connection between patients and the project team. The project team is therefore often not the first to inform potential participants about the project. Instead, the team members contact patients after they are already familiar with the goals and scope of the research study.
Limited experience with technology does not appear to hinder patient participation in health IT studies. Based on the experience of this group of grantees, people who are less familiar with technology in general are not considerably less able to participate in health IT research studies. Grantees noted that, even when potential participants have limited experience with technology, the vast majority could be trained to use computers and the Internet within a relatively short period. However, depending on the complexity of the health IT application, training for the project can be a major component of the enrollment process. There may be a difference between the effort required to train a patient to use a computer sufficiently enough to access a personal health record (PHR) and the effort required to use that PHR productively. Training participants on the project's health IT application can absorb a large amount of time during enrollment, and researchers should plan for this accordingly.
External forces may affect whether people are willing and/or able to enroll in research projects. Grantees agreed that even when using exemplary recruitment methods and strategies, many eligible participants who could benefit clinically from these projects still are not interested in joining, and those who join often cannot be retained through the course of the project. In the majority of cases, potential participants do not to respond to recruitment efforts because of a general lack of interest. In addition, for projects that are studying services provided through a specific private health care organization, the participant may need to be able to afford health care services from that location to participate in the project. In those cases, other factors such as insurance status and income can prevent those interested from being able to participate in the study.
Conclusion
Recruiting participants into research studies is a challenging task that often requires a higher level of effort than anticipated. AHRQ health IT grantees funded under the Ambulatory Safety and Quality Program involved in patient recruitment for their health IT projects encountered issues similar to ones other researchers have faced in the past. Grantees noted the approaches they are using and identified the variety of challenges they face when recruiting participants into health IT research studies. By talking directly with other researchers who are also actively recruiting participants, the grantees have had the opportunity to share their concerns, hear suggestions from those most equipped to provide them, and make connections for future collaboration.