Between September 2004 and September 2005, a total of 243 informants representing affiliations from a broad collection of settings that varied with respect to practice type, size, ownership, and access to technologies such as electronic medical records and web-based research tools, were interviewed. Interview participants were diverse by advanced degree (MD 70%, PhD 10%, DDS 6%, MD PhD 4%, RN nurse practitioner 3%, Master's degree 3%, and unknown 3%). Thirty percent of participants were female. Participants came from 35 different states.
For phase one interviews, we identified 106 potential informants and attempted interviews with 97. Amongst those attempted, we completed 73 (response rate 73/97, or 75%). Of those not interviewed, six declined to participate, one was unavailable, and 17 did not respond. Additional File 2
shows the different categories of informants and the number of each type interviewed during phase one. In many cases, respondents could be placed in more than one category, but for the purposes of this report we list only their primary role.
For phase two interviews, we identified 237 potential informants and attempted interviews with 204. Amongst those attempted, we completed 170 (response rate 170/204, or 83%) including interviews with 112 active clinicians. Fifteen informants participated in more than one interview. Of those who invited to participate, 16 of 33 community providers referred to the research team by a practice-based research network (PBRN) contact specifically for interview, declined because of their busy schedules. Additionally, four other invited participants were unavailable, and 19 did not respond. Additional File 3
shows the different categories of informants and the number of each type interviewed during phase two.
Across both phase one and phase two interviews, there were many cases in which respondents could be placed in more than one participant category. However, we categorize respondents only according to their primary role. For example, in addition to the 44 phase two providers listed below, 10 more clinicians with active practices were interviewed but are categorized as PBRN leaders, rather than clinicians. Several academic medical center leaders (>8) also maintain active clinical practices.
Interview themes documenting factors impeding clinician participation in research based in community settings
The factors identified that impede clinician participation in community-based research fall into three categories: the need for greater attention on the part of the research community to address concerns unique to community practitioners; the absence of necessary infrastructure; and the current alignment of financial incentives. Below, we introduce each of these categories with a quote from informants and then provide multiple dimensions of the factors as identified by informants. We address each of these perceived barriers along with possible strategies to overcome them in Table .
Barriers clinicians have identified regarding participation in clinical research
Category one: Need for greater attention to concerns of community practitioners
'If clinicians are recruited to participate in research activities and their participation is seen as valuable, as opposed to just being a passive partner for a study, then they'll come to the table to help with the conceptualization and contribute to the science and all of the project.' Quote from a clinician involved with research in community clinical practices settings.
Addressing community practice concerns
Clinicians as a group repeatedly expressed the belief that without acknowledgement of their potential contribution (via non-fiscal or fiscal recognition), they have little stake in clinical research and will not contribute in a sustained manner. When clinicians believe their voices are heard and responded to, they have more of a stake in clinical research and are more willing to respond to the inevitable challenges that arise.
Some of the mechanisms that were suggested by respondents to engage clinicians included: reframing research questions and study designs to increase meaning for community clinicians; attending to the complexities of the relationships between community and academics, which can become magnified in research studies; and addressing clinician and patient distrust of research.
Both clinicians and research leaders indicated a mechanism is needed to identify and focus on research questions that are of interest to community-based clinicians and patients, which could help close the gap between the existing research enterprise leadership and clinicians.
Physician participation in clinical research ultimately depends upon their belief that the research will benefit them and their patients.
Community clinicians repeatedly voiced their views that they have important contributions to make about which study designs are likely to be feasible in their practice settings.
Community clinicians indicated their interests would likely be better captured when study designs generate evidence to inform the complex clinical decisions practitioners like themselves make in their practices.
Table compares two categories of study designs pertinent to clinical research in community settings: explanatory and participatory (or practical) research trials. Combining explanatory and participatory trials may be an effective strategy for including community practice settings in research that aims to bridge the gap between basic science (e.g.
, the mapping of the genetic code) and clinical applications. Placebo controlled studies, may not be a viable option in some community practice clinical situations. However, trials can be designed to compare two different treatments of the same modality (e.g.
, comparing an investigative medication to a standard treatment for the same indication), two treatment modalities (e.g.
, medication vs. counseling), or one modality versus both [24
]. Trials also can be designed to inform the extraordinary challenges associated with translating the results of clinical research into clinical practice, ranging from the need to evaluate whether treatments shown to be efficacious in clinical trials continue to be effective in real-world practice, and the need to better understand the best ways to implement effective treatments across a wide variety of settings. The following sections describe strategies that are available to respond to concerns clinicians voiced regarding the evaluation of the implementation of effective strategies in community settings.
Table 4 Explanatory and practical clinical trials: Two options for clinical trials in community settings [7,42]
Concerns pertinent to the clinician's relationships with the clinical and scientific communities
Community clinicians expressed concerns that research participation might result in shifts in patient management from their own to academic practice settings.
To alleviate such concerns, study designs can specify that patients identified for research participation will maintain their relationships with their own clinicians for the bulk of their care.
Clinician and patient distrust of research
Despite the extraordinary advances in clinical care and outcomes that have emanated from research, longstanding distrust has also accumulated. Years after the Public Health Service (PHS) Syphilis Study in Tuskegee, Alabama, practicing clinicians expressed concerns that many patients remain mistrustful of research, fearing that risks are not fully disclosed, benefits may be exaggerated, and health information may be mishandled [27
]. Accordingly, many clinicians are reluctant to discuss potential research opportunities with their patients.
Expansion of research into community settings will need to build on recent efforts to improve trust and safety for patients engaged in community research efforts [30
]. Facilitating equitable and diverse involvement of clinicians and patients in clinical research can help achieve these goals as described by the Council of Public Representatives (COPR)[31
Category two: Absence of infrastructure
'A lot of the research can't necessarily be turned over entirely to the practices and the practitioners because they just don't have the manpower or the infrastructure. When we're able to send people out into the practices it's actually very helpful. What an expanded, national network can do is identify a study coordinator for each participating practice in their network. That person may or may not be one of the physicians. More often it's a staff, possibly a nurse or a clerk. They're responsible for the day-to-day operations of the project. When they are long-term, that provides a critical infrastructure ingredient for success.' Quote from an experienced research administrator who has led research in a variety of community settings.
Clinicians expressed concern regarding the lack of a permanent infrastructure to assist with identifying and choosing appropriate research opportunities, acquiring necessary data collection and other research skills and equipment, working with institutional review boards, and handling other challenges they would face if they agreed to participate in research.
Organizational supporting structures will need to become flexible enough to allow clinicians a voice in essential practice and other local decisions, while remaining durable enough to assure continued high-quality research endeavors. Resources for such an infrastructure will need to be allocated in advance, and initial costs will be high. Both central (e.g.
, NIH, national specialty societies) and local organizations could provide support, but in exchange, a cost-effective, stable research infrastructure will need to be established [7
Flow of information to clinicians
Clinicians cited not only a lack of knowledge about pending research opportunities but also a lack of information needed to make an informed decision about participation (both the appropriateness of their practices and their ability to accept such a commitment). To alleviate these concerns, the clinical research enterprise could provide a reliable and sustained flow of information about research opportunities and eligibility requirements. NIH has already launched their landmark effort to provide information on NIH-funded clinical trials to patients and clinicians [34
When clinicians have access to study questions and protocols before agreeing to participate, they can choose protocols they can successfully implement. A trial registry modeled upon the NIH Clinical Trials registry is one mechanism for providing potential clinician participants with regular updates, including study-specific enrollment and other information that would help them gauge the quantitative and qualitative value of their participation in studies. A clinical trial registry would also provide a venue for sharing trial data among participants of ongoing trials, potentially improving the quality of data reporting and ultimately the quality of research. Improving research quality, in addition to ensuring human protection and safety, may renew public confidence and trust in the clinical research enterprise. Quality control of community practice-based research will have to be rigorous throughout the research process. Inclusion of clinicians in a quality assurance system will make use of their knowledge of their own patients and their familiarity with their practice's operations, thus enabling design of a reliable and valid data collection and monitoring process that will work within their practice, while also demonstrating transparency to outside stakeholders. Uniform standards of training, credentialing, and quality oversight will be key.
Easing the burdens of IRBs and the Health Insurance Portability and Accountability Act (HIPAA)
Currently, all studies must undergo IRB approval, and high-risk studies receive ongoing review by a data safety monitoring board. Interviewees expressed many concerns about the challenges of navigating IRB requirements. To address such concerns, several leadership groups have called for the formation of a more standardized and centralized IRB, which has now occurred at the National Cancer Institute [33
Translational research will require more clinical training and mentoring on the research process and data integrity. Additionally, we will need to implement an extensive and rigorous system of accountability to assure proper human protection and safety. A centralized web-based process to provide HIPAA and IRB training and certification and a web-based registry of certified clinicians has already become standard in many settings. Ensuring accountability at the local level will involve the research sponsor, the principal investigator, data centers, laboratories, pharmacies, and clinicians as well as an organization with responsibility to assure real time accountability. Providing clinicians with standardized training in general and study-specific principles of research with a particular focus on reporting adverse outcomes will enhance the function of this process.
Category three: Current (mal)alignment of financial structures
'I think it does probably come down to being able to at least offset their time that's involved. And as you pointed out before, there may be some additional personnel costs that go above and beyond what the typical office personnel can handle in terms of patient education, monitoring forms, all the rest. So that's got to be part of the cost of doing credible clinical research.' Quote from an experienced clinician who has led research in community settings.
Realigning financial incentives
Voiced repeatedly in our interviews was the complaint that in the competitive and productivity-driven environment in which clinicians practice medicine today, research participation that disrupts patient flow, decreases staff efficiency, or otherwise threatens the economic viability of a practice, discourages further interest in participation.
Although non-financial incentives such as prestige, personal satisfaction, and improved patient care motivate clinicians to participate in research, these incentives cannot entirely substitute for financial compensation for services [36
]. When clinicians are only minimally involved with research, removal of the strongest disincentives (e.g.
, burdensome inclusion/exclusion criteria and inefficient data collection strategies) is likely to be appreciated. However, as the level of clinician effort and participation increases and clinicians become more familiar with clinical research, interviewees stated that monetary compensation becomes a stronger incentive than clinical research support measures.
Fair market compensation of clinicians' time
Currently, NIH-funded research relies heavily on clinician volunteerism and non-financial incentives [36
]. Developing and adopting a transparent and equitable system for compensation of clinicians' time and efforts in clinical research participation could result in a more stable cohort of research clinicians. Compensation could be guided by a principle of replacement value, compensating clinicians as if they were engaged in clinical care instead of research (for the equivalent effort and time) [38
Cost reimbursement for clinical activities
Clinicians considering research opportunities cited uncertainties regarding costs that they or insurers would be asked to bear. Setting and revealing reimbursement schedules for most common research tasks could greatly reduce financial uncertainty and help clinicians make informed participation decisions. A relative value scale (RVS)-like system with associated clinical payment structure [38
] that accounted for cost variation by geographic region, specialty, and clinical and prior research experience would allow sponsors, principal investigators, and research-associated organizations (such as PBRNs or academic research organizations) to better understand the financial costs and benefits of a research protocol before making a commitment. Further, protocols, budgets, and expected payments that estimate the actual cost of the activity would allow community clinicians to anticipate whether research participation would fit the patient care flow in their practice settings.
Improving predictability of research activities
A related concern clinicians cited was the lack of predictability of research activities both within one study and from one study to the next, resulting in problems with budgeting, management of time, space, staffing, and monitoring. If a practice can anticipate long-term research participation, it is more likely to adjust workflow to accommodate research than if it anticipated participating in only a single or occasional study. Table offers strategies to improve the predictability of clinical research.
Strategies to improve the predictability of research
Clinicians emphasized that assuring adequate and appropriate liability coverage is mandatory before they can actively participate in community research [39
]. Such assurances require coordination with clinical malpractice carriers and the coverage strategies used by research sponsors. A plan to address clinician liability coverage is most likely to achieve success if research sponsors and insurers collaborate with the private sector to address the needs of community clinicians. The distribution of costs and risks among research sponsors, clinicians, and patients, will need to ensure that research-related costs will not be born by either the physician or the patient [24