A physician's prescription for healthcare computing leadership in the 21st century


Scot Silverstein, MD

Site home | Healthcare IT failures | Reader opinions | Medical Informatics | Nursing Informatics | Taxonomy of skills | Informatics role | Critical job recommendations | Clinical vs. Management computing | Risks in wrong hands | Mistaking MIS for CS | Unqualified Decisionmaking | Consumer health | Hi-Fi Computing | Testing the MIS cartel | Doctor's prescription for MIS | Advocacy | Tactics and stereotypes | Related writings | Interesting links



A physician's prescription for healthcare computing leadership in the 21st century

I will use the Y2K problem to illustrate the issues concerning healthcare information technology leadership by non-medical personnel. Some suggestions then follow for a more logical and optimized MIS role in healthcare, working in partnership with medical informaticists on clinical computing initiatives.

The so-called Y2K problem is a societal shortcoming that makes the points of this web site stand out strongly. In healthcare, it makes deficiencies in leadership due to insufficient education, skills, and insights stand out like a grade five heart murmur in a newborn. Imagine emergency room physicians putting in cheap, short-life batteries in their defibrillators that become unusable after a few years, surgeons who put in sutures that give way after a period of time, or engineers who design cars that simply stop running on the middle of the highway at 60,000 miles.

The Y2K problem has been touted as a 'technical' problem. Local companies in the Philadelphia region alone have spent near one billion dollars ($1,000,000,000) on Y2K remediation, not counting systems where Y2K was incidentally fixed during other upgrades, in the past several years (Philadelphia Inquirer, Sunday Business, "Local companies have spent nearly a billion on Y2K", May 30, 1999). Rarely mentioned is that sadly, for this billion dollars and for similar expenditures across the world there are no new hospitals, no new inventions, no new technology. Far worse, in a recent report, a U.S. Senate special committee advised that "a potentially monstrous wave of litigation may occur from people who suffer from business and government failures related to the problem. Projections of liability costs go as high as $1 trillion." (Philadelphia Inquirer, February 28, 1999, "Senate panel assesses Y2K risk to the economy".)

In reality, programmers were only doing what they were told by management: "be fast, cost-effective, get the product out the door sooner, rather than later, worry about today, not tomorrow." The origins of the Y2K problem are more a management problem than a technical problem. If the leaders had told the programmers to make sure their code was robust, lasting, and truly space-efficient, there would have been no Y2K problem.

At best, the leadership did not do so since most often the top leaders did not understand even basic concepts of computer and information science, programming, or interdependencies in information systems. This led to a trickle-down effect on people at lower management levels who may have known better. Soft, apologetic excuses and political correctness aside, I might add that the degree of folly behind the Y2K problem might not be easily captured in English, a relatively non-colorful language.

To compound the problem, it has been management that delayed and postponed repair of the Y2K problem in the interest of short-term gain, pathognomonic of deeply myopic thinking. Programmers and the technically-inclined have been warning about this looming problem for many years. Healthcare is particularly behind in Y2K remediation, directly due to these management shortcomings.

Instead of making changes in the way information technology is overseen or managed, the U.S. government is trying to limit liabilities for the Y2K debacle. According to the Associated Press, the U.S. Senate approved legislation on June 15, 1999 that supporters said could save the economy from being crushed by lawsuits against companies in connection with Y2K computer problems ("Senate approves lid on Y2K suits", Philadelphia Inquirer, June 16, 1999). Such legislation, while understandable, does not go far enough in preventing future widespread problems such as Y2K. It might even encourage future shortsighted, reckless information technology practices. The White House is opposed to this legislation in the interest of consumer rights, and the outcome remains to be seen.

In any case, this author believes the Y2K issue to be one of the biggest problems ever to occur as a result of faulty beliefs in a management system where true expertise in the domain area is considered unnecessary. Some people and organizations have stated that executives who opposed or ignored the advice of computer experts on spending maximum effort at the earliest possible time on Y2K remediation should face civil and criminal penalties for negligence, if Y2K-related failures cause significant disruptions in vital services. It is becoming apparent that high technology has become too critical in a technology-dependent society to allow suboptimal management decisions to continue unchallenged. Computers amplify the human mind but do not improve or replace it. It is apparent that computers can amplify human stupidity as much as human intellect. Muddy thinker plus computer does not equal scholar.

Where will the next Y2K-like problem due to similar causes arise? Areas such as air traffic control, genetic engineering, weapons production, and nuclear energy come to mind as possibilities. Fixing the Y2K problem has not been the biggest management challenge of the century as some have stated. Rather, fixing the "paint-yourself-into-a-corner" myopic thinking and mismanagement that created the Y2K problem in the first place has been the biggest management challenge, and so far, addressing that challenge has not yet even begun.

Ironically, a $1 trillion cost of Y2K problems if ever realized might put a big dent in the ROI (return on investment) some organizations have realized from computer automation!

Concerning proper healthcare leadership by true experts, the Mayo clinic, whose doctrine on the correct way to practice medicine ("The best interest of the patient is the only interest to be considered"), has a cohesive and participative environment that was built on philanthropic ideals and that was dedicated to the belief that medicine should benefit all of society. Physicians there are expected to bring their individual talents to bear on the collective treatment of patients. One current trend in the health-care industry involves turning hospitals over to professional administrators; Mayo is governed by physician-led committees and is prospering. Indeed, the bylaws of the Mayo Foundation require its president and CEO to be a physician ("The Agenda - Total Teamwork", Paul Roberts, Fast Company magazine, Apr. 1999). There is perhaps an important lesson to be learned from this.

MIS personnel are predominantly business people, with education and experience whose focus is business. They generally lack grounding in science, biomedicine, medical ethics, law, and other important medical areas in which clinician informaticists are well-versed or experts. It is a severe example of arrogant presumption to believe medical patient encounters are just "business transactions", with informational needs that fall under the leadership of business people. Such a leadership structure is neither very reasonable nor prudent. It is unlikely to lead to greater success and satisfaction for patients, clinicians, or MIS personnel themselves. Such a strategy stems from muddled thinking and habits of the past. As Bob Lewis astutely observes in his book IS Survival Guide, this problem is in part due to management and TQM consultants having successfully sold the false idea that "everything in the known universe works like a factory." This thinking has "contaminated" many disciplines, including clinical medicine. (Unfortunately, the reality is that physicians are not by nature commodities brokers, and clinics are not meat-processing plants.)

In an advertisement in the Wall Street Journal (March 18, 1999), consultant Computer Associates International presents a Dilbert comic strip poking fun at the obfuscation of "business speak" along with the observation that nobody plays the acronym game better than MIS...but all those buzzwords and fancy technical terms are very counterproductive... we always tell it like it is...without the B.S.

Protecting patients is the theme of an increasingly visible advocacy effort, led by such organizations as the American Medical Association (American Medical News, "AMA advocacy efforts protect patients", April 5, 1999). It is also the theme of this Web site. In that tradition, and as a physician used to advising and treating people for bad habits and behaviors that cause poor health by "telling it like it is", I offer the following prescription for improvement of healthcare IT:

I believe that breakthrough levels of "casualty reduction" in healthcare information technology are possible, given the right leadership. This breakthrough thinking is unlikely to come from the "school of hard knocks", as several previously-quoted healthcare IT management recruiters apparently believe.

Collaboration must replace competition. The behaviors of political dysfunction in healthcare institutions must become unacceptable. These institutions are of vital importance to the community, and people working in them must not forget that or put their own interests ahead of the community's. The recent, unprecedented $1.5 billion dollar bankruptcy of the Allegheny Healthcare System in Philadelphia, one of the largest healthcare bankruptcies in the history of the United States, is good evidence that such a prescription is good medicine.

MIS must evolve and grow from a mainframe mentality. Distributed computing will be the norm in the 21st century. The now-crippling 'control mentality' that sprang from the mainframe era must be abandoned. Information technology of different types must be placed in the hands of the appropriate leadership.

Perhaps MIS should heed Bill Gates, CEO of Microsoft. As he recently wrote, "The old saying knowledge is power sometimes makes people hoard knowledge. They believe that hoarding knowledge makes them indispensable. However, power comes not from knowledge kept but from knowledge shared." (Business @ the Speed of Thought: Using a Digital Nervous System, Warner Books, 1999.)

On the need for managed computing: "Managed care" arose in healthcare supposedly due to inefficiency, waste, failed outcomes, overinflated prices, and lack of regulation. Whether these factors existed or not, the idea was that efficiency and outcomes could be improved through quality control and oversight. Clearly, a situation of inefficiency, waste, and unnecessary failed outcomes exists in healthcare information technology. A system of managed computing may be an answer to these problems.

Healthcare informatics personnel need to assume leadership roles in this process of managed computing, applying the best qualities of the clinician, such as rigor, intelligence, dedication, integrity, results orientation, personnel performance and skills evaluation, and metrics to the problems of healthcare IT. Information techology in healthcare settings needs to function with the same attention to detail as clinical medicine (not the other way around).

Healthcare IT seems to be a largely unregulated industry in terms of qualifications and certification, exactly opposite to medicine. (Imagine if anyone could practice medicine, if they were able to get enough "experience" through reading books or attending a varying number of unregulated training classes and "on the job" training.) Further, there are no standardized metrics of performance in clinical computing as there are in delivery of medical care, such as physician report cards showing morbidity and mortality rates.

At the beginning of the 20th century, medical education and practice was largely unregulated. In 1910 the Carnegie Foundation published the Flexner report which depicted a lack of admission and curriculum standards in medical education and accreditation, and poor quality facilities among medical education institutions. Flexner's report stimulated system-wide reform in medical education with programs integrating a scientific base and government implementing rigorous testing and certification boards to protect the public.

In January, 1995, the Standish Group reported that 52% of IT projects overrun their initial budget estimates by 189%, and 31% are canceled before completion. A report published in Computerworld (June, 1998) stated that 63% of large IT projects are delivered late and 58% are brought in over budget. Even more damning is the report in The Wall Street Journal (April 30, 1998) which states, "42% of corporate information-technology projects were abandoned before completion." In medical terms, the mortality rate of IT projects is too high. A sobering thought when one considers that each year in the USA, information technology projects are launched at a cost of $250 billion. (From a report by Gopal K. Kapur, Center for Project Management, Palm Desert, Calif., gkapur@center4pm.com)

A story in the newspaper USA Today, "When Computers Fail" (Dec. 7, 1999), puts the cost of IT failures at $100 billion per year in the United States. Significant, expensive IT implementation failures at varied organizations such as Nasdaq, Hershey Foods, Whirlpool, state agencies, and the Pentagon are cited as just a few examples.

Notes one technology consultant in this story, "There's this mystique that these systems are managed and built by the most qualified pros on the planet. But most [I.S. personnel] are just regular Joes who arenít measured for any standards about what they know. If you can program your Nintendo and can do a good selling job, youíre a hot commodity." Even NASA is experiencing related problems. A November, 1999 report on the loss of the Mars Climate Orbiter, caused by a failure of computer personnel to convert orbital data in English units to the metric system, is critical of "untrained employees and a lack of oversight" (USA Today, Dec. 8, 1999, p. 4A).

Such a lack of standards on education and accreditation for healthcare IT workers contributes to healthcare IT's morbidity and mortality, and is likely to contribute to patient morbidity and mortality as well. At the same time, a milestone Nov. 1999 report "To Err is Human: Building A Safer Health System" from the Institute of Medicine (of the U.S. National Academy of Sciences) describes almost 100,000 deaths each year as a result of preventable medical errors. The report indicates that such errors are often caused by information-related problems (such as illegibility and unavailable or inaccurate records) at the point of care. Corrective measures at a national level are being recommended. Success and quality in healthcare IT are imperatives in correcting this problem.

Considering the critical nature of healthcare, standards of accreditation for those who want to work in clinical computing settings, and metrics for performance, will be of significant benefit. Healthcare accrediting agencies such as JCAHO (Joint Commission on Accreditation of Healthcare Organizations), American Medical Association, etc. should consider requirements for healthcare IT workers involved in clinical computing, and perhaps move towards basing accreditation of hospitals on the presence and proper roles (i.e., authority) of medical informatics-trained clinical personnel.

Healthcare would be wise to adopt a "focused-function" structure that is already in place in other sectors, such as pharmaceuticals and academia. Departments of management computing must co-exist and share authority with specialized computing departments such as bioinformatics computing, scientific & statistical computing, and in healthcare, clinical computing, led by experts in the specialty area. Perhaps in hospitals there should be discrete MIS (management information services), CIS (clinical information services), and HNS (hardware/networking services) sections. HNS would be a resource for the management and clinical information sections.

A CMIO or Chief Management Information Officer should have authority and freedom to manage (that is, lead and administer) Management Computing without undue interference from non-technological senior managers. A CCIO or Chief Clinical Information Officer position must have the same latitude. They would need to collaborate on mutually-relevant issues and report to a top executive. A combined CIO position in healthcare can exist if the holder has expertise in both management computing and clinical medicine.

Such CIO's exist, some being MD's, but are uncommon currently. There is an excellent article about MD/CIO's and the University of Utah's efforts to train informaticists for this role in the April 1999 Health Data Management ("A Physician as the CIO", p. 40). Pierre Pincetl, MD, Informaticist and CIO at University of Utah Hospital and Clinics, believes it's appropriate, advantageous, and highly important for physicians to become CIO's in the future.

Another excellent example of a physician CIO is Mitchell Morris, M.D. at the University of Texas M. D. Anderson Cancer Center in Houston, Texas. Dr. Morris has graciously provided this information: "Since Sept. 1997, Dr. Morris has overseen M. D. Andersonís information services, including telecommunication systems, Internet servers and web sites, and all institutional computer systems, networks and applications. He is also responsible for implementing the comprehensive cancer centerís innovative care management program, consisting of practice guidelines, collaborative care paths, quality improvement, outcomes assessment, and case management. In addition, he has led efforts to develop a computer-based patient record and data warehouse of patient information. In addition to his duties as Vice President for Information Services and Healthcare Systems, Dr. Morris is a surgeon, clinician, and Professor of Gynecologic Oncology. Dr. Morris was recently appointed to the chair of the Department of Health Services Research (ad interim) at M.D. Anderson and is Adjunct Professor of Informatics at the University of Texas Health Science Center. He maintains an active research program in gynecologic cancer and outcomes and has authored over 100 scientific publications."

MIS personnel should take more time to learn about medicine and its nuances and idiosyncrasies. Medical Informaticists have returned to school to obtain cross-disciplinary skills. As one reader pointed out, MIS personnel do not seem to have put in a comparable effort to obtain expertise in clinical medicine. Spending some time as volunteers in the E.D. or patient units of their own hospitals, taking CPR/Basic Life Support courses, attending courses on healthcare policy or medical science at a local university, and many other avenues are open to MIS personnel wishing to become more familiar with clinical medicine.

In Summary

Healthcare is a vital national resource. Information technology is becoming increasingly critical to healthcare, and will become even more critical in the future. Proper management of these resources by people with the proper expertise is an imperative. The optimal leadership structures must be identified and implemented. Political battles due to "control neuroses", self-interest and other dysfunctional motivations must end. Collaboration must replace competition.

Healthcare information technology is too critical and fragile a resource to risk being mismanaged by those who do not have the organization's best interests in mind, or who do not properly understand the culture, complexities, interdependencies, and Achilles' heels of medicine and clinical information systems.


Next: Advocacy in Medical Informatics as a means to progress.


Please send me Comments and related material