Tuesday, October 03, 2006

On Intel's and Walmart's "prescription" for Healthcare IT

A medical "prescription" from microprocessor manufacturer Intel's chairman appeared recently. Some who attended this talk have opined that the AP may have not reported his speech entirely accurately, but my feeling is that the gist is likely correct, as I've heard this type of industry prescription for "revolutionizing medicine through computing" (with design and implementation dominated by non-clinical business and IT personnel, of course) numerous times before. I attemped to contact Intel via its public relations office last week and they promised me a call to clarify the AP article, but I have not heard back from them.

With that caveat in mind, paraphrased from the article below, this latest industry view appears to be: "Healthcare is responsible for offshoring, so healthcare better computerize itself, which is easy since other industries like Wal-Mart did it; computers will solve all of healthcare's ills - and therefore save American industry." Wow...

Intel's Barrett blames broken health care system for U.S. job exodus

WASHINGTON (AP) - Intel Corp. chairman Craig Barrett warned Tuesday that U.S. jobs will continue to move offshore at a rapid pace unless corporate America exerts its power to force the health care industry to adopt systems that will cut costs and improve efficiency.

... ``Every job that can be moved out of the United States will be moved out ... because of health care costs,''

Employers should demand that hospitals select standardized record systems to lower costs or take their company's business elsewhere, Barrett told a crowd of 400 at a hotel ballroom.

``The (health care) system is out of control, it's unstable, it's basically bankrupt, it gets worse each year and all we do is tinker around the edges when what we need are major fixes,'' Barrett said, adding that health care costs make it increasingly difficult for U.S. companies to compete with rivals abroad.

The speech came on the same day that health care research group Kaiser Family Foundation reported that health care premiums rose at a 7.7 percent annual rate last year, double the pace of consumer inflation.

Barrett, who chairs the world's largest computer chip maker, said companies should only do business with health care providers who meet certain standards, including fully electronic patient records and published ``best practices'' for patient treatment. Hospital networks could and should be transformed into ``competitive centers for excellence'' that are paid to keep employees healthy.

... Barrett was joined on-stage by Wal-Mart Stores Inc. Executive VP Linda Dillman. He said the health care industry could learn from the efficiency of the retail giant, which tracks every item in inventory using computers.

``Every other industry has adopted this technology and (the health care) industry continues to sit here and debate,'' Barrett said.


The seeming naivete about the complexities of healthcare and the non-sequiturs in this "prescription" are of significant concern.

First, the most troubling issue is an accusatory tone in placing blame for offshoring and deficient US industrial competitiveness on healthcare, not on industry leadership's short-sightedness in maximizing profit and return to investors and speculators at the expense of American workers through socially-disruptive mass layoffs and exploitation of the Third World. This is not to mention other varieties of managerial incompetence, and yes, good old-fashioned corruption, Enron-style.

Also not mentioned regarding healthcare costs is the profit motive of "managed care" and the corruption in the healthcare industry itself as seen on Healthcare Renewal. These are problems computers are unable to solve and will not solve.

(If one is going to fling mud and accusations, I find the somewhat dated site of the Anti-Intel Association and stories such as this one - also see here and here - regarding an aggrieved Intel employee quite interesting.)

Next, Barrett's prescription brings to mind Donald Lindberg's admonition in 1969:

As far back as 1969, EMR and Medical Informatics pioneer Donald A. B. Lindberg, M.D., now Director of the U.S. National Library of Medicine at NIH, made the following observation. He wrote that "computer engineering experts per se have virtually no idea of the real problems of medical or even hospital practice, and furthermore have consistently underestimated the complexity of the problems…in no cases can [building appropriate clinical information systems] be done, simply because they have not been defined with the physician as the continuing major contributor and user of the information" (Lindberg DAB: Computer Failures and Successes, Southern Medical Bulletin 1969;57:18-21).

Here is Mr. Barrett's bio:

Craig R. Barrett became Chairman of the Board of Intel Corporation on May 18, 2005. He became Intel’s fourth President in May of 1997 and Chief Executive Officer in 1998. He was elected to Intel’s Board of Directors in 1992 and served as Chief Operating Officer from 1993 to 1997. Barrett began his tenure at Intel as a Technology Development manager in 1974. Prior to joining Intel, Dr. Barrett was an Associate Professor at Stanford University in the Department of Materials Science and Engineering.

This will look familiar to Healthcare Renewal readers. Mr. Barrett apparently understands healthcare from the perspective of a lay businessperson, not from a seasoned clinician's perspective.

It also brings to mind the lack of knowledge of such people of the severe difficulties even a state-controlled medical system has encountered trying to "computerize" its GP's:


Not to mention what all too often happens in this country in our own healthcare IT projects.

Barrett's prescription brings to mind my admonition that "clinical computing and business computing are two different specialties of computing":

... Today the IT personnel and non-medical managers (e.g., non-degreed IT staff, BS or MS in computer science, MBA's, even PhD's) who often hold leadership roles in EMR and clinical data research initiatives via control of critical decisions, budgets and resources, often have inadequate or nonexistent clinical experience and insight. Specifically, personnel of an information technology background, with little or no background in the biomedical sciences, often are empowered as enablers, rather than facilitators, of such initiatives. They retain a major say in what is -- and is not -- done, and in the tools provided to perform clinical care and biomedical R&D.

From a dual perspective as both a clinician and computer professional, it is evident that this arrangement is faulty, and that critical clinical computing projects benefit greatly from an alternate approach to project preparation, development, implementation, customization and evaluation as compared to management information systems (MIS) projects. Clinical computing and business computing are different, highly distinct subspecialties of computing.

IT personnel in hospitals often believe that success in implementing management information systems applications ("business computing") supersedes or actually renders unnecessary the mastery of medicine in leading and controlling implementation of clinical computing tools. Yet, mastery of applied IT towards implementing management information systems is in large part mastery of process (e.g., in acquiring and supporting vendor-written software) and repetition, as opposed to the practice of medicine, which requires mastery of complexity.

In other words, applied IT is a field of a relatively small number of principles, a large number of arbitrary conventions and rules, and a narrow body of knowledge applied repetitively and programmatically, often without scientific rigor. This may be illustrated by the fact that most areas of applied IT can be done well, and often are, by those with little or no formal training. This is not to imply that applied IT is itself easy, which it is not. There is no substitute for talent and real-world experience.

In clinical IT settings, however, there must be the right experience. Experts in clinical computing must provide effective solutions via seasoned application of the concepts, techniques, knowledge, and processes of medicine, and display an expert level of critical thinking in applying principles, theories, and concepts on a wide range of issues that are unique to clinical settings. Business IT experience alone does not provide a sufficient background for such responsibilities to be carried out effectively. Further, medicine is a domain of many difficult, nonintuitive principles, experimentally-derived natural laws, and a large body of knowledge applied in a broad, interconnected manner, ideally with critical scientific rigor. It cannot be practiced successfully without significant mastery of an enormous body of biomedical knowledge and significant hands-on patient care. The IT model of "If it's information, we do it" starts to fall apart and impede progress in such organizationally and sociologically-complex environments.

Finally, such industry admonitions bring to mind the fact that effectiveness of clinical IT in reducing costs has not yet been proven - it's still considered an experiment. Such admonitions also bring to mind the fact that the healthcare system has little discretionary capital for IT misadventures, as well as the dirty industry secret that many IT projects across all domains fall severely below their design goals or fail outright:

Research shows high failure rate on IT projects - information technology

A September survey by the Gartner Group, a technology consulting company, revealed that approximately 40% of information technology (IT) projects do not produce their intended results--an indication of how badly some companies need the consulting services of a tech-savvy CPA. Gartner's prediction that global IT spending will reach $3.3 trillion by 2002 highlighted the broader implications of this finding.


(Also see Statistics over IT Failure Rates - ed.)

With these issues in mind, such a prescription from the head of a microprocessor manufacturer and mass-merchandiser would be risible if it weren't so presumptuous and misleading.

This is not to mention that the prescription transfers many of those rare healthcare dollars into the IT industry. That is not simply a small conflict of interest.

Rather than trying to lay blame for U.S. industry's competitive problems and offshoring on the healthcare industry, and bludgeon healthcare and healthcare providers with presumptuous prescriptions, Intel and Wal-Mart should be investing money in improving understanding of how healthcare IT goes wrong, and developing best practices for design and implementation, in concert with medical societies, medical professionals and the informatics and IT vendor community.

Healthcare IT is not so simple. Put in terms Mr. Barrett might understand, just as the semiconductor industry has hit a wall of difficulty in moving CPU clock speeds beyond about 3 GHz due to limitations of engineering, physics, manufacturing, and our imperfect understanding of what goes on in submicron geometries at such high speeds, so there are difficult social and technical issues in healthcare that make computerization extremely challenging and certainly not a panacea as he rather naively suggests.

Semiconductor research is ongoing to overcome these difficulties, as is Medical Informatics research to better understand the difficulties in the healthcare IT world. Bludgeoning healthcare organizations into ill-conceived IT initiatives is not a useful solution.

Finally, as an aside, clinicians generally don't complain about anti-competitive or wasteful practices in the microprocessor industry.

I could complain that tens of millions of person-hours and billions of dollars were wasted dealing with the segmented memory architecture of the Intel 8086/8088 and the resultant fritter, workarounds and bugs, as opposed to other designs such as the Motorola 68000 with linear addressing. This is a design issue that should have disappeared with the likes of the PDP-8. See The Curse of Segments for more on this issue:


... The [Intel] 8086 was an
ill-conceived attempt to provide an address space larger 64K bytes [of the older 8 bit microprocessors] without actually incurring the costs of a larger architecture. The 8086 is basically a 16/16 architecture. It has a 16-bit program counter, a 16-bit ALU, four 16-bit general purpose registers, and some 16-bit index registers. It also has four 16-bit segment registers. The 8086 performs all computation and data transfer in 16-bit arithmetic, with one exception. Immediately before gating an address onto the external address bus, the 8086 selects one of the segment registers, shifts it 4 bits to the left, and adds the address to it, using 20 bit arithmetic. The external address is therefore 20 bits, and the processor has an address space of 1M byte.

... Under normal circumstances, a design so twisted and flawed as the 8086 would have simply been ignored by the market and faded away. However, 1980 was Intel's lucky year. IBM chose the 8086 as the processor for the PC. Backed by IBM's marketing might and name recognition, the IBM PC quickly captured the bulk of the market.

Anyone remember extended v. expanded memory? Programs like EMM386 that used extended memory to simulate expanded memory? The crashes and unreliability of Windows 2.0, 3.0, 3.1, 95, 98 and ME that tried to force parts of a large, complex operating system down the throat of segmented-architecture processors, in effect trying to make a silk purse out of a sow's ear? This is something we still haven't fully recovered from, and those who still use Windows ME or earlier may know what I'm referring to.

But I don't complain about this, because clinicians should not be telling the microprocessor industry how to run their businesses. Perhaps that industry could return the favor.


As I have written regarding IT, a person is either part of the solution or part of the problem. In my opinion, this kind of advice from large corporations is part of the problem.

(Note: I use both Intel-based and non-Intel based computers equally.)


No comments: