Q. Integrated

The author's other life is revealed — consensus on "integration" explained and critiqued — the irony of this consensus is noted — Adrian explains The Big Picture — systems integration is taken to task by Adrian — consulting is probed and prodded — consulting considered as an Important Substance in fluid programmable capitalism — the internet is posed as a solution — — the healthcare crisis is solved by Adrian and the author.

1. 'Representative' would not be the best word for the experience of trying to understand healthcare via an internet startup company in the radiology department of the country's oldest and most well known hospital. It is exceptional at least. My first attempts to figure out the infrastructure of healthcare from this perspective took the form of a research article for a publication called Healthcare International of The Economist Intelligence Unit. The job was serendipitous, like almost all jobs, and rested half on the letters 'MIT' and half on a personal recommendation. The research article that I wrote reads like hundreds of others of greater and lesser specificity, slinging buzzwords and futurological gestures, a result of editing, but also (no matter my disavowals) of the fact that I managed to channel a certain rhetorical consensus that I couldn't have seen through without benefit of greater experience or someone who knew better. Despite the fact that this field and these people might not be representative of healthcare, this consensus is. Less because it represents what people actually think (which, I suppose, is the anthroplogists charge at some level), than the fact that a consensus diluted and misstated by a media industry, pitched to investors and business observers, preached by doctor-reformers actually makes sense to any given participant (and this miracle is verified by its appearance at trade shows and conventions, in magazines and trade publications of all kinds, in interviews with people in the states and abroad, in the performative force that these words and stories have and evoke, in meetings, articles, and interviews). This is, in some way, more than simply "hype." Or this could be disavowal, once again, coming from someone who has contributed to it.

2. In any case, the story I tell in that article, and as it is told in various other articles, in different venues, and most often in lowest common denominator form in advertising, centers primarily on the keyword "integration." It goes something like this: the bad old days of "hospital centered medicine" are giving way to the promised land of "patient-centered medicine" by way of the internet and the creation of standards for the seamless integration of the healthcare industry. The hospital is now the Enterprise, integration based on standards is the sine qua non, and the patient will get her own "bill of rights".

3. "Citadels", "silos", "legacies", "islands." These are the mistakes of the past, even if it is unclear how they were made. No matter, because the combination of we-know-better-now and the magic of technological progress leads to the solution to healthcare's problem of integration. Often this consensus refers to a coming era when government will finally get the hell out of the way and let the proven methods of economic efficiency and organization solve the problem. As in the case of Larry Lessig, this maddening refrain usually has no conception of governence at all, only that corporate principles are the only way to integrate an industry like heathcare that "lags behind" the rest of business. [1]

4. An irony takes shape here concerning integration. The long twentieth century of high corporate-hospital healthcare represented a degree of integration rivaled only by its double, the for-profit manufacturing corporation. Under one roof were specialties of all kinds, diagnostic technologies, fully equipped operating rooms, patient rooms, offices, power plants, heating systems, supply and storage areas, complete record-keeping facilities, cafeterias, chapels and morgues. In teaching hospitals, classrooms, teaching OR's, amphitheatres, and faculty offices. A patient could enter the hospital with any disease, any complaint, any desire, and could be directed to the right place. Of course, some hospitals were better than others, but they all seemed to aspire to this kind of integration. Even some degree of regional integration was achieved by the Hill Burton acts, regional academic medical centers and the planning agencies of the seventies— albeit with some persistent and egregious shortages in both rural areas and cities— that allowed for a certain national accessibility. What was this if not a desire for integration?

5. Sabel's firms were tightly integrated until recently as well: this is where fixed and sunk costs lay making the governing goal of iterated goal setting over the last twenty years disintegration. The decentralization- de-hierarchization— of corporations sought in order to free up idle hands. Fixed costs can be liberated by making production fluid and configurable, sinking costs each night, weighing anchor in the morn. Why wouldn't the same be true for healthcare, why call for integration now?

6. Ambiguity rends the specificity of the word in the direction of competing infrastructures. What exactly— and where— is the infrastructure of healthcare? What authority determines it? what integration—and how big— is necessary for the reform of healthcare? Is it the integration of the hospital— small businesses under cover of voluntarism? Is it the financial integration of the corporation— the 'enterprise' as in the 'integrated delivery enterprise'? Is it the integration of allocation— access to technology and to information in the promise of "patient-centered healthcare"? How is 'learning by monitoring' allowed in healthcare, especially if there are no standards. Once again, a response refers to the problem of scale and convention— that is, the integration of an entire industry according to standards that allow such fluidity and programmability, that allow for the liberation of technology that is currently allocated by hospital or corporation. Somehow, this complexity of interests and constituencies, and the lack of an unspecifiable standard can produce the greatest anxiety, and even in the most perspicacious observers, like Adrian, a strong sense that heathcare 'lags' behind other industries.

7. The consensus that infrastructure and integration are the 'problems'— the consensus I channeled into The Economist Intelligence Unit— demands solutions from an array of people who feel varying degrees of propriety about how things should change, as well as who should change them— this is why Adrian divides the world up the way he does, in order to understand who is out there, what they are doing, who he should approach, and how he should sell Amicas (and, significantly, for how much— according to what pricing model). Adrian offers a big picture of the non-internet model of integration:

The big picture is quite simple, if you look at it. Say that there's a $20B market out there for medical records systems, my guess is that 2/3rds of it is in systems integration and consulting, and 1/3 of it is everything else, hardware, software.... The world as it stands now is these products that require huge amounts of systems integration expense to consulting firms and also a huge amount of expense in house, you know, the customer himself provides IS staff, and administrative staff and help desks and all sorts of things. So when you actually look at the cost to society of doing medical records the way they're being done, 2/3 of it is wasted, kind of in the same sense that administering public services in a communist country where you'd imagine some large fraction of the food spoiling in a warehouse on its way to the dump. This is what my impression is of 'high tech' information systems in healthcare.

8. Just to be clear, the communists in this allegory are not healthcare institutions, they are the systems integrators. Adrian is frustrated that hospital administrators choose to use them, rather than buying the products that already "work." However, even though Adrian's tripartite division of the world might look objectively sound from the outset, it stumbles first on the object of production: product, service, consulting. Defining the difference, finding the tangible discrimination between product and service is difficult— witness the proliferation of companies that now offer 'solutions' rather than products and services or even all three at once if only as categories for things that are both product and service, if not-quite-yet solutions. There is much to be confused about for a small company faced with the prospect of a changing healthcare industry, the internet economy, and the corporate world— not the least of which is the problem of discerning what the difference between a product and a service is— not to mention a "solution." Adrian asks these questions with repsect to the array of companies and interests that he initially divides into three categories, but in good pragmatic mode, has simply revised as his understanding proceeds.

9. Several established companies might fall into Adrian's second category of "application software vendors": Adac, Sunquest (two with whom Amicas has agreements), McKesson-HBOC, Eclipsys, Cerner, along with an ecology of companies who have been responsible for building various components that these larger companies might use (smaller software vendors, hardware vendors, network specialists, etc.). Attending a trade show such as The Healthcare Information and Management Systems Society (HIMSS) gives a sharp sense of the complicated web of companies that exist in this field (though it would hardly be called 'integrated' since 9 out of 10 'partnerships' were undefined, undefinable, or 'marketing' partnerships). Many smaller companies struggle to win partnerships, funding, or outright purchase by the larger companies. Most have small stakes in a specific section of the healthcare industry, often without any sense that they have any role to play in the mysterious "integration" of healthcare at large. But despite their small size, this is not Amicas' position— they remain committed proselytes of the internet vision.

10. Some of the larger companies metioned above have a clear goal in the development of technology, others are 'service' businesses. McKesson-HBOC, Cerner, and now IBM Global Healthcare (IBM-owned Lotus as well) are extremely large ventures, with diversified interests that are often difficult to discern either from the perspective of anthroplogist, or of the journalist. The marketing efforts that these companies engage in sound so similar to each other that differentiation can really only happen after purchase. The only people truly deserving sympathy in this context are the hospital purchasing agents confronted by a blob of marketing that promises instant integration. Adac and Sunquest (which I take as examples, because Amicas has signed partnership deals with them— Adrian's story of signing a deal with Adac is in the interview from November) are mid-sized companies that are clearly focused on Healthcare Information Systems (HIS)— though how clearly is a constant source of concern for Amicas.

11. Further confusing this field is the unstudied explosion in the "systems integration" work of consulting companies like Arthur Anderson, McKinsey &Co. or IBM — and in some cases, HBOC, which despite its origins as a HIS company, now offers a great deal of 'systems integration' services as well, even if it does not call itself a consulting firm [2]. The non-specific non-service called "consulting" offers no specific product or service but only a promise of expertise and reputation packaged as a "solution." Adrian recognizes these companies and reserves for them the category of hype:

"It appears to me like all of these CIO magazines that are throwaways that I get that just seem to me to have ads that are full of buzzwords, with obviously the content behind the buzzwords is a service business that's being offered, not a technology business. But that's as far as I can get. In other words, I can get from them an understanding that there's this world of people selling very expensive services based on what I consider to be very lame or trivial technology, and I'm not able to understand that."

12. This bewilderment, combined with the comparison to the essentially wasteful practice of "systems integration", concerns the location of value in healthcare. Consulting firms and their systems integration efforts are not solving a specific problem, but simply shoring up a "broken model":

You've got outfits like HBOC who are systems integrators that basically say, look, have us do everything and it will be ok. The problem with that is twofold: first of all it's outrageuosly expensive, and second of all it doesn't guarantee for that administrator that it actually meets the needs of 37 different constituencies that they have to keep happy. To the extent that they contract for custom features from these people it becomes even more expensive. It's just a broken model."

13. The reason that Adrian can insist on this is because he has tested these waters as someone who meets the needs of just one of these "37 different constituencies"— radiology. In late 1998, Amicas had been invited with three other image management companies to respond to an RFP issued by Imnet, the largest RIS company in the country, which had recently been acquired by HBOC. Upon making this pitch, Adrian came out more convinced than ever (in the way that confusion convinces when it is insoluable) that the only way to solve the problems of infrastructure and integration was the internet.

And I walked out of this process with total amazement, cause, here was a group that was supposedly trying to license our technology and who we would want to work with, definitely the largest group doing this in the world [Imnet], being bought by the largest HIS vendor in the world [HBOC]. And I walked out of there not being able to tell why anyone would wanna buy their product, or who of the customers in the thirty or forty people that i'm trying to sell Amicas to, would I send to them to buy Amicas from, as opposed to from us or from Adac or somebody else.

14. This chance to perform for HBOC had, at the outset, seemed like a really fabulous opportunity for Amicas. Excitement was high among employees. Had they gotten a deal, it probably would have been very good for them in the short run. Being a start-up, Amicas is perpetually cash-strapped, and at bottom, would not turn down a deal with good terms involving cash enough to make payroll. Adrian's puzzlement over HBOC/Imnet, however, was a sidelong disavowal, a sour-grapes response based on his satisfaction that HBOC is a 'broken model.' Regardless, they did manage to sign a deal during this same time with Adac, another company in the same "market segment" as Imnet, and subject to some of the same puzzling characterisitics of system integration. This ambivalence translates into a mild form of defeatism with respect to this 'channel' of distribution for Amicas.

I see them [Adac and Imnet/HBOC] as basically being service businesses. Let me give you an example: Adac takes 9 months to install a hospital. Adac— our partner— after they've signed a contract. Adac is incapable of putting an Adac server into downstairs so we could test our stuff. Its beyond their thing, they would rather put in a thousand dollar a month T-1 line than ship me an installation of their software. And mind you, Adac is people that I consider technically at the head of the RIS pack, and my friends and the people I deal with and I like them and they like me. So clearly they understand what we're doing, I think I understand what they are doing as a product, so what does that tell you? That basically tells you that these people are a service business, and we're product people, and I have no idea what those people [HBOC] are doing.

15. There are a combination of concerns present in this vague defeatism. One is that Adac does not have a solid and specific technology to sell: "if you have good technology you should be able to shrink wrap it. If you know how to make a car, you shouldn't have to train a chauffeur to drive it around." Adac would rather have an expensive open line directly to Amicas, which indicates that their technology, if they have a technology, is less product and more service— i.e. systems integration of other technologies— in the degenerate sense that reduces to consulting in Adrian's terms: "So you see what I'm saying, just because you are offering service doesn't mean you can be anything more than a consultant." Consultaning per se is not the villain here, however, but rather a corporation that disavows its consultative role by calling it "a service" is Adrian's frustration. Adrian therefore differentiates between some set of things that can be either product or service, but that are specific, not "lame or trivial," and essentially "shrink-wrappable," and consulting— an activity neither good nor bad but thinking makes it so.

16. There is nonetheless something mysterious in the extraordinary success of consulting firms— perhaps it is their vagueness. In an almost parodic mode, Arthur Anderson's motto is the fate of a language snowed under the vagaries of this marketing vagueness beyond control: "Helping clients improve businsess process." These are the offerings of such consulting firms: "knowledge products," "process solutions," "business services," "industry programs," "network solutions," "application management solutions,", "service operations," "functional practices," "operations effectiveness services," "strategic information products management services," and my personal favorite, available from Cambridge Technology Partners "assimilation solutions"[3].

17. A glance at just about any of the large management and IT consulting firms reveals the same. Without denying that the differences here are finely articulated at the level of practice, without being discernable at the level of marketing, it is nonetheless striking that this melange of words always includes 'service,' 'process 'or 'solution.' This confusion is essential to the particular moment in history that Sabel identifies in his firms, and that he then extrapolates backwards to pragmatism and forwards to the possibility of a new form of governance he calls "democratic experimentalism." It is part of a much longer discussion about the fate of manufacturing industries per se, but now includes a new discussion about the role of the service industries as the primary economic activity of North Atlantic nations.

18. It may be that "consulting" holds Sabel's flexible workgroups together. They are the caulk of fluid programmable capitalism. When it is unclear what one should be doing, or with what technology, or even, at some level, for what purpose, consulting firms are willing to risk suggestions, reforms, re-engineerings, and redesigns. After all, their stake is limited to the provision of a service, the suggestion of a solution. There is no identification with an industry or its overarching goals vis-a-vis the division of labor. Consultants, in this crazed light, are pure parasites. The disease they carry from industry host to industry host is relatively benign, a kind of institutional isomorphism that promotes similar solutions to diverse problems. They are the funnel for management strategy handbooks and manuals for the transformation of 'process'. All of the latest great ideas are first tried in consultancies. They will jump fast onto the bandwagon of internet proselytism, and preach it convincingly even if they don't practice it at customer sites. They are the excited early adopters, the first wave, the cutting and bleeding edge, the "I liked them way before they were cool" of capital.

19. Even if this is so, it happens with none of the obviousness or inevitablity that one might guess it would from reading Sabel on automotive firms. Even if "learning by monitoring" slowly becomes obvious to some constituencies in healthcare, it is still not obvious enough to cause the kind of reform that an integrated corporation would take under threat of bankruptcy. In healthcare, things are deliberatly less hyperactive. Nonetheless, from Adrian's perspective it is the systems integrators and consultants who are struggling, while healthcare acutely senses change without responding. Adrian's senses, however, are tuned to the infrastructural transformation of the internet and to the still somehwhat fuzzy sense that something— "transactioanl networked things"— navigates between product and service (avoiding "solutions" all together):

C: I assume you think that what's going to change is not the systems integration and software industry but the organization of these healthcare entities. Its more likely that its going to break up into 37 different industries.
A: I do believe that. Yes. Part of the reason why these people [system integraters] are struggling is because they don't want to lead the market towards breakup, because that just makes people hate them. So on the one hand people would like to breakup, but on the other these people don't want to be the instrument of that. From the products point of view, or from the software systems point of view, what's going to happen is that people are going to increasingly focus on transactional networked things. So instead of having to buy your front end and your back end from the same people you'll be buying your back end from us and your front end from Abaton.com or whatever and these things will meet happily, will mix and match happily. So for instance, people don't worry about where they are gonna save their files when they [decide to] buy Microsoft Word, as their front end. Nobody would argue that you buy a different word processor or that you worry about whether the institution is running Oracle vs. Sybase before you decide to buy Word vs. whatever. So I think in healthcare people haven't quite come to grips with that. Microsoft hasn't, Lotus, IBM...

20. Companies such as HBOC are not disappearing off the map, however, and that puts Amicas in a strange position. On the one hand Adrian, Barry Gutwillig and Jan Moe must explore the possibility of selling, or licencing, Amicas to these companies— HIS, RIS, Systems Integration, and consultant firms; on the other, these companies are often posed— precisely because of the very confusing melange of companies— as competitors vis-à-vis healthcare institutions. Sometimes they are de facto competitors, since a purchaser could end up being forced to choose, based on insufficient or unavailable knowledge between two providers that promise the same thing, even if, from Amicas' perspective the technology and/or configuration is completely different (Which is why Adrian, in his DI article suggests that purchase agents demand demonstration installations from potential sellers, something which he insists should be possible, but which is rarely asked for). Furthermore, because Amicas is primarily a radiology product, this confusion is reproduced again at the level of radiology consultants, radiology information systems (RIS) vendors and PACS companies. For Adrian the beauty, and the difficulty, of Amicas is precisely that it can be either competitor or partner to each of these firms. This unanticipated versatility captures the transformation of modernity in the space between 'product' and 'service'.

21. The combination of the infrastructural transformation of healthcare that Adrian sees undermining the stability of systems integrators and healthcare constituencies with the strange versatility of Amicas means that the deals with Adac and Sunquest that provide Amicas with Value Added Reseller (VAR) distribution channels are imperfect. They are not the route to Amicas' domination of the market. In the deal with Adac, the crux of the problem lies in the kind of deal that either side saw as acceptable. What ended up being negotiated— the VAR agreement— basically allows Adac to use Amicas as a component in the larger systems they install or integrations that they perform. This depends on Adac to do the integration, for which Amicas gets very little more than they would get if they sold it themselves, but that Adac has a larger sales team. Ideally, Adrian would much rather see his product tend towards one of the other two categories in his ad-hoc trilogy. On the one hand, Amicas could be part of the modality vendors products, and that is a channel that Amicas has yet to explore (even though, Adrian explained that one of the results of the foiled HBOC RFP was that it "made Philips stand up and take notice"— Philips representing the modality vendor selling the devices that "collect the bits"). On the other hand, Adrian imagines Amicas in an internet niche that would be a kind of protected reserve based on its technical specificty. By the 1998 RSNA convention, Adrian was instructing his staff to preface every statement with the word "internet", to make sure that, for any conceivable customer it was clear that Amicas was an internet company with internet technologies for an internet age that would soon be upon healthcare: "If you have to say only one word, say 'internet'. Why is that good? Because it will integrate now and in the future."

22. There are some object lessons about the internet that Adrian has relied on in various stages to articulate how this integration will work. One is 'disintermediation.' The travel agent and the stockbroker are services that— up until the explosion of the web in 1993— owned the gateways to information, and could charge people to discover and serve that information to them. The internet makes such a service irrelevant, or worse, criminal, like selling air to ingenuous breathers. Adrian therefore wants to suggest that there are two people who risk disintermediation by the internet— HBOC-like systems-integrators, and more significantly— doctors. The two are structurally similar in this lesson because they both withhold information from someone to "take advantage of an emotional situation." The case of the doctor is obviously very complex {q.v. below?} but in the case of HBOC it simply means that the combination of the possibility for a kind of commodity connectivity with the openness of internet standards and protocols means that HBOC offers nothing more than a name, has no product or technology that someone else might not be able to build better and faster and plug into the same internet to solve the same problem for HBOC's customers. Here we come closer to learning by monitoring, but only when the healthcare industry agrees to allow its infrastructure to become the internet. But this kind of "controlled chaos" that the internet represents in terms of solutions is usually the farthest thing from the deliberate motions of a healthcare organization and the propriety of its purchasing agents and IS staff who still have major stakes in controlling something that still feels, on the ground, very much like a hospital. I challenged Adrian on this, trying to suggest that what was missing was simply a warm body with promises to make, if not to keep:

C: To me it seems like a lot of that mystery is just this fairly rational desire on the part of healthcare to have someone who is responsible, someone who takes responsibility for this, as opposed to having the sort of free-for-all, where there is a certain control issue over who maintains control over the integration of these things...

23. Adrian explains why HBOC should care:

A: I would be really frightened if I was you, you HBOC. Your question is extremely well posed, but now let me ask you this: Look around at the rest of corporate America, in other words, lets not think of healthcare as a not-for profit charity that doesn't work by corporate standards. Lets think of it as managed care, continuing to head in the direction it has headed, lets imagine that it's gonna happen even if the Democrats control both houses of congress, that we're not gonna go back to the Canadian model, no matter how you and I might feel about the politics of that. Now ask yourself the question: what other large scale software businesses or information systems businesses are industry specific relative to corporate America? Who's doing well out there? Peoplesoft. Well do you think there's a Peoplesoft out there that's different for car manufacturers than the one for lawyers? No. As a matter of fact, Peoplesoft is in at Partners doing their purchasing thank you very much.... Do you think that digital dictation systems from IBM are gonna remain health care specific, or will it just be that within a year that MedSpeak for radiology is simply a new dictionary that costs 69.95. Are you, HBOC, willing to stake the next generation of software on the fact that people wont use Word to type up their reports when its already integrated with IBM and Dragon Dictate, just because you understand how the service business works. I challenge you to think of one example of a large scale software supplier whether they be Oracle, Peoplesoft, SAS, that to corporate America that you can say oh yeah these guys are in the insurance business, and they are making billions of dollars and price/earnings rations of 80:1, so you see what I'm saying, just because you are offering service doesn't mean you can be anything more than a consultant.

24. I took this insight with me to the 1999 HIMSS conference in hopes of discerning how people and corporations were thinking about the internet and the re-integration of healthcare. Posing as a journalist once again, I managed to further confuse myself about the kinds of partnerships, licenses, sales, collaborations, integrations etc. that companies were engaged in. In particular, it was the strategies of the largest software corporations such as Microsoft, IBM, and IBM-owned Lotus that seemed most mysterious, because they seemed to be selling an internet solution, with the kind of integration that Adrian talked about, but in no case would anyone explain the details of any given arrangement[4].

25. It started to feel like everyone thought everone else was a potential partner, or competitor, regardless of what they actually did. I found that no one could be differentiated based on marketing alone (which, along with Amicas, had been thrown in a semantically irresponsible way at the internet, claiming all manner of web and internet-based solutions, without any sense that those words referred to specific technologies that companies did not actually offer. Little in advertising could be so safely called lying as the marketing strategies that I have seen at these trade shows), and certainly not based on the core technical competency of a company. This promiscuity of partnerships and potential negotiations included even me, as journalist and as representative or another large corporation, The Economist (even though my formal ties to them consist of nothing more than an informal agreement to provide research articles), which made for an uncomfortably imprecise understanding of what I was there for (to uncomplicate matters a bit I did not represent myself as an anthropolgist— based on my previous experience at RSNA— but only as a journalist, which was difficult enough to perform)[5].

26. The experimentalism of partnerships and attempted solutions approximates what Adrian imagines re-integration of healthcare on the internet would look like— that good applications cleave to infrastructure (Microsoft Word and Dragon Dictate are good for any industry, healthcare just needs a medicine-specific dictionary)— but he reserves some of his bewilderment for companies like Lotus, who should know better:

It's like Lotus Notes: What does Lotus Notes do for you? Lotus notes does everything for you! That's not the right answer! HBOC already does everything for me! Lotus notes, needs to do something, like for instance, it needs to manage the directory of priveleges and give me a way by email so that if I want to be called at 924-5273 instead of 924-5339 today, I should be able to have a web page where I can go and— if I'm logged in properly and I have a [password], [I] should be able to change my name and that information should be available to Amicas. Does Lotus Notes do that? No. Can it do that? Of course it can!

27. From this, Adrian can extrapolate to a hazy, but entirely possible solution to the re-integration of healthcare:

A: So this is the big picture problem. That the world in healthcare has to have ... the players— whether they be the customers or the big vendors or the little shrimps like us— they all have to have some kind of a vision of how they are gonna interact with each other at the level of, say, twenty services — pick a number, and somehow you can reduce 99 percent of what happenes in healthcare to twenty services that you are gonna have to take care of. Acquiring images in radiology is one of them, dishing out passwords is another, connecting via Virtual Private Networks on the internet is a third. You can think of 17 others, some involve collecting information from doctors, or putting screens in front of people who schedule things, some of them involve putting screens in from of patients, some of them have to do with the interaction with pharmaceutical and medical device companies, some of them have to do with hardware, and the support thereof, outfits like StorageTek [who offer large data archiving systems].
So if you were to draw a list of sort of the essential services... you and I could sit here and based on the little that you and I know and we could probably reach a list of twenty that would have absolute total consensus. C: as far as what needs to be done? A: What needs to be done to meet 99.5% of what goes on out there. If that were ever done such that people basically could say "This is what I sell, this is what he sells..." then the whole thing would fall together very quickly. And all this system integration value would be captured and the customers would feel happy. But there are strong disincentives against doing that from the industry's point of view.

28. Imagine a world where people could simply say: "This is what I sell and this is what he sells?" In a manufacturing industry, this would be obvious (in fact, it sounds a little bit like the division of labor), and Adrian is well aware of it: "If you tried to build cars by building your own relays to control the robots, then you would be out of business in a second, but there is too much of that going on in healthcare." Such an integration would yield tremendous value, not just to the firms involved, but to the "37 different constituencies" of healtcare. This hope— this faith — based on the obscure promise of the internet as infrastructure, is a faith in certain principles of (self)-organization that are commonly referenced simply as "capitalism" or "the free market" even though they may well exist in economies judged by American standards to be socialist. Principles judged to experimentally, incrementally improve a system of "production" (or in this case, perhaps allocation of access to technology) better than the deliberations of hierarchical committees hospital by hospital or region by region. The history of American capitalism is the history of a highly and peculiarly socialized capitalism. The prominence of regulatory agencies, and their presence in every industry is indication enough that the government occasions markets with the faith that it is improving welfare, preventing abuses, or in some manner deliberately directing society. The kind of government this implies tends ever away from deliberate democratic decision making towards a technocratic mode. A gyroscopic government in which movements by business are counteracted by tiny adjustments in standards and regulations, not dramatic and final appeals to constitutional principles. Change is the mean against which justice is measured, not stability.

29. Too much of healthcare, from Adrian's standpoint (and he is by no means alone) is restricted by the demand that the government (or government agencies, or hospital administrators) decide these rules in the legislature in advance of any possible system, that they be implemented and enforced in a hierarchical manner that values complete top-down control, and worst of all, that this is in turn perceived as the least hurtful manner of allocating access to healthcare. For Adrian and Amicas, this relationship to control is opportunistic "taking advantage of an emotional situation," and should be replaced by a system of competitiveness that allows infrastructures to be rapidly improved and adapted, dynamically fulfilling the needs of an industry whose knoweldge base changes with every issue of JAMA or Nature.

30. This faith, however common, is not (yet) a consensus. There are considerable concerns that linger in the world of healthcare policy that, maybe, just maybe, handing healthcare over to the market might produce some particularly egregious radicalizations of some already persistent gulfs in the equality of access. Simplicity obscures the existing interpenetration of governence and market, however, and it may be that there is no such choice to be made, by anyone. From the most local level of the doctor's negotiation of an increasingly baroque system of allocation by managed care corporations all the way to a legistaltive morass of regulatory agencies and the "patient's bill of rights," there is no simple authority that would sever the gordian knot. This desire is a recapitulation of that essential activity of doctoring, the relationship of power in which the doctor plays the subject supposed to know, allowing the patient to become patient, to be acted upon by a force which exists nowhere except in the belief that someone else knows you better than you do. But in this case, there is no doctor— not government, not experts, not doctors themselves. As a result organizations localize to solve the problems that they can, integrate in ways they see fit and do their best to satisfy the impossible desires of the legislature and the administrative state.

31. The re-integration of healthcare as the internet, the creation of a quasi-public fluidly programmable market in healthcare information for allocation, care and organization offers itself as the only solution adequate to the times. The consensus on integration creates its own crises and solutions, dead-ends and failed experiments that look all the more unacceptable as experiments rather than misjudgments. However in the absence of centralized decision-making, there is no other possibility than to ask everyone to examine their own relationship to responsibility, and conceivably, to assume it for the future.

Last Modified 11-Sep-99 9:01 PM ckelty@mit.edu

Go Back to the Start