GDP Expenditure

With the new coalition government in the UK, we are seeing early signs of a serious assault on public spending on the state run NHS. Similar challenges await other European countries with bloated public debt. Part of the debt run up by Greece, for instance, arose from efforts to off-shore hospital debt.

In the end the question remains, as it always has, how much money should a country spend on healthcare. The answer, as has always been the case, is as much as you can afford. Research shows that levels of spending (in terms of percentage of GDP, for example) do not correlate well with health status, outcomes and other key indicators of the performance of a health system. Indeed, it can be said with some degree of confidence that GDP spending is NOT an indicator of health system performance.

What does appear to be a factor though is HOW that money is spent and HOW the system is organised to deliver health services.  Recent OECD work has clarified various characteristics of health systems. What is striking are a couple of already familiar features:

• Not all countries pay 100% of the health bill from the public purse; many, such as France, use co-payments. Countries with socially unacceptable waiting lists have tended to be those with the highest levels of pure public expenditure (such as the UK, Norway and Canada). What this suggests is that there may be important features in how health systems organise themselves to deliver care that is adversely affected when the system is funded from general taxation. Efforts to introduce purchaser/provider separation, for instance, is an effort to create distance between the two quite different objectives, which in tax funded systems have been merged and caused considerable policy confusion, as well as operational difficulties. (I can mention the situation in the Canadian province of Alberta, where the response to funding constraints has been essentially to ‘nationalise’ the system, thus removing key drivers for reform. I can also refer to the Nuffield, UK, study that showed poorer health outcomes in the centralised health system in Scotland compared to now quite devolved purchaser/provider based system in England; and this despite having higher per capita expenditure in Scotland.)
• Most countries have mixed economies of provision and relatively easier ways for new types of providers to emerge. Lower performing health systems seem to discourage new providers of care to enter the health market; this is an element of overall system design, perhaps regulatory over-reach and dated statist thinking.  But perhaps we are becoming smart enough to know how to design more responsive health systems, which in the end are almost chaotic given the nature of human beings and illness (random?) and so need to be understood as complex adaptive systems rather than tightly managed and controlled (think of the tightly coupled banking system which lacked the ability to realign itself quickly and effectively in response to a financial shock; Homer-Dixon’s remarkably prescient work here is worth looking up).  Managed designs usually end in tears, as they fail to deliver the responsiveness and flexibility that is critical for healthcare to respond to changing demand and fluctuations caused by shocks to the system.

There is no right number of doctors or nurses or hospitals or beds. What there is, though, is the right number of these for the design and structures necessary to deliver effective care.  And these can be designed and developed to use human talent differently, and more effectively.

In the UK, we will hear a lot about ‘front line services’ and protecting them from cuts. I have no problem with protecting front line services, but that does not mean that they will not be delivered in different and novel ways, that may be a better use of the expertise available.  The health professions will undoubtedly circle the wagons and predict dire consequences to the public, so called shroud waving. But what is better is a recognition that healthcare systems are highly inefficient; they are weak adopters of revolutionary change, and they are protective of established working practices — part of the reason for this protectiveness arises from the health professions having become co-dependents to the addiction to public money on the one hand and protected ways of working on the other. In a nutshell, they have become resistant to innovation and reform, and in some respects lost control of the their profession and the profession has ceased to evolve to meet the care needs of people — an emergent adaptive response characteristic of complex systems.

Hospitals are artefacts of industrial era organisational design principles — they embody craft mentalities in the organisation of care, and build on public support to protect their infrastructure (from closure, for example), rather than the public demanding better services, which may not require a hospital in the first place. The difficulty people have in unbundling a hospital (it can be done and I can share the algorithm with you in another post if you like) simply reinforces the protected nature of healthcare work. In part, the emergence of e-health (more precisely, the use of digital information and communication technologies, artificial intelligence/neural networks, predictive algorithms, smart devices, etc) offers a serious challenge to established patterns of working, as these various components have the collective effect of redistributing knowledge, embedding knowledge and skill in devices, and altering the use of bricks and mortar infrastructure — a high-tech/low touch outcome is not the necessary outcome if we are clear on our outcomes.

It is also not just a matter of a cost-effectiveness study of whether an e-consultation is better than a face-to-face consultation.  The evidence for this is actually quite easy — when the telephone was invented, businesses might have one, on a stand, which people would queue up to use. Now, a modern business would hardly do a business case to put a telephone on everyone’s desk — indeed, it hardly needs a business case to ensure everyone has a smart phone — yet in healthcare, smart phones are still rare, yet have the potential to radically alter information flows and hence work flows — 25% of US doctors now have one and ePocrates is one of the most downloaded clinical apps from Apple store, so it is coming. You don’t do a business case when the underlying business logic itself is what will fundamentally change and that is really what e-health is all about.

They say, in capitalism, that it works partly through a process of creative destruction. Otherwise, we’d still be riding around in horse-drawn buggies, and you wouldn’t be reading this note on a computer linked to the internet. There is, however, a general reluctance to apply that process to publicly funded institutions, and by extension to publicly funded ways of working.  The words government and entrepreneur are an oxymoron for many people. But that does not have to mean that public funding cannot be used to incentivise new ways of working and new forms of healthcare delivery. The challenges, in the end, lie in our heart and willingness to change, to create and innovate.

And so to austerity. There is little to fear, except our ability to resist change, protect legacy ways of working, and failing to grasp the real prize, that of doing things better and more effectively.  We will, no doubt, hear the opposite.

Beware digital errors as they can bite

We all know accidents (unusual occurances in healthcare) can happen. Where systems are involved, errors can arise from how a system works, the way the various bits mesh, the knowledge and training of everyone involved working together.  It is no real surprise that some errors arise from the technologies that we use. In particular, health information technology systems can cause new types of errors and mistakes, beyond just not working properly.

In the US, the Health IT Policy Committee has proposed establishing a database to track potential safety risks related to IT systems.  These risks include:

• hardware and software failure and bugs
• workflow interactions between staff and users
• interoperability problems
• implementation and training deficits.

Since healthcare work is complex, the workflow risks are particularly complex and can arise from, for instance, inaccurately understanding how a manual system achieves its results, and thereby designing a software-based system that fails to do just that. There is a funny little thing that happens when a patient sees a doctor; the doctor often will use writing a prescription to terminate the patient encounter — tearing the piece of paper off the tab, a swirl of signature and handing the slip to the patient leads to the patient leaving, a neat way to end the consultation.

In an automated system (electronic prescribing, for instance), the consultation is not terminated in this behavioural manner, but involves essentially hitting the return key on the keyboard to enter the required prescription data in the system, and perhaps handing (or not) the patient a copy — but the Rx is off on electronic wings to the pharmacy for dispensing. There is an error that can occur if the doctor does not hit the return key between patients — the Rx list builds up, from patient to patient, until the return key gets hit (unless some sort of failsafe has been built in); this error actually happened and it was an alert pharmacist commenting to the patient that the doctor had added a lot of new drugs that the alarm was raised. Perhaps the patient should have been more distrustful, too.

We must be mindful of risk and error in any kind of technology, but particularly in systems where it is very hard to look inside the black box of software code.

I wrote a paper on digital risk some years ago, which can be found here: Patient Safety and Digital Risk. I have also raised the issue of risk in the even blacker box of predictive algorithms used to data mine record systems and profile risk of patients and this can be found here: Predictive Health. This second paper suggested that software may need to be subjected to comparable regulatory review like a medical device.

Just because you can’t drop it on your foot, doesn’t mean something can’t be dangerous.

There is trouble in e-health land, at least in Ontario’s funny notion of what they might mean.  EHealth Ontario has been subject to an emergency audit of its procurement or not of an electronic health record [EHR] by the Auditor General of the province.  Apparently, somewhere approaching C$1 billion has been spent with virtually nothing to show for it.  The problems lie in a bad ehealth strategy, and inappropriate use of consultants.

There are lessons here for other jurisdictions, as they seek to embrace the benefits of EHRs, and ehealth more widely, in particular. Of course, what is an EHR for, is the core question.

One of my alma maters, McMaster University, has sprung into the fray saying it has an EHR called OSCAR that could be implemented for perhaps 2% of the estimated cost of a provincial EHR.  Their argument being that a lot of doctors are using it.

EHRs are not a tool for doctors, though.

EHRs are an integrated information repository to facilitate better healthcare.  Doctors are not the only oranges, and nurses, physios, social workers, pharmacists, OTs, oh, yes patients and parents, informal carers, too, need access to health records. In my view, patients should own and hold their own health record, to ensure high audit standards (would you let an error remain on your health record if you knew about it?).

Servicing the specific needs of doctors alone is not an EHR strategy worth having, and doctors themselves should be the first to say this. It is time they showed leadership within the wider healthcare system, and rejected self-serving models, such as McMaster’s, which automate obsolete information models. McMaster, too, should have known better.

The Ontario Ministry of Health has wisely rejected OSCAR’s offer, but for the wrong reasons.  Citing the need for doctors to choose their own systems, just shows their continuing logic of catering to the needs of a particular health profession, rather than addressing the systematic provision of patient information within an integrated decision-support system.

All this is being driven by beleagured officials who really need to think again about their priorities and why they really need an EHR.  Perhaps they are afraid to admit to having made a mistake.  Such hubris.

Clearly more work is needed to define the purpose of the EHR and the goals for an ehealth strategy in Ontario (and other jurisdictions of course), before more taxpayers’ money is spent on ehealth.

Oh yes, apparently Ontario are going for a tender on a diabetes registry. NYC has one. I fear the worst.

The elephant in the room in healthcare is the hospital, about which I have suggested that we will build the last one in 2025.  What will “smart hospitals” look like, and why should we care?

Why should we care?

Hospitals are expensive and complex labour intensive organisations originating in industrial era thinking.  Little has been done to modernise the institution itself, although much has been done of course to improve what hospitals do. We also know that hospitals account for a considerable carbon burden and consume a huge amount of energy since they operate 24 hours a day. We know that as labour intensive institutions they suffer from the challenges all such organisations face as they try to improve operating practices and reduce running costs. Healthcare delivery is characterised by regulated cartels, which serve both to protect the public, and protect professional practice from incursion by other health professionals.  A bit like an early 20th century factory with craft guilds.

We should care because these institutions need to become smarter in the use of modern technologies and practices, but this process is slow and cumbersome, and while they evolve, the taxpayer is faced with paying the costs of institutions which in many cases should be replaced. This is not to say that those who lead hospitals are not focused on these issues, but only to say that their job is not easy and with the many vested interests around, challenged.

What would be refreshing would be leadership for clinical workflow change to come from the professions themselves, due recognition of their need to evolve and reform rather than simply protect the status quo.  We need these groups to drive change in healthcare, rather than waiting for politicians or Ministries of Health to set the agenda. Of course, informed and empowered patients will eventually not put up with much of the nonsense that confronts them when they seek healthcare, but that is another story.

What will they look like?

We are left with wondering how to improve how they do what they do.  Enter ‘smarts’. This brings together a constellation of forces currently abroad in the world, ranging from automated building management systems, smart grids, energy recovery systems, to wireless technologies in hospitals to remove the wires.

Coupling smart systems together creates networks that can link patients in their home to monitoring facilities and first-responder capabilities. With the added advantage of wireless, we have untethered remote monitoring.  In the end, we have real-time healthcare.

Smart hospitals will not need to define themselves in terms of their geography or location, that is in terms of buildings. They will define themselves in terms of two factors:

1. their capabilities and
2. how they deliver these capabilities.

Indeed, the organising logic of the modern hospital will be replaced with one akin to a dating agency — it will link people with needs to capabilities to meet those needs — built on a sea of clinical, and patient information, and connectivity to various organisations that can deliver the services (healthcare) that is needed.  This breaks the current approach to vertical integration (based on the industrial conglomerate model) and replaces it with the virtual hospital, a network of focused and tasked organisations.

I had scoped such an approach to a redesign effort for a teaching hospital, which would have replaced a campus model (mainly an old building and some attached add-ons) with a distributed and electronically-linked (ehealth stuff here)  network of perhaps 24 centres scattered across a city of a million or so.  But industrial era logic prevailed and they went with the single building.

I guess we won’t get smart hospitals until we have smart planning.

We have had years of reform efforts in healthcare, and despite what country one picks, the themes are depressingly familiar: cost-containment, more health professionals, patient empowerment, more primary care, value for money, and so on.  These types of reforms are rarely revolutionary, despite the claims, and the benefits not as readily forthcoming as forecast. For instance, we have had perhaps 20 years of integrated care pathways, yet such simple knitting together of care is still elusive.  What is clear, though, is that you can’t continue to spend good taxpayers’ money on unreformed health systems.

Reform models reflect the history of our healthcare (and other) systems, deriving from organisational and service delivery models of the industrial age.  Hospitals are really just 1030s conglomerates, and the claims that vertical integration likely to improve care and drive down costs, are simply copying the corporate models of General Motors, General Electric, GEC, Westinghouse, some of which are no more.  We don’t really live in that sort of world anymore, and despite the vast amount of money spent on healthcare, it is still the least information-enabled of all sectors of our economy, even though healthcare floats on an ever-changing sea of knowledge and clinical/patient information.  Our current notion of healthcare is wedded to the brains of individuals (i.e. health professionals), not the collective intelligence of many people working together (dare I call this cloud cognition, hive minds, or distributed cognitive systems…?).

I think we need to take a different look at reform models, and embrace a new terminology, one built on disruption.  Disruptive technologies in particular are game-changing, they alter our modes of interaction with other people, change how we manage information, make decisions, perhaps even think. They, of course, produce winners and losers, as these sorts of changes often are zero-sum. Keep in mind that health reform has tended to be non-zero-sum; there has been a fear of creating losers while at the same time trying to reward winners, so-called protection of legacy providers, and we see this in the most recent UK Department of Health plans to allow failing NHS providers two tries to improve performance before alternative providers will be allowed to take over the work. Disruption says enough is enough, and we must do things differently.

We don’t know that much about disruption except by what its effect is on us, but there are efforts to understand  disruption.  But this work has been weakly connected to both the policy space in which these insights can achieve some measure of meaning, and the real-world.  Healthcare systems can go to great lengths to frustrate innovation and change.  It is, therefore, timely and pleasing to see efforts of understand disruption, and the forthcoming report on disruptive forecasting from the US Committee on Forecasting Disruptive Technologies, National Research Council, may offer a renewed impetus not just to the forecasting work, but to its utility.

I like disruptive technologies for their ability to shift our thinking away from industrial age paradigms to information age paradigms.  In this way, we break the logic of physicality that defines, for instance, hospitals, and leads to new approaches anchored around the health information value chain, which unites patients and all actors in health systems (payers, providers, industry, academe).  Ehealth is one of these potentially disruptive technologies, as it achieves a couple of key disruptions, in terms of decoupling patients from physical location, and of the potential pooling of knowledge in distributed cognitive systems with machine intelligences through smart/remote diagnostics, predictive modelling and in time physical models of disease.

But disruptions are a much harder sell, but it seems to me that difficult public finances does offer an opportunity for rethinking: one should not waste a perfectly good crisis as it is an opportunity to evolve. (with apologies to Rahm Emanual who said “never waste a good crisis”.

READ an interview I gave on ehealth here: [LINK to Euractiv ehealth interview]

From across the Atlantic comes news of apparent financial maladministration at E-Health Ontario, the body charged with implementing the province’s e-health strategy.  It seems to be the usual nonsense of untendered contracts, friends in high places, and chums helping chums.  It is also an example where no one seems to have asked the simple question, “why would you do that?”  — the strategy is a nonsense, and I am surprised that no-one challenged this before the policy had gone this far in implementation.

I would, naturally be more inclined to be concerned if the province’s e-health strategy were actually about e-health, or likely to deliver results worth having, but the $700 million or so per year will be spent on things like a diabetes registry, wait times, electronic prescribing/electronic health records.  Only the last have anything really to do with e-health.  The last can also be procured, so there really isn’t a need to make a supplier meal out of putting something in place.  I will concede though that an EHR is a critical component of e-health, but it isn’t quite the same as e-health — it is a bit like confusing the foundation of a house with the home it will become.  But having worked on eRx,  the province’s failure to prioritise some sort of a patient-held smart card is a mistake as without this it is difficult to deal effectively with identity.

Without system redesign in the province, the e-health strategy is really just throwing good money away and given the current economic (and political) climate, this is no longer an option, if it ever really was.

Two things are of critical importance.  First the province needs to have a thorough-going governance review of e-health Ontario, mainly to determine how to make sure it is fit for purpose in actually providing the leadership for development of an e-health infrastructure service delivery platform.  Secondly, and this is the challenge, it is necessary to make sure that the e-health services are ones that the public will use and value.  The province has failed on both counts.  The next challenge though will be to find people to review e-health Ontario who haven’t been tainted by this scandal and benefited from the feeding frenzy e-health Ontario created. It may require looking further afield, to interested, but uncontaminated parties.  They may even not live in Ontario — golly gosh, so much for made-in-Ontario mediocrity.

So, having vented on that last point,what would an outline e-health strategy look like for Ontario, assuming that some governance arrangements are put in place,.  These are really just illustrations as certainly I would want to get a good understanding of priorities from interested patient groups:

• There are about 90 rural and small hospitals in the province.  A good plank in an e-health strategy would be to enable them to become a single, integrated, but distributed healthcare provider, perhaps with some sort of local and shared corporate governance.  A distributed healthcare provider, using e-health infrastructure technology would deliver specific outcomes to rural people, such as access to networked diagnostic imaging technologies, electronic prescribing and remote access to health records.  I would certainly save people in Thunder Bay a lot of trouble getting down to Toronto for a scan.  With a little bit of imagination and thought, this could work.
• About 60% of diagnostic facilities are located in Toronto, but which has only about 25% of the population; these are licensed clinics which often only offer a single procedure.  Using networked imaging technologies, remote diagnostic telecare booths (you can buy one from Cisco) many of these suboptimal centres could be relocated either to the rural network, in the previous plank, or provide a more accessible urban service across the provinces main urban centres.
• Smart card technologies (whether a smart card or an electronic secure passport) would give a better reason for constructing electronic health records than ones focused on improving data access for health professionals alone.  Patients, when given access to their health information, will have a vested interest in ensuring that the information is correct (my Ontario health record when I lived there had an error showing I had a condition affecting women, but I am a man — I still don’t know if the error was corrected; in an electronic system, that error would have been a problem, but I would have made certain that it was corrected, too).  As an ‘auditor of one’ patients can make sure information is correct, and drive substantial service quality improvements.  This is not to say that health professionals can’t do that, just that the evidence shows it comes slowly and is complicated by cartel-like professional practice barriers.  Start by putting the e-health card in the hands of the heavier users of the health system, to better manage their healthcare, access to information, and gradually as people see their family doctor, or get born, migrate the whole population over.  Of course, this will mean that family doctors, clinics, pharmacies will have to adopt some sort of information system.
• Don’t do what the English NHS is doing with Connecting for Health, by creating a large-scale government-led initiative.  E-health Ontario’s predecessor took a look at Denmark, but failed to learn the lessons despite what they wrote in their sham of a consultation document — they missed the point partly because they appeared to have another agenda heading toward a particular solution.  Denmark has shown how disparate stakeholder groups can work together to create an information system that works, and does things people value.   Better that than spend vast amounts of money on a grand plan to nowhere.

The general plan is to build an infrastructure that starts with the patient/family as user.  My experience in developing an interactive health television channel showed me the importance of starting there, and defining the benefits from that perspective.  Change will drive from that end too.  Finally, engage all the stakeholders (like the Danes did), find commercial partners with interesting technologies that do things that people value (rather than whizzy technologies), look for alternative systems to pay for healthcare services, as failure to develop a suitable and workable reimbursement system for e-health services is a barrier ( just ask Norway).  Oh yes, don’t forget political will.

Gaming and simulations plus modelling are health markets that look very interesting and offer considerable opportunity for disruption of existing knowledge processes in healthcare:

Predictive modelling

• to help people understand and manage their health better by using modelling to visualise health states using avatars and body-image
• moving beyond the use of predictive modelling and data mining to find high- or at-risk individuals for case management purposes
• link modelling to powerful mapping visualisation technologies to enable better decision-making and planning

Simulations and games

• engaging health professionals and consumers in simulated environments using gaming methods

There is a tension between health systems and their need for innovative technologies and the absorptive capacity of the health system itself to both adopt an innovation and modify existing clinical practice to release the full benefits of the innovation.  From a policy perspective, this presents a variety of problems not the least of which is that it is largely pointless to put more money into unreformed health systems.  From a commercial perspective, how is business to decide what are priorities, what are the innovations to back and what markets will adopt them.  Governments and payers can do much to signal markets what their priorities are and back that with appropriate reimbursement policies to enable these technologies to earn their way in the world.

But it is not that simple, and there is a clear need for policy makers and ‘the market’ to interact productively, so both win.  In the absence of this, we will have the continuing saga of the medico-industrial complex driving technologies forward but with no payers.

Competing interests characterise what people think are healthcare technology priorities.  With the often overbearing weight of government, healthcare technologies often reflect preferences that emerge from the policy priorities of governments  locked in an iron triangle with industry and (usually) doctors.  This medico-industrial complex leads to technologies that are sought by doctors, and when companies seek guidance for their own product development priorities, they consult doctors, and around we go. There is some good reason to do this, as it is widely argued that it is doctors who decide what services, medicines and devices patients will end up using, so it is sensible to ask them what they would like.  The problem with this is obvious, as doctors are not consumers of the functions of the medicines or devices they prescribe.  That countries are invariably forced into some form of economic evaluation of health technologies and the use of prescribing guidelines offer some evidence that doctors, in this case, cannot in the main be trusted to make appropriate decisions in this respect.

Let’s take e-health as a case in point.  Often confusingly called ‘telemedicine’, the priorities range from devices and services that patients may actually use, to technologies to facilitate consultations and information exchange between health professionals.  The latter, though, is really just the automation of existing clinical practice.  The former is far more interesting, and far more disruptive of existing practices — perhaps that is why we don’t have much of it?  Then there are technologies that really have a major impact on disease diagnosis, but which are expensive, but through elaborate clinical protocols are restricted or limited — why not adopt ‘best technology first’ and stop wasting the patient’s time.

Some priorities for further thought:

1. Following work by Christensen and others, how can health systems identify technologies that will have the positive benefit of disrupting in the nicest possible way stale clinical practices and yield an order of magnitude improvement in health system productivity (with a corresponding decrease in per-capita costs)?
2. What technologies are most effective from a patient/end-user’s perspective and that they will actually value and use?
3. What commercial realities are needed to enable sensible reimbursement of e-health services by payors?
4. How do we research, invest in and commercialise winning technologies and move them very quickly into use?

That is to say,

• how much do we really want to reform health care delivery using innovative technologies, and what implications will that have on our current approaches and assumptions — this is as much about clinical change as political will;
• what technologies can we have now, today or soon;
• how can we use reimbursement/payment systems to encourage use and uptake, and
• why is healthcare so slow to adopt new technologies?

Progress in healthcare can come from changes to the way clinical work is done.  An example is interventional radiology, which combines radiological investigation with treatment, in a single step.  It moves radiological technologies, such as MRI, CT, Ultrasound, from being mere diagnostic technologies to integration into the surgical work itself.

So why the slow uptake in the UK where a couple of years ago the Healthcare Commission, in one of its investigations, noted that this approach to treatment would have probably saved lives?

The NHS is a slow and late adopter of technologies.  Difficulties giving the necessary clinical freedom to health professionals means that important leading edge, but proven technologies, are slow to be adopted.  The exploration of novel approaches to offering clinical services, outside of hospitals, for instance, in free-standing “theranostic” (therapy and diagnostic) clinics would not only advance the cause of patients, but achieve a step change in service delivery by NHS providers.  Why aren’t the newly freed Foundation Trusts getting on the business of developing services wrapped around this approach to care?

People are obviously of good intent by urging reviews of funding to elected officials in the suitably hushed setting of the House of Commons, but in the gritty reality of healthcare delivery, creative solutions are needed to address not only the timely implementation of interventional radiology, but also overcome the fear of change, of novel technologies and of changes to  clinical practice that change and technology brings.

Central to all healthcare systems is the notion of the hospital.  Are these remnants of industrial-age or can they be rethought and refreshed for the post-industrial and information world we are likely to inhabit for some time?  Foucault spoke of the birth of the clinic (hospital); I will write about its demise.

The logic of hospitals has a lot to do with aggregation of technologies and brains.  It is easier to move the patient to the hospital where integrated systems kick in and provide care, than to have all that expertise go to the patient.  That paradigm is getting tired, but yet our thinking is still hospital-oriented.  What is the way out?

Evolution of artificial intelligence systems, for instance, points to the possibility of remote locations having access to clinical brains, either embedded in portable diagnostic technologies, or through distributed intelligent systems, or even more mundanely at the end of a telephone.  Perhaps it will take time to be comfortable with robotic surgeons, but remote manipulation of robotic surgical equipment is not inconceivable in daily use.

A rather interesting book from the early 1970s, by Maxmen, The Post-Physician Era, offered thinking about the direction of travel.  While getting many things wrong — we still don’t have shopping malls on the moon, he did, given the thinking of the day, accurately identify AI as a challenge to human diagnosis, and saw the obsolescence of the pharmacist through robotic dispensing.

The overall forces at work here are the migration of specialist human knowledge into devices and into software, that can be used by less-skilled people (i.e. not necessarily clinical professionals).  Self-diagnostic testing kits are just a primitive example.  Roll the clock forward with electronic health records, Web  2+.whatever, and advances in materials science, etc, and we have a constellation of factors which form a new pattern for healthcare service delivery.

And when will we build the last hospital?

It takes perhaps 3-5 years to plan a hospital and a couple to build one.  It is also critical in the design to take into consideration the evolution of use, changing demography, etc, to perhaps 20 years into the future.   I think by 2025 we will acknowledge that the existing hospital infrastructure should not be replaced, but slowly wound down as useful clinical environments.  Given the average useful lifespan of anything from 25 to 100 years, we need to be thinking the thoughts about the last hospital within the next 5 to 7 years.  There are, no doubt, hospitals in the early planning stages, that when built will be instantly obsolete.

Tempus fugit.