A stroke of genius

'Four years ago the mortality rate among stroke patients was one of the worst in the country; now it is one of the best.  The unit was spending more on stroke care than it received - today it is in the black.

How was this done?  To begin with, Dr Steve Allder, a Consultant Neurologist, who managed the turnaround, carried out detailed research into the incidence and variability of strokes in the area.  Then, using this data, he tailored the systems' response not to targets set by Whitehall but to the care that patients actually needed.  Response times improved and the provision of immediate specialist treatment meant patients were released from hospital sooner.  Costs were halved and death rates slashed.  It did not involve spending more money, or building a new unit or shaking up the structure, but analysing what patients needed and designing a system to meet it.  The problem was not a lack of resources; it was the way they were used'

This article by Simon Caulkin was first published in The Telegraph, 14 March 2011.  It remains a remarkable story of transformation and a true economic benchmark as well as one of the NHS's best kept secrets.  In focussing on delivering what mattered from a users point of view, the team in Plymouth improved the lives of countless people and their families and saved the lives of countless others.

Hear Dr Allder talk in his own words about the counterintuitive insights he and his team discovered when they went about studying their system.  The numbers speak for themselves (see below); a 23% cost benefit to the acute hospital; a 17% reduction in externally commissioned rehabilitation beds; quicker care and in the right place.

Its an inspirational story of genuine medical leadership.  How did he do it?  He started by changing thinking.  He ended by saving lives.

Domain	Before	Now
% stroke patients treated in ASU	60%	100%
Average time to transfer to ASU	1.6 days	2 hours
% patients spending at least 90% of their time on ASU	30%	80%
Average LOS	16 days	6 days
Bed base	56 beds	39 beds
Cost per patient	£6k	£3k
Profit per patient	(£2k)	£1k

Domain	Cost Benefit
Net cost benefit released in Acute	23% Neurology Budget
Cost reduction in externally commissioned rehab beds	17%

A product of our thinking

The Nuffield Trust are running a summit today.  They may be talking about another report they have produced which tracks spending and labour productivity over the years of the reforms.  It concludes

‘If we are to maintain care free at the point of use, only three things can fill the widening gap between rising demand for health care and limited budgets. Either the public pays more for care; or the NHS rations the care provided; or it becomes more productive. Whether or not the last option can ever fill the hole entirely, taxpayers and patients have the right to expect the NHS to search for every opportunity to improve efficiency.’

The evidence from our studies of health systems does not support such a conclusion.  We find:

• No evidence that value demand is rising.  None
• No evidence that an aging population is the problem described
• No evidence that the NHS has a productivity problem – although the fact that it thinks it does is a problem

To conclude what the Nuffield Trust has is, therefore, illogical and unnecessary.  Worse, it may, in fact, add to costs and cause harm. 

The NHS and many of its commentators do not understand demand at all.  Instead of viewing all work that is done as work that needs to be done, better is to understand the difference between value demand – the things people want help with – and failure demand – the failure to do something or do something right.

Our evidence is that value demand is actually stable.  Failure demand on the other hand is increasing as the system generates demand in its plausible but ultimately wrong headed attempts at managing demand.  That is where the real lever for change is.  But how can you leverage that change if you don’t understand what demand actually is?

As for their ‘only’ three things to save the NHS.  Well, I don’t want to have to pay for someone to repeatedly assess me but not understand and solve my problem.  I don’t want to pay for a system that is designed to deal with ‘fix me’ demands when the biggest demand people make is ‘help me’.  I don’t want to pay for a system that prioritises access over success.  Nor do I want to pay for a system whose leaders think that age is a driver of health care costs when all the evidence contradicts that conclusion (Zweifel et al (1999), Shamani and Gray (2004), Zweifel et al (2004) and Werblow et al (2007).  The OECD (2006) estimated that ageing contributed just 0.4% of the average 4.3% annual growth in healthcare spending across all OECD countries between 1970 and 2002.  Breyer et al 2011 estimate the range of impact to be between 1.2% and 1.5%.  Cutler (1995), Newhouse (1992) and Smith et al (2000) all argue that the evidence suggests that the ageing of the population contribute only 2% to the overall increase in cost.)

In that context the idea that the NHS needs to ration care is a myth of biblical proportions.  Don’t ration the care you offer me.  Provide me instead with what I need to solve my problem.  It’s better for me - and cheaper too.

And please, don’t focus on making a system that is not designed to solve my problem more productive at not doing it.  Doing the wrong thing, but faster is still the wrong thing.  It leads people to manage their costs even though managing costs leads to costs going up.  A focus on productivity is counter productive if the aim is to create a more effective and affordable system.  It won’t work.

It concerns me that the Nuffield Trust believes there are ‘only’ three things that can be done to keep the NHS free at the point of use (leaving aside the glaring irony that the public – that’s you and me – pay more for it…).

They would have us believe that economies of scale is the answer, when its economies of flow that is needed; that the system should be standardized when it should be designed to absorb variety; that the system should be designed for complexity (or special cause), not frequency; that extrinsic incentives work when the real challenge for leaders is to design a system that people are intrinsically motivated to work in; that quality can be inspected in when it can only be designed in…

I do not believe the Nuffield Trust is right to focus on the three things they describe.  Better would be to apply these 12 words and 2principles.  Some systems already are.  And they are proving that there is a better way.

Health IT - an elegant exercise in wishful thinking

Health Information Technology has been in the headlines this past week with Jeremy Hunt committing the "NHS to be paperless by 2018".  Why: because it will "save billions, improve services and help meet the challenges of an ageing population".  £4.4bn of savings in fact according to a PwC report.

It all sounds perfectly plausible, but the savings will be completely illusory.  Take a look here for a brilliant exposition by Systems Thinking for Girls as to why.  In fact, I predict that the investment in health IT, as currently proposed will cost, not save money.  Given that successive governments have been going at this 'problem' since 1998 and the £12bn debacle that was NPfIT (the Public Accounts committee concluded that electronic records system was 'unworkable' and the government admitted that it had wasted the money and was closing down the programme) making such a prediction has to be one of the closest things to a 'dead cert' bet.  The expected savings from these programmes can never be realised because the underlying logic about how savings are generated is flawed.  These wrong headed expectations are probably exemplified best by a 2005 report by RAND. 

In it RAND predicted that adoption of an Electronic Medical Record would generate savings of $81bn PER YEAR for an investment of $20bn.  These savings are based on extrapolating the savings generated from the reduction in time people spend 'doing' things.  The approach to health IT is a manifestation of the dominant command and control management paradigm.  The logic is:

• identify how much work is coming in
• work out how many people there are doing the work
• assess how long it takes for people to do things

Under this production-driven view activity equals cost.  With this logic efficiency becomes; the more we reduce activity times the more we cut costs.  Leaders then focus on three things:

• standardising work
• reducing activity times (because activity = cost)
• driving out waste

From this logic, if Health IT reduces the steps in a process, or prevents duplication, or speeds up the time it takes to get information, this all saves peoples time and, as a result, will save money.  These savings can be extrapolated across the whole system and, voila, several billions can be saved.  With respect to the RAND study the President was convinced and the investments were approved.  It sounds unbelievable that such extrapolations would be made - but it is unsurprising given the above perspective about how the work works.

Equally unsurprising, the savings failed to materialise as a revised Rand study shows.  Rather than saving money it seems the investment may have cost money .  Healthcare spending in the US has, in fact, risen by $800bn since the 2005 report.  It seems that the billions invested in IT have made it easier for providers to bill for services.  How ironic.

From our studies, massive investment in Health IT, aside from there being no evidence of their benefit, is the wrong problem to solve.  Developing and tending to relationships, end-to-end over time, understanding what matters to people in the context of how they live their life is what the system needs to orientate itself to.  But there will be no time for people to do this because they will be spending countless unproductive hours either implementing a new electronic system or dealing with the problems the system will cause.  There will be much gnashing of teeth as NHS staff will appear to look even more unproductive.  'We need another technological solution' will be the cry.  And so the cycle will continue...

The argument for Health IT is that by putting all the information relating to all a persons transactions in one place better clinical decisions will be made.  It will be safer; cheaper.  But here's a test.  Take a persons current paper medical records.  Place all the assessments end to end.  Would reading them give you a sense of the person as a whole and help you to understand them and the context in which they live their life and thereby help the system to solve their problem?  Our evidence is that it will not.  That evidence alone should be enough to prevent this headlong dash into doing more of the wrong thing faster.  But that is not the only evidence.

A long term US study of people going online to view their clinical records concluded that online access, rather than reducing demand on health services, was associated with more use of clinical services.   “Contrary to expectations and the results of some prior studies,” the study authors found “a significant increase in the per-member rates of office visits and telephone encounters” by online patients. There was also a significant increase in clinic visits after hours.

But stark evidence of failure is not enough to convince supporters of the redemptive powers of Health IT that their logic is wrong.  Instead they argue that what is needed is more and better application of the same flawed logic.  To be fair to PwC, despite making grand claims about the savings to be realised, they at least say “significant further work is required to further substantiate some of the evaluations of potential benefit, and especially the evaluations of potential financial benefit.”

This will not stop leaders:

• ignoring negative reports
• committing millions to 'incentivising' providers and doctors to use new systems or
• estimating quality and productivity benefits that do not (and cannot) materialise

It's why Groopman and Hartzband call investment in Health IT an 'elegant exercise in wishful thinking'.  I wonder if Health IT and the benefits claimed for it are more like the management equivalent of homeopathy, something close to the Secretary of States heart.

The way Health IT is currently conceived will cost not save money.  Worse it will not solve peoples' problems, the real point of leverage in the system.  There is a better way, one that might end up with an electronic patient record, but certainly doesn't start there.

By the way, it's not all bad news.  RAND’s 2005 report was paid for by a group of companies, including General Electric and Cerner Corporation, that have profited by developing and selling electronic records systems to hospitals and physician practices. Cerner’s revenue has nearly tripled since the report was released, to a projected $3 billion in 2013, from $1 billion in 2005.