virtualgryphon.com

A colleague sent me a pointer to an article on drbicuspid.com (Electronic health records: Part I — Boon or boondoggle?) and asked for my view.  The article reports — generally positively — on how a a proponent of EMR/EHR is trying to influence the Obama administration to make faster progress on a combined medical/dental EMR/EHR.

The proponent, Valerie Powell, PhD, is reported to have said “Our nation needs excellent chronic care, and to make this happen means ensuring that the dentists and physicians can communicate. People are dying. This is urgent now.”

There is, of course, nothing new in this extreme sort of rhetorical flourish. But the exchanges between real practitioners and the proponent that follow the article are, well, testy. Dentists in practice — at least those who left comments — are not especially in favor of EMR/EHRs. They do not believe that IT will necessarily lead to safer care and find the arguments of the proponent — someone with only a peripheral connection to dentistry — unconvincing (an excerpt is included below).

One thing is clear, though: the proponents of IT see a grand opportunity to advance their cause. The Obama administration has embraced the idea of healthcare IT as a key element in its goal to get safer, better, cheaper healthcare. The healthcare IT program has been stalled in the U.S. and the U.K. for some time — for a variety of reasons — and the commercial and academic interest groups are eager to get what they consider to be their share of the Federal spending pie.

The claims made for EMR/EHR are in the realm of technofantasies, about which more later. But here are some of the comments I made to my colleague regarding the DrBicuspid articles.

1) The proposition that a comprehensive IT solution is essential to safe care is almost entirely driven by IT-affiliated people.

2) Few of the proponents know much about caring for patients.

3) Very few of the promised benefits of IT implementation in hospitals have been obtained. The cost of the IT systems now in place is much greater (usually acknowledged to be by a factor of 2 or more but often much, much more) than was expected and the systems themselves are far less comprehensive than was promised.

4) The presence of elaborate IT in healthcare has had a substantial negative effect on efficiency. Medical clinics in one hospital, for example, have had a 15 to 20% reduction in the number of patients that can be seen in a single day because of the burden of interacting with IT. This is not an isolated finding.

5) EMR privacy is a big problem. Privacy issues are deeply entangled with the use of IT in ways that generate new problems.  Although privacy problems may well come to be the predominate issue in the use of IT for healthcare, it is likely that the degree of commitment to IT-based approaches will be so advanced by then that there will be no practical way to restore privacy to healthcare. It is highly likely that enhanced privacy will be a feature of the most expensive, private treatment plans and that privacy will become a key differentiator between what the rich and poor get.

6) Weak counterfactual reasoning marks a weak case. The author claim that a person died because we have not yet implemented a comprehensive system is itself strong evidence that the proposal does not have strong merit.The resort to weak counterfactual reasoning (x happened because not y ==> if y then x) marks the absence of strong data in favor of the proposal.

7) The claims made are techno-fantasies. Techno-fantasy started with the Greeks  (see deus ex machina) and has been a constant theme in the 20th and 21st century U.S.  In general, techno-fantasies are wishful thinking about future trajectories for technological innovation and its social impact and rely on extrapolation and counter factual reasoning. Techno-fantasies are not limited to IT, of course. “Too cheap to meter”was predicted to be the future of electrical energy in the atomic age.

8 ) TFs are social. Techno-fantasies are usually generated to achieve specific socio-political (rather than narowly technical) ends. The need to convince economic and political elites to invest in lines of research or development puts a premium on the ability to articulate a crisp, concise, and attractive vision of the future. Such visions are simplistic, optimistic, and hedonistic to varying degrees. By their own nature, such visions are unrealistic and cannot be achieved.

9) What happens when TFs don’t pan out? Committing resources to techno-fantasies inevitably leads to cyclic disillusionment and revitalization of the fantasy that continues, sometimes for decades. Failure to produce the desired (and promised) benefit creates tension between the proponents and the elites. This, in turn, leads to retrenchment of the fantasy in stereotypical ways. Most frequently, the failure is initially blamed on incomplete implementation, “growing pains”, or narrow, technical issues.

[In the case of artificial intelligence (AI) for example a decade of excuses were supplied: “not enough memory”, “too slow computation”, “limitations of the programming language”, and so forth. By claiming that the underlying problem (intelligence) had been solved and that the failure to achieve the fantasy was the result of narrow technical factors, the AI community was able to keep the government funding stream intact. It was only in the 1990’s that it became clear that the program was truly a fantasy.]

In other cases the failure is blamed on active interference with the implementation. Opposition by individuals or groups — often those who correctly characterized the plans as fantasies — is identified as the reason that the program has not met its goals. Presently, in healthcare, the culprits are physicians and nurses who have not managed to “get with the program” or who are opposed for personal or professional reasons (e.g. “fear of the loss of prestige”). [Active interference by wreckers as the proximate cause for failure of a ‘good plan’ echoes the Stalinist purges of the 1930’s…see Solzhenitsyn, The Gulag Archipelago, Pt.1,Ch.10: “The Law Matures”.]

[The process of creating and sustaining a techno-fantasy in the face of setbacks, failures, and disillusionment is well documented in military systems. For the case of nuclear weapons, see Donald MacKenzie’s Inventing Accuracy: A Historical Sociology of Nuclear Missile Guidance (MIT Press,  1993).  Here, claims about the accuracy achievable with missile technology allowed fantasies about counter-force rather than counter-value nuclear exchanges that would leave cities untouched. This techno-fantasy has, thankfully, never been completely debunked. Another example, with truly horrific consequences, was the effort to develop “precision” bombing prior to World War II. The techno-fantasy that surrounded the Norton bombsight (succinctly expressed as the expectation that the system would place “a bomb in a pickle barrel” from ten-thousand feet) remained an idee fixe’ for much of the prewar period and was only abandoned in favor of a statistical approach after first-hand experience in the mid-war period showed it to be a fantasy. See Stephen McFarland’s America’s Pursuit of Precision Bombing, 1910-1945 (Smithsonian Institution Press, 1995) for details.]

10) Fertile ground for TFs: Intractable problems. The proposal has obtained political support because of lobbying by interest groups and the willingness of politicians to embrace magic bullet approaches to ‘fixing’ the healthcare financial problem. Intractable problems are unacceptable in fast-paced politico-industrial contexts. The current situation in healthcare is intractable for a variety of reasons. Techno-fantasies allow the full weight of an intractable problem to be ignored, at least for a while.

11) The mileage you get may vary. A mark of techno-fantasy is the wide range of quantitative predictions that accompany it.  According to an October 2008 Healthcare IT News article , candidate Biden claimed that advanced healthcare IT will provide “$180 billion potential savings” and that this money would be available to fund “universal” healthcare.  In 2005, a group from RAND estimated that the the “potential health benefits, savings, and costs” from the adoption of “interoperable EMR [electronic medical record] systems” could be “$142 - $371 billion” (Health Affairs, 24(5): 1103) (the study relied heavily on data from the IT industry).  A GAO report in the same year was considerably less optimistic and noted that the savings were not well defined. (In 2005, according to that report, the Secretary of HHS estimated that the savings from healthcare IT would be $140 billion per year. At the same time that he appointed David Brailer, MD, PhD, as the National Coordinator. Dr. Brailer is now the Chairman of Health Evolution Partners, a healthcare investment company, and a board member of Prematics, an IT company.)

12) Claim victory. The simplicity and breadth of language that make them attractive do not fit well with the complexities of real world systems or the conflicted, disputed historical development of socio-technical systems. Scapegoats for failure are usually easy to find and the embarrassment of losing so much  money, time, and prestige gives impetus to rationalizations about the outcomes. Often the scale of investment is so large that the programs that accompany the techno-fantasies become “too big to fail” and are treated as (qualified) successes. This is especially true for national prestige programs that fail, e.g. the U.S. space shuttle functioning as a cheap, reliable, and reusable “space truck” or the present space station Freedom. Such systems never fail, they just fade away.
———————–
Excerpt from the DrBicuspid.com exchanges following the article..

1/6/2009 8:20:50 PM
D. Kellus Pruitt DDS wrote:

“The President referred to an oft-cited 1999 report in which the Institute of Medicine (IOM) estimated that between 44,000 and 98,000 Americans die each year due to medical errors.”
How many of those were dental errors?
Kolodner has never once mentioned dentistry.  He may be a VistA genius, but he doesn’t know squat about my job, Dr. Powell.
hat is the problem with lots of people who would mandate help.
1/7/2009 12:18:53 PM
Valerie Powell RT(R) PhD wrote:

Dr. Pruitt, if you were better informed about the VA EHR (VistA), you would know that Dr. Kolodner, when he was practicing in the VA, already had at his disposal an integrated medical and dental record of the type I am recommending. The dental systems were being introduced by 1987, over 20 years ago. According to Prof. Peter Groen, Shepherd University, who formerly worked for the VA, the VA dental package offered a tailored template/screen depending upon the provider selected (e.g., Prosthodontic, Endodontic, Oral Surgery, Periodontic, General). If you have an intergrated system, it is not necessary for an endocrinologist or cardiologist to be a dentist or to know about dentistry in the manner you so delicately portrayed, but only to know about the importance of oral health for systemic health and to know what aspects of oral disease are significant for the respective disease states they treat.

1/7/2009 12:33:28 PM
Valerie Powell RT(R) PhD  wrote:

Dr. Pruitt, please recall that last year one dental error in which, significantly, a significant cause was not having taken an adequate medical history, result in death of the patient. I could say, if the two dental providers involved had had access to an integrated medical/dental record, they would have had a medical history available to them. This incident was duly reported in DrBicuspid as “Chicago dentists settle out of court in sedation death,”…