Research Activities asks the experts
The new AHRQ evidence report Making Health Care Safer II: An Updated Critical Analysis of the Evidence for Patient Safety Practice described on this month’s cover, identifies the top 10 patient safety strategies (see box on page 3) that can be implemented immediately by health care providers. If widely implemented, these patient safety strategies have the potential to vastly improve patient safety and save lives in U.S. health care institutions.
Research Activities interviews four of the many patient safety leaders involved in the development of the safety report: James Battles, Ph.D., AHRQ Task Order Officer for the report; Peter J. Pronovost, M.D., Ph.D., Senior Vice President for Patient Safety and Quality at Johns Hopkins Medicine; Paul G. Shekelle, M.D., Ph.D., Director of the Southern California Evidence-based Practice Center; and Kaveh Shojania, M.D., Director of the Centre for Patient Safety, University of Toronto.
Research Activities (RA): AHRQ has developed clinical tools to help clinicians implement a number of the patient safety practices identified in the new report. What impact have AHRQ’s clinical tools already had in implementing the patient safety strategies identified in the report?
James Battles, Ph.D. (JB): Many of AHRQ’s tools and resources designed to help institutions implement the safe practices outlined in this report have been extremely effective in reducing harm to patients. For example, our CUSP (Comprehensive Unit-based Safety Program) tools and support programs have had dramatic results on the outcome of care. Use of CUSP reduced central line-associated bloodstream infections (CLABSIs) in intensive care units by 40 percent nationwide. We are seeing similar results from our work with catheter-associated urinary tract infections (CAUTIs), with over a 47 percent reduction in CAUTIs in 800 participating hospitals in 33 States so far. Our tools to improve teamwork in health care like TeamSTEPPS® have been adopted by over 30 percent of hospitals nationwide. Improved teamwork has been shown to reduce surgical mortality and to reduce birth injuries. Improved communication and teamwork can also reduce medical liability claims.
RA: Given that several of the identified patient safety practices are targeted toward "common" safety problems that occur once per 100 hospitalized patients, such as potential adverse drug events, falls, and blood clots, do you anticipate a large impact on patient safety if they are implemented?
JB: If these safe practices are adopted and used, they will dramatically reduce harm to patients and lower the cost of care. However, they must be used and adopted by all care providers within the organization and units within the institution in order to achieve results. Implementation of safe practices is hard work and often requires significant behavioral change and changes in culture within institutions. When implementation is successful, the impact on outcome of care is significant.
RA: What would you say is the biggest take-home message of the report?
JB: Adoption of safe practices can have a dramatic impact on the care provided by institutions, hospitals, nursing homes, and medical offices. It is also important to not only look at the evidence about the effectiveness of a given patient safety practice, but also examine the evidence around adoption and uptake of these practices. You cannot expect changes and improvements in health care if a given safe practice is not implemented to improve care.
RA: With AHRQ funding, you developed CUSP to reduce healthcare-associated infections (HAIs) such as CLABSIs in hospital units. This bundle of safety practices has saved thousands of lives and hundreds of millions of dollars and is now being implemented across the United States and in several other countries. CUSP can be applied to several of the strongly encouraged patient safety practices in the report, such as prevention of CLABSIs, catheter-associated urinary tract infections, and ventilator-associated pneumonia. Do you anticipate that CUSP will be used to implement these patient safety practices and in your experience how easy is it for health care facilities to implement CUSP?
Peter J. Pronovost (PP): Yes, we believe CUSP can apply to many of the safe practices identified in the report. In our efforts to prevent CLABSI and other preventable harms, we learned that successful efforts have three components: (1) a model to implement evidence-based practices that includes a checklist of best practices, identifies and mitigates barriers to implementing the checklist items, and ensures patients receive the checklist items; (2) feedback of performance data on both the outcome and, when possible, process; and (3) an intervention to improve culture and engage local staff called CUSP. Change happens at the bedside. If improvement efforts are to be successful, they must engage and empower clinicians in the care areas. CUSP is that generic strategy and can be applied to any area, inpatient or outpatient, and coupled with efforts to reduce any type of preventable harm.
RA: A core part of several patient safety practices identified in the report is an emphasis on teamwork and communication to reduce patient safety problems. How have these areas improved in the 12 years since AHRQ’s initial safety report and what barriers remain?
PP: The second step in CUSP is to ask staff how the next patient will be harmed. Poor teamwork is the most common answer to this question. Yet poor teamwork is not one behavior; it could be many, and the teamwork interventions should target the specific teamwork challenges. Luckily, AHRQ supported the development of TeamSTEPPS, a robust teamwork training program. We have worked to link specific teamwork challenges identified through CUSP with the specific tools in TeamSTEPPS to provide clinicians the tools that best address their needs. While broad teamwork training is beneficial, it is also effective and efficient to offer specific training to address clinicians’ specific concerns. Still, more work is needed to better diagnose teamwork problems and map that diagnosis to specific therapies (training).
RA: Did anything surprise you about the report’s findings?
PP: I was impressed by how much the science had advanced. In the first report, the science was so immature and most of the recommendations were supported by weak evidence. Since then, the number of research studies evaluating patient safety interventions has increased dramatically, as well as the number of interdisciplinary teams working on safety and the quality of the studies. Yet many studies still lack a clear program theory, too many studies still use a "pre-post design," and many studies lack valid outcome measures and provide little information about context. Nevertheless, it takes time to mature a field. Given that preventable deaths are likely the third leading cause of death, the research investment in patient safety is far too small for the magnitude of the problem. Greater investments would further accelerate the science of patient safety and save lives. It is encouraging that patient safety has finally become a respected scholarly discipline and that an ever-growing cadre of new faculty are focusing their careers on reducing preventable harm. The future of patient safety is indeed very bright.
RA: Successful implementation of safety practices has been shown to be highly context-dependent, often working effectively in some hospitals but not in others. You worked on an early AHRQ report that examined the role of context in patient safety. Could you give an example of how context can facilitate or impede the implementation of patient safety practice?
Paul G. Shekelle (PS): The role of context and implementation effectiveness is still early in its maturation as a field of scientific inquiry, so most studies to date have been descriptive and not hypothesis-testing. But one of the most commonly found contextual factors facilitating successful implementation is leadership support, either at the top level or the unit/program level or both. If the leaders are 100 percent behind making something happen, it most often does. Implementation of a patient safety practice might still succeed even if the leaders aren’t all-in on it. But when problems arise during an implementation, as they almost always do, without the leaders’ active support it is harder to overcome the barriers to success.
RA: How has the patient safety environment (context) evolved in the past decade in hospitals and other health care facilities both in recognizing problems such as preventable medical errors, taking steps to address them, and making patient safety a priority?
PS: I think the first big difference is that we now recognize that patient safety is a big problem, bigger than was previously thought before the Institute of Medicine report, and that making care safer is mostly not a matter of exhorting individual clinicians to do a better job or check their work more carefully, but rather building systems that will produce safer care. Years ago most hospitals did not have a chief safety officer or the equivalent, yet now almost all do. The importance of patient safety as an institutional concern has been elevated. And this concept of building safer systems is gaining ground.
RA: What will be the biggest challenges for health care organizations in implementing the patient safety practices identified in the report and what will help them implement these practices more quickly?
PS: I think the biggest challenge will be the recognition that this will take a lot of active effort to make it work. We’d like it if we could make care much safer simply by buying a better technology or some other highly standardized, off-the-shelf solution. But the reality is that that’s not going to be the answer in almost all cases. What is going to work is looking at what has worked in other institutions, and then not leaping to the conclusion that if you could just implement that exact solution at your institution then all will work out fine. Instead, you need to look at the individual components of that patient safety strategy that was successful somewhere else, and see if they are likely to work in your practice environment. This will require talking to frontline staff, the people who are going to be most affected by the implementation. If something doesn’t seem like it will "fit" in your environment, is there another way you might try to achieve the same or similar effect? Try it and see. Collect data. See if your care processes or outcomes are improving. See what can be further improved. Above all, get started.
RA: You have led many educational initiatives in patient safety and you and Dr. Robert Wachter authored a book on what you termed the "terrifying epidemic of medical mistakes." Do you see this report as a significant step forward in guiding hospitals toward fewer medical mistakes by identifying encouraged patient safety practices?
Kaveh Shojania (KS): We know that some 5–10 percent of hospitalized patients suffer harm from their medical care, i.e., "adverse events," so it isn’t actually as sensational as the title—a concession Dr. Wachter and I made to the publisher—would imply. Yet, in recent years, a few studies have shown that the rates of preventable adverse events have not decreased over time. In order for rates of adverse events such as bloodstream infections from central lines to go down, three things have to happen: we have to have patient safety practices that we know reduce common adverse events, these practices have to have been disseminated quite widely, and we have to have a measurement tool sensitive to changes in specific types of adverse events. This AHRQ evidence report speaks to the first of those necessary developments—the identification of effective patient safety practices. If one compares patient safety to a major area of biomedical research, such as cancer, we’re not doing that badly, as Dr. Eric Thomas and I point out in a recent editorial in BMJ Safety & Quality. The "war on cancer" has been going on for 40 years and has consumed probably over $1 trillion, orders of magnitude more than has been spent on patient safety. Plus, cancer research had a substantial head start in the form of decades of relevant existing research and a huge scientific workforce. The fact that we have even some effective patient safety practices (and many more promising ones) after only one decade, much less financial investment, far less existing relevant research, and a much smaller scientific workforce, seems to me like a very reasonable achievement.
RA: The March Annals of Internal Medicine features a series of articles included in the report. You are editing a special issue of the BMJ Quality & Safety that will include additional articles based on chapters in the report. Do you see these special issues of the journals extending the impact of the report to more clinicians and researchers?
KS: Very much so. AHRQ has excellent dissemination processes that will reach many health care decisionmakers and researchers actively engaged in patient safety work. But medical journals will reach a broader audience of clinicians and researchers interested in topics covered in the new report. Annals of Internal Medicine has a very large audience of clinicians and researchers in the United States and abroad. BMJ Quality & Safety has a broad readership internationally, including researchers and health care managers actively involved in patient safety work. So I think the publication of many of the articles from the safety report in two medical journals—one a U.S.-based general medical journal and the other an international patient safety journal—substantially adds to the dissemination of this work.
RA: What changes in how we view evidence for safety practices have taken place between the first safety report and this report?
KS: First of all, the changes are the same as would occur for any area of research—some things we thought were very promising turned out not to be. Evidence changes over time. One of the more clear-cut examples of this was a very clinical patient safety intervention—the use of peri-operative beta-blockers to reduce post-operative cardiac complications. At the time of the previous report, it received the second highest ranking in terms of the recommendation to implement. But due to later evidence, we no longer can endorse this patient safety practice to the extent that we felt was appropriate in 2001. Signs of progress in the new report include evidence that provides robust support for intensive teamwork training in terms of improvements in hard clinical outcomes. More generally, though, I think the updated evidence report highlights the degree to which it is often difficult to distinguish a patient safety practice from the strategies required to implement it successfully. We are looking closely at the role of context and the degree to which we can identify safety practices that work in some settings but not others. The next step will be to develop measures of specific aspects of context that can be reported in individual studies and applied in syntheses of the studies addressing a given topic.