Accelerating Change Today A.C.T FOR AMERICA’S HEALTH

FEBRUARY 2000 Reducing Medical and Improving Safety

Success Stories from the Front Lines of

THE NATIONAL COALITION ON THE INSTITUTE FOR HEALTHCARE IMPROVEMENT Accelerating Change Today (A.C.T.) FOR AMERICA’S HEALTH

A.C.T. is a collaborative initiative of the National Coalition on Health Care and the Institute for Healthcare Improvement. It aims to improve the quality of health care in the through the identification of “best practices” and administrative and clinical innovations that are: (1) yielding better out- comes; (2) making the delivery of care more efficient; (3) increas- ing access to timely medical care; (4) making the health system easier to use; (5) lowering costs, and (6) reducing medical errors and inappropriate care. The initiative seeks to accelerate the spread of best practices and innovations throughout the health system by publishing them and through presentations at medical meetings and health care and business symposia. A central pur- pose is to make a broad range of health care stakeholders, including consumers and those who pay the health care bill, About This Publication more aware of cutting-edge efforts to improve the quality of health care. The initiative will actively encourage the replication

This report presents the stories of of best practices in health care facilities. institutions and organizations that made a commitment to change and innovation in the field of and medical reduction. THE NATIONAL COALITION ON HEALTH CARE The profiles herein reflect some of NCHC is the nation’s most broadly representative alliance working to improve America’s health and health care. It is comprised of 90 member organizations. They include some the most pioneering and innovative of the nation’s largest businesses, labor unions, health care providers, consumers groups efforts underway in this field. The and religious organizations. The Coalition was founded in 1990. It is non-profit and health facilities and organizations non-partisan. Its members are united in the belief that America needs better, more affordable health care and that all Americans should have health insurance. Former discussed in this report were identi- Presidents George Bush, Jimmy Carter and Gerald R. Ford serve as the Coalition’s fied by a team of six experts (see Honorary Co-Chairs. Former Iowa Governor Robert D. Ray and former Congressman credits and acknowledgements). Paul G. Rogers of Florida are the Coalition’s Co-Chairmen. NCHC is in Washington, DC. Founder and President Henry E. Simmons, M.D., M.P.H., F.A.C.P., is a widely respected Their choice is in no way meant to pioneer in the field of health quality assessment and improvement. imply that other institutions and health care facilities are not under- THE INSTITUTE FOR HEALTHCARE IMPROVEMENT taking meaningful and laudable IHI is an independent, non-profit education and research organization based in Boston, efforts to reduce errors and MA. It was founded in 1991 with the goal of fostering collaboration among health care organizations to improve the quality of health care. IHI each year holds a wide array of enhance patient safety. educational forums, symposia and workshops, and demonstration projects for medical professionals and health care administrators. IHI’s co-founder and President, Donald M. Berwick, M.D., M.P.P., a practicing pediatrician and clinical professor at Harvard Medical School, is one of the nation’s leading authorities on health care quality. Accelerating Change Today A.C.T FOR AMERICA’S HEALTH

FEBRUARY 2000

Reducing Medical Errors and Improving Patient Safety

Profiles of institutions and organizations that made a commitment to change… and a difference

THE NATIONAL COALITION ON HEALTH CARE THE INSTITUTE FOR HEALTHCARE IMPROVEMENT Credits and Acknowledgements

LUCIAN L. LEAPE, M.D., (Team Leader) is an Adjunct Professor of Health Policy at the Harvard School of Public Health and a member of the Health Sciences Division at RAND. At RAND, Dr. Leape has directed several studies of overuse and underuse of cardiovascular procedures and the development of medical practice guidelines. He is a member of the Board of Directors of the National Patient Safety Foundation and the Committee on Quality of Care in America at the Institute of Medicine. In the past, he has served on the Health Services Research Review Committee of the Agency for Healthcare Research and Quality. Dr. Leape’s research has focused on medical error prevention, quality of care, unnecessary , and the development of practice guidelines. He recently led the Institute for Healthcare Improvement’s initiative in prevention of adverse drug events.

DAVID BATES, M.D., is Chief of the Division of General Medicine at Brigham and Women’s Hospital, Medical Director of Clinical and Quality Analysis for Partner’s Healthcare Systems and an associate professor of medicine at Harvard Medical School. Trained as a clinical epidemiologist, Dr. Bates’ main interest and area of research is the use of computer systems to improve patient care. He has also done extensive work on evaluating the incidence and prevention of adverse drug events.

DONALD M. BERWICK, M.D., M.P.P., is President and CEO of the Institute for Healthcare Improvement in Boston, MA. He is a practicing pediatrician, a clinical professor of pediatrics and health care policy at Harvard Medical School, and an associate professor of health policy and management at the Harvard School of Public Health. Dr. Berwick currently serves as Chairman of the National Advisory Council at the Agency for Healthcare Research and Quality. He was one of 32 appointed members in 1997-98 to the President’s Advisory Commission on Consumer Protection and Quality in the Health Care Industry. He is a member of the Committee EDITOR on Quality of Care in America at the Institute of Medicine. Dr. Berwick is a co-author Steven Findlay of Curing Health Care: New Strategies for Quality Improvement (Jossey-Bass, 1990). ASSISTANT EDITOR MARILYN SUE BOGNER, PH.D., is President and Chief Scientist at the Alyssa Keefe Institute for the Study of Medical Error in Bethesda, MD. She is Chair of the International Workgroup on Medical Error and an advisor to the National Patient WRITERS Safety Foundation of the American Medical Association. She has served as a con- Caralee Adams, Lisa Krieger, sultant to numerous groups and worked with organizations including the Institute Lani Luciano, Christine Paul, of Medicine of the National Academy of Sciences and the Food and Drug Kris Rebillot, Shelley Reese, Administration. She edited and contributed to the book Human Error in Medicine Jane Roessner, Jan Ziegler (Lawrence Erlbaum Associates, Inc., 1994).

STEVEN MEISEL, PHARM.D., is Assistant Director of for Patient DESIGN Care at Fairview Southdale Hospital in Edina, MN. He is Co-Chair of Fairview’s Pensaré Design Group, LTD Integrated Formulary & Drug Use Committee and serves as the Chair of the Institute for Healthcare Improvement’s Breakthrough Series on Adverse Drug Events and Medical Errors. Prior to his affiliation with Fairview, Dr. Meisel was Clinical This report was funded by a grant Pharmacy Coordinator at St. Joseph’s Hospital in St. Paul, MN, and Chief of from The W.K. Kellogg Foundation Pharmacy, Keams Canyon Indian Hospital in Keams Canyon, AZ. Dr. Meisel is a of Battle Creek, Michigan. member of the American Society of Health System , and is a contribut- ing editor for the Institute for Safe Practices’ Medication Safety Alert. 2000 by the National Coalition ELLISON C. PIERCE, JR., M.D., is Executive Director of the Anesthesia © on Health Care and the Institute Patient Safety Foundation in Boston, MA. Dr. Pierce founded the group in 1985 and served as its president until 1996. He is an executive committee member of the for Healthcare Improvement National Patient Safety Foundation. From 1968 to 1996 he was Chairman of the Department of Anesthesia at New England Deaconess Hospital in Boston. Prior to that, from 1960 to 1968, he was the Vice Chairman of the Department of Anesthesia at the Brigham Hospital in Boston. He has been an adjunct professor of anaesthesi- ology at Harvard Medical School from 1960 to the present. 1 Table of Contents

Can We Make Health Care Safe? By Lucian L. Leape, M.D. 2

Back from the Brink: Making Chemotherapy Safer The Dana-Farber Cancer Institute 4

A Government Health System Leads the Way The VA Health System 9

Making Doctors Computer Literate Brigham and Women’s Hospital 12

What’s in a Name? When it Comes to Drugs—Lots Bristol-Myers Squibb 16

Tackling Medication Errors Head On Luther-Midelfort Hospital 18

Making Mistakes Harder to Make Fairview Health Services 21

A Dummy Helps Pave the Way to Better Medicine VA/Stanford Simulation Center 24

A Medical Specialty Blazes a Trail was first in addressing medical errors 26

Resources and Suggested Reading 29 2

challenge. The urge to punish is deeply Can we make health care entrenched. And it is reinforced by the sin- cere belief of many doctors and nurses that errors result from individual carelessness. If there is any lesson to be learned from safe industries, it is that fear, reprisal, and SAFE?SAFE? punishment produce not safety but defen- BY LUCIAN L. LEAPE, M.D. siveness, secrecy and personal anguish. Third, health care professionals and MEDICAL ERRORS—the very term core, this shift means redefining how to staff are anxious to deliver safe care. sounds ominous. And well it should. When accomplish one of the basic tenets of med- When shown effective methods for you enter a hospital or a nursing home, or ical practice: “First, do no harm.” changing systems to make them safer, go to a doctor’s office, you don’t always The vital first task is to recognize that, doctors, nurses, and pharmacists eagerly expect to be cured, but you certainly don’t as in other complex fields of science, and energetically embrace them. expect to be hurt! Yet, when the presti- technology, engineering, and human en- Fourth, the problem of medical errors gious Institute of Medicine (IOM) in deavor, errors in medicine are caused by is not due primarily to a lack of knowl- November 1999 released the landmark failures in the systems and organizations edge. Rather, the chief culprit is in- report, To Err Is Human: Building a Safer that human beings build. adequate dissemination and implemen- Health System, many Americans were The evidence is overwhelming. Medical tation of ideas and practices we know shocked. The report contained this sober- errors most often result from a complex work. Recently, for example, the National ing, if not new, statistic: each year more interplay of multiple factors. Only rarely Patient Safety Partnership (NPSP) released than one million people in the U.S. suffer are they due to the carelessness or mis- a list of 16 proven and accepted best prac- from preventable medical and conduct of single individuals. Yet, in the tices in medication safety. Many have 100,000 die from them. As the media wide- past, rather than addressing those under- been known for years. Yet, most are not in ly reported at the time, that’s more than lying system design faults, error preven- place in the vast majority of hospitals, and died in 1998 from breast cancer, AIDS and tion strategies have relied almost exclu- no hospital in the country uses them all. motor vehicle accidents combined. sively on enhancing the carefulness of Fifth, safety pays. A single, preventable The IOM panel, on which I served, caregivers, heavily reinforced by fear of adverse drug event in a teaching hospital called on health care organizations, punishment for failure. But punishment has been estimated to cost over $4,000. A doctors’ groups, regulators, government drives reporting of errors underground, serious can cost hundreds of thou- agencies and Congress to make patient preventing the very systems examination sands of dollars. Reducing injury rates by safety a priority. Within weeks, Congress that is needed to discover and correct the even 10 to 15 percent can yield large-scale launched hearings, President Clinton underlying causes. As the stories that fol- savings. Most systems-level changes for held a Rose Garden ceremony to appoint low show, when punishment is removed, safety require few extra resources. And a task force, and a host of other groups reporting of errors sharply increases— even those that do require a significant launched initiatives. Health insurer Aetna often by 10 or even 20-fold. up-front investment produce a full pay- said it would spend an extra $1 million Discontented, even horrified, by the back rapidly. For example, the most this year on patient safety research. The status quo, a small band of health care expensive system to allow doctors to write American Hospital Association said it leaders and organizations began to coa- their orders at computer terminals may would join forces with the Institute for lesce in the mid-90s around the mission require an initial investment of $2 million Safe Medication Practices to provide hos- of dramatically reducing medical errors. to $3 million. But the proven ability of pitals with proven strategies to reduce Their persistence and willingness to try such a system to reduce medication errors errors. The Joint Commission on Accredi- new solutions has borne fruit, created by 50 to 80 percent within a year or two tation of Healthcare Organizations (JCAHO) models that have reduced some types of means most hospitals recoup the invest- announced it will begin making random medical errors by 80 to 90 percent in just ment in a few years. inspections at hospitals to probe errors. a couple of years, and yielded many Finally, there are significant external And a business group led by, among oth- important lessons. barriers to improving patient safety. ers, General Motors, said it would pres- The first is that leadership is essential. Hospital efforts to restructure their sure hospitals and doctors it contracts In its absence, efforts will be fragmentary, approaches to error prevention may be with to reduce errors. uncoordinated, and have only minor overruled by state boards of nursing, This new momentum is profoundly effects. The leaders of the institutions pro- pharmacy or medicine, departments of welcome. After all, numerous past warn- filed in this publication understand that. health, or the courts. Reason: these insti- ings, based on research results, went rela- They made highly visible commitments to tutions are still locked into a blame and tively unheeded. A critical mass appears make their institutions safer. punish approach to errors and a focus on now to have been reached. Indeed, it may Second, change is difficult. Even with individual culpability. In the name of not be hyperbole to say that a deep cultur- the most enlightened leadership, creating accountability, a nurse loses her license al shift on this issue is underway. At its a non-punitive atmosphere is a major because of a medication error or a doctor 3 THE INSTITUTE OF MEDICINE’S KEY RECOMMENDATIONS is sued for an error-induced injury. The fear of malpractice litigation thus • Congress should create a federal Center for Patient Safety that sets national goals becomes a major barrier to openly dis- for medical error reduction and tracks progress in meeting those goals. cussing or reporting errors. • Congress should authorize the creation of a nationwide mandatory medical error WHAT WE NEED TO DO reporting system to collect data on errors that result in death or serious harm to patients. • Federal and state laws should encourage the development of voluntary medical Reducing the risk of error in health care error reporting systems in all health care facilities, to include reporting of all events will require a substantial and sustained that could have harmed patients. effort at all levels of the health care sys- tem. It must become a priority goal wher- • Data and information on medical errors should be legally protected when used by ever care is given—the doctor’s office, the professional peer review organizations to improve health care quality. hospital, and the nursing home. That goal • Medical professional societies and health care licensing bodies should focus greater must be supported by the commitment of attention on patient safety. both human and financial resources. The chief executive officer—or other leader— • The Food and Drug Administration should increase attention to the safe use of drugs. must articulate the vision of safe care that • Health care facilities and organizations should make continually improved patient he or she calls upon others to work toward. safety a declared and serious aim and implement proven medication safety practices. Leadership is also needed to redefine accountability. Some have been concerned that not punishing people for making errors relieves them of responsibility. ies should be held accountable to the pub- establish a Center for Patient Safety. One Actually, the converse is true. In a safety lic and the professions to set and enforce of the key functions of such a Center oriented organization, every employee safety standards. The plain fact is that would be to convene groups to set such feels and accepts a personal responsibility many hospitals do not initiate error reduc- standards. to identify unsafe conditions, report them, tion programs because no one demands it. Finally, regulators and accrediting and work toward their elimination. Inertia prevails. If, for example, the JCAHO bodies need to become much more posi- Health care organizations must hold or the Health Care Financing Administra- tive forces for safety in health care, both workers accountable to perform their tion (which runs Medicare) were to require in terms of preventing errors and in duties carefully, conscientiously, compe- full implementation of the NPSP safe med- the way they react to them. When mis- tently, and safely. But they will still ication practices for accreditation or pay- takes occur, they need to shift their focus make errors. Who is responsible? The ment, I’m quite sure nearly 100% compli- from individuals to organizations, requir- organizations and systems. Who is ance would occur within a short period ing both a thorough investigation of the responsible for the systems? Managers. of time. Likewise, if the Food and Drug underlying systems failures and a plan Consider a nurse who makes a serious Administration required pharmaceutical that addresses those failures. medication dose error. If one of the fac- companies to label in a stan- The publication you hold in your tors leading her to make the error is that dardized, readable manner, in bar-coded hands presents case studies of hospitals, she is working a second shift, or has a packages, and if they only approved new health facilities and organizations that doubled patient load, or is inadequately drug names that would not be confused have made the commitment to systemati- trained for her responsibilities, whose with other , medication errors cally reduce the risk of patient harm in responsibility is that? would decline substantially. health care. Their stories show the dra- Of course, if a doctor or nurse has Safe industries don’t shy away from matic results that can be achieved, often injured a patient through an error caused standards that reduce the risk of injury. in a short period of time. The profiles also by egregious misconduct, neglect, or They embrace them, recognizing the only shed light on the obstacles that get in criminal activity, he or she must be pun- way to achieve safe practice is to clearly the way of reducing errors, the mecha- ished. But if such a person had a prior spell out what it is and require it. We have nisms by which change occurs most effi- history of reckless behavior and disregard strict rules governing working hours, pro- ciently in organizations, and the very of safe practices, why has he or she been cedures, equipment, and maintenance for human process of shifting the prevailing permitted to continue working? our airlines and railroads. But we have paradigm on this issue. The stories will Simply put, management must “man- few such rules in health care. The need is enlighten you. But more than that, I hope age” for patient safety just as they man- obvious, and opportunities abound: stan- they will motivate you to act—to push age for efficiency and profit maximiza- dards for maximum working hours, work your institutions to get involved in the tion. And safety must become part of loads, staffing ratios, expertise (pharma- struggle to make sure that health care, what a hospital or health care organiza- cy, intensivists, ER specialists, etc.), safe first and foremost, does no harm. tion prides itself on. medication practices, and many others. But accountability needs also to work The IOM panel called for the Agency The author is an adjunct professor of health in reverse. Regulators and accrediting bod- for Healthcare Research and Quality to policy at the Harvard School of Public Health. 4 BACKBACK FROMFROM

THETHEMaking BRINKBRINK Chemotherapy Safer

BY CHRISTINE PAUL

Stunned by Every medical facility’s worst nightmare changed forever. Reeling from the bad became a reality for Boston’s prestigious publicity and glare of the spotlight, Dana- the death of a Dana-Farber Cancer Institute in the early Farber undertook a painstaking internal winter months of 1995. Two patients examination of what went wrong. State being treated in an experimental drug authorities also probed the case. The Joint patient because study for advanced breast cancer, the Commission on Accreditation of Health- Institute belatedly discovered, had care Organizations (JCAHO) placed the of an error, received massive overdoses of chemother- Institute on temporary probation. Five apy in November 1994. On December 3, years later the investigations and lawsuits one patient had died as a direct result of are only beginning to wind down. a prestigious the error. The other suffered permanent But Dana-Farber didn’t stop at just heart damage. Complicating a tragic situ- uncovering what went wrong. Seeking to cancer center ation for all, the patient who died—Betsy wrest from the tragedy something of last- Lehman, 39—had been a medical reporter ing value not only for Dana-Farber but for the Boston Globe. The paper trumpet- for health care facilities and patients commits to a bold ed the death when it learned that a med- everywhere, the Institute’s leadership ical error was the cause. And it dug into boldly committed themselves to making every crevice to ferret out what had hap- Dana-Farber a model of patient safety new approach to pened and affix blame. The case quickly and error prevention. attracted national media attention and Today, many experts in the field think error prevention was followed closely for months. Indeed, they have succeeded. Dana-Farber has the Dana-Farber incident became the pioneered new systems to prevent mis- highest profile medical mistake in the U.S. takes in the delivery of chemotherapy in decades, perhaps ever. agents—systems that are being emulated Stephen Sallan, M.D., a senior oncolo- by others. And in making error preven- gist at Dana-Farber when the overdoses tion a top priority throughout, Dana- occurred and now chief of staff, recalls Farber has raised the profile—in a positive that difficult time: “It shocked and sad- way—of the very issue that brought the dened us in very profound ways, both facility to its knees just a few short years professional and personal. For months, a ago. “We wanted to find and fix the root Dana-Farber Cancer Institute pall hung over the Institute. Everyone causes of our mistakes. But we also want- was deeply affected, because after all, we ed to make our systems the best in the Boston, Massachusetts are caring and conscientious people in the world at preventing errors while provid- business of saving lives.” ing the latest in cancer care,” says James As in the wake of all tragedies, life Conway, M.B.A., Dana-Farber’s chief goes on for the survivors—but it is also operations officer. 5

A rechecks the drug order with a technician as part of Dana-Farber’s to create an institutional culture that drug order entry system. (Photo by Steven Gilbert, StudioFlex Productions) encouraged the reporting of errors, no matter how small. When you walk into Dana-Farber’s clinical areas today, you’re very likely to see something you would not have seen just a few short years ago: a at a computer terminal. In many hospi- tals and clinics, computers are, of course, indispensable for administrative purpos- es. But doctors and nurses still order treatments and make care decisions based on “hard copy”—written patient charts and other paper-based forms. Dana-Farber is changing all that. “We basically asked which pieces of our system could be taken out of human hands and given to a computer with built- in safeguards,” says Lawrence Shulman, M.D. Shulman oversees Dana-Farber’s clinical service for adult patients, for which drug ordering is now completely computerized. (The hospital’s pediatric service is reviewing a transition to a sim- ilar system.) “As for the components that remain in human hands, we designed a catch such an error. Nurses involved with comprehensive system of double and High-tech the patients’ care were able to check the triple cross-checks,” he added. order against the protocol to ensure that Gone are slips of paper, handwritten safeguards the right drug, dosage, frequency, route scribbles and lined paper that can easily and time were being followed per the hide a critical decimal point. Instead, for Like many hospitals that have traced the doctor’s written order. In this case, how- adult cancer patients the doctors must fill “root cause” of a fatal medical error, ever, a protocol written and approved by out an electronic form. After typing in Dana-Farber’s doctors, nurses and admin- research scientists, rather than the clini- the patient’s name, weight, and height istrators discovered that behind the cians administering it, seemed to confirm (critical to proper dosing), the physician human error that caused the death were a that the high dosage level was appropri- indicates the intended drug, dose, and host of “systems errors.” The saga began ate. “The fact that and scien- number of days it is to be dispensed. The when Betsy Lehman and Maureen tists think of dose in terms of course of computer compares this data with a pre- Bateman, 52, were admitted on November , and nurses and administrators programmed set of drug-specific guide- 14 and 16, respectively, for treatment of think in terms of daily dose, played a sig- lines with established high-dose limits recurrent breast cancer. The physician in nificant role in the overdose,” says and route of administration. Dana- charge of their care was fully aware of the Conway. At the time, there was no com- Farber’s staff developed these guidelines protocol of the clinical study under which puter system in place to monitor the by searching the scientific literature and they were to be treated. It was designed to amount of the drug ordered—and signal a drawing on their own expertise. The pri- test whether cimetidine, a common anti- possible error. As a result, both women mary focus is on the “five rights”: give ulcer drug, could boost the tumor-killing received four times the daily dose. It was the right drug to the right patient at the effects of cyclophosphamide, a powerful a toxic blow to their hearts. Lehman died right dose through the right route at the anticancer drug. The drug therapy was to three weeks later. Bateman survived, but right time. If the doctor makes a mistake— be administered over four days, followed with a damaged heart, and died of cancer for example, typing in 300 milligrams by a bone marrow transplant to restore several years later. instead of 30 milligrams—the computer the patients’ blood-forming and immune Dana-Farber’s first corrective action signals a possible error. The only way to cells, which unavoidably are destroyed by was to put in place a computerized system override a dosage that exceeds the pre- high doses of chemotherapy. for drug and chemotherapy orders. It also established limit is for a senior physician Unfortunately, the doctor wrote the launched a series of backup systems and to provide a reason for the deviation. drug’s total four-day dose to be given set in motion a many-pronged effort to Such changes are then double-checked by daily for four days. The backup system in strengthen its “team approach” to treat- nursing and pharmacy staff. The system place at the time was not designed to ment and care delivery. Finally, it sought handles more than anticancer drugs. It 6 A WAKE-UP CALL

Dana-Farber’s staff and patients as well as academic experts in medical-error prevention Double and are often asked to share their insights at local, regional and national conferences. Below are key messages directed to health care stakeholders. triple checks

P A TIENTS When it’s time to dispense medicines, “The best thing you can do is let your health care providers know when something isn’t the checking continues. Before ordering a going well and how they might help. Once it’s on their radar screen, they can start deal- drug from the pharmacy, a nurse reenters ing with it.” the patient’s weight and height into the Judie Beard, cancer survivor, member of Patient-Family Advisory Council computer. If they differ by more than 5 percent, the computer returns the order PHYSICIANS to the senior physician for reverification “We need to acknowledge that we’re all human beings and that error is always a possi- and signature. The discrepancy may be bility. We need to work together to create an environment and culture that minimizes the caused by a measuring or reporting error, likelihood of error.” or it may be due to a change in the Lawrence Shulman, M.D., Vice Chair of Clinical Services, Adult Oncology patient’s weight, which is not uncommon in cancer therapy. NURSES At the pharmacy, there’s another “Listen to the users of the system—the nurses, doctors, pharmacists and patients—and computerized review—for drug-drug, encourage them to identify ways to simplify and standardize the system.” drug-food, or drug- reactions. The Susan Grant, R.N., Chief of Nursing and Patient Care Services drug order is checked against the patient’s medical history, including PHARMACISTS , current lab results, verification “Don’t just focus on the individual person or individual departments. Instead, work togeth- of registration in a clinical study, and er to look at the whole patient-care system.” the protocol itself—anything that might Sylvia Bartel, R.Ph., Director of Pharmacy signal possible error. Once okayed and prepared by the RISK MANAGERS pharmacy, the drug is then delivered to “Promote a non-punitive environment for reporting error. It encourages the reporters to the nursing station. Two nurses check the identify system problems and assists in preventing errors.” drug’s label, as well as the patient’s wrist- Maureen Connor, R.N., Director of Risk Management band, to ensure it’s going to the right patient. For experimental , EXECUTIVE OFFICERS there’s a triple-check. The patient is then “The effort to prevent medical errors can only be successful if executive leaders make closely monitored to detect and respond patient safety one of their priorities.” to any adverse reactions such as weak- James Conway, M.B.A., Chief Operations Officer ness and nausea—which often become even more pronounced with higher doses TRUSTEES of chemotherapy. Betsy Lehman, for “The function of Dana-Farber’s board of trustees has changed in fundamental ways. They example, had voiced complaints. But such now see error prevention and quality control as central to their role.” side effects were not unexpected. And in Richard Bohmer, M.B.Ch.B., M.P.H., Senior Lecturer, Harvard Business School the absence of the comprehensive check- ing system that is now in place, the physician’s error in writing the order went undetected. In fact, her overdose also automatically signals for auxiliary tutions—Dana-Farber, Brigham and wasn’t discovered until three months later treatments, such as anti-nausea drugs, to Women’s, and Massachusetts General when an assistant data manager was be included in the patient’s treatment. Hospital—provide collaborative services recording research statistics. Dana-Farber’s customized computer in adult oncology. Today, the majority of system cost $1.7 million. It’s an integra- care provided at Dana-Farber consists of tion of two computer checking systems— chemotherapy done in an outpatient set- Cultural shift a pharmacy checking system developed ting. Surgery, radiation therapy, and at Dana-Farber and an order-entry sys- other procedures requiring hospitaliza- That lapse got Dana-Farber’s doctors and tem adopted from Brigham and Women’s tion are done at its sister hospitals. To administrators to ask some bold ques- Hospital, a sister institution (See story further integrate the drug order-entry tions. Were nurses, pharmacists, and tech- page 12). Since 1996, through a joint system, an even more sophisticated com- nicians reluctant to question a physician’s venture titled Dana-Farber/Partners puter system is now being phased in orders or treatment plan? Did the profes- CancerCare, three premier Boston insti- gradually at all three institutions. sional culture encourage their input? Was 7

everyone working together to ensure that Indeed, Dana-Farber has made ongoing Board of Registration in Nursing for their care was both optimal and safe? The dis- patient education a top priority. A new roles in the incident. Three other nurses couraging answer was that communica- Patient-Family Advisory Council advises do not agree with the Board’s position and tion between these various health profes- staff on ways to enhance care. Some will soon face hearings. sionals around the issue of medical errors Council members accompany the medical The Institute’s current leadership is and patient safety was less than optimal. staff on hospital rounds to get patient understandably reluctant to comment on That led Dana-Farber to commit to nur- feedback. “We are truly welcomed on these personnel matters. They focus turing a “team approach” to the preven- these rounds,” says cancer survivor Judie instead on the future and the importance tion of errors similar to that emphasized Beard. “When mistakes are uncovered, the of establishing a system that will make in, say, doing an operation. staff genuinely wants our input.” firing people unnecessary (except in To that end, Dana-Farber set up a cases of intentional or truly negligent Pharmacy Performance Improvement behavior that leads to patient harm or Team. The team meets twice a month and Reporting errors puts a patient at risk of harm). is now part of the 15-member Pharmacy “Removing the threat of judgment of and Therapeutics Committee. Says Sylvia But by far the toughest job Dana-Farber the individual frees people up to discuss Bartel, Dana-Farber’s pharmacy director, undertook was to ensure errors would in system failures. But you have to be “Each of us can freely question any aspect fact be uncovered—and reported. It set careful to not decrease the incentives for of a proposed treatment plan.” Beyond about to implement a “non-punitive” personal vigilance, accountability and that, management has emphasized that all system of error reporting. But it was responsibility,” says Richard Bohmer, personnel should feel free to raise ques- launching such an effort amid a tense M.B.Ch.B., M.P.H., who teaches health tions if they see anything that doesn’t environment of suspicion in the wake of care management at Harvard Business look right. Physicians are also exhorted to the overdose events. School. At the same time, Bohmer and work with nurses and pharmacists to con- Although Institute leaders realized sys- other management experts say health tinuously review treatment decisions. temic errors were the major factors in the facilities have held a wrong-headed And as new cancer treatment protocols cases of Betsy Lehman and Maureen notion of what would keep errors in are developed, they are evaluated by a Bateman, they had in fact dismissed the check. “The general assumption [before] multidisciplinary team which puts safety resident physician who had made the orig- at Dana-Farber was, if you hire the high on its criteria list. inal error. The senior attending physician in best people, give them the best training This questioning spirit is especially charge at the time was also fired. In addi- and provide the best equipment, that important in shaping the guidelines for tion, 15 nurses recently have reluctantly excellence in care is a happy coalition experimental therapies, which often carry accepted reprimands by the Massachusetts of these organizational elements,” says higher risks. Certain patients, whose dis- ease progression affords them no other options, are willing to receive new drugs, drugs at new strengths, or in novel com- binations with other medications or pro- cedures. Betsy Lehman, for example, had volunteered to be part of a breast cancer study involving high doses of one drug whose powers were enhanced by another. The team’s focus is on anticipating problems and preventing errors in the first place, not just trying to fix things after they occur. Pooling their expertise often leads to unanticipated solutions. “I could never go back to our old clinical practice model,” says Bartel. “Sometimes I wonder why we didn’t think of this before; it makes so much sense. Then I realize it’s because we didn’t have to think about it; the culture wasn’t there to support it.” Patients are also considered “team members.” “The more informed the patient is, the safer we all are—the patient and the system. We encourage them to speak out,” says Susan Grant, R.N., Dana-Farber’s Rather than a handwritten order that can be misread, physicians enter drug orders via chief of nursing and patient care services. a computerized ordering system. (Photo by Steven Gilbert, StudioFlex Productions) 8 CHANGING THE PARADIGM

FROM “EVERYTHING’S FINE” TO

All is Fine! Endless Opportunities Errors Rare Errors Everywhere Tell as Little as You Can Tell Whatever You Can Keep Board Out Actively Involve Board MD’s Don’t Participate Doc’s Actively Involved Middle Management Must Fit Exec’s Will Lead Staff Error Rate is Average No Threshold for Errors

“WHO DID IT! JUST FIX IT! FROM I WANT NO MORE!” TO

Just Give Me a Name Look for the Root Cause Penalize the Reporter Thank the Reporter Add More Checks Simplify the System It Will Pass You Will Carry the Burden Fix as They Happen Set Reduction Priorities

Bohmer. “For years, they got away with lines for the use of the anesthetic were participate in a Harvard study of error pre- this line of thinking—lots of hospitals established within weeks as were a new vention in the ambulatory setting. The have gotten away with it.” set of guidelines for performing bone mar- study’s goal is to define—and ultimately Dana-Farber’s shift to non-punitive row aspirations and biopsies. In addition, prevent—errors that can occur in outpatient error reporting is still evolving and far a new procedure was put in place to more clinics, doctors’ offices, nursing homes and from perfect. But “compared to the past,” extensively train and monitor physicians- centers that do outpatient surgery. says Robert Soiffer, M.D., co-chair of the in-training who are learning the proce- Dana-Farber physicians, nurses, phar- Pharmacy and Therapeutics Committee, dure. At the conclusion of the investiga- macists and administrators are also often on “there’s been a dramatic increase in staff tion, in a memo to his colleagues, the doc- the road, sharing their message with thou- reporting of errors—about 60 to 80 each tor expressed his gratitude for the “profes- sands of people at meetings and confer- quarter.” Most have been minor and not sional respect and treatment he received ences nationwide. “We realized that we had affected patients directly, he says. in a potentially embarrassing situation.” the burden of our action and the responsi- Several “near misses” have occurred Maureen Connor, R.N., who oversees bility to correct it,” says David Nathan, since the overdose incident, although Dana-Farber’s risk-management program, M.D., Institute president since 1995. “But we nothing of that magnitude. All were says the new reporting climate produces also realized we had the power to help oth- caught by the pharmacy staff who ques- results. “Health care providers and ers learn from our experience.” tioned ordered prescriptions. patients are more willing to report every And others are learning. Several can- The Institute’s response to a March error, down to the smallest mistake. With cer centers have switched to a safer way 1999 incident is telling. A physician-in- this information, we can move forward of handling medication orders since the training (a fellow) injected a patient with and deal with it,” she added. overdose tragedy at Dana-Farber. Some, a local anesthetic in preparation for a including Memorial Sloan-Kettering in bone-marrow aspiration. The patient suf- New York, are implementing computer- fered a seizure but later fully recovered. A The Next Steps ized order-entry systems. committee was immediately convened to But no system, whatever its advan- review the incident and interview the Dana-Farber officials say they will sustain tages, will eliminate the need to be con- physician. They probed the incident to see the commitment to error reduction in other stantly on guard, watching for every pos- if any processes should be changed. It ways as well. For example, with funding sible crack in the safety shield. Says Dr. turned out the fellow had used a dose from the Risk Management Foundation Nathan, “Even with the best effort, error slightly higher than standard. New guide- in Cambridge, MA, Institute personnel will prevention requires eternal vigilance.” 9

A Government Health System Leads the Way Two physician leaders with a deep commitment to safe medicine helped make the sprawling VA health network a model for others

BY LANI LUCIANO The VA Health System

When Kenneth Kizer became Under VA hospitals—a report that led to “medical errors” by virtue of being pre- Secretary for Health in the U.S. Depart- Congressional hearings—Kizer decided to ventable if proper care had been rendered. ment of Veterans Affairs (VA) in 1994, he make a virtue of visibility. He testified in The other half were associated with an knew he was taking on a major challenge. 1997 that such errors not only occurred assortment of medical mishaps, includ- Long-maligned for shoddy care, the VA but were far more common than most ing improper insertion of catheters and system is also famous for being an people understood—both at VA hospitals feeding tubes, botched surgery, falls, and entrenched and highly politicized bureau- and other health care facilities. And he patients receiving the wrong drug or too cracy resistant to change. pledged to Congress that he would wage much of the right drug. “I saw Born on the Fourth of July like an all-out campaign against medical The VA report was an almost instant everyone else,” the 48-year-old physician errors in the VA system. “I thought, let’s beacon for other, private health facilities administrator says, “but I also knew that put all the problems out there and use and systems to begin tracking errors the VA’s problems were no worse than heightened awareness to correct them, not more systematically and preventing average. They were just more visible.” just in the VA but throughout health care.” them. Indeed, the watchword now in the Kizer set about making a series of changes It was a bold move that by most hospital industry is “if the VA can do it, at the VA that shook the institution. accounts has paid off. Media attention in we certainly can.” Among them were decentralizing the VA’s late 1999 surrounding a report from the The $18.3 billion VA is the nation’s power and emphasizing outpatient-based Institute of Medicine (IOM) on medical largest integrated hospital and health care primary care. In the process, more than errors included numerous favorable men- system. It includes 172 hospitals, 600 out- half of all acute care hospital beds have tions of the VA’s effort to address the patient clinics, 132 nursing homes, 206 been eliminated. But it was a commit- problem. And the VA’s own report—tally- counseling centers, 73 home health care ment he made in 1997 that has recently ing medical errors in its system for the agencies, and assorted other programs. catapulted the VA into the limelight, first time—was positively perceived as an The VA employs 200,000 people, and and could be judged Kizer’s most endur- exercise in government openness. The VA more than three million veterans a year ing legacy. (He stepped down as the VA’s report documented nearly 3,000 errors seek medical services at VA hospitals. administrator in July 1999 and now heads that resulted in some 700 patient deaths Indeed, to Kizer, the VA’s national pres- the National Quality Forum, a private between June 1997 and December 1998. ence, size, and status as a government non-profit group.) About half (277) of the deaths were entity—plus its vital role in medical After the St. Petersburg Times docu- suicides. Of these, an undetermined num- research and training—made it ideal for mented a series of fatal medical errors at ber—but perhaps most—can be considered setting the national example he had in 10

mind. “We capture a huge amount of In a few seconds, the software verifies data, we’re too big to ignore and, since that the right person is receiving the right government employees can’t be individu- drug in the right dose at the right time. ally sued, our system is freer than most to The program screens for half a dozen admit mistakes.” other potential problems, such as drug The first steps came in January 1998. interactions. If everything checks out, the Kizer gathered the VA’s effort into a single software simply records the event—pre- administrative entity, the Patient Safety serving an electronic record. If not, it Improvement Initiative. Later that year, flashes an immediate warning. Kizer recruited former astronaut James The 350-bed hospital—like almost all Bagian to head the VA’s newly formed other U.S. hospitals—previously had no National Center for Patient Safety, head- formal system of verifying medication quartered in Ann Arbor, MI. Bagian, a prescriptions at the “point of care.” physician, mechanical engineer, and vet- Doctors and nurses were simply exhort- eran of two space flights, was one of the ed to double check, says Edgar Tucker, strangulation, are a third patient safety tar- lead investigators on the 1986 Challenger director of the VA’s Eastern Kansas divi- get. The VA is conducting intensive work- accident. His new mandate was both sim- sion. And that was not proving to be shops to teach staff alternative ways to ple and staggering: identify errors and good enough, Tucker says. safeguard patients without using restraints. potential errors in the VA system and fig- Many of the hospital’s nurses worried The training is paying off. According to ure out how to prevent them. “I heard the at first that the new system was mainly a Judith Salerno, the physician who is the theme from Mission Impossible playing in way to track efficiency and boost work agency’s chief consultant for geriatrics and my head,” jokes Bagian, whose playful loads. But the very quick decline in errors extended care, two-thirds of all VA nursing demeanor belies his rock-hard resolve. dispelled skepticism, Tucker says. Medi- homes are currently restraint-free. The initiative aims to reduce medical cation error rates at the hospital plunged Beefing up the science base of error errors through changes in everything from from an average 34 per month in 1993 to reduction is also a priority. Four VA cen- administrative procedures to contracting, six per month by mid-1999. Similar ters—in Palo Alto, CA, Cincinnati, OH, technology support, nurse training, and results were obtained at the VA hospital Tampa, FL, and White River Junction, VT— medical practice itself. A few parts of the in Leavenworth, KS. have been designated Patient Safety multi-faceted effort are well underway— A similar bar-coding system eventual- Centers of Inquiry. Each will get $1.5 mil- such as the new program to track and ly will be required for blood products lion over the next three years to conduct report errors. But Bagian is quick to point used in transfusions. Says Kizer: “This is research and identify safety techniques out that most are still evolving. a relatively low-cost, high-benefit safety and technologies other industries are using The first big test of a systemwide improvement that all hospitals can and which may have health care applications. change, he says, will be a requirement that should invest in.” But both Kizer and Bagian say the all hospitals put in place a bar-coding sys- biggest challenge in reducing medical tem similar to, but more sophisticated than, errors and making hospitals safer places is the kind grocery store check-out clerks use Other Initiatives changing the culture of medicine. As Kizer to read the price of foods. All VA hospitals A second innovation the VA undertook encapsulates it: “Can you imagine a nurse must have the system by June 2000. was to revise its medication storage proce- yelling ‘Doctor, stop!’ or even mentioning Bagian says the decision was accelerated dures. Following the advice of the National ‘Are you sure you want to do that?’ No, because an initial test at two VA hospitals Patient Safety Partnership (NPSP)—a group you can’t, and that needs to change.” in Kansas proved the technology dramati- the VA founded in 1997 with 11 other Bagian says the initiative aims to cally successful. It reduced the medication organizations—all VA facilities must replace medicine’s fault-finding hierarchy error rate 70 percent over a 5-year period. remove a host of particularly hazardous with the institutional trust and sense of The Colmery-O’Neil VA Medical medications from areas where patients are personal responsibility typical of “high- Center in Topeka, KS, began testing an cared for. Heading the list is the drug most reliability” organizations—such as his for- early version of the bar-coding system in frequently involved in medication errors— mer employer, NASA. 1994. Now they have 35, $2,700 hand- concentrated potassium chloride. It can Though brimming with ideas he’d love held battery-operated computer termi- cause cardiac arrest when injected in undi- to execute right away, Bagian knows the nals, two per 30-bed ward. luted amounts. Also banished are concen- effort will take time and that he must The system works this way: patients trated epinephrine, digoxin, insulin, lido- move deliberately and cautiously. To that and staff wear bar-coded identification caine, pancuronium and verapamil. Such end, he is field-testing several projects strips, and all medications also have their drugs can cause significant harm or even before trying to implement them at all VA own ID strips. Before medicating a death if delivered to the wrong patient or facilities. One is revamping of VA’s error patient, a nurse or staff member laser- given in inappropriate doses. data repository, known as the Safety scans all three strips—from the patient, Physical restraints, which cause an esti- Events Registry (SER). All VA personnel nurse, and medication—into the computer. mated 100 deaths a year, most due to can report to the SER so-called “sentinel” 11

message or not. If not, we’ll head back there and figure out where we missed the boat and how to do it better.” The VA also has not been afraid to use some tried and true methods of inducing “Close calls show us weaknesses in change to reduce errors. Recognition through awards is important. But putting the system just as harmful events do, some money on the line appears to pay off, too. Individuals who suggest broadly applicable safety improvements qualify but at a much lower cost in human for bonuses of up to $5,000. Institutions can garner up to $25,000. Topeka’s med- misery and wasted resources.” ication scanning system won the maxi- mum prize in 1999. Continuing education for staff mem- bers is also a critical component of the events—incidents or situations that caused patients or their family members if a mis- initiative. In 1999, the VA became the significant harm to a patient. But Bagian take that resulted in patient harm is country’s first health system to require wants to make the system a repository for uncovered. This has made it easier for the that all permanent employees complete close calls, too—events or situations which VA to encourage error reporting. But thirty hours of continuing education each could have resulted in an accident, illness, Bagian doubts that this protection has year, ten of which focus on quality im- or injury, but didn’t. played that big a role. Rather, he says, fear provement and patient safety. Bagian also “Close-calls show us weaknesses in the of job loss or punishment is the obstacle. wants to launch a magazine devoted to system just as harmful events do, but at a “We don’t favor the beatings-will-con- safety, such as those published by the much lower cost in human misery and tinue-until-morale-improves approach,” Navy and Air Force. “They’re filled with wasted resources,” Bagian says. To identi- smiles Bagian. “We want people to under- inspiring you-are-there type stories that fy problems that require the most urgent stand that all problems merit analysis make you consider how you’d handle a attention, the SER uses a safety assessment because they’re opportunities to inter- sudden problem. They’re absorbing to code. It rates individual incidents on both vene. Everyone on the team has to buy read and they stimulate you to think.” their severity of harm to the patient and into that crucial belief.” The new hand- Yet another idea is to increase the use their frequency of occurrence. Incidents book and the system it promotes are cur- of simulators for medical procedures. that cause great harm or occur often are rently being tested in Florida, Puerto Such devices allow doctors and nurses to targeted for immediate intervention. Rico, California and Nevada. This effort practice procedures and emergency drills, Paradoxically, Bagian says, making SER also will test on-site “root-cause” analysis and make mistakes, using equipment— a more useful tool for improving patient training at all VA facilities by September including computerized human dum- safety will depend a great deal on making 2000. The goal of the three-day sessions, mies—before they ever perform them on VA doctors, nurses, and technicians feel he says, is to foster new attitudes, not real patients. (See story page 24.) VA safer using it. “It’s the front-line folks, the impose new hard and fast rules or formu- facilities are also under orders to double clinical and risk management staff, that las. “The old model was, you stressed who the number of they perform. will make this initiative work. We need was at fault. The new model is that we Currently, 15 percent of patients who die their trust and their input,” says Bagian. understand you do not come to work to in a VA hospital get an . Medical To foster that trust, he has rewritten make errors, and we want to minimize the experts have agreed for years that autop- the VA’s Patient Safety Improvement risk that you will do so,” says Bagian. sies are very useful tools for learning handbook. The new version stresses Getting information on errors and about medical mistakes. institutional and interpersonal coopera- potential errors is but half the calculus in Bagian is far from blasé about the tion, not personal consequences. And it improving safety, however. The other half challenge ahead. To build a system where emphasizes that reported errors are sub- is responding. Under a new set of direc- safety permeates the culture will take ject to analysis leading to corrective tives from Bagian, error analysis must years, he acknowledges. “We will need action and preventive measures rather culminate in an action plan that moves to everybody up and down the chain of than punitive action against staff mem- solve the problem. Simplistic remedies, command, from Congress to doctors and bers. The sole exception is a criminal or such as “nurses must pay more attention” patients to believe in this new approach,” deliberately unsafe act. are no longer acceptable, and even he says. “We have to encourage and The VA’s efforts to uncover mistakes thoughtful, plausible solutions are sub- praise and reward all the steps forward, are aided by its exemption from the legal ject to ongoing scrutiny for effectiveness. including baby steps, so that people are liability cloud that hangs over private Says Bagian: “We’ll get monthly progress comfortable taking bigger steps. But once hospitals. Neither the VA nor its doctors, reports and we’ll be able to tell by the we reach critical mass, I think the as government employees, can be sued by content whether the staff really got the momentum will be unstoppable.” 12 Making Doctors Computer Literate BY JANE ROESSNER

Even as the e-health boom gains suit quickly. Serious medication errors at Coaxing physicians to momentum, many health facilities are the hospital have decreased by 55 percent still far behind in adapting to the brave over the past few years. give up pen and paper new world of information technology. Brigham and Women’s began to em- And though they have seen computer- brace the computer age as early as 1989. It ized machines—like CAT scans and built the Brigham Integrated Computing cut errors sharply MRIs—revolutionize medicine, many System (BICS), a clinical information sys- hospitals have not yet put computers to tem running on a network of over 6,000 and saved $5 million use to increase productivity, stream- computers. Physicians, nurses, and admin- line administrative functions, enhance istrators used the system to access lab research, and improve quality. Some results, discharge summaries, and other to $10 million at one health care institutions, though, are clinical data. By 1991, the hospital’s blazing trails in this form of computeri- administrators were casting about for more inner city hospital. zation that the rest of the industry is cer- specific ways to use computers to improve tain to follow. Brigham and Women’s both the delivery and quality of care. Hospital, a 726-bed academic medical A key target emerged soon enough: center in Boston, MA, is one such facili- “iatrogenic injuries”—injuries to patients ty. Among the most powerful changes: that occur during hospitalization—and, Brigham’s physicians today write all more specifically, injuries due to medica- their orders for treatment and medica- tion errors. Studies at the time estimated tions on a computer screen, rendering that “adverse drug events” were far and paper orders a thing of the past. away the most common medical error and The payoffs have been many. Doctors cost U.S. hospitals over $20 billion per and nurses say it saves time and makes year. In addition, the problem seemed to them feel more professional, efficient, and be on the rise as the health system confident. And hospital administrators say became more complex and medicines the system has saved them between $5 and became more plentiful and powerful. Brigham and Women’s Hospital $10 million a year. But it’s the direct ben- With 36,000 inpatients and 610,000 out- Boston, Massachusetts efit to patients—in reduced errors—that’s patients per year seen at four different likely to compel other hospitals to follow sites, the Brigham’s administrators knew 13

they had a significant opportunity to physicians write their medication orders on Too much would overwhelm doctors. Too improve quality and save money. the computer would go a long way to solv- little would undermine physician buy-in David Bates, a fellow in general medi- ing the problem. In addition, Bates became and the hoped-for reduction in errors. cine at the Brigham at the time, couldn’t convinced that the computer and new In the end, the team settled on a menu have agreed more. A pioneer in quality information technology could become a of medications. For each medication, research, Bates had found that both the powerful “decision support” tool for physi- software offered a “pick list” of appropri- existing processes by which physicians cians—giving them pertinent information ate doses. Before any medication order ordered medications as well as the way at the time they were writing orders. would be accepted, physicians were they were administered were fraught with required to enter dosage, method of problems and invited errors. His research delivery (for example, by mouth or intra- found that 56 percent of adverse drug DESIGNING venously), and frequency. events at the Brigham occurred because But the key new element was this: as of errors at the time of ordering. Bates THE SYSTEM soon as the physician completed an order, discerned early on that the primary it would be checked for errors. For exam- source of the problem was the age-old Brigham and Women’s set out building ple, if the patient was found to be allergic tradition of handwritten orders. Doctors the order entry system in 1991. Jonathan to the drug the doctor was ordering or were scribbling their drug and test orders Teich, then director of the hospital’s was already getting another drug that on forms with triplicate copies, one of Information Systems group, believed that didn’t mix with the new one, the comput- which would be torn off and taken to the the most potent application—the one that er would alert the physician. If the doctor pharmacy (or shot through pneumatic would have the biggest impact on care— chose to override such an alert, he or she tubes). Also, about half of the medica- was “the ability to challenge and critique would have to enter the rationale for the tions issued by the pharmacy were and guide the physician‘s order at the override. Orders were then sent directly to recorded as having been written by “doc- time it was placed.” At the same time, he the pharmacy. There, they underwent a tor unknown,” and in an astounding 80 and other designers recognized that a second check for errors. percent of lab orders, the doctor’s name computerized order entry system would could not even be determined. impose formalism onto what had been— And despite exhortations to write rather bizarrely in retrospect—a largely THE ROLLOUT more clearly, physicians’ handwriting informal process. In essence, the new sys- stayed notoriously illegible. tem would be asking physicians to make The commitment of the hospital’s leader- Bill Churchill, director of the Brigham’s a major change in the way they practiced ship proved to be a vital ingredient to pharmacy since 1990 and a pharmacist medicine. Physicians are required to doc- successfully implementing the new sys- there since 1974, says deciphering the ument all of their orders—not just for tem. CEO Dick Nesson was a persuasive ordering physician’s name was the least of medications, but for lab and radiology champion. He personally mustered the the problems. “We simply couldn’t read tests, food, and activity—essentially support of the hospital’s department the doctors’ orders,” he says. “Different everything they do in the process of car- chiefs. Indeed, when one chief threat- people would use different symbols and ing for a patient. At the Brigham, some ened to instruct his residents to refuse to abbreviations for medicines, and the phar- 16,000 orders are written each day, 40 use the system, Nesson made it clear that macy didn’t know what they meant. You percent of them for medications. every department was required to at couldn’t tell if the order was for 0.5 mil- The team’s first decision was to get least give it a try. ligrams or 5 milligrams.” physicians to buy into the system by After a one-month pilot trial on the The problem was compounded by dra- designing it for them. Teich decided that bone marrow transplant unit in January matic advances in the pharmaceutical if one aim was to promote decision sup- 1993, initial rollout began in May 1993 industry. “New drugs were coming on- port, doctors had to use the system them- and continued for the next 18 months. line every day,” Churchill says, “These selves. They couldn’t rely on nurses or The rollout took place, says John Glaser, were potent drugs with profound ability administrative staff to enter their orders the Brigham’s chief information officer, in to help, but also to harm. And these drugs into the computer. To get doctors to use “punctuated bursts”—first in Medicine, interact differently on each patient. the system, Teich knew he had to make it including general medicine, cardiology, Asking physicians to keep all of this easy to use and not time consuming. and oncology; and then, after a two- information in their heads is a lot. Even Physicians told the design team in no month pause, in Surgery, and then in for pharmacists who specialize in this, it uncertain terms that if it took a lot more Obstetrics and Gynecology. can be overwhelming.” time, it simply would not fly. The team Like the design itself, the rollout put the At the same time that the data were made the computer format resemble the doctor front and center. The operative prin- being gathered, a potential solution was written forms doctors were used to, but ciple in the short run was to minimize being developed. Bates and others rea- they also imposed standards on how the problems for the physician, even if it meant soned that a comprehensive computerized screen should be filled out. The final con- putting other departments (for example, “order entry” system combined—and here ceptual challenge was to find the right the lab or radiology) and the nursing staff was the trick—with a requirement that balance of decision support information. in an awkward position—between the paper 14

world and the electronic world—for a peri- the new system was using 10 to 12 per- displaying information at the time the od of time. Indeed, when the new system cent of their time, according to a physician is ordering medications. went into effect, a variety of questions time/motion study. That translated into The Brigham’s system offers five about workflow bubbled up. For example, an average 40 minutes a day longer writ- modes of decision support: if a physician writes orders for postopera- ing orders. Fortunately, as they adapted to • It alerts physicians to potential drug tive care before surgery, how and when are the system, doctors recovered an average interactions, patient contraindications or those orders activated? What happens if of 20 minutes of that time. But it wasn’t allergies; those orders say “continue all medication,” enough. So the designers set about trying • It calculates proper doses—determin- but the medications change in the interim? to make the process more efficient. ing the right amount of a particular med- How is it indicated that orders have been Their first discovery was that a signif- ication for a particular patient at a partic- taken care of? icant amount of time could be saved if ular time. The system’s designers “Lots of things had to be handled doctors wrote orders in groups rather researched the hospital’s database to explicitly that were handled implicitly in than singly. While writing single orders determine the most commonly ordered the paper world,” explains Gil Kuperman, could take as much as five times as long doses of particular drugs. Using this a physician and one of the system’s on the computer as by hand, writing information, they constructed “pick lists” designers. “Humans can be ambiguous. orders in groups was much faster. For of doses, highlighting the dose most com- Computers aren’t.” example, writing orders one by one for monly ordered. • It presents drug substitution infor- mation. The designers recognized the “Lots of things can be handled explicitly that power of the system as a tool to change physicians’ behavior. When physicians were handled implicitly in the paper world. ordered the common ulcer treatment and prevention drugs Zantac® or Tagamet®, Humans can be ambiguous. Computers aren’t.” the system suggested they consider sub- stituting Axid®, a less expensive but equivalent medication. The intervention Physicians were invited to use a “feed- patients who needed a blood thinner was allowed physicians to make the switch back” button on the screen to air griev- time-consuming—at five minutes per by simply pressing an “OK” button. In the ances and suggest improvements in the patient, for forty patients, taking over very first week the intervention was system. Teich recalls that in the first week three hours. To facilitate “group orders,” introduced, physicians elected to substi- of rollout, his department received 122 the system’s designers built one screen tute the less expensive medication an pieces of “hate mail” via the feedback but- with a spreadsheet listing all patients in astounding 94 percent of the time. ton. “We felt better,” he says, “when we certain categories. Doctors could then • It can alert physicians to potentially realized that 99 of them came from the write orders for blood thinner for many dangerous situations. For example, if a same person.” The feedback mechanism patients at once—a process that turned lab test for a patient indicated a low also invited physicians to interact with the out to require only seven seconds per potassium level, the computer would system; they became active participants in patient. It proved to be a critical change, query if the patient was also taking making the system better, not passive vic- improving efficiency all around. digoxin—a potentially dangerous combi- tims of its intrusion into their lives. nation. If so, the computer would identify In time most physicians responded and page the patient’s physician, at any positively, Teich says. They liked the basic CONTINUOUS hour of the day, informing the physician information about proper doses and fre- of the reason for the page and asking for quency for medications. And they liked IMPROVEMENT action. Physicians were overwhelming in the computer’s ability to calculate the their support of this feature, giving it a 96 right dose of potassium for a particular Getting the initial bugs out and weaving percent approval rating. patient. But the single greatest perceived computerized order entry into the day-to- • It can track patients’ test results for benefit by physicians, and the one that day fabric of the hospital took about a kidney function and alert doctors and really saved time, was the ability to write year. The next step was to add the “deci- staff if medication doses exceed recom- orders from remote locations—for exam- sion support” component. mended levels for people with abnormal ple, writing an order for a patient on the Decision support was designed with kidney function. Dubbed the Nephros 12th floor when they were on the 9th two goals in mind: reducing errors (e.g., Project, this feature was made a part of floor; or even writing orders from home. “patient X is allergic to medication Y”) the decision support because so many But, as anticipated, there were com- and reducing costs (e.g., “would you like adverse drug events involved patients plaints. The major one involved the added to switch to this equivalent, cheaper with kidney problems. Doses of quite a time commitment. Before computerized drug?”). The basic notion behind “real few common drugs that would be appro- order entry, doctors spent approximately time” decision support is that the comput- priate for a patient with normal kidney two percent of their time writing orders; er can modify the physician’s decisions by function could be dangerous or even fatal 15

in a patient with low kidney function. As ry care physicians to use such a system cost of medications under capitated con- a result of Nephros, the percentage of than the hospital’s own physicians. tracts by as much as 25 percent for a sav- orders with dosing that is appropriate for The hope is that physicians will buy into ings of an estimated $20 million per year. kidney function has increased from 30 the system because it can help them navi- As for the more distant future, Brigham percent to 70 percent. gate the complex rules of managed care— management is looking closely at adapt- the varying rules plans have that specify ing computer assistance and decision sup- which drugs and which tests are covered, port to emergency room operations and to RESULTS— under which conditions. These can change involving patients more directly in their from month to month and it’s almost own care. For example, on-site computers, AND THE FUTURE impossible for physicians to keep up. with internet links, could help people According to John Glaser, an order entry understand health issues and treatment Computerized physician order entry at the system with decision support information choices for their condition. They could Brigham is now six years and some 35 on every plan’s formulary—updated regu- also help people communicate directly million orders old. It has become an inte- larly—could be highly attractive. Brigham with doctors and other health providers. gral part of providing care at the hospital. and Women’s researchers project that such It’s the brave new world that many Paper orders are a rapidly fading memory. a system could reduce outpatient medical experts envision. And the Brigham is Thirteen thousand orders are entered daily errors by as much as 80 percent and cut the already heading in that direction. at the Brigham, 88 percent of them by a physician. On an average day, 386 of these orders are changed by a physician because BUILD YOUR OWN OR the computer suggested doing something different. An average of four times a week, a physician cancels a chemotherapy order BUY OFF THE SHELF? because the dose is too high. Any way it’s cut, computerized order Building a computerized order entry system cost Brigham and Women’s Hospital about entry has earned its keep. Research by David Bates concludes that the system $1.4 million in 1993; thereafter, annual maintenance has cost about $500,000 a year. has yielded a 55 percent reduction in How many hospitals have the resources—financial, medical, and technical—to build such serious medication errors—those that either had the potential for harm or actu- a system? How does a home-grown system compare with a ready-made one? ally resulted in injury. In all, the system According to the Brigham’s Information Systems team, building a system from scratch saves $5–10 million per year, with most savings coming from use of less expen- makes sense for a hospital if three conditions are met: sive drugs and tests. As well as it has performed, the system • The hospital believes it can move faster than the market. at the Brigham has been due for an upgrade for several years. That is now underway. • The hospital is uncertain as to the best configuration of such a system, and wants The biggest issue has been the network, which has been at capacity for some time, to learn from rapid cycles of testing and fixing features. with about 3,200 linked PC’s. This is being replaced with a faster network. Also, all • Many functions within the hospital are intertwined. applications are being moved to a faster, more flexible software program. Hospitals that meet these conditions can usually develop a system that meets their In addition, plans are underway to extend the drug order entry system to needs better. The system is also likely to be cheaper in the long run than if purchased from the outpatient setting—the 610,000 a vendor. Doing it yourself also gives an organization deeper knowledge on using tech- patients treated annually at the Brigham’s outpatient sites. Even though nology to improve care. the primary care sites already use an However, most community hospitals and smaller hospitals will not have the human electronic medical record, computerizing ordering will be challenging. Primary resources to develop a system on their own, and will likely be better off purchasing one care physicians see many more patients, from a vendor. Although off-the-shelf systems won’t have all the features or the flexibil- each for a much shorter time. Although the range of orders is narrower in the pri- ity of Brigham’s system, most vendors in this sector—such as Cerner, Eclipsys, SMS, and mary care setting, there are many more HBOC—are making important improvements. The cost of off-the-shelf systems varies, orders per day. Moreover, it will be more difficult to “require” independent prima- depending in part on the computer networks already in place. 16 WHAT’S IN A NAME? When it Comes to Drugs—

A pharmaceutical giant tackles a drug mix-up and sets a precedent

Bristol-Myers Squibb LOTS

BY CARALEE ADAMS

harmaceutical giant Bristol-Myers Confusion was apparently occurring cerned. He worried that the inserts and P Squibb Co. (BMS) had a problem. with the generic names (cisplatin and car- warnings would go unread—as such inserts Reports were filtering into the company’s boplatin) as well as the brand names often are. And he was concerned that more Princeton, NJ, headquarters that one of its (Platinol® and Paraplatin®) because they patients might be harmed if BMS did not premier drugs—Platinol® (cisplatin), used all shared the same “stem:” “platin.” move aggressively. The ISMP also was urg- to treat testicular, ovarian and advanced But BMS at that point still was unsure ing the company to take stronger measures. bladder cancer—was being confused by whether they had a systemic problem on The Platinol®-Paraplatin® mix-up was pharmacists and doctors with another their hands—or whether the reports were then and still is symptomatic of a serious anticancer drug it makes called Para- isolated events due to fluke mix-ups. and growing problem: health profession- platin® (carboplatin), used mostly to treat Even so, it took a precautionary first step als make drug dispensing and administra- advanced ovarian cancer. in 1992—putting letters to the editor in tive errors every day—some of them The issue was serious because the various oncology medical, nursing and life-threatening, even fatal—because the appropriate dose of cisplatin is about five pharmacy journals warning of possible names of many drugs are so similar. In to six times less than that for carboplatin. cisplatin mix-ups. the most recent example, the FDA received Mix-ups can be fatal. Indeed, the poten- 53 reports of dispensing errors involving tial danger of a mix-up between the eports of the error kept mounting, the three sound-alike drugs Celebrex® (an two drugs had been noted as early as R though. By the end of 1995, some 32 arthritis drug), Cerebyx® (to treat seizures) 1991. The Institute for Safe Medication medication errors related to cisplatin had and Celexa® (an anti-depressant). Practices (ISMP), based in Pennsylvania, been reported to the Food and Drug The precise magnitude of the problem is had received a report of a woman about Administration (FDA), and eight people difficult to gauge. There is no comprehen- to get a bone marrow transplant who was had died, according to the FDA. That sive record-keeping or data base that tracks to receive cisplatin. Because of a pharma- prompted BMS to add information to the it. According to the FDA, roughly half of cy dispensing error, she received cisplatin label for Platinol® highlighting the poten- the 6,000 medication errors reported to the at the carboplatin dosage. She later died. tial mix-up. It also distributed letters to agency between 1992 to 1997 were due to In another case that year, a nurse tran- oncology health care professionals to labeling or packaging issues. Of that half, scribed a telephone order wrong, writing alert them to the problem and to the dis- some 27 percent were caused by generic or cisplatin instead of carboplatin. The phar- continuance of one form of Platinol®, a trade name confusion. But Jerry Phillips, macy dispensed the drug as written, and powder for injection that had the highest the agency’s associate director of medica- the patient received the higher dose. The potential to be confused with Paraplatin®. tion error prevention, says studies indicate next morning the physician saw the tran- Despite these steps, Frank Pasqualone, the number of medication errors reported scribed order, and the error was discov- then director of policy and planning for to FDA—including mix-ups—represent only ered. The patient died a week later. BMS Oncology-Immunology, was still con- a small fraction of those that occur. 17

t’s a significant and persistent prob- he vast majority of the mix-ups with nearly everything the advisory panel had “I lem,” concurs Bruce Lambert, an asso- T Plantinol® and Paraplatin® were recommended had been put into place. ciate professor in the Department of linked to the wrong container being cho- The company won’t reveal the cost of Pharmacy Administration at the Univer- sen as it was dispensed by a pharmacist. the changes, but Pasqualone says it was a sity of Illinois at Chicago. He says phar- “It was clear that it was a problem “significant investment.” The costliest macists and other health professionals being contributed to by their packaging,” item was the advertising campaign to have been concerned about it for years says Dominic Solimando, a pharmacist educate health care professionals about and have pressed the pharmaceutical and oncology consultant in Arlington, the product changes. industry to change their practices. VA, and a panel member. Indeed, the Notably, one change the panel did not Magnifying the problem today, packaging designs of the original recommend was changing the name of Lambert notes, is the growth in the num- Platinol® and Paraplatin® boxes were the drugs. Doing so would have added ber of medications. There are about extremely similar—which turned out not $5–10 million to the cost of the response, 13,000 prescription drugs on the market to be an accident. The packaging of all estimates Michael Cohen, president of the now, according to IMS Health, a health the oncology products were intentionally ISMP. Although Cohen favored a name care information company. And in the similar for purposes of product line con- change, the panel fully considered that past decade, the number of new registered sistency and brand recognition, a BMS option and the consensus was that chang- trademark drug names has quadrupled, spokesperson says. ing the name would not have added to from 744 new trademarks registered in Working with company staff, the the prevention of the mix-ups. Instead, it 1988 to 3,038 in 1998. This creates a panel came up with a number of recom- would have confused pharmacists and major challenge for an industry always mendations: perhaps even led to other drug mix-ups. striving to come up with catchy drug • Redesign the cartons of both drugs In the end, most agree that the compa- names—such as Prozac® and Viagra®. to de-emphasize the generic name and ny did the right thing. “All in all, we want Not surprisingly, with so much invest- eliminate the confusing similarities—put people to act like Bristol-Myers Squibb did ed in brand names and the marketing that the letters “cis” in bold uppercase and red and fix the problem,” says pharmacy man- surrounds them, drug makers are usually letters with the “platin” in smaller type. agement expert Lambert. “Whether they did very reluctant to change the name or • Add a graphic of a stop sign to the so quickly enough is another question.… package design of their products if a package to warn pharmacists and other When errors come to light, we’d like to see problem arises. users to verify the drug name and dose. (companies) be much more aggressive.” That ultimately was the problem fac- • Redesign the outer carton so the crit- “Bristol-Myers Squibb set a good ing BMS. But first the company sought to ical information is on more than one side example of acting responsibly and using gather the facts on Platinol® mix-ups and of the box, making the warnings visible innovation to correct a problem that led reach out to experts. Pasqualone pressed no matter how the box is placed on the to serious patient injury,” says William his superiors to convene an advisory pharmacy shelf. Ellis, executive director of the American board of outside experts including five • Put warnings of the maximum Pharmaceutical Association Foundation, pharmacists, two nurses and one physi- dosage of cisplatin on the vial. a research group affiliated with a nation- cian. The panel’s central mission was to • Put a set of three stickers on the car- al professional society of pharmacists. review the data and advise on a response ton to help certify the drug had been pre- Cohen says BMS’s actions should to the Platinol®-Paraplatin® issue. But it pared at the proper dosage. serve as a model to others. It was one of was also charged with recommending • Put the same stickers on I.V. bags; the first companies willing to reach out to ways to make the company’s products one is used by the pharmacist and anoth- outside experts and analyze what could safer in the future. er by the nurse. be done to correct a problem with its Pasqualone’s mounting concern, Pasqualone set out to convince the product, rather than blaming outside assertive approach, and BMS’s commit- company to make all the changes. It took forces, he says. But, he adds, the more ment to change their practices eventually time. But Pasqualone was determined. global response to drug name mix-ups is set a precedent for the industry that Indeed, as 1996 progressed, the num- careful testing of drug names, labels, and reverberates today. ber of cisplatin-related problems mount- packaging before they are launched. The independent advisory group gath- ed. Twenty-nine cisplatin medication After the packaging changes were ered in Naples, FL, in January 1996. For errors were reported in that year, with made in 1996 and the advertising cam- eight long hours, and with company rep- another eight deaths—bringing the total paign in 1996 and 1997, cisplatin-related resentatives on hand, it engaged in a since 1991 to 61 incidents and 16 deaths. medication errors and deaths declined process dubbed “failure mode and effects By spring, Bristol-Myers’ management dramatically. In 1997, six cisplatin errors analysis.” The group examined, in detail, was firmly on board with the panel’s and were reported, with one patient death. In the distribution of the two drugs from the Pasqualone’s suggestions. They gave the 1998, only one incident was reported. And time they were loaded onto trucks after go-ahead for rapid action. Some of the by October 1999, two incidents had been manufacture to the moment they were outside experts were consulted as packag- reported. No deaths related to cisplatin injected into a patient in a hospital. ing changes were made. By October 1996 overdoses were reported in 1998 or 1999. 18

Putting one determined doctor in charge of reducing errors made all the difference at one Tackling Medication Errors Midwest hospital HEAD ON Luther-Midelfort Hospital Eau Claire, Wisconsin BY KRIS REBILLOT

Like hundreds of hospitals in hundreds Fortunately, the vast majority of the Midelfort’s initiative is being watched close- of similar towns, 310-bed Luther Hospi- errors uncovered at Luther-Midelfort were ly and eventually could have even broader tal and Midelfort Clinic sit placidly amid minor. But the findings—from a painstak- application. Mayo runs 13 hospitals in 54 the tree-lined streets of Eau Claire, WI, a ing year-long study of charts and proce- communities in and around Rochester. city of 60,000 at the confluence of the dures—frightened Resar, his colleagues, Eau Claire and Chippewa Rivers. And like and the hospital’s administrators. All real- most doctors, Roger Resar thought ized that a process which allows too many The Mayo System Luther-Midelfort—hospital and clinic— small mistakes is the same process that were pretty safe places for the 2,000 leads inevitably to a big mistake. “Maybe and Change patients a day that pass through their we gave a sleeping pill to someone who Two factors have fueled the innovations at doors. As far as Dr. Resar could tell, a ter- shouldn’t have gotten one. That’s not a big Luther-Midelfort—the sheer force of Resar’s rific team of physicians and nurses, deal. But what if someone got penicillin personality and an organization that came backed up by a crackerjack administrative and they were allergic to it,” Resar says. to embrace the cause and open itself staff, were delivering top quality care in a Resar was put in charge of dramatical- to change. “You have to recognize that well-greased system. Resar knew, of ly reducing errors at Luther-Midelfort. His improvement is hard work,” says Dr. course, that errors sometimes occurred, as first task was to take on medication errors— Terrance Borman, the hospital’s medical they do at every hospital. In his 22 years where most of the errors were occurring. director. “It has to be someone’s job. We’ve as a pulmonologist, he’d witnessed his Resar was also asked to build a new system recognized that in order to help our doctors share. But he felt Luther-Midelfort had a of error reporting, which proved to be one improve, we need doctors who are change solid program to prevent serious errors of his toughest challenges. agents.” Ron Hitzke, Director of Inpatient before they occurred, catch them in a The results so far have been highly Pharmacy Services, says Resar has been timely manner when they did, and rewarding, Resar says. Since mid-1998, “the spark plug that keeps us going.” address what went wrong. medication errors at Luther-Midelfort Dr. Patrick Macken, a nephrologist in- Resar has a very different view today. have declined by 82 percent at the key volved in many of Resar’s projects, concurs. And so do his bosses. After authorizing points of patient care—admission, transfer “He’s recognized as an excellent doctor and Resar in 1997 to assess errors and efforts from one department to another within a team player. He’s very passionate about to reduce them at Luther-Midelfort, the the hospital, and discharge. Insulin errors his work.” Macken says the Mayo organiza- hospital’s leaders came to a sobering have been cut by 50 percent after a stan- tion’s commitment to let Resar do the nuts reassessment of the success of their exist- dard dosing regimen was implemented. and bolts work of quality improvement has ing programs. And 90 percent of patients on the blood- made all the difference. “Resar creates the “I was astounded,” says Resar. “We thinning drug Coumadin® are now get- new forms, he looks at the data, he’s the were finding 200 to 230 actual or poten- ting the proper dose, up from 75 percent one who works closely with the nurses.” tial medication errors for every 100 charts before improvements were made. Resar says getting physician buy-in is a we reviewed. That means if you went to As part of the Mayo Regional Health key to the success of the initiative so far. “If our hospital, you’d be at risk of two to System, and thus partner to the famed you’ve been around a hospital long three errors during your stay with us.” Mayo Clinic in Rochester, MN, Luther- enough, you know the guys who carry the 19

big stick are the physicians. They have to review further shaped the initiative as During their initial chart study, for be in on it from the beginning,” he says. He patterns emerged. Data showed, for example, Resar discovered one case involv- also acknowledges the approach used at instance, that 56 percent of the hospital’s ing a disoriented elderly woman who took Luther-Midelfort to implement change: medication errors occurred at the “inter- the anti-coagulant drug Coumadin® on a built a project in small steps—steering clear faces of care”—in other words, during daily basis. The patient was admitted over of making wholesale changes too quickly— admission, in-hospital transfers, and at a weekend by her primary physician’s part- and test changes first on a smaller scale. discharge. Armed with this information, ner, who was unaware that the patient took “We spend five minutes coming up with a Resar and his team began their first for- the blood-thinner. The patient’s regular form,” says Resar. “We don’t do a lot of mal improvement project in 1998 on a doctor didn’t catch the omission on planning because we know it probably 40-bed medical unit. They committed to Monday and as a result, the patient had a won’t be right the first time. Some organi- the creation of a “drug reconciliation stroke during her stay. zations take months to design a form, but process” to address medication errors Justesen’s admitting nurses began to spending too much time thinking about it occurring during the interfaces of care. sit down with patients and their families makes things too complicated.” This system They set a target: they would reduce actu- to detail and verify every drug they were also allows physician “champions” to take al and potential adverse drug events by taking, its dosage, frequency, and when it part in short, 20-minute meetings aimed at 60 percent in three months. was last taken. If needed, clinic records improvements, instead of involving them They met the goal. The team began were pulled, and pharmacists were con- in knock-down, drag out review sessions first by tackling medication mistakes sulted. Everything was compared to a that are common in many organizations. occurring during the admission process. doctor’s orders. If something didn’t make “We started with just five patients and two sense and was time-sensitive, the doctor nurses on one shift,” said Resar. “We took was called. The nurses had 24 hours to get Tackling an idea and tried it. We didn’t refine it to the medication record in order. death. If it wasn’t right, we let it go. If it Once the new admission process was Medication Errors worked we kept moving it to a larger established, Resar tackled medication The efforts at Luther-Midelfort started in cycle.” The first step, he says, involved the errors associated with transfers from the 1997 when the hospital sent a number of simple act of moving the “admission med- hospital’s intensive care unit. That effort staff members, including Resar, to a sym- ication notes”—typically buried in a pile of involved the formidable task of getting two posium on reducing medication errors. paper—to the front of the chart. “It took a computer systems to work with each Resar and hospital leaders became converts lot of work just to get that done,” said other—the Intensive Care Unit had a system to the emerging—but still minority—view Resar. “We had to figure out what form to separate from the rest of the hospital. Now, that errors were a more serious problem at use, and we had to get people to move it.” a process is in place allowing those systems hospitals than generally appreciated. Equally painstaking was the develop- to work in coordination. In addition, new Resar’s first step was simply to start ment of new protocols for admitting forms make it easy for doctors to detail talking to staff about the issue. He queried nurses who became accountable for medications for patients leaving the ICU. nurses, doctors, pharmacists and purchas- unraveling the snarl of drug informa- The discharge process was next. ing agents, asking them blunt questions in tion. Elderly patients, for example, often Typically, hospital pharmacists were spend- informal and off-the-record conversations. are taking 12 to 20 different medications ing 30 minutes to two hours per patient to His most pointed question was always, a day. “It was a nightmare,” says Jane discern precisely what the doctor’s “dis- “Are there accidents waiting to happen?” Justesen, a nurse manager. charge on home medication” orders meant. As these conversations progressed, it became clear to Resar that, one, errors were Measures to prevent medication errors at Luther-Midelfort have involved the entire probably more common than believed at health care team and the patient, as well. (Photo by Rick Gregerson) the hospital and, two, employees had been withholding information about them. Retribution had never been a big part of the corporate culture at Luther-Midelfort. But even in this positive work environ- ment, the threat of punishment held sway. Nurses, faced with the choice of “telling” on colleagues (who could face disciplinary action) were not reporting medication mis- takes—even the most minor ones. Resar’s informal survey yielded a list of key drug delivery issues—and specific medications—of biggest concern. That list came eventually to form the basis for a three-year, 11-part project. Resar’s chart 20

Working closely with the pharmacy staff, Once again, the problem stemmed from ment projects grow naturally as the suc- Resar built a new process within a couple too many options. There are a number of cesses pile up and hospital staff adapts. of months. Now at discharge, the pharma- ways to calculate Coumadin® dosage, That way, says Borman, “we get very little cy’s computer generates a record of every and there are a wide variety of tablets staff resistance.” The change happens medication the patient has taken while in available. In addition, Resar and Macken almost by osmosis, he adds. “Peer pressure the hospital. All the doctor has to do is found that patients were coming in for lab kicks in because the ideas really work.” checkmark the medications he or she wants tests when their primary physician was Innovations in patient safety are mov- continued. Every drugstore in Eau Claire not available to read results. Partner ing faster and faster through the Luther- accepts the watermarked document as a physicians, unfamiliar with the patients, Midelfort system, and momentum is prescription form. would make adjustments and order addi- building for larger systemwide quality “Not only does this save the doctors’ tional tests as a precautionary measure. improvements. The Coumadin® project time,” says Hitzke, “It also prevents the Resar worked with Macken to come up took 18 months while the sliding scale potential for mistakes in transcription.” with a standard nurse-run protocol for insulin project took only two months to Pharmacists also have more time to help outpatients on Coumadin®. The result: a roll out. The “medication reconciliation” patients. “We’ve always wanted to be reduction in unacceptable blood clotting process aimed at the interfaces of care is more clinically involved, and this is a measurements from 25 percent to 10 per- now in place in 80 percent of the hospi- way to do that,” says Hitzke. cent in patients taking the drug. Today, tal. The process has evolved through 14 Resar and the patient safety team also nurses trained in drug interactions write iterations. Most recently, Resar has given targeted specific drugs they found associ- all Coumadin® orders in the majority of himself three months to implement a list ated with errors and delivery problems. clinic departments, and the patients come of innovations recently devised by the Chief among these were insulin for diabet- in for lab tests at a set time. Nurses have Massachusetts Coalition for the Preven- ics and Coumadin®, a blood thinner com- guidelines allowing them to adjust pre- tion of Medical Errors. These involve set- monly prescribed for clotting disorders and scriptions with a requirement that they ting up educational review policies for heart rhythm irregularities. Resar’s research consult with physicians when lab results oral medications given intravenously and found that about 20 episodes of low blood go too far out of range. the color coding of different drug delivery sugar were being identified among diabet- systems, among others. ics each week in the hospital. The problem: And after a year-long effort that start- practically every doctor on staff used his or Affecting the ed out in just one department, a “non- her own “sliding scale” to adjust insulin punitive” error reporting system is in dosages to keep patients’ blood sugar lev- Bottom Line place hospital-wide. As predicted, reports els stable. “We were all trained differently; The payoff for these improvements is not of errors from nurses and technicians everybody had his or her own way of only better quality care but cost savings. have risen dramatically—up seven-fold in doing things,” says Macken. “We identified While Luther-Midelfort administrators the first month after the policy was 14 patients who were being cared for by have yet to measure the actual magnitude implemented in the spring of 1998. Says nine doctors, who used 12 different scales of the savings, there is little doubt they Rupp, “We have worked hard at changing for insulin. How were the nurses supposed are significant. “We know at least 12 peo- the cultural fear of punishment. Now we to know what to do?” ple were admitted for bleeding complica- tell people they’ll get in trouble if they Testing different options with a hand- tions from Coumadin® last year,” says Dr. don’t report an error—and they have a 48- ful of patients under Macken’s care, Resar William Rupp, the hospital’s president hour period to let us know if something worked with pharmacists to come up with and CEO. “By making the changes in our has gone wrong.” a standard scale for insulin dosage. That outpatient clinics, we’ve had no admis- Resar says he has the proverbial “job scale is now being used for all diabetics sions for that this year. Coagulation mea- that is never finished.” But his enthusiasm admitted to the hospital. “The nurses are surements among patients are more stable for the work—and the mission—is un- happy with it,” said Macken. “The phar- because of our nurse-run protocols. We’ve diminished. In years past, “we were macists are thrilled.” The patients also been able to cut back the need for addi- benchmarking ourselves against medioc- won: insulin medication errors declined tional blood tests by 30–40 percent.” rity,” Resar says. “Now, we are trying to by 50 percent. Rupp says other hospital CEOs ask him create an organization that’s among the Problems involving blood-thinning how he can afford to have a pulmonolo- safest in the country. I think we can do it.” drugs appeared even more life threaten- gist do administrative work. “I say we Then he launches into a list of future pro- ing. Clinic patients taking Coumadin® can’t afford not to have him do it. Based jects: standardized protocols for infusion were being admitted to the hospital for on all of his improvement projects, I esti- pumps, better post-operative pain proto- either excessive bleeding, resulting from mate Roger saves us four to six times cols, and simplified forms to track “near too much of the drug, or strokes, a poten- what we pay him each year.” misses” on medication errors. “Some of tial result from getting too little of the Rupp and Borman also champion the my colleagues think I’m wasting my edu- drug. Adverse drug reactions also were “pilot project” approach to change at cation because I’m not ‘doing’ medicine,” being noted in patients taking Coumadin® Luther-Midelfort. The strategy is to test says Resar. “But I know I’m just taking in combination with certain antibiotics. things on a smaller scale and let improve- care of a larger population of patients.” 21

Making

MISTMISTHarderAKESAKES to Make

BY SHELLEY REESE

Like a growing number of hospitals have the most impact. A multi-disciplinary nationwide, staff and administrators at team ended up creating 15 separate pro- Fairview Health Services, a system of jects. All started off on a small scale so seven hospitals based in Minneapolis, they could be implemented quickly and became increasingly aware in the mid- efficiently. The primary testing ground 1990s that too many patients were getting became Fairview Southdale, a 300-bed the wrong drugs or the wrong dose of a hospital in the twin cities suburb of Edina. drug. Steve Meisel, assistant director for The projects involved medications as clinical pharmacy services at Fairview diverse as anticoagulants, chemotherapy, Focusing on the Southdale Hospital, says it was not so insulin and sleeping pills. They also much an epiphany as a dawning con- explored ways to adjust drug doses for management of sciousness and conviction that the prob- people with abnormal kidney function lem had to be made a priority. and took patients out of harm’s way by a few high-risk In 1996, the hospital’s leadership simply removing some of the most dan- launched a program to reduce the risk of gerous drugs from areas where they were drugs, a Minnesota drug errors. After three years, the result has readily available. been a series of innovations aimed at cre- All the projects were grounded in a hospital system ating an integrated medications manage- common—and common sense—premise: ment system with much tighter controls, make it more difficult for staffers at every cut medication fewer handoffs, and increased accountabil- stage of the process to make a mistake. ity. To do this, Fairview brought together A crucial choice was made early on errors and sharply doctors, nurses, pharmacists, and quality and may well prove a popular model for improvement experts to focus on the entire other hospitals. The hospital system’s improved care continuum of drug delivery. pharmacy was selected to spearhead and “The chief thing we had to communi- oversee the initiative. The choice is not an cate to people at first was that they may obvious one. While pharmacists have long be doing their job very, very well but that played a key role in hospitals, their exper- the whole process was just not working,” tise has often been underutilized, and says Denise White, a quality consultant at some doctors are still reluctant to cede Fairview. “We had to get everyone to step pharmacists control over patient care. back far enough to see how they work in The decision in this case was based on the whole context.” the judgement that pharmacy had both The first step, both White and Meisel the needed expertise and organizational say, was to identify clinical areas, , skills and a strong rapport with various Fairview Health Services and medicines where the problems loomed departments throughout the system. Those largest and where pilot projects would ties had been forged over 20 years by the Minneapolis, Minnesota 22

presence of pharmacists at nursing sta- “The results were dramatic to the point tions throughout Fairview hospitals—a that I almost didn’t believe them,” says practice common in Minnesota but not Dr. William Hession, president of the throughout the country. Minnesota Heart Clinic. “It was pretty humbling to see.” The Warfarin Project The initiative’s success is largely attributable, says Meisel, to tracking and Among the first changes Fairview follow-up of patients. But he says it’s also Southdale initiated was to reorganize the due to the staff’s acceptance and embrace management of a drug called warfarin of guidelines that were developed in con- among cardiology patients. Warfarin, an junction with the Minnesota Heart Clinic. anticoagulant, is prescribed for a variety Two full-time pharmacists, Melissa Van of conditions to prevent and treat blood Holland and Caren Allivato, ran the pro- clots. It’s a difficult drug to manage be- gram from within the Minnesota Heart cause numerous factors—including diet, Clinic. (A third pharmacist has since been alcohol use, and other medications—can added but one is part-time.) Patients are affect the amount of drug in the body. seen face-to-face at each visit to identify Patients must be carefully educated potential problems, make adjustments in about warfarin and factors that can affect their therapy, and to provide ongoing its use. Frequent monitoring of blood education about their drug therapy. tests is also necessary to assure the drug Says Hession: “If you look at why peo- is working properly and to prevent poten- ple fail on warfarin, it’s because they get tial problems. These problems include lost. Patients figure, ‘if I’m not bleeding severe and potentially life-threatening and the doctor is not bugging me, I must over a period of months and then adopt- bleeding and clotting. be OK’ and they don’t get their tests. Now ed at some or all of Fairview’s hospitals. The warfarin project was prompted in we have a database, and we know who Reorganizing drug management of the fall of 1996 when data showed that an should be coming in for blood studies. patients with impaired kidney function average of 1.7 patients per week—almost Patients are getting watched more closely, proved particularly successful. Using 90 a year of the roughly 4,000 taking the and that is really paying off.” He says dosage guidelines developed by the hospi- drug—were being admitted for bleeding. patients also better understand the effects tal’s Pharmacy & Therapeutics Committee, The majority, 70 percent, of the admis- diet and drug interactions can have on pharmacists were given permission to sions were thought to be due to an error them and learn to avoid potentially dan- independently change physician-pre- in monitoring or dosing, or to inadequate gerous situations. scribed doses of drugs for patients with follow-up. Warfarin was thus an ideal Both Fairview and the Minnesota impaired kidney function. The purpose: to candidate for more intense management. Heart Clinic were so impressed with reduce the risk of toxicity associated with The project was first launched as a six- results that they extended the warfarin too high a dosage, a common problem. month pilot for 135 patients of the program to other patient populations. The About 70 doses now are being changed Minnesota Heart Clinic, an independent hospital made the service available to its each month. Physicians are notified. cardiology practice affiliated with the orthopedic patients and vascular patients, Fairview Ridges Hospital developed a suburban hospital. A Fairview pharmacist for example. These patients are often pre- similar protocol for sleeping pills, such was assigned to educate patients on war- scribed warfarin after surgery. Results as triazolam (Halcion®) and temazepam farin, oversee dosing, and follow-up with have been similarly promising. After the (Restoril®), to reduce the risk of overmed- those patients who had been started on pharmacists took over the orthopedic ication. Too high a dose of such drugs can the drug in the hospital. patients’ warfarin management, the num- heighten the risk of falls and accidents. The program yielded benefits from day ber of blood clot complications for the Under the new protocol, if a physician one. Cardiology patients with insufficient group fell from an average of 7.6 per year prescribes a dose that is too high for the monitoring fell from 22 percent before the to 4. Today, Van Holland and Allivato, patient, the pharmacist is empowered to pilot to less than one percent. Blood tests and new colleague Kim Graupe, manage change the dose. showed the share of cardiology patients warfarin therapy for about 400 patients. In a related step at Fairview Ridges, with the correct dosage of warfarin in their But warfarin is not the only drug to all patients considered to be at risk of bloodstream increased from 35 percent to come under Fairview’s microscope. The falling are identified on admission. 65 percent. As a result, warfarin complica- hospital also established dosing protocols Nurses then develop plans to prevent tions per 100 cardiology patients fell from for sleeping pills and for patients with falls, for example, by assisting at-risk 12 to two. In addition, patient knowledge impaired kidney function. And the project patients to the bathroom. about the drug, measured by correct team developed a chart with a sliding Meanwhile, the team at Fairview answers on a test, improved from a base- scale for determining correct insulin Southdale tackled the issue of drug errors line of 30 percent to over 80 percent. dosage. All three programs were piloted due to miscommunication. It developed a 23

was dropped because staff objected to the additional responsibility. More troubling, the long-term viabili- ty of some programs now in place remains uncertain. The warfarin project, for example, relies on tenuous financing. It costs $175,000 a year to run, Meisel says, with expenses comprised almost entirely of the two pharmacists’ salaries and a portion of a secretary’s salary. Initially, the hospital funded the entire program. But in January 1998, the Minnesota Heart Clinic agreed to begin covering the costs of its increased usage. Dr. William Hession, left, Minnesota Heart Clinic’s president, reviews a patient’s chart, while Caren The cardiology component of the service Allivato, above, a Clinic Pharmacist, works with a patient taking warfarin. (Photos by Jill Neu) runs about $100,000 per year. Still, the doctors, patients and Fair- pre-printed chart with a sliding scale for hazard drugs might mistakenly be admin- view end up paying for some of the cost: insulin dosing and a corresponding pre- istered. At all Fairview hospitals, such Medicare only pays a portion for face-to- printed form for recording medication medications were simply removed from face visits and does not reimburse for administration information. The aim was to floor drug carts stock. The move reduced phone consults or follow-up. The hospital reduce errors caused by illegible handwrit- the number of stock items in intensive and Minnesota Heart Clinic bear this cost. ing, improper abbreviations, overlapping care units by about 30 percent. In addi- “The onus falls on the wrong parties,” blood-sugar ranges, and failure to treat tion, to prevent patients from suffering said Hession. “Right now, none of the high blood sugar levels in a timely fashion. dangerous reactions to drugs such as insurance companies are willing to come The hospital system also sought to astemizole, an antihistamine, and oral to the table to fully fund this kind of tighten controls on drug prescriptions ketorolac and metformin, which are used service.” He says he hopes insurance by changing the way information is to treat arthritis and diabetes, respectively, companies will eventually come to see the recorded. For example, at Fairview they were simply removed from the hospi- cost effectiveness of warfarin manage- Southdale chemotherapy order forms tal system’s pharmaceutical drug formula- ment as a preventive service. were redesigned to prompt practitioners ry. With the exception of metformin, less Despite such financial uncertainties, to record the level of detail necessary to hazardous medications were substituted. Meisel is brimming with other ways to safely prescribe, dispense, and admin- So far, the success of Fairview’s vari- reduce medical errors. He would like, for ister these powerful drugs. Under the ous programs are confined to distinct example, to expand the more intensive previous system, doctors could place pockets of its complex care systems. And medical management of warfarin to neu- orders verbally or over the telephone. there is no single, overarching statistic rology and primary care patients and, The new template provides consistency describing how much patient safety has eventually, to the entire patient popula- and reduces the opportunities for poten- improved, says Meisel. It is impossible to tion. And he’d like to apply similar proce- tial errors. Fairview Ridges Hospital has measure progress in such a tidy way, he dures to drugs such as amiodarone, which since adopted the system, as well. says, because there is no baseline figure is prescribed for cardiac rhythm disorders, “We tried to design the template to to show how often mistakes are made. and digoxin, which is prescribed for heart reflect exactly the way people thought “If you ask me what’s the overall med- failure, and other high-hazard drugs. about dosage,” said Paula Welford, a ical error rate in our hospital, I would not Finally, Meisel would like to see the nurse manager formerly in the oncology have a clue,” Meisel says. “It’s not like hospital invest in technology to automate unit. “So there is no chance of mistaking measuring the mortality rate after bypass and streamline drug dispensing with com- a one-time only dose for a daily dose.” surgery. You don’t always know when an puters and lower the risk of errors with a The same logic was used to develop error occurs, and people don’t always bar-coding system that tracks all drug new procedures for copying information agree on what constitutes an error.” orders. Having witnessed the results a onto patient medication records. By Fairview’s successes have not been handful of small-scale initiatives can yield, establishing criteria to govern the way without accompanying failures. Some he’s eager to tackle more ambitious mea- information is recorded—such as the use pilot programs have been dropped sures. “When you think of the contribution of abbreviations and brand versus gener- because of lack of staff support. For pharmacy can make to improving safety ic names—Fairview Ridges reduced the example, a measure intended to ensure and how far things can go, it’s enormous,” number of places information might be patients are not being over-medicated Meisel says. “There are several hundred misinterpreted by 60 percent, says Meisel. would have required doctors to review more things we’d like to try. We’d like to Finally, Fairview implemented a key their orders prior to a patient’s transfer get to the point where no patient again change to reduce the likelihood that high- from a critical care unit. That initiative would be the victim of a medication error.” 24 AA DummyDummy HELPS PAVE THE WAY TO Simulation technology, like that used to train pilots, can help BetterBetter doctors learn from their mistakes MedicineMedicine VA/Stanford Simulation Center Palo Alto, California BY LISA KRIEGER

In the operating room, things are not VA Medical Center in Palo Alto, CA, and keep skills sharp. And pilots frequently going well. An electronic monitor record- a professor of anesthesia at Stanford, the learn to use new equipment and fly unfa- ing a patient’s heart rate registers a sharp PatientSim simulator has become a cen- miliar planes in a simulator. rise from 70 to 100 and then to 160 beats tral learning tool at the VA/Stanford Simulators are credited with playing a per minute. His blood pressure is crash- Simulation Center for Crisis Management key role in the enormous reduction in air ing. His airways spasm. Warning lights Training in Health Care. Both PatientSim crashes and fatalities over the past few flash red. Alarms sound. and the Center were inspired by Gaba’s decades. Simulation-based training is also All in a few minutes, routine knee firm belief that the failure of doctors to used in automobile driving, shipping, surgery has turned lethal. The anesthesi- learn to cope with urgent crises and react military command and control, and the ologist reaches for a syringe, fumbles, in a systematic, timely way is a major operation of nuclear power plants. Gaba then drops the vial. Time is running out. factor in preventable anesthesia and believes simulator training in health care The doctors scramble to figure out what is operating room medical mishaps. can and should play a similar role. going wrong. They are acutely aware that Early forerunners of PatientSim at the Like pilots, he says, surgeons and anes- the combination of rapid heart rate and VA/Stanford Center were the nation’s only thesiologists can face crisis events that are plummeting blood pressure could kill a hi-tech human simulators until just a few infrequent but potentially lethal. Both person in minutes. years ago. But PatientSim now has some pilots and doctors must absorb a prodi- But fortunately, not this patient. Lying 150 “friends” in hospitals around the world gious amount of technical knowledge. And before this medical team is literally a made by two competing manufacturers. the bias in their education and training dummy, a simulated emergency care The growing number has occurred because leans heavily towards mastering a defined patient made of plastic, wires and comput- simulator-based training has gradually and complex set of skills. But unlike er chips. His formal name is “PatientSim,” gained an important foothold in medical today’s pilots, doctors are often poorly but he usually goes by Sam, Johnny, or education. And with interest in reducing trained in “crisis management”—how to Jessica (add a wig and some other features medical errors and improving safety rising respond when something goes wrong, and he becomes a she). Even if Sam’s doc- rapidly, medical safety experts predict sometimes terribly wrong, Gaba says. tors made all the wrong decisions, he medical simulation training to spread more Many anesthesia mistakes, for example, would survive; a simple touch of a “reset” widely in coming years. are caused by technical problems, such as button brings him back to life. The technology is modeled to a large incorrect administration of a drug. But But the doctors who are tending to degree on the experience in aviation. many others are caused by behavioral Sam are keenly aware that they’ll not After World War II, the rapidly expanding problems and poor communication among have that luxury with a flesh and blood commercial airline industry knew it had a doctors, nurses, and technicians. For exam- human being. They are “practicing” on safety and liability problem on its hands. ple, a doctor barking an order into thin air the simulated patient to learn how to Simulators were one way the industry in an operating room could go unheeded if avoid and overcome medical errors and to sought to enhance training and reduce it is not addressed to a specific individual. respond in crisis situations—where studies the risk of pilot error—which had risen Gaba also thinks the system trains doc- consistently show the risk of medical sharply as planes got more and more tors poorly to do new tasks. It is not errors is particularly high. complex. Today, airline pilots are required uncommon for a doctor performing a new Since its creation in 1986, based on to learn to fly in a simulator before they technique or using a new piece of technol- work by Dr. David M. Gaba, Director of take the controls of a real plane. Hours of ogy to try it out for the very first time on the Patient Safety Center of Inquiry at the additional simulator time are required to a real patient. That, Gaba says, is unwise. 25

In the practice of anesthesia, Gaba complete model of cardiovascular, pul- Most amazingly, Sam is programmed to notes that deaths from errors have monary, fluid, acid-base electrolyte and respond as a healthy human would to more declined sharply over the last 30 years thermal physiology. It includes computer- than 70 medications. When he gets an anes- (see story page 26). But “we need to, and controlled electromechanical lungs, heart thetic, for example, his eyelids stop moving can, make it even safer,” he says. Deaths and breath sounds, changeable airway and his breathing slows; doctors must from anesthesia errors have declined to anatomy, and a palpable pulse. quickly stick a tube down his windpipe to as low as one per 100,000 to 200,000 give him air. If the tube does not go down “healthy” patients (those whose physical HOW THE correctly and he fails to get enough oxygen, condition can not have contributed to the SIMULATOR WORKS his heartbeat speeds up and carbon dioxide death). But the risk of dying in a domes- makes his blood pressure rise. By mimicking tic commercial jet flight in the U.S. is far, The VA/Stanford Simulation Center, an actual patient during surgery as much as far lower—about one in 8 million. opened in 1995, looks and feels like a fully possible, doctors know quickly if they’ve Gaba says the idea for the patient sim- functional operating room, complete with injected the wrong drug, erroneously inter- ulator came to him when he was reading assorted high tech medical equipment, preted vital signs, or made other errors. the 1985 book Normal Accidents by Yale surgical lights, and tension in the air. An After every simulator session, doctors sociologist Charles Perrow. The book ana- adjacent control room contains a comput- and nurses can watch their performance on lyzes the social and human side of tech- er that executes all Sam’s “physiologic” video monitors in a debriefing room. These nological risk. It concluded that system functions—from the cardiovascular and sessions, Gaba says, provide profound complexity, and the human response to pulmonary systems, to metabolism, fluid insight into glitches in the system that can emergency, make error and failure and electrolyte balance, and thermal reg- lead to patient harm. inevitable in many high-risk industries. ulation. For example, the system moves “It’s not about finding bad doctors,” The relevance to anesthesia and medi- “mathematical blood” from one heart Gaba says. “It’s about helping doctors learn cine was clear to him immediately, Gaba chamber to the next and from blood ves- to work together better to prevent prob- says. “I wanted to test how anesthesiolo- sel to blood vessel—a process requiring lems and errors.” gists respond to unexpected events. And thousands of computations every second. The simulator also is being used for it dawned on me that we could build The simulator can mimic dozens of research. Gaba and his team have studied something to measure and monitor this.” complex and realistic clinical crisis sce- decisions made by intensive care unit clin- But building a patient simulator posed narios. And instructors can modify these icians, comparing their responses to those challenges for Gaba not faced by the scenarios in hundreds of ways. Thus, Sam of a prototype computer-based decision designers of flight simulators. An artifi- can be programmed to act like a healthy support system. They also are measuring cial environment that mimics the human young truck driver or a frail 80-year-old the influence of fatigue on cognition. body is enormously more complicated woman. Instructors can induce a heart “Within the next few years, we can than simulating flight. The human body attack, respiratory trouble, an allergic expect to see continued, incremental presents thousands of variables that make reaction to medicines, and other kinds of improvement of simulator technology,” even fickle weather conditions seem easy. distress. For each event, myriad settings Gaba says. Beyond that—in 10–20 years— The first electromechanical anesthesia are possible to mimic severity and respon- the next generation of simulators will simulator had been developed in the late siveness to treatment. And reactions to likely use virtual reality technology to 1960s. Though ahead of its time, the drugs and procedures vary with each com- replicate the operating room’s environ- machine was flawed and drifted into obliv- puter-programmed change in physiology. ment. And beyond that...well, anything is ion. In 1986, Gaba and a colleague began The simulator is programmed to show possible. It’s likely that simulators will be developing the first generation of realistic the same distress symptoms as a human used broadly in training and perhaps even simulators, using new computer capabili- patient. For example, a working blood become so commonplace that they allow ties to provide signals to actual clinical pressure band is strapped to Sam’s biceps doctors to practice particularly tricky instruments. This model was improved and an IV is placed in one arm. His pulse operations before they do them. with a modified mannequin which allowed can be felt at both wrist and neck. The Already the crisis management training mask ventilation, intubation and monitor- heart beat can be heard with a stethoscope pioneered in anesthesiology has been ing of breath sounds. But it was still rela- and the heart rhythm can be made to be extended to delivery room crisis and tively crude, with no pulse, spontaneous abnormal. Inside Sam’s mouth are teeth, neonatal resuscitation, emergency room breathing or human-like physiology. tongue, uvula and windpipe. Electro- and trauma care, intensive care medicine, A major redesign in 1989 incorporat- mechanical, computer-controlled lungs are cardiac arrest response (“code”) teams, and ed a physiologic model of the cardiovas- embedded in its chest; they “breathe” spon- other health care settings. Although Gaba cular system. Then in 1992, the CAE Link taneously, as well as by hand or mechani- believes that simulators will improve med- Corporation, a manufacturer of military cal ventilation. Incoming gases are detect- ical safety, he’s quick to say they are com- aviation and space flight simulators, ed and quantified and their concentrations plementary to, not a replacement for, con- licensed the technology from Gaba’s fed to the physiologic and pharmacological ventional means of medical education. team to develop a commercial patient computer models. Oxygen sensors ensure After all, for all his marvelous technical simulator. Its mannequin contained a that if Sam quits breathing, alarms go off. sophistication, Sam is still just a dummy. 26

A MEDICAL SPECIALTY BLAZES A TRAIL Deeply disturbed by patient deaths and facing skyrocketing malpractice premiums, the field of anesthesia became the first to seriously tackle medical errors. It’s now a model for others.

It’s just a small device, resembling a was placed by mistake into the esopha- plastic clothespin with some wires that gus. It’s an “operator error,” and when it connect to a monitor about the size of a happens, irreversible brain damage and portable phone. Without looking closely death can occur in minutes. In the past, you might think it was part of an elec- by the time the mistake was detected, it tronic game. But the elegantly simple was usually too late. But the clothespin- device is among the most important like device, developed in the early 1980s, in medicine. It determines whether a discerns the levels of oxygen circulating patient under general anesthesia is in the patient’s bloodstream. receiving enough oxygen—oxygen that The oxygen pulseoximeter, as the will keep the patient alive through the device is called, is one of dozens of procedure and, if all goes as expected, important advances in the field of anes- permit her to awaken in the recovery thesiology in the past 20 years. What BY JAN ZIEGLER room with no ill effects. makes the field unique is that many of Oxygen, of course, is vital to all living the advances were either specifically tissue. In decades past, oxygen depriva- aimed at reducing the risk and rate of tion while a patient was under anesthesia errors, or did so as an important sec- was one of the most common errors in ondary benefit. As a result, undergoing medicine. Lack of oxygen, or hypoxia, anesthesia has gone from being one of occurred most frequently because the the most notoriously dangerous proce- tube that was supposed to deliver the pre- dures in medicine to one of the safest. cious gas through the trachea to the lungs In the 1950s, the death rate from anes- 27

thesia was roughly one in 3,000 to 4,000. allowed to leave the operating room for a At the same time, outside forces were By the 1970s, it had come down to rough- cup of coffee after the patient was impinging. Malpractice insurers were tak- ly one in 10,000, says John Eichhorn, “under.” Ether was the primary agent ing a proactive stance towards anesthesia M.D., professor and chairman of anesthe- used, despite the availability of other errors. Eichhorn says the major malprac- siology at the University of Mississippi gases. They monitored the patient by tice carrier for the nine Boston hospitals Medical Center. But by 1990, Eichhorn’s simply observing signs such as breathing affiliated with Harvard Medical School own meticulous research among patients rate and pulse. “It was really very rudi- made the rounds of anesthesia depart- in Boston found just one non-fatal mis- mentary,” Pierce says. ment chiefs in the mid-1980s. Their mes- take among 392,000 patients undergoing In 1960, Pierce became vice-chairman sage was simple and blunt, he says: “You anesthesia. Other recent data indicate a of anesthesia at Peter Bent Brigham guys are costing us too much money.” rate of one death per 200,000 to 300,000 Hospital in Boston. “As I collected cases Spurred by the insurer’s warning, anesthetics administered. over the next 10 years,” he says, “it Pierce, Eichhorn and others helped form “It has been a true revolution in health became obvious there were a lot of deaths the Harvard Department of Anaesthesia care,” says Eichhorn. that could be prevented.” Risk Management Committee. The group That revolution and the field of anes- The public was beginning to surmise embarked on creating a comprehensive thesia is now being held up as an exam- that as well. No longer willing to accept set of practice standards for the field. ple for other medical specialties, and the “old Joe” explanation, people began They would be the first ever. The stan- medicine in general. to sue anesthesiologists for malpractice dards were issued in 1985 and published Most notable, say experts, is that anes- in record numbers in the late 1970s and in the Journal of the American Medical thesiologists—led by their professional early 1980s. As a result, malpractice Association in 1986. They quickly gained groups—went about making the necessary premiums for anesthesiologists sky- broad currency and remain today the changes themselves. They set strict inter- rocketed, becoming among the most core set of practice standards in the nal standards and organized initiatives. expensive in medicine. field—widely adhered to. Regular updates No outside government body—state or The media poured fuel on this fire. have taken place. For example, pulse federal health authorities, Congress, or The risks of anesthesia became a common oximetry, the mechanism that uses a the American Medical Association—made story in print media, and television was small clip or bandage-like device on fin- them do it. There was of course another, not far behind. In 1982, the ABC news- gertip, ear, nose or toe to measure oxy- baser motive: malpractice liability. And magazine show 20/20 aired a now- gen in the blood, became part of the eventually, some in the profession also famous segment highlighting shoddy standards in 1990. Capnography, a com- pushed for state regulations. anesthesia practices at several institu- panion device that measures blood levels tions. The show warned viewers that of carbon dioxide, was added to the stan- 6,000 would die or experience brain dards in 1991. The early years damage that year because of avoidable anesthesia error. Among the early leaders of this effort, Pierce and others knew a major effort Redesigning Ellison “Jeep” Pierce is widely considered was now in order or the speciality would to have been particularly indispensable. suffer long-lasting damage. In 1983 the machinery His Spencer Tracy-like appearance and Pierce became the vice-president of the manner inspired allegiance and trust. American Society of Anesthesiologists Even as doctors were finding better and His occasional gruffness hid compassion- (ASA). He used his position there— safer ways to use the technology at their ate commitment to the improvement of bolstered by the gathering outrage over disposal, engineers like Peter Schreiber medical practice. What’s more, Pierce is malpractice premiums—to establish the and Jeffrey Cooper were trying to make still involved in the fight. He serves ASA Committee on Patient Safety and the technology more user-friendly. today, at 71, as executive director of the Risk Management. The group set stan- Schreiber, who moved to the U.S. from Anesthesia Patient Safety Foundation dards of care and safety and used every Germany in 1970 to start his own compa- based in Boston. means available to get the word out— ny, says the equipment even in the 1970s Pierce began his in 1954 at including the production of a series of was “very imperfect.” It left too much the Hospital of the University of educational videos that remain popular. room for operator error and machine fail- Pennsylvania. “Our practices back then Two years later in 1985, Pierce and other ure. The machines consisted of a metered were abysmal,” he says. “I can remember leaders organized the first International gas cylinder and a vaporizer for the other we had one cardiac arrest a week as Symposium on Preventable Anesthesia gas or gases that would be administered a direct result of the anesthesia. We Mortality and Morbidity. The meeting concurrently, such as nitrous oxide or would go to the family and say, `We’re was a turning point for the wider adop- oxygen. There was no way to measure the sorry, old Joe just couldn’t take it.’ That tion of more sophisticated patient safety patient’s responses to the anesthesia. was it. They accepted it.” measures in the field. The creation in Anesthetists simply observed their breath- Pierce says it was common in those 1984, of the Anesthesia Patient Safety ing and appearance and took their pulse, days, for example, for anesthetists to be Foundation also added momentum. much as as they had in the 1850s. 28

Schreiber knew that monitors were eral Hospital and Brigham and Women’s new rule in the law required board certi- needed to keep track of cardiac, lung and Hospital in Boston. The “other stuff” was fication for the post. Hospitals were other functions. But he says there was the human error involved in using the given six months to junk out-of-date tremendous initial resistance to this idea. technology. His research demonstrated machines. If they didn’t, they had to Anesthetists at the time considered what that inadequate experience and familiar- close the operating room. All anesthesia- they did an art form. Many, he says, ity with equipment were major factors in related deaths and other adverse events flatly told him: “I do not trust any anesthesia errors. His ground-breaking had to be reported to the state health mechanism or electrical equipment that is papers, published in 1978 and 1984, led department by hotline immediately, and between me and my patient.” to training programs that are widely in writing in 30 days. Failure to comply Schreiber’s most important insight— believed to have played a major role in carried a number of penalties, including and one that resounded later through anesthesia error reduction. loss of license. much of medical equipment design—was But Moss did not stop there. In the that humans were not wired well to mon- early 1990s, he pushed for regulations itor “boring events” for hours at a time. It’s the law in New Jersey that would govern hospi- And he defined surgery as a boring tals and facilities—primarily doctor’s event—from the anesthetists point of view The latest step in this evolution began a offices—with just one operating room. at least. Attention can easily slip during decade ago. Even as anesthesiologists “They were hiding behind the idea that hours of watching a patient. were eagerly embracing the new equip- they were simply offices,” Moss says. His argument slowly began in the ment and adhering to voluntary stan- “But a lot of surgical procedures are 1970s to make a dent. Manufacturers dards, some felt more was needed. One done in a doctor’s office.” rolled out a plethora of monitors that of these people was Ervin Moss, M.D., a It was a long fight but Moss, again, would provide information on bodily practicing anesthesiologist in New Jersey ultimately prevailed. The New Jersey leg- functions. Then, realizing the sheer num- and New York for 45 years. Moss islature in June 1998 passed a law regu- lating office-based anesthesia and surgery. But New Jersey is so far the only state with such a law on the books. California has regulations but they are not nearly as strong, Moss says. A few other states are looking at the issue. Moss believes a major push from anesthesiologists and “We can not rest on our consumers will be needed to get the issue on to the front burner. laurels. That’s the bottom line. Meanwhile, the anesthesia profession continues to wage an aggressive cam- There will always be room for paign to further reduce errors. One initia- tive has yielded important data. It’s called more improvement, and we the “Closed Claims Study.” It originated at the University of Washington in Seattle in simply must achieve it.” 1985. The project uses data from settled malpractice claims to analyze anesthesia accidents. Among its findings to date are ber of these devices was impractical, they believes voluntary standards are com- that death and brain damage represent a designed a system that contained all the mendable and useful, but lack the teeth declining proportion of malpractice necessary monitoring components. needed to go the rest of the way towards claims—31 percent in the late 1990s, Cooper, meanwhile, was focused on reducing anesthesia errors. down from 56 percent in the 1970s. another element of anesthesia technol- He began lobbying heavily for state Inadequate ventilation represented 22 ogy. As a bio-engineer at Massachusetts rules in the mid-1980s. His first opportu- percent of all claims in the 1970s but just General Hospital in the 1970s, he began nity came when his own state of New seven percent in the 1990s. Pierce says to use a method popular in the field of Jersey set about changing its medical the data are more concrete documenta- aviation safety—called critical incident licensing regulations in 1988. Moss tion that the most serious, and fatal, kinds technique—to evaluate the interface insisted that anesthesiology be governed of anesthesia accidents have declined. between machine and human in anesthe- separately. The result was passage of the That these event still occur at all sia. He did not like what he saw. Anesthesia Regulation Bill. The new law, though, is very disturbing, Pierce says. “The equipment was no longer the which covered any facility with two or “We can not rest on our laurels. That’s the biggest problem; it was all this other more surgical operating rooms, had swift bottom line,” he says. “There will always stuff,” says Cooper, now chairperson of impact. Three hospital chiefs of anesthe- be room for more improvement, and we bio-engineering at Massachusetts Gen- siology were forced to resign because one simply must achieve it.” 29 Resources and Suggested Reading

SELECTED BOOKS Perrow, C. Normal Accidents. (New ORGANIZATIONS York, NY: Basic Books) 1984. Bogner, M.S., ed. Human Error in A fascinating book that opens the world of National Patient Safety Medicine. (Hillsdale, NJ: Lawrence complex systems and introduces the reader to Foundation (NPSF) Erlbaum) 1994. concepts of human error. The contrasts between 515 N. State Street, 8th Floor This collection of presentations from a symposium the effectiveness of error prevention in aviation Chicago, IL 60610 on errors in medicine provides a good overview of and the merchant marine are striking. PHONE 312-464-4848 many aspects of health care errors. A number of the FAX 312-464-4154 key contributors to current thought concerning Reason, J. Human Error. (New York, EMAIL [email protected] error in medicine are represented. NY: Cambridge University Press) 1990. WEBSITE www.ama-assn.org The “bible” of error theory, written by the leading Cohen, M. Medication Errors. (Washing- authority. Reason reviews the literature and estab- The NPSF maintains a website that provides links to ton, DC: The American Pharmaceutical lishes a theoretical framework for understanding a variety of sites with information on patient safety. Association Foundation) 1999. error. A treasury of important concepts. This book provides an overview of the types of Institute for Safe Medication medication errors, processes that lead to errors, Practices (ISMP) research on medication errors, and suggestions ARTICLES 300 West Street Road for minimizing the potential for future errors. Warminster, PA 18974 Alibhai, S.M.H.; Han, R.K.; and Naglie, PHONE 215-956-9181 Corrigan, J.; Kohn, L.; and Donaldson, G. “Medication education of acutely FAX 215-956-9266 M., eds. To Err Is Human: Building a hospitalized older patients.” J Gen Safer Health System. Committee on Intern Med, 1999, 14:610-616. ISMP provides an independent review of medica- Quality of Health Care in America, The results of this study indicate that even though tion errors that have been voluntarily submitted Institute of Medicine. (Washington, DC: patients perceive the education they receive from by practitioners to a national Medication Errors Institute of Medicine) 1999. their physician or pharmacist to be satisfactory, Reporting Program (MERP) operated by the U.S. The report lays out a comprehensive strategy many patients still make errors when they take Pharmacopeia (USP). for government, industry, consumers, and health their medications. providers to reduce medical errors, and it calls on US Pharmacopeia (USP) Congress to create a national patient safety center Bates, D. W.; Cullen, D. J.; Laird, N.; 12601 Twinbrook Parkway to develop new tools and systems needed to Petersen, L. A.; Small, S.; Servi, D.; Rockville, MD 20852 address persistent problems. Each chapter of the Laffel, G.; Sweitzer, B.; Shea, B. F.; PHONE 800-4-USP-PRN (800-487-7776) report contains a reference list, allowing the reader Hallisey, R.; Vander Vliet, M.; Nemeskal, WEBSITE www.usp.org to select additional material based on a specific R.; and Leape, L. L. “Incidence of adverse area of interest. drug events and potential adverse drug USP Medication Errors events: Implications for prevention.” Reporting Program Kahneman, D.; Slovic, P.; and Tversky, JAMA, 1995, 274:29-34. PHONE 800-23-ERROR (800-233-7767) A., eds. Judgment Under Uncertainty: A study of 4301 inpatient admissions of medical National Coordinating Council Heuristics and Biases. (New York, NY: and surgical intensive care units at two tertiary for Medication Error Reporting Cambridge University Press) 1982. care hospitals. The authors concluded that adverse and Prevention The classic collection of studies of how humans drug events (ADEs) were common and often pre- PHONE 800-822-8772 make judgments under (usual) conditions of ventable and that serious ADEs were more likely uncertainty. The authors coined some of the key to be preventable. USP has a Medication Errors Reporting Program. terms in behavioral science. USP’s Practitioner Reporting Network (PRN) Bates, D.W.; Leape, L.L.; Cullen, D.J.; et publishes the USP Quality Review, which often Leape, L. L.; Kabcenell, A.; Berwick, al. “The impact of a physician computer- discusses medication error topics. D.M.; and Roessner, J. Reducing Adverse ized order entry system on prevention Drug Events: Breakthrough Series Guide. of serious medication errors.” JAMA, National Patient Safety Partnership (Boston, MA: The Institute for Healthcare 1998, 280:1311-1316. Office of the Under Secretary for Health Improvement) 1998. To evaluate the impact of computerized physician Department of Veterans Affairs Institute for Healthcare Improvement Guides are order entry (POE), patients admitted to three medical 810 Vermont Avenue NW based on the real-life experiences of health care units were studied for seven-to ten-week periods in Washington, DC 20420 organizations that have made dramatic changes four different years. The baseline period was before PHONE 202-273-5807 while participating in a Breakthrough Series implementation of POE, and the remaining three were FAX 202-273-7090 Collaborative. Each contains a synopsis of the after implementation. During the study, the non- Collaborative goals and results, the model for missed-dose medication error rate fell 81 percent. accelerating improvement, change concepts that VHA National Center for Patient Safety Collaborative participants used successfully, addi- 2215 Fuller Road tional resources, key contacts and a bibliography. Ann Arbor, MI 48105 The book is based on a 1996-97 Collaborative. PHONE 734-930-5890 FAX 734-930-5899 30

Bates, D.W.; Spell, N.; Cullen, D.J.; Cullen, D.J.; Bates, D.W.; Small, S.D.; et Leape, L.L.; Cullen, D.J.; Clapp, M.D.; et al. “The costs of adverse drug events al. “The incident reporting system does Burdick, E.; Demonaco, H.J.; Erickson, in hospitalized patients.” JAMA, 1997, not detect adverse drug events: A prob- J.I.; and Bates, D.W. “Pharmacist partici- 277:307-311. lem in quality assurance.” Jt Com J Qual pation on physician rounds and adverse Analysis of the costs of adverse drug events identi- Improv, 1995, 21:541-548. drug events in the intensive care unit.” fied in Adverse Drug Event Prevention Study (see The yield of spontaneous reporting of ADEs by the JAMA, 1999, 282:267-270. Bates et al., JAMA 1995 above), using a case-control hospital incident reporting system was compared to To measure the effect of pharmacist participation methodology. Preventable ADEs cost twice as much intensive, confidential data collection by means of on medical rounds in the ICU on the rate of pre- as nonpreventable ADEs and averaged $4,685 each self-report and daily record review. Ninety-four per- ventable ADEs caused by ordering errors, the at one teaching hospital. Authors estimate that cent of ADEs discovered by record review and confi- authors compared between phase 1 (baseline) and hospitals in the study are losing $2.8 million each dential report were not reported as incident reports. phase 2 (after intervention implemented) and phase year on preventable ADEs. 2 comparison with a control unit that did not Dansky, K.H.; Gamm, L.D.; Vasey, J.J.; receive the intervention. The rate of preventable Bedell, S.E.; Deitz, D.C.; Leeman, D.; and and Barsukiewicz, C.K. “Electronic ordering ADEs decreased by 66 percent from 10.4 Delbanco, T.L. “Incidence and character- medical records: are physicians ready?” per 1000 patient-days before the intervention to 3.5 istics of preventable iatrogenic cardiac J Healthcare Manage, 1999, 44:440-455. after the intervention. In the control unit, the rate arrests.” JAMA, 1991, 265:2815-2820. Electronic medical records are often presented as one was essentially unchanged during the same time Landmark study that demonstrates that most cardiac solution for reducing medical error. This article out- periods: 10.9 and 12.4 per 1000 patient-days. arrests that occur in hospitals are caused by misuse lines some of the barriers in successful acceptance of drugs and are, therefore, potentially preventable. of this mode of operation within the clinical setting. Lesar, T.S.; Briceland, L.L.; Delcoure, K.; et al. “Medication prescribing errors in Bordage, G. “Why did I miss the diag- Ko, R. “Adverse reactions to watch for a teaching hospital.” JAMA, 1990, nosis? Some cognitive explanations and in patients using herbal remedies.” West 263:2329-34. educational implications.” Acad Med, J Med, 1999, 171:181-186. One of the first descriptions of prescribing errors 1999, 74(10):S138-S143. This review article looks at patient safety issues in hospitals. A review article that has several purposes: 1) to regarding the use of alternative medicines and describe diagnostic errors in general; 2) present offers several advisory lists for clinicians whose Lesar, T.S.; Briceland, L; and Stein, D.S. details about two specific types of diagnostic patients use herbal remedies. “Factors related to errors in medication error; and 3) apply lessons learned to the educa- prescribing.” JAMA, 1997, 277(4):312-7. tional process so medical students will learn how Kronz, J.D.; Westra, W.H.; and Epstein, This study quantified the type and frequency of to avoid similar errors. J.I. “Mandatory second opinion patholo- identifiable factors associated with medication pre- gy at a large referral hospital.” Cancer, scribing errors. Systematic evaluation of every third Chassin, M.R.; Galvin, R.M.; et al. “The 1999, 86:2426-2435. prescribing error over a one-year period at a tertiary Urgent Need to Improve Health Care Johns Hopkins Medical Center was the site of this care teaching hospital revealed total errors of 2103, Quality.” JAMA, 1998, 280:1000-5. study. A mandatory second opinion program was with 696 errors meeting a study criteria of errors This article explained the framework of health care put in place, and it was found that a significant with the potential for adverse patient effects. quality defined as “underuse, overuse, and misuse,” portion of initial diagnoses reviewed within the and concluded that these problems exist in large study had discrepancies with the conclusions McDonald, C.J. “Protocol-based comput- and small health care organizations in all regions drawn through the second review. er reminders, the quality of care and the of the country. non-perfectibility of man.” N Eng J Med, Leape, L. L.; Bates, D. W.; Cullen, D. J.; 1976, 295:1351-55. Classen, D.C.; et al. “Adverse Drug Cooper, J.; Demonaco, H. J.; Gallivan, T.; Pioneering paper on the potential use of computer- Events in Hospitalized Patients.” JAMA, Hallisey, R.; Ives, J.; Laird, N.; Laffel, G.; ized reminders to aid medical decision-making. 1997, 277:301-6. Nemeskal, N.; Petersen, L. A.; Porter, K.; This study examined the excess length of stay, extra Servi, D.; Shea, B. F.; Small, S.; Sweitzer, McNally, K.M.; Page, M.A.; and costs, and mortality attributable to adverse drug B.; Thompson, B. T.; and Vander Vliet, Sunderland, V.B. “Failure mode and events (ADEs) in hospitalized patients. The authors M. “Systems analysis of adverse drug effects analysis in improving a drug found that ADEs are associated with a significantly events.” JAMA, 1995, 274:35-43. distribution system.” Am J Health-Syst prolonged length of stay, increased economic bur- The landmark article on systems ananlysis of Pharm, 1997, 54:171-177. den, and an almost two-fold increased risk of death. adverse drug events. The authors define the Description of use of such analysis in pharmacy underlying problems and identify the “proximal practice. Classen, D.C.; Pestotnik, S.L.; Evans, R.S.; causes” of medication errors. and Burke, J.P. “Computerized surveil- Melmon, K.L. “Preventable drug lance of adverse drug events in hospital reactions—causes and cures.” Sem Med patients.” JAMA, 1991, 266:2847-51. Beth Israel Hosp, 1971, 284:1361-7. Description of the continuous monitoring system One of the earliest papers on preventable adverse at LDS Hospital in Salt Lake City, Utah. This is an drug events. A classic. impressive system that integrates clinical, laborato- ry, drug, and cost information to identify ADEs both Thomas, E. J.; Studdart, D.M.; Newhouse, prospectively and retrospectively. J.P.; et al. “Costs of medical injuries in Utah and Colorado.” Inquiry, 1999, 36(3):255-64. The authors estimate the direct and indirect costs of adverse events. Extrapolation to all hospital admissions in the U.S. indicates national costs P REPARED BY LAUREL SIMMONS, INSTITUTE FOR HEALTHCARE IMPROVEMENT for preventable adverse events of $17 billion. FOR ADDITIONAL INFORMATION ON THE QUALITY IMPROVEMENT PROJECTS PROFILED IN THIS REPORT PLEASE CONTACT:

Back from the Brink: What’s in a Name? When it A Dummy Helps Pave the Way Making Chemotherapy Safer Comes to Drugs—Lots to Better Medicine The Dana-Farber Cancer Institute Bristol-Myers Squibb VA/Stanford Simulation Center James Conway, M.B.A. Frank Pasqualone David Gaba, M.D. Chief Operations Officer Vice President, Marketing Director, Patient Safety Center of Inquiry Dana-Farber Cancer Institute P.O. Box 4500 at VA Palo Alto Health Care System 44 Binney Street Princeton, NJ 08543 Professor of Anesthesia Boston, MA 02115 PHONE 609-897-2511 Stanford University School of Medicine PHONE 617-632-2158 FAX 609-897-6399 Patient Safety Center of Inquiry, 112A/PSC FAX 617-632-2161 EMAIL [email protected] VA Palo Alto HCS EMAIL [email protected] 3801 Miranda Avenue Palo Alto, CA 94304 Tackling Medication Errors PHONE 650-849-0421 A Government Health System Head On FAX 650-849-0285 Leads the Way Luther-Midelfort Hospital EMAIL [email protected] The VA Health System Roger Resar, M.D. James P. Bagian, M.D., P.E. Change Agent A Medical Specialty Director Luther-Midelfort—Mayo Health System Blazes a Trail VHA National Center for 1400 Bellinger Street Anesthesiology was first Patient Safety (10X) Eau Claire, WI 54703 in addressing medical errors 2215 Fuller Road PHONE 715-838-3566 Ann Arbor, MI 48105 FAX 715-838-3812 Ellison C. Pierce, Jr., M.D. PHONE 734-930-5890 EMAIL [email protected] Associate Clinical Professor FAX 734-930-5899 of Anaesthesia EMAIL [email protected] Harvard Medical School Making Mistakes 770 Boylston Street, Apartment 10C Harder to Make Boston, MA 02199-7709 Making Doctors Fairview Health Services PHONE 617-450-0242 Computer Literate FAX 617-450-0247 Steven Meisel, Pharm.D. Brigham and Women’s Hospital EMAIL [email protected] Assistant Director of Pharmacy Gilad Kuperman, M.D. Fairview Health Services Instructor in Medicine 6401 France Avenue South Harvard Medical School Edina, MN 55435 850 Boylston Street, Suite 202 PHONE 612-924-5894 Chestnut Hill, MA 02467 FAX 612-924-1511 PHONE 617-732-9069 EMAIL [email protected] FAX 617-731-3690 Caren Allivato, Pharm.D. David Bates, M.D., M.Sc. Clinical Pharmacy Specialist Division of General Internal Medicine Fairview Health Services Brigham and Women’s Hospital 6401 France Avenue South 75 Francis Street Edina, MN 55435 Boston, MA 02115 PHONE 612-924-5840 PHONE 617-732-7063 EMAIL [email protected] FAX 617-732-7072 EMAIL [email protected] The National Coalition on Health Care The Institute for Healthcare Improvement 1200 G Street NW, Suite 750 135 Francis Street Washington, DC 20005 Boston, MA 02215 PHONE 202-638-7151 PHONE 617-754-4800 FAX 202-638-7166 FAX 617-754-4848 EMAIL [email protected] EMAIL [email protected] WEBSITE www.nchc.org WEBSITE www.ihi.org