Continuing Medical Education

Physician Practice Change II: Implications of the Integrated Systems Model (ISM) for the Future of Continuing Medical Education Mark Albanese, MA, PhD, George Mejicano, MS, MD, George Xakellis, MBA, MD, and Patricia Kokotailo, MD, MPH

Abstract In a companion paper, the authors provide the development and description of the Integrated Systems Model (ISM). In this article, they describe 14 general implications of the ISM for continuing medical education (CME). They discuss how applying the ISM would change CME by describing (1) how CME and the larger health care environment would be restructured if they were based on the ISM and (2) how the ISM would impact CME under the current environment of health care in the United States. They close by describing how the ISM can be

used as CME moves to address the long lag between discovery and practice and begins to decrease its dependence on pharmaceutical companies. The ISM helps not only explain why the current health care system in the United States (or anywhere) produces what it produces, but also predict what that system would produce if it changed. At present, the ISM is a conceptual model, but with more research into measures of its various elements, it could become a more quantitatively predictive model. In its present form, however, the ISM can

In a companion article in this issue (pages 1043– 1055), the authors describe the ISM.

the interpretation of study results.”2 That said, the medical education community has proposed no shortage of different models.3–5 One potential limitation of these models is their focus on only a portion of the factors that determine the effectiveness of CME; factors such as motivating conditions (e.g., Pay for Performance [P4P]3) or the stages through which change occurs (Rogers,4 Prochaska5). In our companion paper, we provide a model we call the Integrated Systems Model (ISM) that incorporates various models that have had successful applications in order to provide a system-wide model that addresses all of these aspects of CME, avoiding the aforementioned common shortcomings of other CME research.6 The ISM frames physician practice as a form of complex adaptive system (CAS),7 which is built of interacting microsystems and in which alignment of change with the flow of the system is critical to adoption. The superstructure of the ISM is built on Stufflebeam’s Context Input Process and Products (CIPP)8 evaluation model and augmented by several models from human factors engineering.9,10 The context is represented somewhat differently than in the typical CIPP model as it envelopes the system and provides the environment in which the system operates. Another deviation is that the

A

consistent criticism of the research in continuing medical education (CME) is the lack of either a “sound conceptual model of what influences the effectiveness of CME”1 or “a theoretical base to support the choice and development of interventions as well as

Dr. Albanese is professor, Department of Population Health Sciences, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin. Dr. Mejicano is associate dean for continuing and professional development and associate professor, Department of Internal Medicine, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin. Dr. Xakellis is clinical lead physician for geriatrics and extended care, Tomah Veterans Affairs Medical Center, Tomah, Wisconsin. Dr. Kokotailo is associate dean for faculty development and faculty affairs and professor, Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin. Correspondence should be addressed to Dr. Albanese, Department of Population Health Sciences, University of Wisconsin-Madison School of Medicine and Public Health, 610 Walnut Street, 10010 WARF, Madison, WI 53726-2397; telephone: (608) 263-4714; fax: (608) 263-2820; e-mail: ([email protected]).

1056

serve Marinopoulos’s call for a “sound conceptual model of what influences the effectiveness of CME” and address Grimshaw’s concern that current research lacks “a theoretical base to support the choice and development of interventions as well as the interpretation of study results.” The statistician George Box said, “All models are wrong, some models are useful.” The authors believe that the ISM is useful and that maybe it will prove Box wrong. Acad Med. 2009; 84:1056–1065.

inputs box has a region at the top called the reserve in which inputs are sequestered specifically for evaluating and implementing new initiatives. Change is represented by an upper loop between the reserve portion of the inputs box and the processes box; resources from the reserves power this loop. The alignment of the change with the ongoing processes will determine whether the change is symbiotic or parasitic to the system with parasitic changes being extremely difficult to push through to implementation. The products initiate a loop that brings resources back to the inputs, completing the circuit, but not before penetrating metrics and incentives boxes that can augment resources from external sources using mechanisms such as P4P. The interactions among the parts and its general behavior are governed by principles drawn from CAS.6 The ISM and CME

The ISM is a relatively radical departure from models that typically describe CME in the literature because of (1) the characterization of physician practice as a CAS, (2) the concept of resource reserves needed for change, and (3) the need for alignment of the change with the ongoing operation of the system in order to facilitate the implementation and

Academic Medicine, Vol. 84, No. 8 / August 2009

Continuing Medical Education

maintenance of the change—all three of which are unique to the ISM. Additionally, the integration of the different models in one “organism” represents a clarification of the different types of change models and how they relate to one another. Despite its unique characteristics, ISM melds well with findings from CME research. In this article, we first describe 14 general implications for CME indicated by the model. In the two sections that follow, we discuss how applications of the ISM might change the approach to a CME activity in more specific terms, by describing (1) how CME and the larger health care environment would be restructured if they were based on ISM and (2) how the ISM would impact CME under the current environment of health care in the United States. Implications of the ISM for CME The following are general strategies for promoting physician change through applying the ISM to CME. CME providers can use these strategies to enhance the impact of their offerings in changing physician behavior. They are an amalgam of research in the literature and a logical extension of the ISM to CME. 1. Changes that are mission critical and build system reserves are more likely to be implemented than those that do not. This is a basic operating principle of the ISM. The system operates to achieve its mission and to build reserves that allow it to respond effectively to a complex and changing health care environment. Problems arise when a change contributes to only one of the two. A change that is mission critical but draws down reserves will need to be weighed against other mission critical activities. A decision then needs to be made as to whether the system can afford the continued draw on reserves, still achieve its other mission critical processes, and continue to have at least satisfactory reserves. In Maslow’s hierarchy of needs, survival is a most basic one and it will be a primal driving force for change.11 A more insidious problem is when the change creates substantial building of system reserves, but is not mission critical or may even be detrimental to the mission. The implications for CME are that the promoted changes need to be important to the physician practice and that CME providers must address, at least

Academic Medicine, Vol. 84, No. 8 / August 2009

to some degree, draws—financial and otherwise—that a change will make on practice reserves. To not pay such attention will leave the physician to investigate these issues, creating one more obstacle to implementation. 2. The less disruptive a change is to the physician’s current practices, the more likely the change will be put into practice. To make a change happen requires a draw on reserves. The less draw on reserves that a change requires, the more likely a physician can make the change without impacting the various other ongoing processes at the practice. For example, switching a patient from one type of beta blocker to another is going to be much easier than engaging the patient in strategies for making lifestyle changes. While the first change is totally under the physician’s control and can be made with no assistance from any other staff member and/or with little new knowledge or need for additional services, engaging a patient in making lifestyle changes is likely to be much more complicated. The physician is likely to have to learn a new skill set, make the time to implement the change, work with insurers to clarify reimbursement, and more. Each step required to make a change is a draw on reserves and will be a potential block to making the change happen. CME providers, therefore, should provide a realistic assessment of the effort that a change requires on the part of the physician and provide an opportunity during the CME activity for the physician to plan for the change. If a substantial amount of ground work needs to be done to make the change, the foundation for that ground work needs to be laid as much as possible during the CME activity. 3. The fewer providers in the practice who need to change their behavior, the more likely the change will be made. Each individual in the practice is operating with reserves that are at various levels of depletion. Some individuals are more completely deployed in ongoing processes than others and their reserve is lower on a daily basis. As the number of the individuals who need to participate in making a change, especially those with lower reserves, rises the more likely one of them will not have the available reserves to make the change. If the change is implemented without addressing the lack of individual reserves,

the desired change will cease when that individual is needed, or there will be variable implementation as the individual juggles the new activity along with other ongoing activities and responsibilities. Changes promoted through CME that involve a health care team as opposed to only a single physician will need to be fully explored and various implementation issues will need to be addressed during the CME activity. In cases where the change will involve a team, the CME activity should include the entire team. 4. The more compelling and durable the evidence in support of the change, the more likely that the change will be made. A change that has produced impressive evidence for efficacy and has withstood the test of time will more easily motivate physicians. Such a change taps into the need for providing quality care, but also provides some confidence that multiple further change recommendations will not occur. Exceptions to the need for efficacy and durability occur when no acceptable alternatives are available or when the available alternatives are much more invasive, toxic, or costly. In the latter situations, physicians may be desperate for alternatives, and their motivation already exists as a felt need. CME providers must be certain that a promoted change is evidence-based and that the evidence is solid, has shown clinically significant effects by reasonable criteria, and appears unlikely to be supplanted by new recommendations any time in the near future. There are particular dangers in using CME to promulgate practice guidelines that are based on expert judgment without providing data-based clinical evidence in their support. 5. The further a physician moves along the stages of change, the more likely he or she is to adopt the change. The ISM adopts Roger’s stages of change4 for systems and programs and Prochaska’s Transtheoretical Stages of Change model5 for individuals. According to Rogers, the five stages that lead to change adoption and integration are (1) knowledge, (2) persuasion, (3) decision, (4) implementation, and (5) confirmation.4 In the ISM framework, CME providers would first assess a doctor’s readiness to change. CME provided to a physician actively searching for solutions to a

1057

Continuing Medical Education

problem (the decision stage) will be more likely to result in change than an activity provided to a physician who has to be approached about implementing accepted guidelines. The former is already sufficiently motivated and has diverted resources from his or her reserve to begin the change process while the latter may not be even at the first stage of change and will need to be motivated to begin the change process. Generally, CME activities that are designed for a physician’s particular stage in the change process will be more successful than broad spectrum efforts. 6. CME activities on topics about which physicians are already feeling a need for change will be more successful in producing change than those for which physicians feel no such pressure. For cases in which the physician already feels motivated to change, galvanizing forces arising within the physician’s immediate environment are pushing for the change, providing substantially greater pressure than can be externally applied. Needs assessments can be a good source of identifying such topics, but they need to be structured to distinguish felt needs (needs generated from practice outcomes) from just interests. To make such a distinction, needs assessments should include questions such as “Why would you choose to attend this CME program?” that ask physicians about their motivations for seeking information about a topic. Felt needs will generate answers such as “I am seeing an increasing number of patients with X, and I really am not happy with how my methods of management turn out,” or “I had a patient whom I thought I managed well, but her case did not go well at all, and I want to know if there is a better way to manage similar problems.” On the other hand, simple interest will elicit a response like, “I heard about new ways of managing X, and I want to learn more about them,” or “I’ve always wondered how to manage X, and I finally found the time to learn how to do it.” A felt need has a sense of urgency, and often a patient event may precipitate a physician’s participation while an interest tends to be linked more to curiosity and a desire to learn in case a need may arise in the future. 7. The more motivating conditions a CME activity activates, the more likely a

1058

physician will change. The following are examples of the types of motivating conditions that a CME activity can employ: a. credible sources supporting the change such as Cochrane systematic reviews and professional organizations. Another credible source is data on physician performance and patient outcomes that are derived from the target audience and published in peerreviewed journals. As an example, a CME program on managing stroke among patients in rural areas would target physicians in rural areas (likely primary care physicians) and incorporate outcomes specific to these patients; b. information detailing how the change will improve the quality of patient care in terms of reduced patient morbidity and mortality, reduced disease burden, improved quality of life, reduced side effects, minimized number needed to treat, etc.; c. models illustrating the financial benefits or efficiencies that the change will achieve; d. statistics showing the prevalence of the use of the new approach, especially by respected peers or physicians in prestigious practices; and e. testimonials corroborating how the approach has positively changed physician practices and/or patient lives. The implication of the need to activate motivating conditions for CME is that chosen topics need to have substantial evidence supporting their incorporation into practice. 8. Providing CME to physicians who are predisposed to adopting innovations is more likely to result in change. Physicians whom Rogers categorizes as either innovators or early adopters4 are more likely to implement practice changes from an educational intervention compared to physicians who are not so predisposed. Early adopter and especially innovator physicians value being at the cutting edge of medical practice and are likely to adopt innovations because of either intrinsic needs (e.g., peer respect and acknowledgment) or extrinsic motivation (e.g., a marketing advantage in the local health care market)— or both. The value in considering early adopters for CME is the potential to

create local “demonstration projects” that can be disseminated to practices. The outcomes of the demonstration project can motivate later adopters, and the early adopters can provide their assistance, based on their experience with the innovation. The availability of local evidence and local adopters/experts can provide a compelling set of motivators for physicians who are less likely to embrace change. 9. Enlisting the assistance of respected colleagues in either implementing the CME activity or in adopting a change early will promote broader and more rapid change by the physician community. Respected colleagues, especially local opinion leaders, who adopt the change will encourage other physicians to change12; however, advocates for the change need to exercise care in determining which physicians to involve. Rogers cautions that innovators’ colleagues do not always view them positively, and as a consequence, enlisting their assistance can actually be a detriment to broader adoption. Distinguishing between respected colleagues and not-so-respected innovators is critical in implementing this strategy; however, doing so is an imperfect science. One way is to contact a few physicians in the area and ask them who among their colleagues they would most trust regarding the management of patients with X. If the same individuals are repeatedly identified, they are most likely the opinion leaders. National leaders who are recognized for their work in the area can also motivate practice change, but it is important that the audience be able to relate directly to them. For example, physicians working in level 1 trauma centers or academic health centers (AHCs) may have difficulty relating to the resource limitations experienced by rural physicians. Physicians promoting the change being unable to relate to their target physician audience is a fairly fatal weakness. 10. Employing small groups that include at least one member who has made a change will promote adoption. Having a group member who has made the desired change allows the other members of the group to explore the logistics of implementation. A physician making a practice change could potentially encounter a large number of problems.

Academic Medicine, Vol. 84, No. 8 / August 2009

Continuing Medical Education

In discussions of the logistics surrounding making a change, often questions or “what ifs” arise. A member of the small group who has made the change can explain how he or she handled the “what ifs,” allaying any concerns of the other group members. 11. Providing realistic descriptions of the costs of both making the change and maintaining it long-term will promote implementation and maintenance. Ideally, the CME activity will model the costs, including time and money, based on the actual practice characteristics of the participating physicians. The model should account for the fact that change through CME activities may increase costs while reimbursement lags behind. If a CME activity is going to promote practice change, it is critical to the program’s credibility that the cost burden that the change will create be addressed. The program providers can often be more creative in addressing the cost burden than a practicing physician who may not realize the true burden of change until he or she is mired in the complexities of implementation. If physicians invest a substantial amount of their reserves in making a change to find out that it will be parasitic to a high degree, they are likely to abandon the change. This will make them less likely in the future to follow recommendations for practice change from the CME provider. Further, if a physician’s staff is unhappy about investing time and energy into making a change that is abandoned, it may be more difficult for that physician to make changes in the future. 12. Helping physicians mentally picture how a new procedure will affect their practice will assist adoption. Mental imaging is an effective method of working through the logistics of making a change without investing substantial resources to the effort.13,14 The mental picture should include projections of how other staff will be involved and what impact the change will have on their work load. A CME activity needs to allow time for the participants to think about how the change will look from the perspective of every person in their practice— both those who will be actively involved in implementing the change and those who will be impacted even if they will not be specifically doing something to make the change happen. Often, another person must take on additional

Academic Medicine, Vol. 84, No. 8 / August 2009

responsibilities when a colleague becomes involved in a new activity. Anticipating these kinds of collateral effects of implementing change can help minimize any damage they might do. 13. Identifying the likely barriers to implementation and maintenance and providing an opportunity for physicians to develop solutions for the barriers in their own practices will promote adoption. Unanticipated barriers to implementation can be deal breakers as they extend the length of time to implementation, both draining the will to change and requiring unplanned diversions from reserves. The more physicians identify barriers and develop a plan for dealing with them in advance, the more likely the process will result in implementation. Providing a contact with whom a physician can follow-up either for support in overcoming challenges or for additional motivation will also promote adoption of change. 14. Obtaining a strong commitment from physicians to implement the promoted change before they leave the CME activity will provide additional motivation to push the change process through to completion. Physicians making a personal commitment to implement plans created at the CME activity will improve their chances of actually making the change because they make their intent public. Peer respect is important to professionals, and if they make a commitment to change in the presence of their peers, they are likely to follow through. Signed contracts have led to some success as a means of encouraging doctors to commit to change.15,16

CME in an ISM World

In this section, we describe how the ISM would change the approach to a CME activity in more specific terms. We begin by describing how using the ISM would restructure CME and the larger health care environment. For purposes of this discussion, we do not allow the current reality to constrain our depiction. Thus, we have taken liberties, creating structures such as centralized databases and the pervasive use of integrated electronic health records that are not yet generally available in this country, and inventing fictions such as physician-level

classifications based on performance criteria. CME in an ISM world would occur either in the physician’s practice (or home) or at a multidisciplinary continuing professional development center. Practice-based CME Practice-based CME would be the primary type of education in the ISM world and would ideally occur during the work day in the form of flexibly scheduled 20-minute (or so) periods. Practice-based CME would be of five general types: (1) professional fulfillment, (2) practice-based profile optional modules, (3) practice-based profile mandated modules, (4) elective modules, and (5) consultant visits. Professional fulfillment. The first type of practice-based CME, what might be termed professional fulfillment, would be an extension of some of the new types of CME activities that are currently developing. A professional fulfillment activity could occur in a 20-minute (or so) block built into the physician’s day, but would not necessarily be completed during a single work day. An example is providing CME credit for applying what a physician finds in literature to problems he or she encounters in patient care. This activity applies the first and sixth of the implications of ISM for CME (referred to simply as Implication X hence forth) in that it enhances patient care yet requires little if any draw on the reserves especially if this is something the physician already does during normal patient care activities (Implication 1) and deals with a felt need of the physician (Implication 6). Further, providing CME credit to physicians who apply evidence from the literature to their practices acknowledges desirable behavior that they are already doing. Reviewing a paper for a professional journal is another example. Again, this activity does not necessarily require a draw from reserves because it could occur during the time set aside for CME in the normal work day. Reviewers are often selected for their expertise in an area, so this would also be likely to tap into both Implication 6 —the physician’s expertise indicates that he or she already sees the issue as a felt need—and Implication 1— enhancing the physician’s expertise as well as the recognition of that expertise. Collaboration in type II translational

1059

Continuing Medical Education

research projects with researchers at an AHC would also be a CME professional fulfillment option in an ISM world. Such a collaboration would contribute to the mission of the practice by meeting the larger needs of society for advances in medical practice (Implication 1). Practice-based profile optional modules. The second type of practicebased CME would be developing modules built around the patients that a physician sees. This approach addresses Implications 6 and 12; physicians could address a felt need and mentally picture how a change would affect their practice. It also requires creating a new fiction, a CME educator, a person trained in the management of electronic patient databases and educational methods. (However, the CME educator could become an electronic entity some time in the future). The CME educator would monitor and search the electronic records of a particular physician’s patients, tracking the diagnoses that the physician encounters. For this type of CME, the top 10 (or so) most frequent diagnoses the physician encounters would serve as the source for modules. The CME educator would search and compile updates on practice recommendations for these diagnoses (drawing on only credible sources of information) and develop a separate module for each diagnosis. The Cochrane databases would be especially useful for this purpose. The expense of doing this in an ISM world could be relatively manageable. Access to medical records would be available from a central database, and software could be developed that automatically compiles the information in a predetermined manner. The major unknown cost at this time is the development of modules; however, if a bank of regularly updated modules were available, the whole process for an individual physician could be automated and the cost would be minimal. Practice-based profile mandated modules. The third practice-based CME would use the same modules that served its optional cousin. The primary difference is that adverse practice events (e.g., patient deaths, unnecessary repeat visits, nosocomial infections, excessive use of narcotics, departures from accepted protocols) would trigger a mandated module. A CME educator would design a refresher course on the

1060

diagnosis, treatment, and management of patients who, according to an analysis of the physician’s practice, either receive care that is a departure from accepted practice or experience adverse outcomes. Again, the electronic medical record and other databases (such as those that archive patient satisfaction and complaints) would dictate the need for such a module. A practice mandated module would supersede all optional forms of CME. This type of CME would clearly reflect Implications 6 and 7, meeting a felt need and invoking multiple motivating conditions. However, it would be most reflective of Implication 1, being mission critical and adding to reserves. Poor practice outcomes are highly detrimental to achieving the practice mission and to building reserves. Elective modules. The concept behind the elective modules is to broaden physicians’ understanding and awareness of new developments in biological and medical theory. Elective modules could be practice-related, they could deal with basic biological science, or they could provide learning opportunities beyond either of these. An elective module would be an option a physician could select from among many in an extensive library of continually updated CME modules. CME providers could also design elective modules for relatively new developments or for somewhat esoteric specific interests. The Cochrane Collaboration databases would serve as a rich source of information, especially on the outcomes of clinical trials, for building elective modules. In addition to the CME educator pulling material from updates and the like, physicians could design their own learning experiences (e.g., visiting a gene sequencing laboratory to better understand how the process works or visiting the local public health department to gain insight into how outbreaks are tracked and managed). Elective modules (as well as mandated ones) could exploit all forms of media. Elective modules would use compelling evidence (Implication 4), enroll physicians already interested in change (Implication 5 and 6), and attract physicians likely to adopt innovations (Implication 8). Consultant visits. Change can be difficult to implement. As the ISM depicts,6 numerous agents (e.g., any member of the practice who is needed to implement

the change, but who is fully deployed and has no reserve capacity) can impede or even block changes from occurring. For cases in which a physician attempted to implement a change, but failed, the practice could bring in outside consultants to provide recommendations and support for implementing changes. This approach capitalizes on physicians’ movement along the stages of change (Implications 5– 8), enlists the help of respected colleagues (Implication 9), helps physicians identify barriers to making a change (Implication 13), and encourages greater commitment to implementing the change (Implication 14). Continuing professional development (CPD) center-based programming Some CME activities (e.g., conferences; centralized data management; and monitoring, testing, and assessing skills) will need specialized facilities. In the ISM world, CPD centers located in each major region would accommodate these activities. All CPD centers would receive base funding from the “Institute for Health Quality and Safety” (IHQS), an umbrella institute formed from what currently are the Agency for Health Resources and Quality, The Health Resources and Services Administration, and the National Library of Medicine. (In addition to supervising the regional CPDs, the IHQS would be responsible for developing and updating modules for CME as well as analyzing the national electronic medical record database to identify high-need areas for CME.) The regional centers would typically be affiliated with AHCs in order to provide a large enough pool of physician facilitators to serve the needs of physicians in the region. Each center would have conference facilities that host regional meetings, an assessment center, and an education center—all with integrated state-of-the-art information technology as well as data management capabilities. The centers would be responsible for four major charges: (1) updating physicians with current health care information, (2) disseminating new practice guidelines to the relevant health care professional communities, (3) assessing individual health care professionals and health care teams, and (4) providing individualized CME for reentry, retraining, and remediation. They also may employ the CME educators who track needs and support CME for individual physicians. Note that we are changing some terms. In

Academic Medicine, Vol. 84, No. 8 / August 2009

Continuing Medical Education

the ISM world, CME would expand beyond physician education to include that of all health care professionals, including nurses, physician assistants, medical assistants, reception staff, technicians, etc. Health care professional updates. In an ISM world, update courses would bring health care professionals together annually to learn about new developments and new skills (which would be disseminated by the IHQS). The purpose of the updates would go beyond improving health care capabilities, to also reinforcing the basic values of the profession. In addition, these yearly conferences would provide social opportunities for participants to strengthen their networks and to meet and greet new members of the profession. These social and professional components of the course would be especially important for professionals in rural and other isolated practices. Also, the continual reinforcement of the basic values of the profession would help to broaden all health care professionals’ sense of purpose and to cement their belief in the importance of the roles they play. This approach addresses mission critical issues and builds the systems reserves of the physician practices in the region (Implication 1) by identifying the most compelling and durable practices to disseminate (Implication 4 and 7). It also capitalizes on the social aspect of the physician community to motivate and promote change and professional commitment (Implication 9). Dissemination of new practice guidelines to the relevant physician communities. This national IHQS would be responsible for developing new practice guidelines. It would develop these guidelines by funding consensus conferences, systematically reviewing the literature, mining the national database (via automated statistical searches for practices producing either positive or negative health outcomes), and heading other efforts to identify best practices such as the ones endorsed by the National Committee for Quality Assurance. Consensus conferences would allow experts to review evidence on the effectiveness of procedures under consideration and to issue recommendations that the regional centers would then disseminate. The dissemination process would depend on

Academic Medicine, Vol. 84, No. 8 / August 2009

the complexity of the recommendations and the costs involved in their implementation. Reimbursement to practices for implementing guidelines would be adjusted depending on whether the cost was limited to startup or whether ongoing additional costs must be borne. The national IHQS would institute incentives for making changes with the most important incentive being the addition of the new guidelines to the criteria for advancing to a higher level of health care professionals and team certification (see the next section). This system would (1) help physician practices define and achieve mission critical goals and create methods for building system reserves aligned with best practices (Implication 1); (2) help establish and build a compelling base of evidence (Implications 4 and 7); (3) bring opinion leaders together and recognize their expertise (Implications 7 and 9); and (4) outline the guidelines for actually implementing the change in physician practices (Implication 11). Health care professional and health care team assessment. In the ISM world, automated assessments would be constantly evaluating the care that individuals and their teams provide. Electronic search engines would continually abstract patient charts to compare actual notation to what practice guidelines suggest should be there. Assessments would monitor patient outcomes for either unwarranted repeat visits or adverse outcomes linked to earlier visits. Statistical process control methods would evaluate whether events warranted interventions in order to assure quality and safety. Practice data would also help ascertain the quality of care provided. Given that much of the responsibility for outcomes emanates from the entire team, a team would be evaluated similarly to an individual physician. One might envision a variety of levels for grading performance (the number of levels is a relatively arbitrary choice). In this new ISM CME world, the CPD center would monitor physicians and their teams and confer a graduated level designation to high-performing individuals and teams in recognition of their expertise. Higher level designations would yield more complex referrals and a higher reimbursement rate. This approach directly addresses a felt need of the

practice (Implication 6), provides strong motivating conditions for adopting changes (Implication 7), and maintains mission-critical activities (Implication 1) aligned with centrally adopted guidelines (Implication 4). Individualized CME for remediation and reentry. The flip side of monitoring physician and team performance in order to award rising level designations is monitoring in order to detect performance deficits. A centralized database could track many indicators of performance, both positive and negative. Many of the negative indicators (e.g., unwarranted repeated visits, patient complaints and/or requests to change providers, visits to repair damage from previous procedures, nosocomial infections, patient nonadherence) can occur even with the best of care. However, with a centralized database and monitoring, tools like statistical process control can track indicators to determine if they occur with uncommonly high frequency. Statistical process tools are also useful for determining if deficits are limited to certain diagnoses, procedures, or skills. The assessment center housed in a CPD center would diagnose deficits with greater precision than is possible through only practice data surveillance. The assessment center would feature state-of-the-art simulations and a pool of standardized patients for assessing different skills and abilities. A CME educator would design an individualized remedial program based on the assessment of a physician’s practice. Remediation might involve anything from a review of updates on accepted standards of care to a return to supervised practice for a specified period of time. The indicators could identify problems within the health care delivery system, and interventions could be directed at nonphysician staff if practice deficits involved other members of the practice. This approach promotes changes that are mission critical (Implication 1) and maximizes the motivating conditions for change (Implication 7). A second situation that would lend itself to individualized CME would be health care professionals who wish to return after a period of absence to practice in a field in which they had previously received training. The assessment of their retention of knowledge and skills could

1061

Continuing Medical Education

be accomplished via cognitive testing and objective structured clinical examinations patterned after either the United States Medical Licensing Examination Step 2 clinical knowledge and clinical skills examinations or the National Board of Medical Examiners Post-Licensure Assessment System. Extensive testing could reveal a returning professional’s capacity for a basic level of practice or for more independent practice or could reveal a need for remedial work. In contrast to reentry into a field in which one has already received training, entering a new field would require formal residency training over a period of several years. Both of these situations ensure that physicians and all medical personnel are capable of achieving mission critical goals and building reserves (Implication 1).

ISM-Based CME in the Current World

Currently CME operates as an unfunded mandate imposed on practitioners both by the profession and by many state licensing agencies as a cost of doing business. CME might be considered analogous to research and development in the business world: it is intended to keep the physician’s skills vital, adapting to ever improving health care practices. It is a means of implementing new advances and best practices that will ultimately prove beneficial to the health of the public as well as the practice of medicine. Currently, the resources for physicians’ CME come from the reserves of either the practices in which they work or their own personal reserves. Critics have labeled the traditional CME— often held in an attractive location—as a paid vacation in disguise, and literature reviews have generally found it’s primarily didactic approach to be ineffective.1,2,17 However, this type of CME recognized the fact that the funding and time came from the physician or his or her practice, so to motivate attendance, it provided the opportunity for a subsidized vacation. If the physician took away one or two nuggets of useful information, the activity was a success. In the new world order, this inefficient and costly approach has fallen out of favor. Health providers also recognize new critical needs for education, such as improving patient safety and the quality of patient care. In this new world, CME is increasingly held

1062

accountable for producing measurable, positive changes in physician practice. The maintenance of certification (MOC) system established by the American Board of Medical Specialties has components that are consistent with ISM-practice-based CME in the current world. In the sections below we first describe MOC and how it complements the ISM. Then we conclude by describing ISM implications for larger programs that are the primary mainstay of the current CME world. Performance-improvement-modules The American Board of Internal Medicine (ABIM) recently moved to a MOC model from a certification–recertification model.18 In this current model, physicians must work to maintain their certification whereas previously they were certified for life once they earned their initial certification. One of the four components that compose the maintenance protocol is practice-based assessment and improvement. Among the options to meet the requirements for practice-based assessment and improvement are performanceimprovement modules (PIMs). One example of a practice-based assessment and improvement PIM involves a Web-based, self-evaluation tool that guides physicians through collecting data from their own practices. Currently these Web-based, selfguided PIMs are available for more than 15 topics. Each topic has variable requirements, but the majority require chart reviews (from a minimum of 10 to the recommended 25) and a Web-based practice system inventory. Some PIMs also require patient or peer surveys (from, as with the chart reviews, a minimum of 10 to the recommended 25). Some PIMs require didactic sessions, and others entail handson requirements that are specific to the setting (e.g., hospital-based patient care). The PIMs provide instructions and forms for collecting the chart reviews, patient surveys, and practice system inventories. Data collected during the PIM are submitted to the ABIM, which completes an analysis and summary of the practice data. Physicians then reflect on the data to determine potential areas for improvement and identify ways to change their systems of care to achieve these improvement goals. The physician then develops a quality improvement

plan, implements a change, and performs appropriate remeasurement to determine if the change resulted in an improvement. The PIM is complete when the physician records the improvement plan and the results of the change in the Web-based tool and submits them to ABIM. Completion of a PIM contributes up to 20 points out of 100 needed during a 10-year period for MOC. The PIM concept is consistent with various elements of the ISM, specifically with implications 1, 5– 8, and 11–14. The relatively large amount of CME credit that a physician can earn from a PIM (20 category 1 American Medical Association credits for some PIMs) exceeds the yearly requirement in some jurisdictions. Some PIMs may promote practice changes that are externally motivated through P4P incentives, directly improving the return on investment to the practice as well as addressing MOC needs (Implications 6 and 7). The PIMs are also aimed at improving the physician’s practice outcomes (Implication 1), so the potential for PIMs to produce positive incentives through both financial and quality increases are quite high. Finally, the PIM constitutes a written document submitted to a professional organization, which encourages physicians to put their plans into action (Implication 14). The PIMs also encourage physicians to use a systems-based analysis of their practice to implement changes. This type of an analysis would benefit from the ISM, especially in conceptualizing the practice as a CAS. The ISM enables physicians to develop their improvement plan using Chen’s mental model14 as a framework for considering their practice (Implications 11, 12, and 13). A physician could develop an analysis of the reserves available and the draw-down needed to make and then sustain the change. An analysis could also help physicians determine how they will integrate the change into the ongoing operations of the practice. In the ISM, physicians would give clear attention to resource costs, outcome expectations, and the sustainability of potential changes at the outset of a PIM. However, they would need to be careful that using the ISM does not lead to “paralysis by analysis” because too much attention to detail during planning can subvert the process. The physician planning a change needs to be cognizant of potential obstacles, but

Academic Medicine, Vol. 84, No. 8 / August 2009

Continuing Medical Education

not hypersensitive to worse-case scenarios. Under ISM, proposals for PIMs would include estimates of the cost of conducting a PIM, including both the time needed to conduct the practice assessments and to implement the resulting practice recommendations. CME activity under the ISM For a CME activity in our current world, the ISM would begin with selection of one or more topics for which, as much as possible, the physicians feel a compelling need (Implication 6). The mechanism for topic selection could vary from national dissemination of new practice guidelines to grassroots needs assessments for a target audience. The needs assessment would be designed to give a picture of current practices and identify where target physicians perceive a need for change. From the topics garnering the greatest positive response and strongest evidence of felt need, those with both the greatest amount of evidence from the literature in their support and the most compelling potential outcomes would be selected for developing CME activities (Implications 4 and 7). CME educators would compile evidence in the most compelling terms, including reduced mortality and morbidity, reduced disease burden, reduced number and length of hospital stays, improved quality of life, reduced number needed to treat, reduced costs, etc. (Implications 1 and 7). CME educators would estimate the costs of implementing the change, including both time and money for startup and for maintenance, based on prevalence data and expenses from selected collaborating physician practices (Implication 11). Among the factors considered in the models would be costs for staff training (time, salaries registration/tuition, travel, etc.), equipment, tests, and referrals. Projections, including analyses of number needed to treat, would allow physicians to run estimates using the input of the prevalence of at-risk patients from their own patient population. After completing the ground work, the activity would be marketed to the target population through direct mailings, listserve e-mails to relevant professional groups, and advertisements in professional journals. Physicians who enroll in the activity would provide their

Academic Medicine, Vol. 84, No. 8 / August 2009

current method of managing patients with the condition in question as well as the accessibility of any specific diagnostic tests or treatments (e.g., the availability of an echocardiogram for a CME program on congestive heart failure) (Implications 5, 11, 13). Enrollees who were already performing at a high level would be offered reduced tuition and a certificate of recognition in return for working with small groups as an expert consultant (Implications 9, 10, and 13). In their presentation of the change program to physicians, CME presenters would provide an evidence-based session on the health benefits of change for patients with the condition (Implications 1 and 4) and on the logistical costs and benefits for the practice (Implications 7 and 11). The CME presenters would model cost recovery from insurance and other health care plans as well as the net financial cost/benefit of the change. A practical description of what would be needed to make the change would be provided next (Implication 11). Physicians would then break into small groups to talk about the new method of care, and an expert consultant physician (a physician who is currently using the recommended method) would be assigned to each group (Implications 9 and 10). The purpose of these initial small groups would be to discuss the new method, its value, and what it would mean for the physician’s practice (Implications 11, 12, and 13). The availability of a physician currently and successfully using the new method would help to establish its credibility and its feasibility in the real world. Next, each physician would write an individual plan for incorporating the new method that would include the shortterm costs of making the change (time, money, and other resources) and then the long-term costs of maintenance. Ideally, physicians would be able to input data from their own practice into a model that provides an estimate of the resource costs of the change (Implications 11, 12, and 13). They would also include the “chain of change”; that is, the other individuals in the practice who would also need to modify what they currently do in order to implement the change. The physicians would also estimate the potential for these other individuals to make the

change and maintain their current responsibilities (Implications 2, 3, and 11–13). If any of the individuals in the chain of change are already fully or overdeployed, the physicians would need to include in their plan a strategy for overcoming that as well as any other barriers to making the change. After the physicians completed their individual plans, the small groups would reform and all members of the group would present their plans and participate in a discussion. The main focus of the discussion would be on methods to overcome the barriers identified (Implications 10 and 13). The final steps would be to revise the plan based on the discussion, to set a date for implementation, and to sign the document as a commitment to implementing the change (Implication 14). The physician would keep a copy for his or her records. The CME presenters would give each physician a tool kit of materials to assist in the implementation. The tool kit would include a bulleted list explaining the process through which the physicians document their outcomes. This document would not only help win the hearts and minds of each person in the chain of change at the home practice, but also serve as a starting point for planning how each individual will contribute to implementing the change. The physicians would also receive the contact information (e.g., a phone number and e-mail address) of an expert in case they encounter any problems (Implication 9). Finally, the CME staff would remind the physicians that a member of their faculty will contact them after a specified period of time to assess the progress of the implementation (Implication 7). Impact of changes in medical knowledge What if another trial is released with results that run counter to current recommendations? Making practice changes requires resources and sometimes financial risk. Vacillating practice guidelines, especially those that subsequent research reveals were harmful or unnecessary, are not uncommon and can demoralize physicians and their staff, dampening enthusiasm for making future changes. A major question that CME needs to address is how can a physician know when a change needs to be made? A secondary question concerns a change that involves substantial financial risk to

1063

Continuing Medical Education

the practice: should various stakeholders—rather than just the practice— share the burden of the risk.

As CME Changes, How Can the ISM Help?

Diffusion of innovation into physician practice has been lamentably long. CME has been severely criticized for its dependence on funding from pharmaceutical companies and the corrupting influence this has had. The problem is that systems generally do what they are designed to do. The U.S. health care system is uniquely built on a forprofit model. As if to punctuate this point, medical students are now graduating with potentially catastrophic levels of debt, so survival imperatives generated by the for-profit system can quickly overwhelm altruistic tendencies. As Jim Collins states about visionary companies in Built to Last: “Profit is like oxygen, food, water, and blood for the body; they are not the point of life, but without them, there is no life.”19

the return to the practice is greater than another less expensive therapy. In this situation, and if the therapy is covered by insurance, the physician does not need to choose between the health of the practice and patient care quality, but must fulfill his or her obligation to society to provide cost-effective care. But our society has chosen to make health care a for-profit industry. How can we expect the practice to sacrifice its ability to survive or thrive in the for-profit system in order to save the insurer, public, and/or patients money? In this situation, ISM predicts that physician practices will choose the option with the higher return to its reserves. And we get exactly what we designed the system to do: return of money to the system through taxes on the income of the medical providers (i.e., physicians) and makers of the therapeutic agents and the creation of the incentives that fuel the drive for new and better therapies. In this scenario everyone (except the insurer) wins . . . unless the profit motive becomes too great and bleeds from the quality care imperative. The ISM predicts that this bleeding is most likely to occur in a reservedepleted practice where the CAS enters survival mode.

Although physician change has been the object of substantial study, few studies take into account that much of the failure to make and sustain changes from CME is a function of the system design and its overriding profit motive. Symbiotic changes, those benefiting patients and the practice, are likely to see the greatest implementation. When the change is in the best interest of the patient, but it jeopardizes the health of the practice (potentially parasitic), the situation is much more complex. Physicians must sometimes choose between therapies that are fully reimbursed by insurers, but thought to be less effective than therapies that are only partially reimbursed (or not at all reimbursed). The practice must absorb the difference in cost if the physician chooses the therapy that is believed to be more effective. New therapies are typically more expensive and less likely to be fully reimbursed as it takes time to conduct enough studies to demonstrate their superiority. This actually slows the pace of translating discovery into practice and may create a bottleneck in moving evidence from the laboratory to the clinic.

What the ISM provides is a framework that shows how the very different models of change relate and interact with one another (change,4,5 P4P,3 CIPP8/human factors engineering9,10), and the ISM incorporates principles that will help medical educators understand the dynamics of change (the primacy of resource reserves, the alignment of change with ongoing processes, and the impact of CAS operating principles). The ISM also provides guidance about how to influence medical practice and may explain many of the disparate findings in the research literature on the effectiveness of CME. In addition, the ISM helps medicine and its practitioners understand what the current U.S. health care system creates and it helps predict what that system will produce if it changes. But, the system is a CAS, and predictions from applying the ISM are likely to be less than perfect. Further, much research is needed to develop the ISM, so medical educators can develop more precise measures of each of its elements.

A more problematic situation is one in which the therapy may not be superior to another therapy in terms of quality, but

In its present form, however, the ISM answers Marinopoulos’ call for a “sound conceptual model of what influences the

1064

effectiveness of CME” and addresses Grimshaw’s concern that current research lacks “a theoretical base to support the choice and development of interventions as well as the interpretation of study results.” As the statistician George Box said, “All models are wrong, some models are useful.”20 We strongly believe that the ISM is useful and that maybe, as it develops over time, the ISM will prove Box wrong.

References 1 Marinopoulos SS, Dorman T, Ratanawongsa N, et al. Effectiveness of Continuing Medical Education. Evidence Report/Technology Assessment. No. 149. Rockville, Md: Agency for Healthcare Research and Quality; January 2007. AHRQ Publication No. 07-E006. 2 Grimshaw JM, Eccles MP, Walker AE, Thomas RE. Changing physicians’ behavior: What works and thoughts on getting more things to work. J Contin Educ Health Prof. 2002;22:237–243. 3 Meterko M, Young GJ, White B, et al. Provider attitudes toward pay-for-performance programs: Development and validation of a measurement instrument. Health Serv Res. 2006;41:1959 –1978. 4 Rogers EM. Diffusion of Innovation. 5th ed. New York, NY: Free Press; 2003. 5 Prochaska JO, Velicer WF. The transtheoretical model of health behavior change. Am J Health Promot. 1997;12:38 – 48. 6 Albanese MA, Mejicano G, Xakellis G, Kokotailo P. Physician practice change I: A critical review and description of an integrated system model. Acad Med. 2009;84: 1043–1055. 7 Pisek P. Redesigning health care with insights from the science of complex adaptive systems. In: Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: The National Academies Press; 2001:309 –322. Available at: (http:// books.nap.edu/openbook.php?record_id⫽ 10027&page⫽309). Accessed May 3, 2009. 8 Stufflebeam DL. The CIPP model for evaluation. In: Stufflebeam DL, Madaus GF, Kellaghan T, eds. Evaluation Models. 2nd ed. Boston, Mass: Kluwer Academic Publishers; 2002. 9 Karsh BT, Holden RJ, Alper SJ, Or CK. A human factors engineering paradigm for patient safety: Designing to support the performance of the healthcare professional. Qual Saf Health Care. 2006; 15:59 – 65. 10 Carayon P, Schoofs Hundt A, Karsh B-T, et al. Work system design for patient safety: The SEIPS model. Qual Saf Health Care. 2006;15: i50 –i58. 11 Maslow AH. A Theory of Human Motivation. Available at: (http://www.advancedhiring. com/docs/theory_of_human_motivation. pdf). Accessed April 8, 2009. 12 Stross JK, Hiss RG, Watts CM, Davis WK, MacDonald R. Continuing education in pulmonary disease for primary-care physicians. Am Rev Respir Dis. 1983;127: 739 –746.

Academic Medicine, Vol. 84, No. 8 / August 2009

Continuing Medical Education 13 Fox RD, Mazmanian PE, Putnam RW. A theory of learning and change. In: Fox RD, Mazmanian PE, Putnam RW, eds. Changing and Learning in the Lives of Physicians. New York, NY: Praeger; 1989. 14 Chen DT, Mills AE, Werhane PH. Tools for tomorrow’s health care system: A systemsinformed mental model, moral imagination, and physicians’ professionalism. Acad Med. 2008;83:723–732. 15 Mazmanian PE, Waugh JL, Mazmanian PM. Commitment to change: Ideational roots, empirical evidence, and ethical

implications. J Contin Educ Health Prof. 1997;17:133–140. 16 Mazmanian PE, Daffron SR, Johnson RE, Davis DA, Kantrowitz MP. Information about barriers to planned change: A randomized controlled trial involving continuing medical education lectures and commitment to change. Acad Med. 1998;73:882– 886. 17 Davis D, Thomson MA, Oxman AD, Haynes RB. Changing physician performance. A systematic review of the effect of continuing medical education strategies. JAMA. 1995; 274:700 –705.

18 American Board of Internal Medicine. Improve Your Practice With PIMs. Available at: (http://www.abim.org/pims/choose/ module/hypertension.aspx). Accessed April 13, 2009. 19 Collins J, Porras JI. Successful Habits of Visionary Companies. New York, NY: HarperCollins Publishers; 2002. 20 Box GEP. Robustness in the strategy of scientific model building. In: Launer RL, Wilkinson GN, eds. Robustness in Statistics. New York, NY: Academic Press; 1979.

pride in my ability to counsel patients on their options, giving each a clear conscience that he or she was making an informed, empowered decision. Yet I was also struggling to learn my craft with the handicap of impaired vision. As a result, I found myself in a Toronto hospital at the sharp end of the scalpel for a change.

career was resting on the results of the surgery.

Teaching and Learning Moments Dull to Sharp . . . My left eye opens to a curtain of blue paper obscuring the operating room lights above. When I close my left eye again, I am surprised to see movement before my right eye, like short streams of mercury flowing across my field of vision. As the sedative wears off, I remember that an ophthalmologist is in the process of replacing my misshapen cornea with that of a donor. I relax my tense muscles a bit with the doctor’s reassurance that everything is going well and that he is close to finishing his sutures. The start of my clinical years in medical school was heralded by an interest in surgical specialties. By then, the impairment in my depth perception was clear. Every suture held taut for me to cut represented a challenge. I lived in the fear that my mentors would discover my condition. I explored the possibility of corrective lenses but lacked the perseverance to wear the uncomfortable contacts. I finally entertained the thought of surgical correction. Before my surgery, I had finished two years of orthopedic residency. I took

Academic Medicine, Vol. 84, No. 8 / August 2009

I now realize that no matter how seamless my procedural descriptions were I could not adequately prepare patients for what they were about to experience. I thought that the combination of medical training and the precise description of the corneal transplant would give me an accurate sense of the operative and rehabilitative phases of my treatment. Despite this preparation, though, I was completely unprepared for the process. Like many of the patients I had counseled, I planned to begin normal activities soon after the surgery. Then I saw a time interval familiar to orthopedic surgeons quoted on my postoperative instruction pamphlet: six weeks. Now I was the one relegated to walking for exercise, the one whose

This experience taught me that it is not possible to empathize with patients unless you have had a similar experience. Despite our advanced training, surgeons cannot stand sideby-side with patients as they undertake the risks of surgery and follow postoperative instructions. Although I can try, I will never truly grasp the effect that injuries and illnesses have on my patients’ lives and the impact of my recommendations on their dreams and aspirations. As my vision sharpens, so does my perspective. When I return to the other end of the scalpel, I will be more able to appreciate what it is that my patients are experiencing and, with what I have learned in mind, able to tailor their care. My eye surgery has given me more than just clarity of vision. Jesse Shantz, MD, MBA Dr. Shantz is a resident, Section of Orthopedic Surgery, University of Manitoba, Winnipeg, Manitoba, Canada; ([email protected]).

1065