- Email: [email protected]
Implementing Research Findings into Practice Using Clinical Opinion Leaders: Barriers and Lessons Learned
Evidence-Based Medicine
Implementing Research Findings into Practice Using Clinical Opinion Leaders:
Geoffrey M. Curran, Ph.D. Carol R. Thrush, M.A. Jeffrey L. Smith, M.A. Richard R. Owen, M.D. Mona Ritchie, L.C.S.W. Dale Chadwick, M.B.A.
Barriers and Lessons Learned here is abundant literature on the effectiveness of various strategies for changing provider behavior,1,2 implementing clinical practice guidelines,3–7 and improving the quality of care.8 Beginning in March 2001, a project was designed to implement research findings about antipsychotic medication treatment for schizophrenia into routine practice.9 The project represented an effort to determine the effectiveness of a multicomponent intervention to improve clinician adherence to schizophrenia treatment guidelines, which recommend using moderate antipsychotic doses and the use of newer “atypical” antipsychotic agents for patients who fail to respond to conventional antipsychotics.10–12 Increasing the rates at which providers adhere to such key guideline recommendations is very likely to improve clinical outcomes for patients with schizophrenia.13–15 The project was supported by the U.S. Department of Veterans Affairs (VA) Quality Enhancement Research Initiative (QUERI), a national quality improvement (QI) program to promote the implementation of research discoveries and innovations into better patient care and systems improvement.16–18
T
Opinion Leaders: Conceptual Overview The concept of using opinion leaders as disseminators of research evidence and internal agents of change has been widely discussed.19–23 Opinion leaders are usually defined as persons who are respected sources of information and are able to exert influence on others’
700
December 2005
Article-at-a-Glance Background: An opinion leader-driven intervention to improve practice guideline–based medication management for patients with schizophrenia was tested at four Department of Veterans Affairs health care facilities. The concept of using opinion leaders as disseminators of research evidence and internal agents of change has been widely reported. Project Overview: Each intervention site received an intensive, multicomponent intervention during the course of one year. The project’s process evaluation included ongoing brief surveys of physicians’ attitudes and behaviors, logs of reports from opinion leader conference calls, and interviews with the opinion leaders toward the end of the implementation period. Barriors or Issues and Potential Solutions: Several barriers or problematic issues surfaced: (1) physicians do not always agree on who is an opinion leader; some sites may have no opinion leader; (2) some sites had poorly developed formal and informal social networks among physicians; (3) a focus on physicians only as agents of change; and (4) how much directive should be given to the opinion leaders concerning how to influence attitudes and behaviors? Discussion: Four major problematic issues encountered during the project offer potential solutions for addressing them.
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
decision making.19–20,24 They are usually defined as informal leaders, not necessarily serving in formal administrative roles.19,24,25 Certain physicians within larger medical practices and systems serve as informal educators and models for behavior on the basis of their knowledge, position or “standing” in the medical community, and interpersonal skills.26–28 This research lent support for social influence theory29 and social learning theory30 such that an individual’s clinical decision-making was at least partially contingent on peers’ values, beliefs, and behaviors (Mittman et al.31 provide a detailed discussion of social influence strategies in medical provider behavior change.). More recent research has addressed strategies to harness the influence on local opinion leaders in support of implementing evidence-based practices.23,32–34 A systematic review of randomized controlled trials using opinion leaders found mixed results in terms of their effectiveness on improving health care practices.23 Since this review, other randomized control trial evidence has been published,35,36 also with mixed results. Evidence from qualitative evaluations of nonrandomized studies suggests positive impacts from QI interventions involving opinion leaders.20 Recent efforts to solidify the conceptualization of change-agent strategies have sought to more firmly differentiate strategies using internal agents of change— opinion leaders—from strategies using external agents such as academic detailing and other “educational outreach” approaches.24,25 Academic detailing is consistently conceptualized as educational training by an expert, usually external to a practice group, who meets one-on-one with providers.31,25 More general educational outreach approaches may involve an external expert who provides a seminar or training session to a group of providers. Educational outreach is often used to train opinion leaders, providing them with additional expertise in targeted practices. In spite of the inconsistent conceptualization and somewhat mixed results from randomized trials associated with the literature on opinion leaders and other change agents, the theoretical promise of these change agents and strategies remains. Perhaps as a result, interventions using these models of change are numerous and growing. As indicated by Locock et al.,20 the
December 2005
literature will continue to face a “methodological conundrum: randomized controlled trials are the definitive way of proving whether an intervention is effective, but are not designed to test the complex, heterogeneous interventions…” (p. 755) of the kind currently being used in clinical QI. Qualitative research is better suited to examining the effectiveness and inter-relationships of change agent strategies that are so frequently “bundled” in complex interventions involving other change strategies. The intervention tested in the current study is just such a model, with numerous strategies and tools integrated into one intervention.
Conceptualization of the Opinion Leader’s Role in the MH QUERI Project The Mental Health QUERI (MHQ) team designed the opinion leader component of the intervention premised on a train-the-trainer model. The opinion leaders were envisioned to be internal facilitators, while the study investigators would be external facilitators working with/through them. To maximize the opportunity for success, MHQ sought to identify and recruit physicians in the settings who were already identified as opinion leaders in the treatment of schizophrenia, specifically those who were identified by their peers as being sought after for advice about clinical decisions. To be an active agent of change, the opinion leaders would: ■ Disseminate MHQ-compiled background information and evidence about the targeted clinical behaviors ■ Disseminate tools developed by MHQ to facilitate change such as pocket cards with recommended medication dose ranges ■ “Champion” the behavior change and the use of the tools with their peers, both formally (for example, in staff meetings) and informally (for example, in consultation with other providers); disseminate and discuss performance data provided by MHQ; monitor progress at their site ■ Report observations, issues and problems back to MHQ staff at regular intervals. MHQ’s role as external facilitators would be as follows: ■ Develop the materials for dissemination ■ Provide training to opinion leaders on the target clinical practices and issues of dissemination/social influence
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
701
■ Monitor the internal facilitation by the opinion leaders
and initiate change in the implementation strategy based upon their feedback (and review of performance data) MHQ’s plan was to limit its use of specific “directives” to the opinion leaders concerning their support of the implementation. Although more general issues of “disseminating” and “championing” would be discussed during their MHQ-facilitated training for the project, the opinion leaders would be encouraged to follow their own ideas for supporting implementation on the basis of their knowledge of the local clinical culture and structure.
Project Overview A multidisciplinary team of MHQ researchers set out to implement the project at a sample of Veterans Affairs medical centers in the South Central VA Healthcare Network, which includes a region spanning Arkansas, Louisiana, Mississippi, Oklahoma, and parts of Alabama, Texas, Missouri, and Florida. Using a quasi-experimental study design, the MHQ team selected four pairs of health care facilities in the South Central VA Network that were matched to the extent possible on baseline antipsychotic prescribing practices, number of patients with schizophrenia, number of mental health providers, and academic affiliation. From each matched pair, one facility was randomly selected to serve as an intervention site, with the other facility serving as a comparison (that is, control) site. Each of the four intervention sites received an intensive, multicomponent intervention during the course of one year, including the following: ■ Identification and training of local clinical opinion leaders ■ Dissemination of educational materials ■ Implementation of computerized clinical reminders ■ Monthly monitoring and feedback of data on antipsychotic prescribing practices ■ Ongoing support and monitoring from MHQ staff The four comparison sites were mailed educational materials and quarterly performance reports but received no other intervention. MHQ identified opinion leaders via surveys of physicians to identify (1) their current knowledge/support of practice guidelines for schizophrenia and (2) the peers to whom they turned to for information and advice. In sites
702
December 2005
where opinion leaders were clearly identified through this process, MHQ approached them directly and sought their involvement. In cases (discussed in more detail below) where leaders were not clearly identified, MHQ sought advice from the local mental health chiefs. The opinion leaders were instructed on the project’s goals, received education on the clinical practices to be addressed, reviewed and gave feedback on proposed intervention tools (for example, feedback reports and pocket cards), and discussed possible barriers and facilitators. The opinion leaders and MHQ staff discussed behaviors and opportunities for opinion leaders to employ in their formal and informal attempts at supporting the project and influencing their peers’ behavior. The leaders were informed that they were expected to present performance data on their facility at monthly staff meetings and were encouraged to champion the project with peers and to lead small group and one-on-one information/motivation sessions as they felt necessary. After the training, the intervention tools were modified based on opinion leader feedback. The intervention tools were then mailed to the opinion leaders, who disseminated them to the relevant providers. With the opinion leaders’ help, each site then hosted a kick-off meeting with psychiatrists and other staff members to discuss the project, the implementation tools, and the performance monitoring that would be taking place. After the kick-off meeting, the opinion leaders maintained regular contact with MHQ staff through weekly or twice-monthly telephone conference calls. MHQ monitored the activities of the opinion leaders, sought feedback from them on “how the project was going,” identified new or ongoing barriers to implementation, and explored potential solutions in partnership with the opinion leaders. The project’s main outcomes—for example, rates of changes in prescribing practices—are discussed elsewhere.9 The results were mixed. Some intervention sites significantly improved care as hypothesized, and some did not. Overall, there was improvement in antipsychotic dosing practices, and a substantial pharmacy cost reduction was observed. The project’s process evaluation included ongoing brief surveys of physicians’ attitudes and behaviors, logs of reports from opinion leader conference calls, and interviews with the opinion leaders
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
toward the end of the implementation period. Data from the process evaluation revealed that the sites appreciated and used the clinical support tools provided by MHQ, especially the laminated pocket cards containing the recommended antipsychotic medication dose ranges. Participants found the performance reports useful, but over time the sites’ enthusiasm for the project seemed to fade. Some physicians stopped looking at the performance data, and some opinion leaders struggled to maintain contact with peer physicians. Several barriers or problematic issues surfaced in the implementation of the opinion leader-driven intervention, some of which were not anticipated by the MHQ team. We now describe the key implementation barriers we faced and offer suggestions on how their impact might be reduced in future implementation projects.
Barriors/Issues and Potential Solutions 1. Physicians do not always agree on who is an opinion leader; some sites may have no opinion leader. As described earlier, MHQ’s opinion leader selection strategy was designed to identify existing opinion leaders at participating sites. The key indicators for leadership status that MHQ used were knowledge and support of VHA schizophrenia guidelines and evidence of influence on peer decision-making. Additional criteria might include, for example, “willingness to share knowledge” and lack of “talking down” to colleagues.19,32 Given the relatively small practice groups in these mental health settings (10–15 practitioners, many part-time), MHQ used more inclusive criteria for opinion leadership limited to knowledge, support of specific guideline recommended practices, and influence on peer decisionmaking. Interviews with key informants did not clearly identify knowledgeable, influential leaders at all four sites, and the results of the surveys were mixed. Physicians at one site clearly selected the lead physician administrator in the mental health service (that is, the mental health chief), another site had a tie for two leaders with five votes each, and no obvious leader was identified by the survey at the other two sites. Several physicians indicated that they themselves were the best resource of information, and some responded that they would just “look up the answer myself.” (At the sites with no clear leaders, the mental
December 2005
health chiefs were asked to nominate possible candidates based on their assessment of the training, knowledge level, clinical competence, and leadership abilities.) Other interventions using similar (or more stringent) techniques have encountered similar problems when identifying opinion leaders.20 For example, Young et al.24 reported that most physician respondents (surgeons) to their survey could not identify an influential opinion leader in their own hospital or specialty. Potential Solutions: When opinion leaders are not easily identified through survey techniques or key informant interviews, additional steps could be taken. For example, when surveys reveal no clear leaders, it may be useful to interview the range of individuals who received at least a few votes. Through the interviews, the candidates can be directly compared on issues such as knowledge in the area being addressed, support for change, connectedness both inside and outside the organization, and perceived influence on peers. If a lack of suitable opinion leadership still ensues, the situation begs the question, “Can opinion leaders be created?” Although it may be difficult to train a person to be a source of knowledge and peer influence across a wide spectrum of clinical expertise and behaviors, it may be possible to train a person to be an opinion leader in more narrowly focused areas of expertise. 2. Some sites had poorly developed formal and informal social networks among physicians. An important element of MHQ’s opinion leader strategy was that the opinion leaders would have both formal (staff meetings) and informal opportunities to meet with peers to discuss the project and focus on the implementation process. The MHQ team believed that opinion leaders should or would have these opportunities or at least that geographic proximity would ensure contact with their peers on a routine basis. Despite the opinion leaders’ assurances about their availability, it became clear during the project, however, that the organizational, and in some cases, physical structures of the sites created barriers to the use of opinion leaders as agents of change. Only one of the four participating locations was structured to accommodate routine contact among the group of physicians and other caregivers in the mental health clinics (all clinical staff participated in daily formal communication briefings, and the proximity of physician
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
703
offices in one area was conducive to informal contact). In some of the other sites, however, physicians were not only geographically separated (for example, different hospital wings or different buildings spread out across many miles), but there were few mechanisms in place whereby opinion leaders and physicians could meet among themselves, either formally or informally. Most sites had monthly medical staff meetings, but not all physicians were able to attend. At one site, some physicians “called in” to the meeting but appeared to not be full participants. At another site the appointed time for the meeting did not coincide with the scheduled hours of many of the parttime physicians, and as a result they hardly ever attended these meetings. In general, the part-time physicians seemed less connected to the existing structures in place in the clinics. As a result of these organizational and physical barriers, the level of contact between the opinion leaders and the physicians in their clinics was suboptimal. Potential Solutions: The goal in this situation is to create more opportunities for interaction and networking among providers. These opportunities can potentially occur in different formats on site and in person, as well as on the phone and/or via the Internet. To varying degrees, the intervention sites were able to arrange kickoff meetings that drew together sizable portions of the participating providers. Such meetings could be replicated at intervals during the project, adhering to the notion of “booster sessions” common in prevention and sustainability work. The on-site opinion leaders could lead these gatherings. Efforts could be made to include as many part-time staff as possible, including having rotating meeting times to accommodate more people. These forums could (1) provide continued exposure to the project’s goals, (2) provide an opportunity for participants to offer “testimonials” about successes, (3) provide an opportunity for leadership to express support, (4) provide an opportunity to present and discuss performance data and to brainstorm and discuss ideas for improvement, (5) provide a forum for questions and comments from staff members, (6) provide an opportunity to openly discuss any problems or additional barriers arising during the implementation of the project, and (7) provide an opportunity to allow new staff members to be oriented to the project (as turnover is common in many health care settings).
704
December 2005
It might also be helpful to use various technological solutions to address the problem of poor contact among project participants either through use of the Internet and/or e-mail to connect participants within a site (or across project sites). One possibility would be a listserv whereby project staff, participants, and opinion leaders could disseminate project information and seek input and present ideas for discussion. Another possibility would be an Internet “community of practice” to allow for training and education, disseminating/ archiving data and reports, chat rooms for real-time discussion, and electronic note boards for questions or comments. Some projects are currently using such technologies with apparent success.27–29 Before considering any of these ideas, it would be important to assess the current means/quality of communication and consult with the participants to get a sense of their preferences and capabilities. Also, it is important to learn of any existing means of communication and contact that is successful so that it can be incorporated into planning the implementation project. 3. A focus on physicians only as agents of change. The project lacked: (1) sufficient and timely attention to the administrative, monitoring, and reporting functions associated with the implementation of the change model as devised; and (2) sufficient involvement from other key stakeholders in the process of medication management. As the project was devised, the physician opinion leaders were to be involved in multiple tasks, as described earlier. During implementation, however, it became clear that most of them did not have sufficient time to devote to all of these tasks. Scheduling training meetings, conference calls, and other project-related activities with them was particularly challenging. The physician opinion leaders spent most of their time seeing patients, and many of the clinics in which they worked were understaffed. Time available for peer interaction, dissemination, and monitoring of progress was limited. It became clear that other clinical team members, such as pharmacists and nurses, could have been helpful in providing support to the medication managementfocused project. Potential Solutions: The most obvious course of action in this case is to involve more clinic team members in the implementation process. Pharmacists (especially
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
at the Pharm.D. level) could be helpful in disseminating background information on medications, providing oneon-one educational sessions (academic detailing) on medication management issues, and championing the case for change. Nurses or social workers could be helpful in monitoring the change process in the clinic and discussing observations with study staff and/or clinic leadership. Nurses may also be effective in encouraging and supporting providers to use evidence-based practices (depending on the degree of collaboration experienced). Administrative assistants or clinic clerks could help disseminate tools and performance reports from the study to clinical staff. With this kind of support, the physician opinion leaders could focus on the peer-to-peer interactions, monitoring of performance, and leading discussion in formal staff meetings. This team-based approached might also have the added effects of increasing visibility and buy-in of the project throughout the organization. 4. How much directive should be given to the opinion leaders concerning how to influence attitudes and behaviors? During the pre-implementation training session for the opinion leaders, the MHQ team discussed general dissemination and implementation ideas regarding the project, but MHQ did not provide specific instructions for the opinion leaders. The reasoning was to preserve the personal style of each opinion leader and to allow some flexibility in implementation at each site. As noted earlier, our intervention was based on the notion that existing opinion leaders would already have been exerting influence on their peers and the system of care through their knowledge, connectedness, and their own style and methods of “persuasion.” Local opinion leaders were encouraged to be flexible in implementing the project according to local needs and barriers while working with their clinical service leadership. During implementation, however, it became apparent that more specific instruction and training for the opinion leaders would have been helpful. At times, some of the opinion leaders asked for direction from the research team, for example, “I’ve disseminated the materials; what else would you like me to do?” The research team attempted to be more directive about ways they could increase the project’s visibility and to help fellow physicians to incorporate evidence and guideline recom-
December 2005
mendations into their thought processes for care delivery. These efforts may have helped to some extent, but in some instances it was likely too little, too late. Potential Solutions: To help address this issue, we believe that a strong pre-implementation training program for opinion leaders in behavior change, adoption of innovation, and opinion leadership would be a good first step. MHQ recommends that such a session (conceived here for a multisite project) should provide some brief theoretical background in: (1) behavior change at the individual level—such as social cognitive theory30 and the theory of planned behavior,37 (2) behavior change at the organization/systems level—such as recent work in the areas of complex adaptive systems38 and organization change,39 and (3) the adoption of innovations.40 The session should provide an overview of the effectiveness of the range of tools and strategies that have been used in QI efforts in similar settings (for example, clinical reminders, performance feedback, academic detailing, and use of opinion leaders). Specific strategies of effective opinion leadership should be covered in detail. The training session might then focus on the local circumstances of the opinion leaders. They could share their assessments of their sites’ readiness to change, discuss current efforts in QI in other areas, and outline projected barriers and facilitators. In preparation for this, they should be given a list of questions in advance so that they can come to the training prepared to discuss these issues. The training session should then focus on specific ideas for influencing change concerning the proposed innovation. Dissemination and implementation plans could then be made in the group setting, with the team of opinion leaders devising strategies together. It may also be helpful if the interaction among the opinion leaders at different sites is continued after the training. Brief conference calls (no longer than one patient appointment) held periodically could allow leaders to share their implementation experiences and provide advice and support. If possible, a second face-toface experience at the project’s mid-point would be ideal. As always, process evaluation41 should be undertaken to understand how the training and monitoring processes worked or did not, and how best to modify efforts for future work. Indeed, this article is the direct product of just such an evaluation.
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
705
Discussion We have detailed our experiences relating to facilitation—both at the internal level (the opinion leaders) and at the external level (MHQ facilitation of opinion leader selection/training and overall monitoring of the project). The project encountered a host of challenges related to evidence and context—the ever and rapidly changing evidence base, conflicting interpretations of published findings, staff turnover, and conflicting demands at many levels within the clinic system, to name several. Indeed, implementation work is highly complex, and while there are several theoretical or conceptual models to pursue for guidance, there remains a need for the literature to document the field-level successes and failures of these models and the components inspired by them. The process of identifying, understanding, and addressing barriers is an essential component of implementation research. Our experiences will continue to inform our future work and, hopefully, that of others who use opinion leaders as agents of change in health care settings. The VA system and sites studied in this project may have some unique characteristics that would limit the generalizability of our findings; other systems of care will undoubtedly present their own unique barriers and facilitators to the use of opinion leaders. One area that may be ripe for future research is in the area of training and development of opinion leaders. We posed the question, “Can opinion leaders be created?” in response to some sites’ inability to clearly identify opinion leaders. These sites pursued a clinical-champion strategy—the use of a personal motivator for change. Such change agents may not necessarily have followers or skills of influence—but may
be willing to learn the clinical and social influence skills necessary. Another promising research area concerns the optimal mix of disciplines and team members for an opinion leader-driven intervention. How many opinion leaders are needed? Is it possible to have too many opinion leaders at a given site? How best should the leadership team operate? Can opinion leaders from multiple disciplines work synergistically? Another issue in need of greater investigation is how to facilitate social network interactions to improve opinion leader effectiveness. This work has been supported by a grant from the Department of Veterans Affairs, Health Services Research and Development Service (MNT 01-033-1), by the Mental Health Quality Enhancement Research Initiative (QUERI) and by the South Central Mental Illness Research, Education, and Clinical Center (MIRECC). Dr. Curran is supported by a Career Development Award from NIDA (K01 DA15102). The authors thank the opinion leaders and the providers who willingly participated in this project as well as the MH QUERI project team who helped make this work possible. The views expressed in the article are those of the authors and do not necessarily reflect views of the Department of Veterans Affairs.
Geoffrey M. Curran, Ph.D., is Research Health Scientist, Central Arkansas Veterans Healthcare System, Little Rock, Arkansas. Carol R. Thrush, M.A, is Assistant Professor, Office of Educational Development, University of Arkansas for Medical Sciences, Little Rock. Jeffrey L. Smith, M.A., is Implementation Research Coordinator, Mental Health QUERI Coordinating Center, Central Arkansas Veterans Healthcare System. Richard R. Owen, M.D., is Research Coordinator and Mona Ritchie, L.C.S.W., is Project Coordinator, Health QUERI Coordinating Center, Central Arkansas Veterans Healthcare System. Dale Chadwick, M.B.A., is Director, Lowell Treatment Center, Lowell, Massachusetts. Please send reprint requests to Geoffrey M. Curran, Ph.D., [email protected].
References 1. Rubenstein L.V. et al.: From understanding health care provider behavior to improving health care: The QUERI framework for quality improvement. Quality Enhancement Research Initiative. Med Care 38:I129–41, Jun. 2000. 2. Wyszewianski L., Green L.A.: Strategies for changing clinicians’ practice patterns: A new perspective. J Fam Pract 49:461–464, May 2000. 3. Davis D.A., Taylor-Vaisey A.: Translating guidelines into practice: A systematic review of theoretic concepts, practical experience, and research evidence in the adoption of clinical practice guidelines. CMAJ 157:408–416, Aug. 15, 1997.
706
December 2005
4. Drake R.E., et al.: Implementing evidence-based practices in routine mental health service settings. Psychiatr Serv 52:179–182, Feb. 2001. 5. Gates P.E.: Think globally, act locally: An approach to implementation of clinical practice guidelines. Jt Comm J Qual Improv 21:71–84 Feb. 1995. 6. Grol R., Grimshaw J.: Evidence-based implementation of evidencebased medicine. Jt Comm J Qual Improv 25:503–513, Oct. 1999. 7. Solberg L.I., et al: Lessons learned from experienced guideline implementers: Attend to many factors and use multiple strategies. Jt Comm J Qual Improv 26:171–188, Apr. 2000.
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
References, continued 8. Weber V., Joshi M.S.: Effecting and leading change in health care organizations. Jt Comm J Qual Improv 26:388–399, Jul. 2000. 9. Owen R.R., et al.: Use of electronic medical record data for quality improvement in schizophrenia treatment. J Am Med Inform Assoc 11:351–357. Sep.–Oct. 2004. 10. Mental Health Strategic Health Care Group, The Psychosis Working Group: Veterans Health Administration Clinical Guideline for Management of Persons with Psychoses. Washington, D.C.: Department of Veterans Affairs, 1997. 11. Lehman A.F., Steinwachs D.M.: Translating research into practice: The Schizophrenia Patient Outcomes Research Team (PORT) treatment recommendations. Schizophr Bull 24(1):1–10, 1998. 12. Practice guideline for the treatment of patients with schizophrenia. American Psychiatric Association. Am J Psychiatry 154:1–63, Apr. 1997. 13. Dixon L.B., Lehman A.F., Levine J.: Conventional antipsychotic medications for schizophrenia. Schizophr Bull 21(4):567–577, 1995. 14. Baldessarini R.J., Cohen B.M., Teicher M.H.: Significance of neuroleptic dose and plasma level in the pharmacological treatment of psychoses. Arch Gen Psychiatry 45:79–91, Jan. 1988. 15. Owen R.R., et al.: Performance measurement for schizophrenia: Adherence to guidelines for antipsychotic dose. Int J Qual Health Care 12:475–482, Dec. 2000. 16. Feussner J.R., Kizer K.W., Demakis J.G.: The Quality Enhancement Research Initiative (QUERI): From evidence to action. Med Care 38:I1–6, Jun. 2000. 17. Demakis J.G., et al.: Quality Enhancement Research Initiative (QUERI): A collaboration between research and clinical practice. Med Care 38:I17–25, Jun. 2000. 18. Fischer E.P., et al.: Quality Enhancement Research Initiative in Mental Health. Med Care 38:I70–81, Jun. 2000. 19. Borbas C., et al.: The role of clinical opinion leaders in guideline implementation and quality improvement. Chest 118:24S–32S, Aug. 2000. 20. Locock L., et al: Understanding the role of opinion leaders in improving clinical effectiveness. Soc Sci Med 53:745–757, Sep. 2001. 21. Lomas J., et al.: Opinion leaders vs audit and feedback to implement practice guidelines: Delivery after previous cesarean section. JAMA 265:2202–2207, May 1, 1991. 22. Gates P.E.: Think globally, act locally: An approach to implementation of clinical practice guidelines. Jt Comm J Qual Improv 21:71–84, Feb. 1995. 23. Thomson O’Brien M.A., et al.: Local opinion leaders: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev (2):CD000125, 2000.
December 2005
24. Young J.M., et al.: Role for Opinion Leaders in Promoting EvidenceBased Surgery. Arch Surg 138: 785–791, Jul. 2003. 25. Harvey G., et al.: Getting evidence into practice: The role and function of facilitation. J Adv Nurs 37:577–588, Mar. 2002. 26. Greer Al: The state of the art versus the state of science. Int J Technol Assess Health Care 4:5–26, 1988. 27. Coleman J.S., Katz E., Menzel H.: Medical Innovation: A Diffusion Study. New York: Bobbs-Merrill, 1966. 28. Wenrich, J.W., et al.: Informal educators for practicing physicians. J Med Educ 46:299–305, 1971. 29. Zimbardo P.G., Leippe M.R.: The Psychology of Attitude Change and Social Influence. Philadelphia: Temple University Press, 1991. 30. Bandura A.: Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall, 1986. 31. Mittman B.S., Tonesk X., Jacobson P.D.: Implementing clinical practice guidelines: Social influence strategies and practitioner behavior change. QRB Qual Rev Bull 18:413–422, Dec. 1992. 32. Hiss R.G., MacDonald R., David W.R.: Identification of physician educational influence in small community hospitals. Proceedings of the Seventeenth Annual Conference on Research in Medical Education. Washington D.C., 1978, pp. 283–288, 1978. 33. Soumerai S.B., et al.: Effect of local medical opinion leaders on quality of care for acute myocardial infarction: A randomized controlled trial. JAMA 279:1358–1363, May 6, 1998. 34. Elliott T.E., et al.: Improving cancer pain management in communities: Main results from a randomized controlled trial. J Pain Symptom Manage 13:191–203, Apr. 1997. 35. Gifford D.R., et al.: Improving adherence to dementia guidelines through education and opinion leaders: A randomized, controlled trial. Ann Intern Med 131: 237–246, Aug. 17, 1999. 36. Berner E.S., et al.: Do local opinion leaders augment hospital quality improvement efforts? A randomized trial to promote adherence to unstable angina guidelines. Med Care 41:420–431, Mar. 2003. 37. Ajzen I.: The theory of planned behavior. Organizational Behavior and Human Decision Processes. 50:179–211, Dec. 1991. 38. McDaniel R.R., Driebe D.J.: Complexity science and health care management. In Fottler M.D., Savage G.T., Blair J.D. (eds.): Advances in Health Care Management vol. 2. Greenwich, CT: JAI Press, 2001, pp. 11–36. 39. Burke W.W.: Organizational Change: Theory and Practice. Thousand Oaks, CA: Sage Publications, 2003. 40. Rogers E.M.: Diffusion of Innovations. New York: Free Press, 1995. 41. Hulscher M.E., Laurant M.G., Grol R.P.: Process evaluation on quality improvement interventions. Qual Saf Health Care 12:40–46, Feb. 2003.
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
707
Implementing Research Findings into Practice Using Clinical Opinion Leaders:
Geoffrey M. Curran, Ph.D. Carol R. Thrush, M.A. Jeffrey L. Smith, M.A. Richard R. Owen, M.D. Mona Ritchie, L.C.S.W. Dale Chadwick, M.B.A.
Barriers and Lessons Learned here is abundant literature on the effectiveness of various strategies for changing provider behavior,1,2 implementing clinical practice guidelines,3–7 and improving the quality of care.8 Beginning in March 2001, a project was designed to implement research findings about antipsychotic medication treatment for schizophrenia into routine practice.9 The project represented an effort to determine the effectiveness of a multicomponent intervention to improve clinician adherence to schizophrenia treatment guidelines, which recommend using moderate antipsychotic doses and the use of newer “atypical” antipsychotic agents for patients who fail to respond to conventional antipsychotics.10–12 Increasing the rates at which providers adhere to such key guideline recommendations is very likely to improve clinical outcomes for patients with schizophrenia.13–15 The project was supported by the U.S. Department of Veterans Affairs (VA) Quality Enhancement Research Initiative (QUERI), a national quality improvement (QI) program to promote the implementation of research discoveries and innovations into better patient care and systems improvement.16–18
T
Opinion Leaders: Conceptual Overview The concept of using opinion leaders as disseminators of research evidence and internal agents of change has been widely discussed.19–23 Opinion leaders are usually defined as persons who are respected sources of information and are able to exert influence on others’
700
December 2005
Article-at-a-Glance Background: An opinion leader-driven intervention to improve practice guideline–based medication management for patients with schizophrenia was tested at four Department of Veterans Affairs health care facilities. The concept of using opinion leaders as disseminators of research evidence and internal agents of change has been widely reported. Project Overview: Each intervention site received an intensive, multicomponent intervention during the course of one year. The project’s process evaluation included ongoing brief surveys of physicians’ attitudes and behaviors, logs of reports from opinion leader conference calls, and interviews with the opinion leaders toward the end of the implementation period. Barriors or Issues and Potential Solutions: Several barriers or problematic issues surfaced: (1) physicians do not always agree on who is an opinion leader; some sites may have no opinion leader; (2) some sites had poorly developed formal and informal social networks among physicians; (3) a focus on physicians only as agents of change; and (4) how much directive should be given to the opinion leaders concerning how to influence attitudes and behaviors? Discussion: Four major problematic issues encountered during the project offer potential solutions for addressing them.
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
decision making.19–20,24 They are usually defined as informal leaders, not necessarily serving in formal administrative roles.19,24,25 Certain physicians within larger medical practices and systems serve as informal educators and models for behavior on the basis of their knowledge, position or “standing” in the medical community, and interpersonal skills.26–28 This research lent support for social influence theory29 and social learning theory30 such that an individual’s clinical decision-making was at least partially contingent on peers’ values, beliefs, and behaviors (Mittman et al.31 provide a detailed discussion of social influence strategies in medical provider behavior change.). More recent research has addressed strategies to harness the influence on local opinion leaders in support of implementing evidence-based practices.23,32–34 A systematic review of randomized controlled trials using opinion leaders found mixed results in terms of their effectiveness on improving health care practices.23 Since this review, other randomized control trial evidence has been published,35,36 also with mixed results. Evidence from qualitative evaluations of nonrandomized studies suggests positive impacts from QI interventions involving opinion leaders.20 Recent efforts to solidify the conceptualization of change-agent strategies have sought to more firmly differentiate strategies using internal agents of change— opinion leaders—from strategies using external agents such as academic detailing and other “educational outreach” approaches.24,25 Academic detailing is consistently conceptualized as educational training by an expert, usually external to a practice group, who meets one-on-one with providers.31,25 More general educational outreach approaches may involve an external expert who provides a seminar or training session to a group of providers. Educational outreach is often used to train opinion leaders, providing them with additional expertise in targeted practices. In spite of the inconsistent conceptualization and somewhat mixed results from randomized trials associated with the literature on opinion leaders and other change agents, the theoretical promise of these change agents and strategies remains. Perhaps as a result, interventions using these models of change are numerous and growing. As indicated by Locock et al.,20 the
December 2005
literature will continue to face a “methodological conundrum: randomized controlled trials are the definitive way of proving whether an intervention is effective, but are not designed to test the complex, heterogeneous interventions…” (p. 755) of the kind currently being used in clinical QI. Qualitative research is better suited to examining the effectiveness and inter-relationships of change agent strategies that are so frequently “bundled” in complex interventions involving other change strategies. The intervention tested in the current study is just such a model, with numerous strategies and tools integrated into one intervention.
Conceptualization of the Opinion Leader’s Role in the MH QUERI Project The Mental Health QUERI (MHQ) team designed the opinion leader component of the intervention premised on a train-the-trainer model. The opinion leaders were envisioned to be internal facilitators, while the study investigators would be external facilitators working with/through them. To maximize the opportunity for success, MHQ sought to identify and recruit physicians in the settings who were already identified as opinion leaders in the treatment of schizophrenia, specifically those who were identified by their peers as being sought after for advice about clinical decisions. To be an active agent of change, the opinion leaders would: ■ Disseminate MHQ-compiled background information and evidence about the targeted clinical behaviors ■ Disseminate tools developed by MHQ to facilitate change such as pocket cards with recommended medication dose ranges ■ “Champion” the behavior change and the use of the tools with their peers, both formally (for example, in staff meetings) and informally (for example, in consultation with other providers); disseminate and discuss performance data provided by MHQ; monitor progress at their site ■ Report observations, issues and problems back to MHQ staff at regular intervals. MHQ’s role as external facilitators would be as follows: ■ Develop the materials for dissemination ■ Provide training to opinion leaders on the target clinical practices and issues of dissemination/social influence
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
701
■ Monitor the internal facilitation by the opinion leaders
and initiate change in the implementation strategy based upon their feedback (and review of performance data) MHQ’s plan was to limit its use of specific “directives” to the opinion leaders concerning their support of the implementation. Although more general issues of “disseminating” and “championing” would be discussed during their MHQ-facilitated training for the project, the opinion leaders would be encouraged to follow their own ideas for supporting implementation on the basis of their knowledge of the local clinical culture and structure.
Project Overview A multidisciplinary team of MHQ researchers set out to implement the project at a sample of Veterans Affairs medical centers in the South Central VA Healthcare Network, which includes a region spanning Arkansas, Louisiana, Mississippi, Oklahoma, and parts of Alabama, Texas, Missouri, and Florida. Using a quasi-experimental study design, the MHQ team selected four pairs of health care facilities in the South Central VA Network that were matched to the extent possible on baseline antipsychotic prescribing practices, number of patients with schizophrenia, number of mental health providers, and academic affiliation. From each matched pair, one facility was randomly selected to serve as an intervention site, with the other facility serving as a comparison (that is, control) site. Each of the four intervention sites received an intensive, multicomponent intervention during the course of one year, including the following: ■ Identification and training of local clinical opinion leaders ■ Dissemination of educational materials ■ Implementation of computerized clinical reminders ■ Monthly monitoring and feedback of data on antipsychotic prescribing practices ■ Ongoing support and monitoring from MHQ staff The four comparison sites were mailed educational materials and quarterly performance reports but received no other intervention. MHQ identified opinion leaders via surveys of physicians to identify (1) their current knowledge/support of practice guidelines for schizophrenia and (2) the peers to whom they turned to for information and advice. In sites
702
December 2005
where opinion leaders were clearly identified through this process, MHQ approached them directly and sought their involvement. In cases (discussed in more detail below) where leaders were not clearly identified, MHQ sought advice from the local mental health chiefs. The opinion leaders were instructed on the project’s goals, received education on the clinical practices to be addressed, reviewed and gave feedback on proposed intervention tools (for example, feedback reports and pocket cards), and discussed possible barriers and facilitators. The opinion leaders and MHQ staff discussed behaviors and opportunities for opinion leaders to employ in their formal and informal attempts at supporting the project and influencing their peers’ behavior. The leaders were informed that they were expected to present performance data on their facility at monthly staff meetings and were encouraged to champion the project with peers and to lead small group and one-on-one information/motivation sessions as they felt necessary. After the training, the intervention tools were modified based on opinion leader feedback. The intervention tools were then mailed to the opinion leaders, who disseminated them to the relevant providers. With the opinion leaders’ help, each site then hosted a kick-off meeting with psychiatrists and other staff members to discuss the project, the implementation tools, and the performance monitoring that would be taking place. After the kick-off meeting, the opinion leaders maintained regular contact with MHQ staff through weekly or twice-monthly telephone conference calls. MHQ monitored the activities of the opinion leaders, sought feedback from them on “how the project was going,” identified new or ongoing barriers to implementation, and explored potential solutions in partnership with the opinion leaders. The project’s main outcomes—for example, rates of changes in prescribing practices—are discussed elsewhere.9 The results were mixed. Some intervention sites significantly improved care as hypothesized, and some did not. Overall, there was improvement in antipsychotic dosing practices, and a substantial pharmacy cost reduction was observed. The project’s process evaluation included ongoing brief surveys of physicians’ attitudes and behaviors, logs of reports from opinion leader conference calls, and interviews with the opinion leaders
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
toward the end of the implementation period. Data from the process evaluation revealed that the sites appreciated and used the clinical support tools provided by MHQ, especially the laminated pocket cards containing the recommended antipsychotic medication dose ranges. Participants found the performance reports useful, but over time the sites’ enthusiasm for the project seemed to fade. Some physicians stopped looking at the performance data, and some opinion leaders struggled to maintain contact with peer physicians. Several barriers or problematic issues surfaced in the implementation of the opinion leader-driven intervention, some of which were not anticipated by the MHQ team. We now describe the key implementation barriers we faced and offer suggestions on how their impact might be reduced in future implementation projects.
Barriors/Issues and Potential Solutions 1. Physicians do not always agree on who is an opinion leader; some sites may have no opinion leader. As described earlier, MHQ’s opinion leader selection strategy was designed to identify existing opinion leaders at participating sites. The key indicators for leadership status that MHQ used were knowledge and support of VHA schizophrenia guidelines and evidence of influence on peer decision-making. Additional criteria might include, for example, “willingness to share knowledge” and lack of “talking down” to colleagues.19,32 Given the relatively small practice groups in these mental health settings (10–15 practitioners, many part-time), MHQ used more inclusive criteria for opinion leadership limited to knowledge, support of specific guideline recommended practices, and influence on peer decisionmaking. Interviews with key informants did not clearly identify knowledgeable, influential leaders at all four sites, and the results of the surveys were mixed. Physicians at one site clearly selected the lead physician administrator in the mental health service (that is, the mental health chief), another site had a tie for two leaders with five votes each, and no obvious leader was identified by the survey at the other two sites. Several physicians indicated that they themselves were the best resource of information, and some responded that they would just “look up the answer myself.” (At the sites with no clear leaders, the mental
December 2005
health chiefs were asked to nominate possible candidates based on their assessment of the training, knowledge level, clinical competence, and leadership abilities.) Other interventions using similar (or more stringent) techniques have encountered similar problems when identifying opinion leaders.20 For example, Young et al.24 reported that most physician respondents (surgeons) to their survey could not identify an influential opinion leader in their own hospital or specialty. Potential Solutions: When opinion leaders are not easily identified through survey techniques or key informant interviews, additional steps could be taken. For example, when surveys reveal no clear leaders, it may be useful to interview the range of individuals who received at least a few votes. Through the interviews, the candidates can be directly compared on issues such as knowledge in the area being addressed, support for change, connectedness both inside and outside the organization, and perceived influence on peers. If a lack of suitable opinion leadership still ensues, the situation begs the question, “Can opinion leaders be created?” Although it may be difficult to train a person to be a source of knowledge and peer influence across a wide spectrum of clinical expertise and behaviors, it may be possible to train a person to be an opinion leader in more narrowly focused areas of expertise. 2. Some sites had poorly developed formal and informal social networks among physicians. An important element of MHQ’s opinion leader strategy was that the opinion leaders would have both formal (staff meetings) and informal opportunities to meet with peers to discuss the project and focus on the implementation process. The MHQ team believed that opinion leaders should or would have these opportunities or at least that geographic proximity would ensure contact with their peers on a routine basis. Despite the opinion leaders’ assurances about their availability, it became clear during the project, however, that the organizational, and in some cases, physical structures of the sites created barriers to the use of opinion leaders as agents of change. Only one of the four participating locations was structured to accommodate routine contact among the group of physicians and other caregivers in the mental health clinics (all clinical staff participated in daily formal communication briefings, and the proximity of physician
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
703
offices in one area was conducive to informal contact). In some of the other sites, however, physicians were not only geographically separated (for example, different hospital wings or different buildings spread out across many miles), but there were few mechanisms in place whereby opinion leaders and physicians could meet among themselves, either formally or informally. Most sites had monthly medical staff meetings, but not all physicians were able to attend. At one site, some physicians “called in” to the meeting but appeared to not be full participants. At another site the appointed time for the meeting did not coincide with the scheduled hours of many of the parttime physicians, and as a result they hardly ever attended these meetings. In general, the part-time physicians seemed less connected to the existing structures in place in the clinics. As a result of these organizational and physical barriers, the level of contact between the opinion leaders and the physicians in their clinics was suboptimal. Potential Solutions: The goal in this situation is to create more opportunities for interaction and networking among providers. These opportunities can potentially occur in different formats on site and in person, as well as on the phone and/or via the Internet. To varying degrees, the intervention sites were able to arrange kickoff meetings that drew together sizable portions of the participating providers. Such meetings could be replicated at intervals during the project, adhering to the notion of “booster sessions” common in prevention and sustainability work. The on-site opinion leaders could lead these gatherings. Efforts could be made to include as many part-time staff as possible, including having rotating meeting times to accommodate more people. These forums could (1) provide continued exposure to the project’s goals, (2) provide an opportunity for participants to offer “testimonials” about successes, (3) provide an opportunity for leadership to express support, (4) provide an opportunity to present and discuss performance data and to brainstorm and discuss ideas for improvement, (5) provide a forum for questions and comments from staff members, (6) provide an opportunity to openly discuss any problems or additional barriers arising during the implementation of the project, and (7) provide an opportunity to allow new staff members to be oriented to the project (as turnover is common in many health care settings).
704
December 2005
It might also be helpful to use various technological solutions to address the problem of poor contact among project participants either through use of the Internet and/or e-mail to connect participants within a site (or across project sites). One possibility would be a listserv whereby project staff, participants, and opinion leaders could disseminate project information and seek input and present ideas for discussion. Another possibility would be an Internet “community of practice” to allow for training and education, disseminating/ archiving data and reports, chat rooms for real-time discussion, and electronic note boards for questions or comments. Some projects are currently using such technologies with apparent success.27–29 Before considering any of these ideas, it would be important to assess the current means/quality of communication and consult with the participants to get a sense of their preferences and capabilities. Also, it is important to learn of any existing means of communication and contact that is successful so that it can be incorporated into planning the implementation project. 3. A focus on physicians only as agents of change. The project lacked: (1) sufficient and timely attention to the administrative, monitoring, and reporting functions associated with the implementation of the change model as devised; and (2) sufficient involvement from other key stakeholders in the process of medication management. As the project was devised, the physician opinion leaders were to be involved in multiple tasks, as described earlier. During implementation, however, it became clear that most of them did not have sufficient time to devote to all of these tasks. Scheduling training meetings, conference calls, and other project-related activities with them was particularly challenging. The physician opinion leaders spent most of their time seeing patients, and many of the clinics in which they worked were understaffed. Time available for peer interaction, dissemination, and monitoring of progress was limited. It became clear that other clinical team members, such as pharmacists and nurses, could have been helpful in providing support to the medication managementfocused project. Potential Solutions: The most obvious course of action in this case is to involve more clinic team members in the implementation process. Pharmacists (especially
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
at the Pharm.D. level) could be helpful in disseminating background information on medications, providing oneon-one educational sessions (academic detailing) on medication management issues, and championing the case for change. Nurses or social workers could be helpful in monitoring the change process in the clinic and discussing observations with study staff and/or clinic leadership. Nurses may also be effective in encouraging and supporting providers to use evidence-based practices (depending on the degree of collaboration experienced). Administrative assistants or clinic clerks could help disseminate tools and performance reports from the study to clinical staff. With this kind of support, the physician opinion leaders could focus on the peer-to-peer interactions, monitoring of performance, and leading discussion in formal staff meetings. This team-based approached might also have the added effects of increasing visibility and buy-in of the project throughout the organization. 4. How much directive should be given to the opinion leaders concerning how to influence attitudes and behaviors? During the pre-implementation training session for the opinion leaders, the MHQ team discussed general dissemination and implementation ideas regarding the project, but MHQ did not provide specific instructions for the opinion leaders. The reasoning was to preserve the personal style of each opinion leader and to allow some flexibility in implementation at each site. As noted earlier, our intervention was based on the notion that existing opinion leaders would already have been exerting influence on their peers and the system of care through their knowledge, connectedness, and their own style and methods of “persuasion.” Local opinion leaders were encouraged to be flexible in implementing the project according to local needs and barriers while working with their clinical service leadership. During implementation, however, it became apparent that more specific instruction and training for the opinion leaders would have been helpful. At times, some of the opinion leaders asked for direction from the research team, for example, “I’ve disseminated the materials; what else would you like me to do?” The research team attempted to be more directive about ways they could increase the project’s visibility and to help fellow physicians to incorporate evidence and guideline recom-
December 2005
mendations into their thought processes for care delivery. These efforts may have helped to some extent, but in some instances it was likely too little, too late. Potential Solutions: To help address this issue, we believe that a strong pre-implementation training program for opinion leaders in behavior change, adoption of innovation, and opinion leadership would be a good first step. MHQ recommends that such a session (conceived here for a multisite project) should provide some brief theoretical background in: (1) behavior change at the individual level—such as social cognitive theory30 and the theory of planned behavior,37 (2) behavior change at the organization/systems level—such as recent work in the areas of complex adaptive systems38 and organization change,39 and (3) the adoption of innovations.40 The session should provide an overview of the effectiveness of the range of tools and strategies that have been used in QI efforts in similar settings (for example, clinical reminders, performance feedback, academic detailing, and use of opinion leaders). Specific strategies of effective opinion leadership should be covered in detail. The training session might then focus on the local circumstances of the opinion leaders. They could share their assessments of their sites’ readiness to change, discuss current efforts in QI in other areas, and outline projected barriers and facilitators. In preparation for this, they should be given a list of questions in advance so that they can come to the training prepared to discuss these issues. The training session should then focus on specific ideas for influencing change concerning the proposed innovation. Dissemination and implementation plans could then be made in the group setting, with the team of opinion leaders devising strategies together. It may also be helpful if the interaction among the opinion leaders at different sites is continued after the training. Brief conference calls (no longer than one patient appointment) held periodically could allow leaders to share their implementation experiences and provide advice and support. If possible, a second face-toface experience at the project’s mid-point would be ideal. As always, process evaluation41 should be undertaken to understand how the training and monitoring processes worked or did not, and how best to modify efforts for future work. Indeed, this article is the direct product of just such an evaluation.
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
705
Discussion We have detailed our experiences relating to facilitation—both at the internal level (the opinion leaders) and at the external level (MHQ facilitation of opinion leader selection/training and overall monitoring of the project). The project encountered a host of challenges related to evidence and context—the ever and rapidly changing evidence base, conflicting interpretations of published findings, staff turnover, and conflicting demands at many levels within the clinic system, to name several. Indeed, implementation work is highly complex, and while there are several theoretical or conceptual models to pursue for guidance, there remains a need for the literature to document the field-level successes and failures of these models and the components inspired by them. The process of identifying, understanding, and addressing barriers is an essential component of implementation research. Our experiences will continue to inform our future work and, hopefully, that of others who use opinion leaders as agents of change in health care settings. The VA system and sites studied in this project may have some unique characteristics that would limit the generalizability of our findings; other systems of care will undoubtedly present their own unique barriers and facilitators to the use of opinion leaders. One area that may be ripe for future research is in the area of training and development of opinion leaders. We posed the question, “Can opinion leaders be created?” in response to some sites’ inability to clearly identify opinion leaders. These sites pursued a clinical-champion strategy—the use of a personal motivator for change. Such change agents may not necessarily have followers or skills of influence—but may
be willing to learn the clinical and social influence skills necessary. Another promising research area concerns the optimal mix of disciplines and team members for an opinion leader-driven intervention. How many opinion leaders are needed? Is it possible to have too many opinion leaders at a given site? How best should the leadership team operate? Can opinion leaders from multiple disciplines work synergistically? Another issue in need of greater investigation is how to facilitate social network interactions to improve opinion leader effectiveness. This work has been supported by a grant from the Department of Veterans Affairs, Health Services Research and Development Service (MNT 01-033-1), by the Mental Health Quality Enhancement Research Initiative (QUERI) and by the South Central Mental Illness Research, Education, and Clinical Center (MIRECC). Dr. Curran is supported by a Career Development Award from NIDA (K01 DA15102). The authors thank the opinion leaders and the providers who willingly participated in this project as well as the MH QUERI project team who helped make this work possible. The views expressed in the article are those of the authors and do not necessarily reflect views of the Department of Veterans Affairs.
Geoffrey M. Curran, Ph.D., is Research Health Scientist, Central Arkansas Veterans Healthcare System, Little Rock, Arkansas. Carol R. Thrush, M.A, is Assistant Professor, Office of Educational Development, University of Arkansas for Medical Sciences, Little Rock. Jeffrey L. Smith, M.A., is Implementation Research Coordinator, Mental Health QUERI Coordinating Center, Central Arkansas Veterans Healthcare System. Richard R. Owen, M.D., is Research Coordinator and Mona Ritchie, L.C.S.W., is Project Coordinator, Health QUERI Coordinating Center, Central Arkansas Veterans Healthcare System. Dale Chadwick, M.B.A., is Director, Lowell Treatment Center, Lowell, Massachusetts. Please send reprint requests to Geoffrey M. Curran, Ph.D., [email protected].
References 1. Rubenstein L.V. et al.: From understanding health care provider behavior to improving health care: The QUERI framework for quality improvement. Quality Enhancement Research Initiative. Med Care 38:I129–41, Jun. 2000. 2. Wyszewianski L., Green L.A.: Strategies for changing clinicians’ practice patterns: A new perspective. J Fam Pract 49:461–464, May 2000. 3. Davis D.A., Taylor-Vaisey A.: Translating guidelines into practice: A systematic review of theoretic concepts, practical experience, and research evidence in the adoption of clinical practice guidelines. CMAJ 157:408–416, Aug. 15, 1997.
706
December 2005
4. Drake R.E., et al.: Implementing evidence-based practices in routine mental health service settings. Psychiatr Serv 52:179–182, Feb. 2001. 5. Gates P.E.: Think globally, act locally: An approach to implementation of clinical practice guidelines. Jt Comm J Qual Improv 21:71–84 Feb. 1995. 6. Grol R., Grimshaw J.: Evidence-based implementation of evidencebased medicine. Jt Comm J Qual Improv 25:503–513, Oct. 1999. 7. Solberg L.I., et al: Lessons learned from experienced guideline implementers: Attend to many factors and use multiple strategies. Jt Comm J Qual Improv 26:171–188, Apr. 2000.
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
References, continued 8. Weber V., Joshi M.S.: Effecting and leading change in health care organizations. Jt Comm J Qual Improv 26:388–399, Jul. 2000. 9. Owen R.R., et al.: Use of electronic medical record data for quality improvement in schizophrenia treatment. J Am Med Inform Assoc 11:351–357. Sep.–Oct. 2004. 10. Mental Health Strategic Health Care Group, The Psychosis Working Group: Veterans Health Administration Clinical Guideline for Management of Persons with Psychoses. Washington, D.C.: Department of Veterans Affairs, 1997. 11. Lehman A.F., Steinwachs D.M.: Translating research into practice: The Schizophrenia Patient Outcomes Research Team (PORT) treatment recommendations. Schizophr Bull 24(1):1–10, 1998. 12. Practice guideline for the treatment of patients with schizophrenia. American Psychiatric Association. Am J Psychiatry 154:1–63, Apr. 1997. 13. Dixon L.B., Lehman A.F., Levine J.: Conventional antipsychotic medications for schizophrenia. Schizophr Bull 21(4):567–577, 1995. 14. Baldessarini R.J., Cohen B.M., Teicher M.H.: Significance of neuroleptic dose and plasma level in the pharmacological treatment of psychoses. Arch Gen Psychiatry 45:79–91, Jan. 1988. 15. Owen R.R., et al.: Performance measurement for schizophrenia: Adherence to guidelines for antipsychotic dose. Int J Qual Health Care 12:475–482, Dec. 2000. 16. Feussner J.R., Kizer K.W., Demakis J.G.: The Quality Enhancement Research Initiative (QUERI): From evidence to action. Med Care 38:I1–6, Jun. 2000. 17. Demakis J.G., et al.: Quality Enhancement Research Initiative (QUERI): A collaboration between research and clinical practice. Med Care 38:I17–25, Jun. 2000. 18. Fischer E.P., et al.: Quality Enhancement Research Initiative in Mental Health. Med Care 38:I70–81, Jun. 2000. 19. Borbas C., et al.: The role of clinical opinion leaders in guideline implementation and quality improvement. Chest 118:24S–32S, Aug. 2000. 20. Locock L., et al: Understanding the role of opinion leaders in improving clinical effectiveness. Soc Sci Med 53:745–757, Sep. 2001. 21. Lomas J., et al.: Opinion leaders vs audit and feedback to implement practice guidelines: Delivery after previous cesarean section. JAMA 265:2202–2207, May 1, 1991. 22. Gates P.E.: Think globally, act locally: An approach to implementation of clinical practice guidelines. Jt Comm J Qual Improv 21:71–84, Feb. 1995. 23. Thomson O’Brien M.A., et al.: Local opinion leaders: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev (2):CD000125, 2000.
December 2005
24. Young J.M., et al.: Role for Opinion Leaders in Promoting EvidenceBased Surgery. Arch Surg 138: 785–791, Jul. 2003. 25. Harvey G., et al.: Getting evidence into practice: The role and function of facilitation. J Adv Nurs 37:577–588, Mar. 2002. 26. Greer Al: The state of the art versus the state of science. Int J Technol Assess Health Care 4:5–26, 1988. 27. Coleman J.S., Katz E., Menzel H.: Medical Innovation: A Diffusion Study. New York: Bobbs-Merrill, 1966. 28. Wenrich, J.W., et al.: Informal educators for practicing physicians. J Med Educ 46:299–305, 1971. 29. Zimbardo P.G., Leippe M.R.: The Psychology of Attitude Change and Social Influence. Philadelphia: Temple University Press, 1991. 30. Bandura A.: Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall, 1986. 31. Mittman B.S., Tonesk X., Jacobson P.D.: Implementing clinical practice guidelines: Social influence strategies and practitioner behavior change. QRB Qual Rev Bull 18:413–422, Dec. 1992. 32. Hiss R.G., MacDonald R., David W.R.: Identification of physician educational influence in small community hospitals. Proceedings of the Seventeenth Annual Conference on Research in Medical Education. Washington D.C., 1978, pp. 283–288, 1978. 33. Soumerai S.B., et al.: Effect of local medical opinion leaders on quality of care for acute myocardial infarction: A randomized controlled trial. JAMA 279:1358–1363, May 6, 1998. 34. Elliott T.E., et al.: Improving cancer pain management in communities: Main results from a randomized controlled trial. J Pain Symptom Manage 13:191–203, Apr. 1997. 35. Gifford D.R., et al.: Improving adherence to dementia guidelines through education and opinion leaders: A randomized, controlled trial. Ann Intern Med 131: 237–246, Aug. 17, 1999. 36. Berner E.S., et al.: Do local opinion leaders augment hospital quality improvement efforts? A randomized trial to promote adherence to unstable angina guidelines. Med Care 41:420–431, Mar. 2003. 37. Ajzen I.: The theory of planned behavior. Organizational Behavior and Human Decision Processes. 50:179–211, Dec. 1991. 38. McDaniel R.R., Driebe D.J.: Complexity science and health care management. In Fottler M.D., Savage G.T., Blair J.D. (eds.): Advances in Health Care Management vol. 2. Greenwich, CT: JAI Press, 2001, pp. 11–36. 39. Burke W.W.: Organizational Change: Theory and Practice. Thousand Oaks, CA: Sage Publications, 2003. 40. Rogers E.M.: Diffusion of Innovations. New York: Free Press, 1995. 41. Hulscher M.E., Laurant M.G., Grol R.P.: Process evaluation on quality improvement interventions. Qual Saf Health Care 12:40–46, Feb. 2003.
Volume 31 Number 12
Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations
707