JAMES P. BAGIAN, M.D., PE
DIRECTOR, NATIONAL CENTER FOR PATIENT SAFETY
VETERANS HEALTH ADMINISTRATION
DEPARTMENT OF VETERANS AFFAIRS
VA PATIENT SAFETY PROGRAM
COMMITTEE ON WAYS AND MEANS
SUBCOMMITTEE ON HEALTH
February 10, 2000
Mr. Chairman and Members of the Committee,
I am pleased to appear before you to discuss VA’s ongoing activities and initiatives to ensure the safety of patients who receive care from VA
I am pleased to appear before you to discuss VA’s ongoing activities and initiatives to ensure the safety of patients who receive care from VA. In December 1999, the Institute of Medicine (IOM) released a report "To Err is Human: Building a Safer Health System." The report reviewed existing studies and concluded that as many as 98,000 preventable deaths occur each year in United States’ healthcare due to error. The IOM recommended creating a new National Center for Patient Safety that would focus on research and policy related to errors in healthcare, improved error reporting systems, improved analysis/feedback methods, performance standards for healthcare organizations and individuals, and other specific governmental actions. Importantly, they cautioned that the focus must be on creating a culture of safety that will require improving systems, not assigning blame.
VA interpreted the IOM report as a validation of our commitment to improving patient safety in our healthcare system. All of the IOM recommendations applicable to VA have either been in place or were in the process of being implemented prior to the release of the report. While VA has had quality and safety related activities ongoing for many years, it was in 1997 that our formal patient safety program was launched (see Attachment 1). Leaders in the field of patient safety and medical error outside VA have participated in the design of our system and recognize VA as a pioneer in these efforts.
During 1997, VA intensified its already extensive efforts in quality improvement by launching a major initiative on patient safety. We recognized that programs to improve quality and safety in healthcare often share purpose and corrective actions. However, we believed that patient safety required a new and different approach. We set out to create a new culture of safety in which our employees detect and tell us about unsafe situations and systems as part of their daily work. Once we know about unsafe situations and systems, we are committed to design and implement new systems and processes that diminish the chance of error.
Highlights of Patient Safety Activities at VA: 1997-Present
VA recognized that patient safety is not a VA-specific issue, therefore we asked other health care organizations to join us in an effort to understand the issues and to act for patient safety. As a result, the National Patient Safety Partnership (NPSP), a public-private consortium of organizations with a shared interest and commitment to patient safety improvement, was formed in 1997. The charter members, in addition to VA, included the American Medical Association, the American Hospital Association, the American Nurses Association, the Joint Commission on Accreditation of Healthcare Organizations, the Association of American Medical Colleges, the Institute for Healthcare Improvement, and the National Patient Safety Foundation at the AMA. Five additional organizations have subsequently joined the charter members in the Partnership: the Department of Defense - Health Affairs, National Institute for Occupational Safety and Health, the Food and Drug Administration, Agency for Healthcare Quality and Research, and the Health Care Financing Administration. This group addresses high impact issues that are of importance to a broad cross section of the healthcare industry. An example of the Partnership’s activity was the establishment of a clearinghouse for information related to the effect of Y2K computer issues on medical devices. The NPSP also called public and industry attention to Preventable Adverse Drug Events and promulgated simple actions that patients, providers, purchasers and organizations could take to minimize their chance of an adverse drug event. (See Attachment 2) The partnership serves as a model of what a private-public collaboration can do to improve patient safety.
In 1998, VA created the National Center for Patient Safety (NCPS) to lead and integrate the patient safety efforts for VA. As the IOM report advises, VA created this center as a commitment to patient safety as a corporate priority with a direct reporting relationship to the Under Secretary for Health. The NCPS employs human factors engineering and safety system approaches in its activities. The first task for the Center was to devise systems to capture, analyze and fix weaknesses in our systems that affect patient safety.
We sought to design reporting systems that would identify adverse events that might be preventable now or in the future. In addition, we sought systems to identify and analyze situations or events that would have resulted in an adverse event if not for either luck or the quick action of a healthcare provider – we call such events "close calls." We believe that "close calls" provide the best opportunity to learn and institute preventive strategies, as they will unmask most system weaknesses before a patient is injured and avoid the liability issues implicit in investigation of injury. This emphasis on "close calls" has been employed by organizations outside of healthcare with great success.
VA consulted with experts (Expert Advisory Panel for Patient Safety System Design) obtaining advice to enhance the design of VA’s reporting systems. These experts in the safety field included Dr. Charles Billings, one of the founders of the Aviation Safety Reporting System, as well as other experts from NASA and the academic community. They advised us that an ideal reporting system a) must be non-punitive, voluntary, confidential and de-identified; b) must make extensive use of narratives; c) should have interdisciplinary review teams; and d) most importantly, must focus on identifying vulnerabilities rather than attempting to define rates of error. VA has used these principles to design the patient safety reporting systems we have in use or in development.
Based on the expert advice and on lessons learned from our first generation mandatory adverse event reporting, the NCPS has developed a comprehensive adverse event, close call analysis and corrective action program which includes an end-to-end handling of event reports. This system not only allows for the determination of the root causes, but also captures the corrective actions as well as the concurrence and support of local management for implementation. The system includes a number of innovations such as algorithms and computer aided analysis to determine the root cause of adverse events and close calls. The Joint Commission on Accreditation of Healthcare Organizations and the American Hospital Association are currently evaluating parts of the system for use.
The improved event reporting system is being pilot tested in VA's VISN 8. Extensive training is used as the new system is introduced to assure full understanding of the search for the root cause and redesign of the system. To date, response from the pilot site is positive. The quality managers and clinicians using the system believe that the new methods analysis of error will make a significant difference in the care of veterans.
A complementary, de-identified voluntary reporting system is in the process of being implemented. It is patterned after the highly successful Aviation Reporting System that NASA operates on behalf of the FAA. It will be external to VA and will allow employees and patients to report unsafe occurrences without fear of administrative or other action being taken against them.
Based on lessons learned, VA has promulgated specific procedures and policies aimed at reducing risk of error. These include such things as restricting access to concentrated potassium chloride on patient care units, use of barcode technology for patient identification and blood transfusions in operating rooms, and for verification procedures prior to injection of radio-labeled blood products. (Attachments 3-6) Based on the observation of a VA nurse when she returned a rental car, VA developed a system for using wireless bar coding to improve medication administration. That system was piloted at the Topeka VA Medical Center and will be in all VA hospitals by June of this year. At least two-thirds of medication errors can be prevented with this system.
In 1999, VA established four Patient Safety Centers of Inquiry. These Centers conduct research on critical patient safety challenges. Activities at the Centers of Inquiry range from fall prevention and operating room simulators to understanding the role of poor communication in patient safety. The Center in Palo Alto, which is affiliated with Stanford University, is a recognized leader in the area of simulation and has been featured prominently in the media. Their simulated operating room allows surgeons and anesthesiologists to train and do research without endangering a patient. VA expects to create additional simulation facilities to train its physicians and other healthcare professionals. One simulator with appropriate staff could train about 600 anesthesiologists and residents-in-training per year. This means that virtually all VA anesthesiologists/anesthetists can be trained in a year on clinical situations that could not be simulated safely in patients. As a result of analyzing common variations during simulated operations, the center has developed a checklist card of facts that should be kept close at hand. These checklist cards will be attached to all anesthesia machines across VA.
VA is partnering with the Institute for Healthcare Improvement to build learning collaboratives aimed at reducing medication errors, a major issue identified in the Institute of Medicine report. IHI collaboratives will affect several hundred VHA personnel each year. Other IHI collaboratives have resulted in measurable improvements and similar results are anticipated with medication errors.
Another key VA strategy to reduce medical errors involves the development of a new curriculum on safety. VA is moving forward with plans to provide education and training relevant to patient safety not only to those already in practice but also at the medical, nursing, and health professional school level. This will be the first time an extensive safety curriculum will be developed and broadly implemented. VA is particularly well situated to lead the educational effort due to the extensive role it plays in the education of healthcare professionals in the United States. (VA is affiliated with 105 medical schools and up to one-half of all physicians train in a VA facility during medical school or residency.) Additionally, we have instituted a performance goal and measure to provide VA employees 20 hours of training on patient safety this year.
VA instituted a Patient Safety Improvement Awards Program to focus interest on and reward innovations in identifying and fixing system weaknesses. Not only does this produce ideas for patient safety improvements that might otherwise go unnoticed but it further reinforces the importance that VA places on patient safety activities. (Attachment 7)
In 1995, VA instituted a Performance Measurement System that uses objective measures of patient outcomes to set goals and reward achievement. Since 1998, VA has incorporated a performance goal and measure for its executives for accomplishment in patient safety activities. Last year, each network had to implement three patient safety initiatives to be fully successful and six initiatives to be outstanding.
Other performance goals and measures assess the use of Clinical Practice Guidelines. By holding entire medical centers and geographic networks responsible for measured outcomes, we are able to institute reminder systems and redundancies that lead to dramatic improvements in performance. For example, patients who receive medications known as "beta-blockers" following a heart attack are 43 percent less likely to die in the subsequent two years and are rehospitalized for heart ailments 22 percent less often. A goal of providing this therapy to 80 percent of eligible patients has been set in the private sector, and recent medical literature reports rates of use as low as only 21 percent in some settings. In the VA, over 94 percent of heart-attack patients receive this life-saving medication.
Another example of the power of using systems rather than relying on individual adherence to clinical guidelines is in immunization. It is estimated that 50% of elderly Americans and other high-risk individuals have not received the pneumococcal pneumonia vaccine despite its demonstrated ability to minimize death and hospitalization. VA’s emphasis on preventive healthcare has led to achieving pneumonia vaccination rates that exceed standards set for HMOs by almost 20% and nearly double published community rates. Similar accomplishments have been achieved in providing annual influenza vaccinations.
We believe that patient safety can only be achieved by working towards a "culture of safety." Patient safety improvement requires a new mindset that recognizes that real solutions require an understanding of the "hidden" opportunities behind the more obvious errors. Unfortunately, systems’ thinking is not historically rooted in medicine. On the contrary, the field of medicine has typically ascribed errors to individuals and embraced the name-blame-shame-and-train approach to error reduction. Such an approach by its very nature forecloses the opportunity to find systems solutions to problems. Other industries such as aviation have recognized the failings of this approach and over many years have succeeded in transitioning from a similar blame and faultfinding approach to a system-based approach that seeks the root causes of errors. VA realized how pivotal culture is to improving safety and in 1998, conducted a culture survey of a sample of employees. Of interest, the shame of making an error was a more powerful inhibitor of reporting than was fear of punishment. Employees readily forgave mistakes in others but were intolerant of their own. We plan to survey culture broadly in VA for several years to track the progress of our efforts.
VA created a database of adverse events and asked our Medical Inspector to review it. The report has been widely, yet often inaccurately, quoted or critiqued in the media. The database was created to discover common and important adverse events in order to focus our efforts in patient system redesign. Commonly, the media assumed that all the adverse events (and deaths) were due to error. They were not. Neither the report nor the database cataloged which adverse events were preventable with today’s state of knowledge and therefore could be characterized as errors. For example, most of the adverse events were falls, suicides and parasuicidal events (attempted suicides, suicide gestures), or medication errors. It is not possible with today’s knowledge to operate a national system of nursing homes and acute-care hospitals treating the elderly and chronically ill without a number of falls. Yet, we know that it is important to look for common factors to allow us to reduce the frequency of falls in the future. Similarly, psychiatrists have tried unsuccessfully to predict which patients will commit suicide. By looking at our data we hope to be able to predict high-risk patients in the future and therefore be able to prevent suicides. We have already learned that men with a recent diagnosis of cancer, who live alone and who own a gun, are more likely to commit suicide. We plan to study the use of additional interventions in this subgroup of patients at high risk of suicide.
With no successful models in large healthcare systems to guide us, VA turned to other high risk, high performance industries to learn principles for safety. We have borrowed both methods and people from safety-conscious settings such as aviation and space travel and from underutilized disciplines like human factors engineering. These efforts have already produced significant improvements in VA, and we believe will do the same in all healthcare settings.
We would prefer that all of healthcare had begun to address the issue of patient safety long ago. For too long, the emphasis has been on holding individuals accountable and hoping that well-intended and well-educated professionals wouldn’t make human mistakes. As the IOM aptly states in the title of its report: "To err is human." We are pleased to be on the leading edge as healthcare takes a systems approach to patient safety. We are anxious to discover new ways to make VA and all healthcare safer. We appreciate your support of these efforts and intend to keep you fully informed of our progress.