Statement of Donald M. Berwick, M.D., President and Chief Executive Officer, 
Institute for Healthcare Improvement, Boston, Massachusetts

Testimony Before the Subcommittee on Health
of the House Committee on Ways and Means

Hearing on Health Quality and Medical Errors

March 7, 2002

Since the IOM Report, To Err Is Human, released in November, 1999, the nation has made a great deal of progress in confronting the burden of injury from health care.  We have made the issue explicit and visible. I think most consumers are now aware of the problem, which is a step toward building will for change.  Some hospitals have promised to improve patient safety. A small, but significant, minority of hospitals have begun to look for injuries and to try to measure their frequency.  Skillful experts, including some from other industries, have started to help us. 

Let me give you just a few highlights:

• A consortium of all stakeholders in the Pittsburgh area has publicly announced that they will reduce medication injuries and post-operative staph infections to zero.  Similar consortia in Whatcom County, Washington, and in Florence, South Carolina, have made plans to improve medication safety dramatically.

• The Children’s Hospital Corporation of America, involving several dozen children’s hospitals, has chosen patient safety as its major care improvement agenda.

• Fifty-seven hospitals from all over the US in a collaborative called “Quantum Leaps” are now openly measuring their medication injury rates, comparing them, learning from each other, and aiming for ten-fold reductions.

• A newly launched medical journal called “Quality and Safety in Health Care” is providing researchers and clinicians in the US and the UK with an outlet to publish their work.

• The National Academy of Engineering and the Institute of Medicine have launched a project to see how new engineering approaches can be applied rapidly and directly to safety in health care settings.

• The VA and NASA have begun a fascinating collaboration on reporting systems and on human factors engineering.

The Federal government has been in the lead in this sea change.  I think you already know that the Veterans Health Administration, the Department of Defense, the Health Services and Resources Administration, and, above all, the Agency for Health Care Research and Quality have invested heavily in new care initiatives, technologies, and research to make patients safer.  The Quality Interagency Coordinating Task Force (QuIC), which has bridged two administrations, is a wonderful example of cooperative learning and action among agencies that too often in the past have not worked closely together on common issues.  We would be nowhere near as far along as we are without this Federal leadership, and I commend both Congress and the Executive Branch for your commitment and bipartisan constancy on this issue.

Along with this important progress, we have also now become aware of some obstacles.  None of them are insurmountable, and in some cases, the Federal government can clearly help to accelerate change.  Here are a few of the problems that have surfaced:

• First, persistent denial. Many clinicians and organization still seem to think, despite the evidence, that the problem of safety is not as large as we know it to be.  This denial is fading fast, and I think we are well into a new era of recognition and aims for improvement.  It remains important for Congress, Federal agencies, and other leaders to keep the pressure on, and to insist on improvements.

• Second, capital costs.  Health care systems are facing – or at least say they are facing – investment barriers to adopting some important technologies that make care safer, most importantly computerized physician order entry – CPOE.  Correctly adopted, CPOE can reduce medication errors by as much as 80%, but it requires a level of computerization that the health care industry largely lacks.  The technologies cost hospitals a lot up front – millions of dollars -- $7.5 million on average according to a recent estimate, and some hospitals say they lack sufficient sources of capital.  CPOE in the long run saves money; one study said it reduces hospital costs 13%.  The American Hospital Association has found that hospitals with good Information Technology support generally cost less and have better outcomes than others.  The barrier is getting from here to there; it’s a transitional problem.

• Third, paying for defects, instead of for safety and quality.  Our system of financing still often pays well for poor care.  A hospital that reduces patient injuries can today experience a significant decline in revenue, since it would often otherwise get paid for managing the consequences of those injuries.  This is not a stopper for ethical hospitals, and most are ethical, but it is a problem.

• Fourth, lack of standardization.  Facing high capital costs, health care organizations are nervous about the lack of standards for coding and procedures, fearing that they could invest today in systems that could rapidly become outmoded or incompatible in the future.  Lack of standardization is an impediment to progress, and I see no force other than the Federal government able to create such standards rapidly and at a national scale.

• Fifth, and perhaps most important, fear.  For many reasons – threats of lawsuits, embarrassment, loss of market share, and others – organizations and individuals in health care are frightened to discover, reveal, and openly address defects in their care.  This fear is incompatible with learning and with safety, itself.  Safe systems are open systems, and a frightened workforce cannot properly address safety improvement. 

Along the way, we have all been a little distracted by the very contentious issue of public reporting of patient injuries.   It is contentious because, on the one hand, it seems only right that the health care industry should be disclosing its performance to the people who depend on it and pay for it.  I hope and believe that we are emerging overall from the era of secrecy about performance of care systems, and the more recent IOM report, Crossing the Quality Chasm, called unequivocally for a whole new level of commitment to transparency, not just to inform consumers, but to allow health care systems, themselves, a better chance to learn from each other.  Safety is important, and it is illogical to exempt it from the rule of transparency.

On the other hand, we do know that people in health care are running scared, and that a frightened workforce hides its defects instead of learning from them.

And so, we have gotten a little stuck since the IOM safety report.  A few courageous health care organizations have just gone ahead and become open about measuring safety, and, I must say, they are none the worse for it.  Most, however, are still pretty timid about it.  They fight disclosure, and they fight CMS when they propose that safety should be openly measured and discussed.

In my opinion, it is high time to leap over our shadow on this one.  I simply do not believe that a risky, complex, stressed industry will have the will or the knowledge it needs to move beyond traditional assumptions about achievable performance unless and until it faces facts and data on its own work.  Safety should be a topic openly discussed, openly assessed, and openly explored, and I hope that CMS, like other important purchasers, will do what it can to assure that that open dialogue becomes more and more widespread.  CMS should insist that the health care organizations it pays must assess, study, and learn about their own patient injuries, disclose those injuries to the patients who are harmed, and continually and demonstrably reduce the risks of injury.

Congress can help this maturation to occur by a few, simple, persistent steps:

1.  Continue to support the investments of the VA, HRSA, DoD, and, most crucially, AHRQ, in the agenda of learning and improvement of patient safety, for the benefit of their patients, and for the instruction of all.

I am deeply concerned about this year’s proposed reductions in the budget of AHRQ.  It would be wise to expand, not to cut, our nation’s meager investment in studying how our $1.5 trillion care system can be made better continually.  Preserving line items for patient safety research is helpful, but not at all sufficient.  Please understand that the proposed AHRQ budget would bring nearly to a standstill new investigator-initiated health services research proposals – bring them to zero – and that means a sudden slowing down of research and investment which ultimately has a major impact on the well-being of our patients.   If you give AHRQ the funds it needs to support investigator-initiated research, even more than in the past, you can count on a high rate of return in health care quality improvement.

2.  Ask CMS to sponsor and allow several market-area experiments to reward quality and safety.  Try to adjust payment streams so that health care organizations that become safer thereby become more viable.  Right now, a hospital or health system that reduces injuries to patients often actually loses financially, because it gets paid for defects.  We have to figure out how to stop paying for defects in care, and to start putting exactly the opposite incentive to work. Pay hospitals more when they reduce their injury rates, and less when they don’t.  Grants, tax credits, or low-interest loans, as some currently proposed legislation would offer, may help some of the less wealthy hospitals move faster into modern information systems.

3.  Create the circumstance in which at least one state or area can test a no-fault malpractice liability system for a few years, so that we can begin to put to rest the most commonly cited obstacle to openness.  Either we will learn that that leads to much more openness, or that it doesn’t, and, either way, we gain knowledge we badly need.  The system we most badly need to test would have the following components:  (a) always letting patients and families know when a patient is injured by care (“extreme honesty”); (b) apologizing; (c) compensating victims of injury fairly and promptly; (d) bearing this liability at the “entity” level (hospital, health care system), not at the personal level (physician, nurse); (e) learning from events, and continually reducing risks within and among organizations; (f) dealing differently and promptly with the small class of criminal and grossly negligent events.

4.  Create some limited privilege for reporting on patient injuries for the individuals who make those reports.  I am not asking for secrecy at the entity level, but rather for some protection for a doctor or nurse who sees something go wrong, or actually falls into some pattern of error, and who can and will talk about it, but only if that does not come back to hurt them. 

5.  Do not worry about constructing some massive national database on errors or patient injuries.  We don’t need it.  It may help if AHRQ has a research database of that type, but regional and statewide reporting systems will be quite enough if we also have ways for the people involved in those systems to meet and talk with each other and to share what they are learning.

6.  Ask CMS to adopt some information technology standards, required of those organizations they pay, for coding of messaging, medications, laboratory tests, and diagnoses.  It is important, and feasible, for CMS to select a single set of such standards from among several good available options, and then to build confidence in the health care system that these will remain stable for a considerable time, so that it becomes prudent and logical to invest in compatible data systems with confidence that they will work well into the future.  That will make CPOE and other safety-enhancing computer technologies much more attractive and feasible.  I also think CMS should join the Leapfrog Group in informing hospitals that, by some deadline, a hospital without either CPOE or a demonstrably better method for medication safety cannot be a Medicaid provider.

We really have a running start now on making health care safer.  We need to keep the heat on the topic, invest in the research to know how to improve the situation, insist on openness and sharing, pay more for quality than for defects, and make it both important and safe for the wonderful people who work in health care to learn about their own errors, about the harm done to patients, and about how to reduce that harm continually.