Section 5

An Agenda for Research in Ambulatory Patient Safety

Synthesis of a multidisciplinary conference to develop an agenda for research in ambulatory patient safety.

Electronic Medical Record

The "electronic medical record" (EMR) appears to offer substantial gains in patient safety through access to and management of patients' clinical information. Implementation of the electronic medical record (EMR) would affect and potentially improve the management of information, support for physician decisionmaking, and coordination of care. The EMR with decision support could also serve as a "tickler" system to remind providers about abnormal tests or need for followup. Understanding the potential benefits and costs of EMR involves many of the same questions posed for CPOE , including the key elements, the changes in physicians' and others' workflow, and cost of purchase and implementation. The perceived risk is further increased by concerns about the continued development of commercially available EMR systems and the ability of vendors to provide continuing service and updates after implementation.

Specific questions for EMR:

  • What are the key elements, functions, and capabilities of an effective EMR?
  • Does an EMR potentially prevent hospitalization or other high risk, high cost events? Does an EMR with decision support improve outcomes of ambulatory care?
  • What is the impact of embedded clinical decision support in the EMR?
  • What are the risks and benefits to an incremental approach to the adoption of electronic records in a practice? Can key functions be implemented, while allowing incremental adoption of the full EMR?
  • What are the costs and benefits of interfacing the EMR with other data systems in a physician practice or other clinical entity? Is this a key to achieving sufficient value from an EMR?
  • How does the ambulatory practices' current processes and systems impact the selection of an EMR system most likely to work for that entity?
  • EMR represents a large investment, both financially and in terms of organizational commitment and change. Are there ways to gain some of the benefits expected from EMR by improving existing processes? Are there ways to achieve some benefits at lower costs by a combination of process improvement and elements of EMR without a large financial investment?
  • EMR is expected to markedly improve the quality of information in the medical record, and the ability to retrieve it for research as well as for patient care. The value of that research would likely be available throughout health care. If there are features of the EMR or of the way a practice uses it that are required for the benefits of research (a "community good") to be realized, are these costs to be borne by the individual entities?
  • The political context for management of individuals' health information is in flux. Until some of these issues are clarified or resolved, there may be increased risk associated with an investment in an EMR. For example, what will be the effect of HIPAA regulations? What proportion of commercially available EMRs are HIPAA compliant?

Other Technologies

The challenge in medicine is to understand how new technologies, such as the Internet and handheld wireless devices, can assist patients and providers in making health care safer.28 For example, use of handheld devices in health care are still in the early stages of development and evaluation. These devices may be used for immediate decision support to avoid drug-drug interactions and promote the proper use of the right medication at the right time. Though there are limited data to support the use of electronic prescribing, many of the conference participants recognized the inherent value of automated prescribing and decision support.

The Leapfrog Group and others are recommending action that they believe will improve patient safety, such as the implementation of physician computerized order entry for medications by hospitals. For what actions do we have sufficient evidence to be reasonably confident that those actions will improve safety in ambulatory care? AHRQ has recently supported $5 million in grants to address the role of clinical informatics to promote patient safety. Many of the funded projects focus on the role of information technology, such as decision support, handheld devices, and electronic medical records.

Recommendation 6: Further research should be supported on the role of information technology to improve ambulatory patient safety, including computerized physician order entry and electronic medical records. These technologies should be evaluated within the larger contexts where they would be implemented and used:
  • The processes within which the technology will be embedded (e.g., the medication process).
  • The systems within which the technology will be used (e.g., the physician office practice).

Return to Contents

Improving Ambulatory Care as a System

Patients and physicians function today within an ambulatory care "system" that does not provide them with the support they need to manage information, ensure effective communication, and coordinate care. Many of the factors that increase patient risk—and many opportunities to reduce risk—are found in clinical processes, clinical support systems, and administrative support systems (Nolan 2000, Barach and Moss 2001). Physicians and their clinical and administrative colleagues will need to better understand the nature of risks and the multiple factors (human, system, and their interaction) that are involved in the occurrence of injuries, accidents, and errors. Improving these systems, such as those to ensure that diagnostic tests are tracked and results promptly acted on, will likely draw on the expertise not only of physicians but also other clinicians and practice administrators.

Safer and more effective ambulatory care will likely require better teamwork among clinicians and patients, perhaps with the exception of patients with simple, acute problems (Helmreich and Schaefer, 1994, Uhlig et al. 2001). Greater teamwork will require a better understanding of the skills and contributions of each clinical discipline, and may require some redistribution of responsibility and accountability from physicians to other clinicians in the "care team." Teamwork among physicians, nurses, and other clinicians may be particularly important in critical care and complex surgery, but teamwork is intertwined with communication and coordination of care more generally. Clinical teams may be particularly effective in managing chronic illness with the patient, who must take an active role in managing her own care as a member of the team (Wagner 2000, Holman and Lorig 2000). There is value in educating patients about their condition and teaching them techniques for self-management such as teaching asthmatics to properly use peak flow devices, and to adjust their medications based on peak flows and symptoms using simple rules. Patients may also benefit from a better understanding of how to navigate and use the care system, and to be more assertive and to raise questions if they do not understand their care, or think something may be going wrong.

Research Questions: What is known about teams and teamwork in health care and in ambulatory care in particular? What can be learned from other industries and fields, and what will be different because of the specific characteristics of patient care? What are the essential features in terms of team composition, team leadership, how the team functions, how decisions are made (Hackman 1990, Katzenbach and Smith 1993, Helmreich and Schaefer 1994, Gaba 1994, 2000, Uhlig et al. 2001)? How are communication and coordination managed? How are tasks matched with expertise? For example, what tasks can be assumed by clinical educators (e.g., for diabetes), clinical pharmacists, nurse care coordinators, by veteran patients?29 For which kinds of patients and patient problems/conditions is "team care" superior to the traditional physician-centered approach? What kinds of education, training, and tools are beneficial?

Recommendation 7: Support research to understand teamwork in health care and how greater teamwork contributes to patient safety. Assess approaches to teamwork in health care to understand their contribution to improving patient safety, particularly ambulatory care, and identify or synthesize approaches that are well suited to health care and its culture. This work would require expertise from a number of fields, including business and the social sciences, to understand the human and cultural aspects.


Beyond the need to improve processes and systems is the importance of a "culture of safety" (Weeks and Bagian 2000, Pizzi et al. 2001). Even with well designed processes and systems and devices, people must have a commitment to safety, and to take best advantage of the skills, knowledge, and commitment of colleagues and work with available processes and systems. With a commitment to safety, people will minimize the potential for error that is always present, and will contribute to improving design of processes, systems, and devices to decrease errors. Achieving a culture of safety isn't a matter of better people, more skilled people, or more committed people, but rather of how people view themselves, their work, and their organization. Physicians and other clinicians are selected, educated, and acculturated to a high level of commitment to safety and quality, but they think of those ideas mostly with respect to their own actions. They generally are not aware of the contributions of processes and systems (and habits) to risk and error, or to reducing risk and improving safety.

Many physicians subscribe to a "culture of individual accountability" in medicine that can be a barrier to achieving risk and error reduction through improvement in processes and systems as well as through improvements in individuals' knowledge and skills, and can inhibit the development of greater teamwork (Paget 1988, Forkner-Dunn 2000, Wu 2000). Physicians who believe that safety depends primarily on acceptance of individual responsibility and that adverse events are usually the result of individual error will not readily perceive opportunities to improve safety by improving processes and systems, or by greater teamwork and shared responsibility and authority.30 Physicians' beliefs about individual accountability are shaped during education and training, and reinforced by our fault-based medical liability system. Beliefs about accountability are also linked to expectations of autonomy and resistance to standardization. If "standardization to excellence" is one key to improving safety, physician resistance to standardization is an inhibiting factor.31

Recommendation 8: Support research to assess the degree to which the prevalent culture is a limiting factor in achieving improved patient safety (and quality of care) through greater teamwork and a systems approach to management, coordination, and communication. Identify ways to enable physicians to more easily understand and see the potential of these approaches to improve safety and quality, and to understand their role in leading work to improve.

Organization of Physician Practices

Most physician practices are relatively small (<10, <5 physicians) and many have very limited access to capital. Data suggest that the most efficient group practice size financially is in the range of 5-15 physicians. Do such groups have the capability to create the culture and infrastructure needed to improve safety? Will optimal size change, particularly if practices will need more extensive infrastructure to improve safety and quality of care?

What organizational structures for physician practice will enable substantial improvements in infrastructure to support ambulatory care? Can management services organizations (MSOs) provide the needed managerial and infrastructural support for independent practices? Do group practices provide a good starting point for building these structures (Luft and Greenlick 1996, Benedict 1996, Royer 1999, Messinger and Welter 2001, Sherer 2000)? How do current group practice structures and practices compare to what we might envision as fully effective in this context? Ultimately, will integrated delivery systems be required?

Return to Contents

Priorities for Research

In addition to recommendations for specific areas of research, several "general principles" for prioritizing research and demonstrations emerged from the conference discussions.

Implementation Is the Rate Limiting Step: Research to Understand How to Apply What We Know

There is considerable knowledge of the characteristics of care processes and systems that improve safety and quality, and a substantial number of good examples from which others can learn (for example, Staker 2000, Wagner 2000). But changes that have been demonstrated to improve care and safety are slow to disseminate, and to be adopted by others (Institute of Medicine 2001). Implementing what is known to work is often the "rate limiting step" for improving patient safety. Therefore, research to understand how to effectively implement changes that will improve patient safety should be given high priority. There is greater value in experiments and demonstrations (which also increase our understanding of the epidemiology of safety) at this point than in observational studies alone.

We know more about the "engineering" side of effective change—the processes, technologies, methods for managing information and communication—than about the "human" aspects of effective change. We need to understand how to effect and manage change, and to apply what has been learned from the existing body of research on diffusion of innovation.

How do persons and organizations learn what would improve care and safety? What are good and trusted sources of information? How do they learn how to implement potential improvements, and gain the confidence to do so? Successful examples—case studies and "stories" from peers—may be the best way for others to learn that safety can be improved, and to learn how to do it. This approach has the advantage of focusing on the positive ("improving patient safety") rather than the negative ("the errors") while pursuing the same goal.

Several conference participants suggested that in health care in general, and physician practices in particular, there is pressure to see more patients, provide more services, and to reduce practice costs—so that most physicians believe they have neither the time nor the resources to focus on patient safety, particularly since most believe that their practices are already safe for patients. It appears that physicians are incurring higher practice costs per physician and per service provided, receiving decreasing payment per service (in real dollars), and increasing the number of services they provide. The increase in volume of services may result from attempts to maintain net physician income in the face of rising costs and lower payment per service. Physician practices are incurring more costs for administrative functions (billing, etc), and clinicians (physicians, nurses) have to spend more time in these administrative tasks. Is there potential to reduce some of these tasks required by payers and regulators, or to standardize them for all payers and regulators, to reduce the administrative burden?

What are the effects on physicians, nurses, and other staff of feeling that they are on a "treadmill" that leaves them with too little time for their patients and their work? Do they find that their work is less satisfying? Are they concerned about making mistakes when overloaded, and does that lead to greater turnover and "burnout"? Would improving infrastructure to support care reduce risk and reduce stress and burnout?

Research Questions: How can we more rapidly learn from the experiences of successful pioneers in safety and quality such as Intermountain Health Care and the organizations participating in the collaboratives managed by the Institute for Healthcare Improvement? How do people decide to take responsibility for making improvements? What is the role of "leadership"? Who can be an effective leader for change in a physician practice, or an ambulatory surgery center? What is the role of recognition and reward? Are there economic factors that support or inhibit change and improvement in ambulatory health care? Is there a "business case" for improved safety and quality for physician practices and other ambulatory care entities? Is there a role for requiring the adoption of proven interventions to improve safety and quality? By regulators and payers? In aviation, adoption of safe practices and safety enhancing innovations can be mandated. What is the role of licensing bodies for physicians and other professionals? Of physician specialty board certification and recertification?

Recommendation 9: Support research and demonstrations in ambulatory care that identifies, develops, and disseminate successful approaches to implementing changes that improve patient safety.

The importance of research to understand how to successfully implement changes that will improve patient safety implies a preference for research that tests or demonstrates improvements to safety in addition to increases knowledge of the epidemiology of safety. That is, where possible, research should include an action component—an experiment or demonstration—and by including a baseline assessment will contribute to our knowledge of "what is" both before and after the experimental intervention.

Sustained Support for Multifaceted Research to Improve Care and Support Systems

Conference participants articulated the principle that while improvements in patient safety will be achievable through focused, specific changes in processes and systems (e.g., improving medication order entry), substantial gains in patient safety will often require interdependent changes in clinical, clinical support, and administrative processes and systems (see discussion of the infrastructure of ambulatory care earlier in this synthesis). Changes of this magnitude and complexity will require sustained effort, and sustained support.

Recommendation 10: Provide sustained multiyear support for projects designed to improve patient safety (and quality) through changes in both clinical processes and in clinical support and administrative support processes and systems. Consider multistage projects, with the hypotheses and experimental design for each stage based on what has been learned in previous stages. Consider projects that also attend to "organizational" aspects of care including structure, culture, and compensation. Provide support for attention to managing change, and for increasing our understanding of how to successfully implement changes that improve safety. Anticipate how "lessons learned" and best practices identified can be disseminated and implemented by others. Attend to aspects of the environment that inhibit or promote improvement in safety, including payment mechanisms and the legal and regulatory environment.

If our focus were inpatient care, we would describe this as "organization wide" research. That description is apt for large group practices and some other ambulatory entities. But much of ambulatory care is characterized by small entities with relatively low degrees of organization and rudimentary infrastructure, yet those settings (e.g., most small physician practices) provide a large fraction of ambulatory health care, and are very likely to continue to do so. It will be challenging to design projects that identify and improve an infrastructure for care in these settings, and that appeal to the physicians, administrators, and others. Perhaps project design would be informed by a greater understanding of the organization and infrastructure of large group practices and other larger, more complex ambulatory care entities.

In addition to such large, multifaceted studies, we will also need many smaller, focused studies and experiments. But they, too, should be designed with an understanding of the larger context of ambulatory care.

Focused Studies within the Larger Context

Focused studies and experiments provide opportunities to rapidly understand critical risks and to quickly design and test specific interventions to improve safety. Focusing on particular aspects or areas of care about which we have considerable knowledge and/or where we know there is substantial potential to make progress should yield good examples of successful improvement in safety, and models for research and improvement useful in other areas. For example, one might focus on areas of high risk or on common and recurring processes and tasks that entail risk, such as areas that generate a substantial fraction of malpractice claims (e.g., failure to or delays in diagnosing breast cancer) or transitions of patients from one provider to another.

For example, focused studies on medication practices provide opportunities for research and demonstrations to improve safety. Understanding and improving the "chain of medication"—prescribing, delivery, and use—may provide a useful model for other areas of patient safety. The use of medications is ubiquitous, and there is evidence of patterns of widespread injuries from medications. Medication safety cuts across many aspects of ambulatory care and safety. Medications are used in most ambulatory sites, and involve most clinicians and patients. Safe medication practices depend on individual knowledge and actions in the context of processes and systems. A large literature of best practices and available tools to improve medication practices can be put to use now. Gains in safety can be expected to be applicable in other areas for patient education, communication, more effective use of technologies. Relatively easy identification of costs avoided through improved safety should enable development of the business case for improving medication safety.

Successful experiments to improve medication safety would provide early examples of success that could in turn build understanding and support for improving other aspects of ambulatory patient safety.

Practice Based Networks for Research, Development, Evaluation, and Dissemination

Practice based research networks may provide an excellent setting for research and demonstrations in ambulatory patient safety, particularly for physician practices (Green and Dovey 2001). There are several models upon which to build, including the experience of the Institute for Clinical Systems Improvement (ICSI) in the Minneapolis-St. Paul area (Reinertsen and Mosser 1998) and the collaboratives organized and supported by the Institute for Healthcare Improvement (IHI) ( Networks or collaboratives provide an opportunity for learning from the experiences of known peers, and should be an excellent "laboratory" for studying the dissemination and adoption of innovations and factors that enable or inhibit that.

One feature of the ICSI experience in Minnesota is particularly notable: Five of the largest health plans and payers in the area have agreed to support and promote implementation of identified best practices (Quality Letter 2001). These five payers will coordinate their actions, so that providers will receive fewer conflicting and confusing messages and be subject to fewer conflicting expectations from different sources.

AHRQ is supporting 19 Primary Care Practice-Based Research Networks and 9 Integrated Delivery System Research Networks at the regional and national level. Many of these networks have received AHRQ funding to address ambulatory safety.

External Environmental Factors in Addressing Patient Safety in Ambulatory Care Settings

Efforts to further understand risks and improve patient safety across the spectrum of ambulatory care settings will occur within a complex social context that is not particularly coordinated or aligned at this point in history. Significant factors in this context include increased media scrutiny of ambulatory care incidents and risks, an evolving tort liability system, accumulating evidence of both ambulatory care risks and successful interventions to manage those risks, Federal and State law initiatives, the responsibilities and authority of professions, the organization and financing of care in both the public and private sectors, the evolving role of consumers as active partners in health care decision-making and management, escalating accreditation activity and complexity in ambulatory care and other aspects of the health care delivery system structure and culture.

These factors are highly interdependent, and several of them have been discussed or alluded to in other sections of this synthesis. The central challenge facing conference participants and others interested in advancing patient safety in ambulatory care is that the ideas and proposals they prioritize and recommend will require support of stakeholders in the external environment. In this section, we summarize the discussion of the conference participants with respect to several of these external factors, identify developments or trends that are emerging now, and explore what might be done to enable change.

Expansion of Collaborative Efforts to Address Patient Safety

The role of collaborative initiatives among health care service providers, organized by the Institute for Healthcare Improvement (IHI), the Institute for Clinical Systems Improvement (ICSI) and others to advance patient safety have already been discussed. Coalition efforts across broader communities of diverse stakeholders also are beginning to emerge, as the result of media attention, the efforts of the National Patient Safety Foundation to hold regional meetings, and a series of "regional summit" gatherings supported by the AHRQ, among other influences.

Who are the principal stakeholders? They include patients, consumers, media; physicians and other clinicians, physician practice administrators, employers, purchasers, payers, product makers, legislators, regulators, lawyers, accreditors, risk managers, liability insurers, educators, and researchers and funders of research. What are their interests and their beliefs? Are there changes for which a consensus can be reached and commitment to joint action forged?

Stakeholder coalition efforts have the potential to disseminate safety science broadly, increase consistency in the use of language/terminology used to discuss risk and safety, advance the alignment of different stakeholder groups behind patient safety priorities and strategies, increase trust among stakeholders and draw new resources to patient safety initiatives. If these coalition efforts can achieve buy-in by leaders of different stakeholder groups, they can serve as important avenues for shifts in attitude and other aspects of cultural change.

The most highly developed coalition initiative to emerge to date is the Massachusetts Coalition for the Prevention of Medical Errors (MCPME), which began to form in January 1997. Although not focused specifically on ambulatory care issues, the coalition's first publication targeted medication use by consumers—an issue with clear application to ambulatory care service delivery and follow-up (MCPME, 1999). The Massachusetts coalition has served as an active forum for discussion of a systems approach to safety, appropriate government action and the accountability of different stakeholders. Legislation introduced into the Massachusetts General Assembly is discussed at coalition-organized meetings and arguably is better informed as a result of those meetings. MCPME leaders have been invited to testify at legislative hearings and the coalition is financially supported, in part, by State government (Barach and Kelly 2000). Similar regional coalition efforts are underway in Virginia and other States as well.

Leaders in the aviation safety movement have emphasized the importance of engagement and buy-in by all stakeholder groups, including those who were highly critical of the movement (Billings 1998). At this time, it is fair to assume that these collaborative efforts will play a crucial culture-carrying role in health care as well (Wilson and Hatlie 2001).

Expansion of Media Scrutiny of Ambulatory Care Safety Incidents and Responses to Incidents

Media attention to safety risks has proven to be a powerful motivating influence in safety improvement, although seldom a comfortable one for providers of the services deemed by media to be unsafe. It is arguable that the national focus on patient safety beginning in the late 1990s is primarily due to media coverage of the Betsy Lehman and Willie King incidents in 1995 and 1996 , as well as the National Institute of Medicine's report, To Err is Human, in 1999. Journalists covering patient safety issues express the view that it is their role to hold the system accountable for ensuring safety, and often demonstrate skepticism or cynicism regarding the effectiveness of professional or institutional self-regulation (Millenson 1997, Tye 1999). Commentators in the patient safety community cite the importance of educating the media and, through the media, educating its audiences, which include consumers, policymakers and regulators. Some have argued that the media itself is accountable to the public for performing this education function, as opposed to merely reporting news (Wilson and Hatlie 2001).

Media attention to a series of adverse outpatient surgery outcomes in Florida in 2000 created a significant government response, including a moratorium on outpatient surgery and the establishment of a increased government oversight and investigation (Prager September 2000 , Rowland 2000). As more care moves from acute to ambulatory care settings, media scrutiny of ambulatory care safety risks and incidents also is likely to increase.

Return to Contents
Proceed to Next Section

Current as of December 2001
Internet Citation: Section 5: An Agenda for Research in Ambulatory Patient Safety. December 2001. Agency for Healthcare Research and Quality, Rockville, MD.