Statement of Herbert Pardes, M.D., President and Chief Executive Officer,
New York-Presbyterian Health Care System, New York, New York

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

Hearing on Legislation to Reduce Medical Errors

September 10, 2002

Summary

• New York-Presbyterian Hospital is one of the nation’s largest academic medical centers and is a center of excellence in the use of information technology (IT).
• IT in the clinical setting can reduce medical errors and increase quality of care through a wise national investment policy.  The Computer-based Patient Record (CPR) is at the center of a technology strategy that would reduce errors.  A complete CPR is impossible without standards for interoperability in healthcare IT.
• Hospitals are functioning at or below margin and are hard pressed to pursue the necessary investments to establish appropriate systems.
• Two major steps are needed: a standard setting process to allow interoperability among diverse systems from different vendors; and federal reimbursement.  Academic medical centers have the expertise and neutrality to lead this process in collaboration with industry and government.
• The interoperability provisions of H.R. 4889 advance the right approach.  They call for the formation of a board of experts from every effected constituency to recognize existing standards and develop or validate new standards.  However, another step is also needed.  Congress needs to allow for the demonstration projects that will assure the efficacy, usability and scalability of standards.  As technology is developed for millions of patients rather than thousands, or even tens of thousands, scalability becomes a major stumbling block.

New York Presbyterian Hospital

New York-Presbyterian Hospital (NYP) is the largest hospital system in New York State and one of the most comprehensive healthcare institutions in the world.  NYP represents a combination of two of the world’s greatest medical resources, New York Weill Cornell Medical Center and Columbia Presbyterian Medical Center with their academic affiliates, the Weill Medical College of Cornell University and Columbia University College of Physicians & Surgeons.  The NYP Health Care System (HCS) has a total of 51 acute care hospital facilities, residential health care facilities and specialty institutions, more than 13,000 affiliated doctors, over 40,000 employees, and annually over 400,000 inpatient discharges and 3.3 million outpatient visits providing healthcare to 5.5 million patients—22 percent of the greater New York metropolitan region.  The Hospital and its affiliated institutions and medical schools are engaged in a large number of medical research projects and receive about $300 million in research funding annually.

NYP is the home of the Columbia University Department of Medical Informatics, which traces it roots to a 1981 National Library of Medicine (NLM) Integrated Academic Information Management System (IAIMS) initiative. The planning and prototype phases of that initiative led to the creation of the Center for Medical Informatics in 1987. In 1994, Columbia University made the Center a full-fledged department in the health sciences campus, with the same rights and responsibilities as other departments such as Medicine and Surgery.

Since the beginning, the Department's focus has been on research, teaching, and service. A phase III IAIMS grant and an IBM contract funded the development of the next generation Clinical Information System (CIS).  CIS has served as the Department's living laboratory for medical informatics research, as a training ground for new informatics researchers, and as New York-Presbyterian Hospital's clinical system. The system is used by 95% of attending physicians and essentially all residents and fellows. It currently has 4000 unique users per month, and there are 2 million patients in the database.

Information Technology and Medical Errors

Computers have great promise in healthcare as tools for monitoring patient care to the most minute detail.  According to the IOM, "The majority of medical errors do not result from individual recklessness but from basic flaws in the way the health system is organized… illegible writing in medical records has resulted in administration of a drug for which the patient has a known allergy…And the health care system itself is evolving so quickly that it often lacks coordination. For example, when a patient is treated by several practitioners, they often do not have complete information about the medicines prescribed or the patient's illnesses."  In short, even highly competent physicians can be stymied by a system that makes timely access to accurate and legible information difficult.

Clinicians are very good at giving care.  Information systems are very good at process.  The errors described in the IOM report were seldom errors of judgment; they were mistakes in process. The man who had the wrong leg amputated; the woman who died from an overdose of chemotherapy; the child with the fatal allergy to anesthesia – if only the clinicians had been aware of that one missing vital piece of information, they would have known the right thing to do.  In many industries, computers track such crucial information.

If all pertinent patient information were made rapidly and seamlessly available to doctors when they needed it, clinicians who are trained to use their judgment based on the available facts would be in a better position to make optimal decisions.  Does the patient have an allergy to anesthesia?  The system would automatically alert the anesthesiologist if an inappropriate drug is prescribed.  On which breast does the mammography report identify a tumor?  The system automatically generates a diagram on the computer's display to verify the location.  Does the clinician administer an excessive dosage of medicine because she has misread the prescription?  The system compares the dosage with information about the patient's condition and alerts the clinician when the dosage is out of range.

And thanks to the World Wide Web, the benefits of this technology are not limited to large academic medical centers.  We have recently seen the introduction of technologies that can monitor patients at great distances over the Web.  From the home or in other settings, patient data are analyzed before being passed on to clinicians, alerting them to changes that need attention and storing information for future reference.

One branch of medical informatics, the science underlying the development of information systems to improve health care and other areas of biomedicine, is called automated medical decision support.  It has been studied for more than 20 years at leading institutions across the nation.  In fact, the standard computer language for generating computer-based medical warnings and reminders is the Arden Syntax, developed in part by researchers at Columbia University's Department of Medical Informatics located at the New York-Presbyterian Hospital.

Yet, developing an integrated technology that meets all the needs of American health care will require research and development support.  Transferring the technology to hospitals most in need will require additional funding support.  Leading health care institutions have the expertise to develop these Clinical Information Systems, but not the funding.

A wise national policy for healthcare Information Technology (IT) should target resources and expertise to increase the application of IT to reduce medical errors and improve health care quality.  The Institute of Medicine suggests that at least 44,000 patients, and as many as 98,000, die each year as a result of medical errors.  Data on medication errors show that 2.8 percent of all hospitalized patients experience a preventable adverse drug event, resulting in increased morbidity and mortality as well as significant added cost to the healthcare system. The cost of preventable medication errors alone is estimated at $2 billion annually.  Emerging technologies to reduce medication errors at the stage of both ordering and administration hold significant promise, particularly if they are able to coordinate all information from the patient’s medical record.  Technologies such as the Computer-based Patient Record (CPR) and secure Web-based communication with patients can enhance the coordination of care, support implementation of evidence-based practice and engage patients more fully as partners in their medical treatment.  It makes sense to focus efforts on the CPR, which lies at the center of a comprehensive hospital IT system, rather than on any single element of the system.  For example, a stand-alone Computerized Physician Order Entry system (CPOE) will not attain the Institute of Medicine's goal of a 50 percent reduction in preventable medical errors by 2005, according to the Stamford, Connecticut. -based Gartner Group's top ten list of IT issues confronting healthcare professionals. A CPOE system that is part of a CPR system linked to clinical alerts and decision support will be the means to secure that objective.

In fact, a broad-based national IT policy would address some of the most difficult problems in medicine in addition to preventing medical errors, including epidemiological data tracking to fight bioterrorism.  It would ensure that patients receive the best care no matter where they live or travel.  It would take advantage of the great global infrastructure—the Internet—and so be broadly available.  It would incentivise industry—insurers, systems manufacturers and software companies—to participate in standards development.

Barriers to Bringing IT to Hospitals

Health care lags behind almost every other major industry segment in investment in Information Technology.  Average IT spending per employee per year among all U.S. industries is about $6,900 per year.  The banking segment spends almost $15,000 per employee on IT.  The insurance industry spends more than $13,000 per employee and telecommunications clocks-in at more than $11,000 per employee.  Health care invests only about $3,000 per employee per year on IT. Another way of casting this is that other industries spend from 7-10% of revenues on IT; healthcare is below 3%.

Yet Clinical Information Services can vastly improve the quality of health care with long-term benefits for overall costs. And it would be centered on the concept of the computerized patient record (CPR).

A computerized patient record would carry a patient’s entire medical history and related information in a secure, privacy-protecting, Web-accessible database.  There are three barriers to implementation of a national health information system policy.

• There is no standard platform technology or terminology for the CPR as there is for business and other applications;

• There is insufficient investment and financial incentive, by the government and the private sector; and

• There is insufficient leadership

The issue of standards, in particular, has been a huge barrier to the wise and widespread use of technology to prevent medical errors.  In the vast majority of healthcare settings, a great deal of disparate information is collected on a patient but that information is not presented to a clinician on a single computer screen.  Thus, the data is only coordinated for the care of that patient when a clinician reviews it, usually when it is brought together on paper.  If a result doesn’t make it to the paper record at the right time, the clinician will not be aware of it when she makes care decisions.  Thus, there needs to be one computerized view of this information.  This requires a large investment in elaborate interface technologies because different vendor systems can’t automatically share all of the information with each other.  Standards would obviate that need. Without standards, hospitals can never be sure of their investment in new technology. With standards, hospitals will be encouraged to invest in IT, knowing they will have a stable platform.

The solutions to these problems are within grasp.  The Federal Government has a clearly definable role in creating solutions that will deliver to Americans the error-reduction promise of the CPR without being intrusive to care providers or the industries that support them.

These things must be done for an effective National Health Information System Policy to be accepted by health care providers, insurers and CIS manufacturers.

• Before standards that support interoperability can be promulgated or driven by market demand, they must be developed.  There is room for much flexibility in such formats, as described in the IOM study For the Record.  Standard setting efforts have been successful in creating protocols for transmitting health data.  One example is Health Level 7 (HL-7), a protocol for transmitting health information that provides a solution to one of the many issues in interoperability of healthcare IT.  HL-7 was supported by academic medical institutions for years in the face of industry skepticism.  As more institutions adopted HL-7, the industry realized the advantage of being able to design to a standard and embraced it most profitably.  The most important thing is to be able to securely share patient information among different caregivers and researchers.  The Government should support research and development for standards by credible institutions through a demonstration process led by centers of excellence in academic medicine, with full collaboration by industry.
• CPR standards must aid hospitals in complying with HIPAA, the Health Information Portability and Accountability Act.  HIPAA rightly mandates extreme care of confidential patient data.  But its enactment has been stymied by a hodgepodge of competing interests and its implementation promises to be one of the largest unfunded mandates ever levied on the health care industry.  A wise national health information system policy will create standards for easier, more affordable compliance.
• Medicare and Medicaid should include incentives for hospitals to comply with the standards of the National Health Information System Policy, and provide disincentives for noncompliance.  It’s cheaper to process an electronic transaction than a paper one—that’s why banks used to reward customers to use ATMs instead of tellers.  The Center for Medicare and Medicaid Services (CMS) has, in the past, viewed technology as a cost-center rather than an investment.  Yet a CPR can reduce difficulty in a variety of reporting and compliance areas.  Medicare can reap the same benefits from technology that private sector industries have, increasing productivity and quality through the strategic use of IT. The Information Technology Association of America projects that an industry-wide investment in IT of $18.1 billion would yield gross savings of greater than $120 billion dollars for the healthcare industry over a six-year period.  But unlike other industries, healthcare providers do not have the profit margins to invest for the future in this area.
• The Federal Government should provide funding to help hospitals invest in IT.  Hospitals and academic medical centers are at the center of an integrated health IT system and will be among the greatest users of the technology.  But they cannot alone shoulder the cost of developing a system that must be able to serve patients, doctors, government agencies and insurers.  Clearly there is justification for a national investment in the development of such a system.  In fact, a nationally functional system will only be built collaboratively.  The system will be potentially useful for those areas of increasing shortage in health care providers—rural areas and the inner city.  Support is given in a variety of underserved locales for aspects of care.

The Patient Safety Improvement Act Of 2002

H.R. 4889, The Patient Safety Improvement Act Of 2002, contains provisions on interoperability of healthcare IT that will allow information technology to fulfill its promise in the area of patient safety.  I want to acknowledge the extraordinary work of several Members of the House in the area of using IT to protect patient health.  First Chairwoman Johnson, who by introducing this bill has advanced the cause of safety to a new level.  This bill is a quantum leap in solving the medical errors problem with tools that have not been used effectively in the past and needed someone of vision to champion it.  Mr. Houghton of New York and Ms. Thurman of Florida, your bill H.R. 3292, the Medication Errors Reduction Act of 2001, paved the way for technology to be applied to patient safety.  Your leadership has been sine qua non in this field.  Mr. Rangel of New York has long understood that a new solution based on technology must be applied to healthcare for us to solve the old problems that have bedeviled us.  Thank you also for your leadership and support of the interoperability provisions.

How does H.R. 4889 actualize the potential of IT in healthcare?  The bill establishes an inclusive process that calls upon the expertise of industry, academic medicine, community medicine, public health and government.  The formation of the Medical Information Technology Advisory Board (MITAB) is necessary so that interoperability solutions can be found that are applicable in the greatest number of healthcare settings, whether government or private.  Those who use IT and those who manufacture it must be aligned from the beginning to design the right solutions.  We have seen too often in the past how good technology fails because what has been designed for one environment does not translate to another.  For example, excellent systems for administrative data exchange do not translate well for clinical care.  The MITAB section of the bill would bring together those who produce healthcare data, those who consume it and those who make the media of storage and exchange around a single effort.

However, there is a step that is at least as important, which has yet to be included in this legislation.  H.R. 4889 needs to recognize that what happens at a theoretical level may not translate well into an operational environment, particularly in a field as complex as healthcare.  For the vision of H.R. 4889 to be realized, it is critical that demonstration projects to test the efficacy, usability and scalability of IT standards be conducted.  Scalability is a key determinant of success.  A technology that appears to be a great idea in the inventor’s garage, and works beautifully in a dozen physician offices, can fail when it has to care for millions of patients.  Medicare and Medicaid’s patients are found in every setting in America, large inner cities, sparse rural counties, suburban tracts and local neighborhoods.  So standards must be tested in a diversity of settings to ensure they will work wherever these patients, and all Americans, receive their care.

Congress has an opportunity rarely presented in history.  It can instigate a paradigm shift by seizing today’s revolutionary technology, and it can save lives by putting it in the hands of the nation’s healers.  It is time to convene the best minds in this field and develop the standards for technology that will be indispensable to our future health.  Academic medical centers can collectively take the lead in developing the technology for patient and clinician needs, and billing and compliance requirements.  But they must be part of a vigorous partnership with the government, physician organizations, insurers and corporations—all of which will benefit from a true national IT system.