Improving care and practice through learning health systems

Darryl W. Roberts

Nursing Management · 2013 · 11 citations · 2 references

Concepts

Abstract

Pressure ulcers and falls are two patient safety problems that nurses can prevent by using the best-available practices. Nurses at the point of care, unit managers, and nurse executives know this; however, best practices have a long road to navigate from university-based studies to the community hospital bedside, home care practice, or long-term-care facility. In fact, the Agency for Healthcare Research and Quality reports that it takes about 17 years to navigate this road.1 In an era when many practicing nurses don't recall a time before Google, this is too long!FigureIn Crossing the Quality Chasm, the Institute of Medicine (IOM) proposed that the nation's healthcare system will improve if all participants engage in a system in which everyone (patients, nurses, physicians, and pharmacists) rapidly shares and learns from each other's successes, failures, mistakes, and remedies.2 In continuation of its mission, the IOM took this concept of a rapid learning healthcare system (LHS) and engaged in further work to identify how to take it from concept to reality, recently publishing the report Best Care at Lower Cost.3 Aided by the phenomenal improvements in information and communication technologies, such a system is now both feasible and practicable. It will, however, take changes in perceptions to make it a reality. Can health systems learn? At the core of an LHS lies the intent to quickly and accurately identify best practices and share them among clinicians, payers, healthcare consumers, patients, and other interested parties. After implementation, an LHS can improve healthcare through standardization, immediacy, security, efficiency, effectiveness, and safety by leveraging the healthcare system's own capabilities and by learning from other industries. The IOM proposes that healthcare can become as standardized as the automobile manufacturing industry, which builds several different models of nearly perfect cars from a single basic core; as immediate and secure as the banking industry, which provides its customers with real-time, secure financial transactions through computers and smart phones; as efficient as the building trades, which leverage a blueprint and a project manager to coordinate the skills of electricians, plumbers, and carpenters in such a way that each trade can efficiently and effectively complete its work without interfering with another trade's work; as safe as the aviation industry, which moves millions of people from place to place each day with so few incidents and such high safety standards that smoke from a rechargeable battery grounds an entire aircraft fleet and makes international news because the industry and its customers expect perfect safety. The IOM and Dr. Charles Friedman, in collaboration with the Joseph H. Kanter Family Foundation, have each published models of how an LHS might look. The IOM model suggests that the use of appropriately placed incentives, in combination with changes in culture supported by forward-thinking leaders, will promote a cycle in which science becomes evidence becomes care, which improves science. All of this is underpinned by technology to make it rapid, secure, and effective. (See Figure 1.) As an example of what an LHS can do, the IOM report proposes that it will increase patient engagement, thus enhancing patient ability to “...find the right care for their individual characteristics, needs, preferences, and circumstances.”3 The IOM approach to an LHS makes clinicians responsible for improving the flow of information to their patients, particularly regarding treatment options. It will also increase patients' responsibilities for voicing their needs and preferences.Figure 1: IOM model of a continuously learning healthcare systemThe model published by Friedman attempts to represent the reality of the multiple stakeholders working interactively through a cloud of technology overlaid with a sense of shared governance and trust, a commitment to patient engagement, and a technologically supported means of iterative and rapid analysis of information and continuous dissemination of knowledge.4 (See Figure 2.) In its proposal, the Kanter Family Foundation suggests that a successful LHS will follow several core values.5 It must be person-focused; protect privacy; include all interested parties in its governance; be transparent and accessible; adapt to changes rapidly; be self-governing, cooperative, and participatory; maintain a high level of scientific integrity; and provide value to its participants.Figure 2: The Kanter Family Foundation model: A learning health system for the nationFor an LHS to work, clinicians, patients, and other interested parties need to share information openly, giving everyone equal access to the knowledge necessary to facilitate sound decision making. But what would that look like? Building an LHS One study reported a patient-centered medical home pilot that used the tenets of an LHS to maximize its effectiveness.6 (See Figure 3.) It began with scanning and surveillance to assess the need for such a system. The authors determined that internal factors, such as patient experience and costs, would benefit from an LHS implementation. From there, they invited key stakeholders to participate in its development through an iterative design process. Part of that process was to identify the technologies available and required for its effective implementation. The implementation process used a pilot versus control format familiar to clinical research and aimed at early adopters in a way familiar to information technology.Figure 3: Group Health Research Institute learning health system pilotThe next three phases are important anywhere, but imperative to an LHS: evaluate, adjust, and disseminate. The evaluation phase used a mixed qualitative/quantitative approach that asked “Did you like it?” and “Did it make a difference?” The evaluation design was pragmatic, but useful, because it was a small pilot/control comparison. Determining whether to use the system held a higher priority than did scientific rigor because it was intended for real use by real people. During the adjust phase, adjustments were made to facilitate the LHS implementation to a broader group and standardize what could be standardized. Finally, the group disseminated their success through a combination of internal and external discussions, webinars, and peer-reviewed journals. The broad scope of the dissemination methods made it possible to reach a wide audience. Clinicians, hospitals, insurers, and patients have to overcome some major cultural issues before building an LHS. Our nation has to accept that mistakes and errors will happen, humans will make them, and systems can cause or prevent them. After this happens, we have to fix the mistakes and share what we've learned so that we can all learn. These steps make an LHS, which may be impossibly simple to create given the available technology and information, and simply impossible given the fear of showing weakness and exposing one's self or organization to litigation. Large healthcare organizations, provider groups, academic hospitals, insurers, and many other participants in the national healthcare system already have and use the technology necessary to identify the systematic weaknesses, assess methods to overcome those weaknesses, implement those methods, evaluate their effectiveness, and disseminate the results. Learning for the future The current state. A nurse at your organization has just made a medication administration error. Fortunately, the result wasn't lethal, but it could've been. The cause was a systematic error in the communication from the prescriber, to the pharmacy, to the nurse. A hospital in the next town experienced a similar error 3 months ago and established an effective protocol to prevent it. Unfortunately, you have no way of knowing this because there's no mechanism for sharing that kind of information and there's no incentive for doing it, either. In fact, liability concerns act as disincentives to sharing error-related information. Because of those facts, nurses and other clinicians in every healthcare setting will continue to make the same errors within their organizations even though similar organizations have found, but don't share, remedies. The future state. A nurse at your organization discharges a patient one day post-total knee replacement. Using the latest evidence on care coordination from a metropolitan hospital 1,000 miles (1,609 km) away shared through an LHS, she integrates his preferences and needs into a plan of postacute care that includes the appropriate referrals; assists him with arranging appointments; and provides access to the knowledge, transportation, and pain management resources necessary to ensure a positive outcome for him and his family. If she identifies any problems, she has an established method of rectifying them and sharing her newly obtained knowledge with others.

References

2