Making the Health Care System Safer
Second Annual Patient Safety Research Conference: Conference Summary
The Agency for Healthcare Research and Quality (AHRQ) hosted Making the Health Care System Safer: AHRQ's Second Patient Safety Research Conference from March 2-4, 2003. Held in Arlington, VA, this conference was designed to do the following:
- Share information on the progress of AHRQ-funded patient safety projects, including accomplishments and early successes.
- Define the challenges and collaborative opportunities within each of the seven patient safety portfolios funded by AHRQ.
- Further the user-driven patient safety research agenda.
- Explore next steps towards translating and implementing research findings into practice.
This report provides a comprehensive summary of the five plenary sessions from the conference: AHRQ's Patient Safety Portfolio, Research-to-Practice Success Stories, Furthering the User-Driven Patient Safety Research Agenda, Challenges in Translating and Implementing Research into Practice, and Where Do We Go from Here?
AHRQ's Patient Safety Portfolio
Research-to-Practice Success Stories
Furthering the User-Driven Patient Safety Research Agenda
Challenges in Translating and Implementing Research Into Practice
Where Do We Go From Here?
C. Andrew Brown, M.D., M.P.H., chairman of the Conference Planning Committee and director of the Division of General Internal Medicine and associate professor of medicine at the University of Mississippi Medical Center, opened the meeting by emphasizing the many activities that have commenced and the many accomplishments that have been achieved over the past year in the field of patient safety. Carolyn Clancy, M.D., director of AHRQ, echoed Dr. Brown's view, noting that recent activities have served to stimulate awareness of the issue of patient safety. Health care systems now understand the magnitude of the problem of patient safety, the seriousness of the risks to its victims, and as a result recognize the need to reduce errors and to improve safety. This awareness represents an important first step in realizing change.
But it is only a first step. In fact, the not-so-good news is that the health care system is probably not much safer today than it was in 1999 when the landmark Institute of Medicine (IOM) study on patient errors was released. The many exciting research findings that identify effective practices for promoting safety have not been incorporated into practice. The industry is not systematically identifying, analyzing, and learning from those errors that are committed. And while the Institute of Medicine (IOM) and other institutions promote the importance of addressing systems' failure to improve safety, recent surveys suggest that neither physicians nor the public understand this message.
Going forward, therefore, the challenge is to translate research findings into practice. To facilitate this effort, findings need to be "packaged" so that they are ready to be used by those on the front lines of medicine. Dr. Clancy urged the researchers in the audience to think of themselves as catalysts for this type of change. While publication may be important, it is not enough on its own. Knowledge must be transferred to those who can change the practice of medicine, including physicians, patients, and hospital administrators. Achieving this type of knowledge transfer requires the formation of partnerships, including collaborative efforts both to educate the community about the importance of systems change and to publicize research findings.
Dr. Clancy emphasized the role of the Federal government in this process. Tommy Thompson, Secretary of the Department of Health and Human Services (HHS) and a strong advocate for patient safety, has made safety one of the most important priorities for HHS. For its part, AHRQ plays a variety of roles in promoting the translation of research into practice, not only by funding projects, but also by publicizing important findings. The fiscal year (FY) 2003 budget allows current projects to continue to be funded, as well as implementation of two new programs: the Patient Safety Improvement Corps (a team of experts to assist States, local governments, and communities) and challenge grants to health care organizations to improve systems by adopting best safety practices. Dr. Clancy also noted that the Federal government has allocated $50 million to assist hospitals (especially small facilities and those in rural areas) with investments in information technology that can be critical to improving patient safety.
AHRQ's Patient Safety Portfolio
The first plenary session included a panel that discussed the current status of each of the seven major areas where AHRQ is promoting patient safety.
An Overview of AHRQ's Activities
Daniel Stryer, M.D., acting director of the Center for Quality Improvement and Patient Safety (CQuIPS) at AHRQ, offered an overview of the agency's activities in the area of patient safety. AHRQ has adopted an "epidemic" model to improve safety. This model tackles the problem in three distinct stages:
- Stage 1: Identify errors, raise awareness, and build capacity.
- Stage 2: Implement proven practices, develop innovative practices, and develop a culture to support action.
- Stage 3: Sustain improvements.
Most of the work thus far has focused on stage 1, with current efforts just beginning to address stage 2. Between FY2001 and FY2003, AHRQ has distributed $165 million to fund approximately 114 grants and contracts with seven different RFAs: Systems-Related Best Practices (SRBP); Reporting System Demonstrations (R-Demo); Centers of Excellence (COE) for patient safety practice; Developing Centers for Patient Safety Research (DCERPS); patient safety research Dissemination and Education (Dis-Ed); effects of Working Conditions (WC) on patient safety; and Clinical Informatics and Patient Safety (CLIPS).
Stage 1 Activities
Many of these grants are related to all three aspects of stage 1—identifying errors, building capacity, and raising awareness.
Identifying and Reporting Errors
To help detect and report errors, AHRQ is participating in the IOM Committee on Patient Safety Data, an effort to standardize data and coding related to patient safety so as to facilitate cross-institutional analysis and learning. AHRQ is also involved in the Medical Error Reporting Integration Project, an interagency initiative designed to integrate agency databases through use of standard data elements.
Within capacity building, AHRQ's activities include the following: developing centers of excellence in patient safety research; distributing training grants; developing a Web-based version of the Mortality & Morbidity Reports; implementing the Patient Safety Improvement Corps; developing international collaborations; and measuring culture.
In addition, AHRQ is bolstering established networks to address patient safety, including the 18-institution HIV Research Network (housed at Johns Hopkins University), the Centers for Education, Research and Therapeutics (CERTs), and the Integrated Delivery Service Research Networks (IDSRNs), a collaboration of nine networks consisting of hundreds of organizations and thousands of physicians who collectively care for over 50 million lives.
AHRQ is raising awareness of patient safety issues through a variety of programs:
- Research dissemination and education grants.
- Small conferences and workshops, such as a recent meeting on the relationship between organizational factors and patient safety.
- User Liaison Program (ULP), which brings research findings to the State and local level.
- Information sheets, including publications aimed at consumers, such as five steps to safer health care and 20 tips to prevent becoming the victim of a medical error.
Stage 2 Activities
AHRQ's efforts are just beginning to move into stage 2. Through the best practices in patient safety systems initiative, the CLIPS RFA, and safe practice reports, AHRQ programs are actively identifying proven best practices for patient safety. The soon-to-be-initiated Patient Safety Challenge Grants will further this effort. But AHRQ faces significant challenges in moving to stage 2 of the epidemic model. According to a study by Blendon and colleagues that was published in the December 12, 2002 issue of the New England Journal of Medicine, only five percent of physicians and six percent of patients believe that medical errors are a serious problem today. Clearly, efforts to raise awareness of the patient safety issue are still needed. Some people "get" the issue of patient safety, but many more do not.
A Detailed Review of AHRQ's Patient Safety Portfolio
Following Dr. Stryer's presentation, steering committee panel members provided a detailed review of the latest activities among grantees in each of the seven RFAs.
Developmental Centers for Evaluation and Research in Patient Safety
Pascale Carayon, Ph.D., professor of industrial engineering and director of the Center for Quality and Productivity Improvement at the University of Wisconsin at Madison, described common themes and challenges within the 18 Developmental Centers for Evaluation and Research in Patient Safety (DCERPS). During the 2002 patient safety meeting, AHRQ's clear message to the DCERPS was to focus intently on building capacity, forming collaborations, and implementation. The stated goal for the DCERPS was to develop new multidisciplinary research teams to improve the Nation's capacity in patient safety research, to expand the patient safety knowledge base, and to assure that new knowledge is incorporated into actual practice.
Over the past year the DCERPS have accomplished much, faced a number of challenges, and uncovered a number of opportunities. Accomplishments include the following:
- Development of cohesive, multidisciplinary, multi-institutional research teams that function at local, regional, national, and international levels.
- Development of a shared vision and new ways of conceptualizing patient safety.
- Reaching out to and collaborating with a wide range of people locally, regionally, nationally, and internationally.
- Developing a variety of outputs, including papers, software, educational products (e.g., courses, training programs), and bibliographies.
- Raising research funding.
Key challenges facing the DCERPS include dealing with Institutional Review Boards (IRBs), especially on cross-organizational initiatives, "reining in" excitement levels, finding the time to balance DCERP activities with other work responsibilities (the people involved in DCERPs have "day jobs" as well), getting buy-in from key constituencies, measuring outcomes, securing research funding, and dealing with an uncertain external environment.
Yet with these challenges come various opportunities, including the potential for collaboration across DCERPs, with other patient safety initiatives, and with other disciplines, care settings, and nations. Opportunities also exist to exchange data collection and education tools, raise additional funding, conduct joint research, and refine or revise "outputs" so as to make them more applicable to potential users.
Dissemination and Education Portfolio
John R. Combes, M.D., senior medical advisor to The Hospital & Health System Association of Pennsylvania (HAP) and the American Hospital Association (AHA), reviewed the activities of grantees within AHRQ's dissemination and education portfolio. He briefly described each of the six projects being conducted in this area.
Neonatal Resuscitation Simulation
This project, underway at Stanford University, tests whether technical and behavioral skills acquired during medical simulation can be transferred to the medical environment. During the pre-intervention phase, team performance is videotaped in the labor and delivery room. The intervention consists of training neonatal resuscitation teams in the medical simulator. Post-intervention measurement will record and reassess team performance in the delivery room.
Patient Safety Curriculum and Tools for Physicians
The American College of Physicians (ACP) and American Society of Internal Medicine (ASIM) have developed a 7-module patient safety curriculum for physicians. Each module covers an important driver of patient safety, using case studies to illustrate key lessons and providing physicians with actionable steps they can take to promote safety. Through active involvement of ACP leadership and the use of ACP infrastructure (e.g., regional chapters), this curriculum is serving to raise awareness of patient safety among physicians to promote positive attitudes, and to facilitate physician behavior change by creating a Web-based "patient safety community." The curriculum is currently being evaluated through questionnaires, with comparisons made between an intervention and control group.
Educating Surgeons in Patient Safety
The American College of Surgeons (ACS) has a program to educate surgeons in patient safety. It consists of Webcasts of sessions on patient safety; an education task force on systems-based practice oriented at medical students, residents, and surgeons; simulation training; and objective, structured clinical examinations and exercises to enhance patient safety.
Patient Safety Through Web-based Education
The National Patient Safety Foundation (NPSF) and the Medical College of Wisconsin (MCW) have developed a Web-based patient safety education program geared at physicians, nurses, and patients. The goal of the effort is to develop a standard method for patient safety education that will reach large audiences. Patient safety resources, including a Web-based patient safety education center, will also be available. NPSF and MCW are still dealing with several challenges, including how to make the information suitable to a lay audience that will be accessing it through the Web, and how to ensure adequate bandwidth to allow quick access to the material.
Error Reduction in Hypertension Treatment
The Harvard Community Health Plan has developed educational interventions to reduce prescribing errors in treating hypertension. Randomized controlled trials in three managed care settings will test the effectiveness of various interventions, including mailed dissemination of educational materials (the control group), mailed dissemination plus academic detailing, and mailed dissemination plus individual academic detailing. The study will compare baseline (pre-intervention) error rates to post-intervention error rates. The HMO Research Network's CERT helped in identifying nine practice sites that are being randomized into the various groups.
Simulation Training and Safety Education
The AHA and Healthcare Research and Education Trust (HRET), in collaboration with the University of Colorado and Geisinger Health System, are developing a patient safety curriculum focusing on simulation training in the cardiac catheterization laboratory. Using standardized performance metrics and six case scenarios, the sponsors will evaluate the effectiveness of simulation training through pre- and post-training tests.
Common Themes and Lessons Learned
Dr. Combes highlighted several common themes from these projects, including a focus on developing and evaluating the effectiveness of both patient safety curricula for specific audiences and various dissemination strategies (including the Internet) for these curricula. In addition, the sponsors of these projects have learned a variety of lessons, including that technology can be difficult (e.g., there were problems in translating case scenarios into simulation and in ensuring adequate bandwidth for Internet-based applications) and that "spin-off" products can provide important opportunities for additional learning and for applying learning to larger groups, including patients. Finally, sponsors have been surprised at how well physicians and patients have accepted these interventions.
Working Conditions Portfolio
Nancy Donaldson, R.N., D.N.Sc., F.A.A.N., serves as clinical professor and founding director at the University of California at San Francisco (UCSF) Stanford Center for Research & Innovation in Patient Care and associate dean for practice at the UCSF School of Nursing. She highlighted the work of the 22 grantees within the working conditions portfolio. She began by noting that although research findings are inconsistent, the preponderance of the evidence links staffing—especially nursing staff levels—with adverse events.
Among the 22 studies being conducted, 14 are evaluating acute care, including six looking at medical-surgical care and three at critical care. Nineteen are making use of primary data. Nine are evaluating care at the unit level. Fourteen are evaluating links between registered nursing staff and patient safety. Among the various outcomes evaluated, five studies are looking at falls, four at pressure ulcers, two at significant clinical events, and one each at the use of restraints and staff workload. To facilitate the ability to generalize study results, three workgroups are currently attempting to achieve synergy and data standardization across these studies.
The grantees have experienced a number of challenges in working with the hospitals, including the following:
- Negotiating entry, as hospitals fear the risk of disclosure and discovery, and have concerns about complying with the privacy and confidentiality requirements of HIPAA (Health Insurance Portability and Accountability Act).
- Meeting Federal Wide Assurance (FWA) and IRB certification requirements. Regulations (and the interpretation of these regulations) vary across IRBs.
- Sustaining site commitment given the dynamic and volatile environment.
- Sustaining target sample size so as to maintain analytic power.
- Capturing data and ensuring data integrity; tactics and logistics vary across diverse systems (as do the costs of capturing data).
- Ensuring consistent, continuous communication across sites.
Future directions for the grantees include efforts to do the following:
- Standardize and institutionalize a set of core measures.
- Study the effects of data source variation on findings across studies.
- Examine the impact of technologies such as personal digital assistants on data entry errors and efficiency.
- Exploit use of the Internet for data transmission and uploading.
- Evaluate the impact of Web data access on strategic quality, safety, and staffing efforts.
- Leverage synergies between research and performance improvement and accreditation imperatives.
Clinical Informatics Portfolio
George Hripcsak, M.D., M.S., serves as associate professor of clinical informatics at Columbia University and vice chair of the Department of Medical Informatics and associate director for medical informatics at New York-Presbyterian Hospital. Several key issues affect the field of clinical informatics, including the following:
- Collection, storage, and structure of data, including access to the electronic medical record or EMR. Good data are needed to make good decisions.
- Acquisition, representation, and application of knowledge.
- Understanding and influencing cognition (how people think and learn).
- Process, including study, design, and deployment of improved workflow systems in institutions.
- Interoperability, or the sharing of data and knowledge across the system.
- Communication among health care providers, via wireless and/or mobile computing technologies.
- Evaluation, or how to demonstrate benefit.
Dr. Hripcsak shared examples of progress being made by grantees on these key issues:
- Speech recognition and natural language processing or NLP are helping to improve data capture. At one hospital, radiology reports are now ready within an hour.
- Ontologies, reasoning, and workflow tools are enhancing knowledge, in part through progress in how knowledge is represented in computer form.
- A better understanding of thought processes and learning is being developed. For example, researchers now understand that different interventions are likely needed for students and attending physicians, since "experts" reason differently than do novices.
- More institutions are installing computerized physician order entry (CPOE) and electronic medical record (EMR) systems, although progress is not as rapid as many would like to see.
- Huge efforts to standardize data, vocabulary, and messaging are improving interoperability and making the sharing of data across institutions a reality.
- Rapidly improving technologies such as the Palm Pilot™ are improving communications, primarily because physicians are willing to use them if they do not disrupt work flow.
- While some evaluation studies are underway, there is general recognition of a need for more studies to evaluate the impact of interventions.
Dr. Hripcsak shared the challenges in using clinical informatics to improve patient safety:
- There is temptation to try any new technology, rather than to analyze systematically what is truly needed. As a result, systems often do not meet user needs and are quickly discarded.
- The benefits of investing in new clinical informatics must be proven, especially to those paying for these investments.
- Clinically rich, coded data are not yet available in real time. Retrospective data does provide real-time decision support for physicians.
- Interventions need to fit within the everyday workflow of those using them, including physicians, nurses, and other caregivers.
- Standards are not yet adequate to allow timely, complete data sharing across institutions.
- Limitations exist on what can be done with data due to privacy and discoverability concerns.
Reporting Demonstrations Portfolio
Nancy Ridley, M.S., assistant commissioner of the Bureau of Health Quality Management at the Massachusetts Department of Public Health, reviewed the progress of the 16 cooperative agreements in the area of reporting of medical errors and adverse events. She began by noting the tremendous variability in these research studies, many of which extend far beyond reporting systems. The studies include both voluntary and mandatory reporting systems, those internal to an organization and those reported to external parties, testing of reporting systems, and evaluations of best practices—i.e., what can be gained from the information being reported. Ms. Ridley reviewed the common issues and challenges facing the grantees.
- Mandatory versus voluntary reporting: This key issue was raised in the IOM report and remains a major issue for many grantees. Hospitals may be reluctant to report errors because of concerns about malpractice liability, confidentiality, and the impact on their brand and reputation (if reports are made public).
- Type of information collected: Ms. Ridley believes there is still a "silo" approach to reporting. Standard nomenclature and taxonomy has not yet been developed. Some studies involve the collection of information only on serious errors or sentinel events, while others also collect information on adverse events and/or near misses.
- Destination of data: Data sometimes are used just for internal purposes, while in other cases the data go to a State agency or other external organization. Concerns about trust can arise.
- Confidentiality: Several studies are addressing the issue of confidentiality and the risks that arise from reporting, including issues of disclosure to patients. The key question is whether adequate information can be provided to individual patients and to the public (in aggregate form) while maintaining confidentiality.
- Role of the media: The media play an important role in determining how information on errors gets reported to the public.
Looking ahead, Ms. Ridley sees a major opportunity for technology to accelerate and simplify reporting, and to improve the ability to use the information collected in a meaningful way. But cultural issues and barriers within organizations remain. She called on institutional leadership and staff to push for the establishment of a "just culture" which combines a systems approach to error reporting and patient safety improvements with individual accountability.
Systems-Related Best Practice and Centers-of-Excellence Portfolios
Eric J. Thomas, M.D., M.P.H., associate professor at the University of Texas Houston Medical School, reviewed the grants within two different RFA areas.
Systems-Related Best Practice Grants
The systems-related best practices grants, which were awarded a year before grants in the other RFAs, include six different tools in various stages of development:
- A chest tube insertion training video based upon video recordings of real insertions, a powerful tool developed at the University of Maryland.
- A safety culture survey for institutions, developed at Stanford University.
- A typology of primary care errors and injuries that was derived from patient interviews, developed at Virginia Commonwealth University.
- An investigation of how the EMR and electronic prescribing can measure and reduce adverse drug events in the outpatient setting, conducted at Brigham and Women's Hospital.
- The time-insensitive predictive instrument information system or TIPI-IS, a system to reduce errors in the diagnosis of acute cardiac ischemia in the emergency department, developed at the New England Medical Center.
- An investigation of how teams can reduce medical errors, based on videotapes of the resuscitation of preterm infants, conducted at the University of Texas at Houston.
Patient safety improvement tools being developed within the three centers of excellence include:
- At Brigham and Women's Hospital: Development of a Web-based reporting system and of surveys to measure attitudes, practices, and behaviors related to patient safety; evaluation of the epidemiology and prevention of medication errors in outpatient pediatric cases and in psychiatric patients; and evaluation of the epidemiology and prevention of intravenous pump errors through internal pump decision support.
- At the University of Pennsylvania: Analysis of the risk factors and costs of errors that lead to hospitalization in elderly patients, with a focus on three drugs (digoxin, phenytoin, and warfarin); determination of the human and system factors (patient- and medical practice-specific) that lead to errors in warfarin adherence; identification of system-based redisposing factors for inappropriate aminoglycoside antibiotic monitoring and dosing that leads to acute renal failure; and description of the effects of workplace stressors and house officers' level of strain on the frequency and types of medication errors.
- At the University of Texas: Analysis of ways to improve the design of infusion pumps and hospital purchasing processes; development of three surveys to measure provider attitudes about incident reporting, professional culture, teamwork, safety culture, working conditions, stress recognition, and perceptions of management; methods to measure and improve teamwork among health care providers; development of a close-call reporting system; and identification of how organizational characteristics in hospitals can facilitate or hinder the ability to learn from errors and adverse events.
Dr. Thomas highlighted the following common themes across these two portfolios:
- A tension exists between the lack of basic knowledge about errors (e.g., why they occur, how to measure) and the pressure to reduce errors. As a result, interventions to address both are being conducted simultaneously—a risky but potentially powerful approach.
- Research into patient safety requires a multidisciplinary approach involving both medical and nonmedical disciplines. While this reality generates tension, it also creates an enormous opportunity to learn from colleagues both within and outside of health care.
- Research must go beyond traditional data sources and methodologies to develop new sources of information and new ways of thinking about patient safety.
- Resources must be invested to train new investigators, particularly within the emerging area of centers of excellence.
- Researchers must take advantage of opportunities that arise during a project. Hospital calls for assistance or advice should, when possible, be answered, as they represent a real opportunity for researchers to improve patient safety.