Hodgson: Cost-effectiveness in NZ Health System
Hon Pete Hodgson
Minister of Health
30 June 2006 Embargoed until midday Speech
Cost-effectiveness in the New Zealand Health System
Christchurch School of Medicine
It is a great pleasure to be invited to contribute to
the Mid-Winter Series Dialogue. I hope my remarks are useful
and I am sure the ensuing dialogue will be very
Tomorrow a new financial year begins and I have set myself the task of making progress in six priority areas [#1]. Here they are:
We must get ahead of the curve on chronic disease, progress the health of older people, introduce a range of new initiatives in child health and complete the Primary Health Care Strategy roll out. The fifth priority I have labelled infrastructure – and it has two parts –information technology and the health workforce.
Each of these five areas is important, large, and any one of them could easily be the subject of this address.
But I have chosen to speak today on the sixth priority – improved cost- effectiveness.
Public discourse on health often runs along two contradictory threads. One is that society wants more health services; the other is that we cannot afford them. Cost-effectiveness brings these two threads together. If we improve the cost-effectiveness of a finite health budget we can improve health care at no additional cost.
That is simply a truism, an obvious fact. Seeking improved cost-effectiveness is a permanent feature of all public health systems, and so are the two contradictory threads – that societies want more health care but can’t afford it. In different countries this debate is shaded differently.
For example, Britain has increased expenditure on health by 7 per cent per annum, after inflation, for nearly a decade. It has secured very significant gains from doing so, yet the public worry that the gains do not adequately reflect investment.
The United States spends 14 per cent of its huge per capita GDP expenditure on health yet their system is regarded by very many US commentators as a partial failure or as somewhat dysfunctional. They spend three times per capita more than we do; yet New Zealanders live both longer and healthier lives.
Since the change of government six and a half years ago, New Zealand has increased expenditure faster than total government expenditure or total economic growth. In the area of elective services, angioplasties have risen by 75 per cent since then, cataract surgery is rising by 50 per cent and major joint surgery is being increased by 100 per cent, to amongst the highest intervention rates in the world, yet there is still a queue.
Health is different. It is different because it is mostly free of charge, because markets can both hinder and help progress, because factors outside the health system impact substantially on health and because a society’s propensity to spend changes as a society becomes richer.
This last point is best illustrated in this graph [#2].
If a nation’s per capita wealth increases by 10 per cent, expenditure on health increases by greater than 10 per cent. In other words the line is not horizontal. That is why health assumes a greater proportion of total economic expenditure as a nation's wealth increases. Note how nations are closely clustered around this line. There is relatively little variation, suggesting that societies choose to apply an ever-increasing proportion of the discretionary spending on health. Note also that the United States is a uniquely expensive outlier. Whatever we do in health policy now or in the future the experience of the United States offers clear evidence of what not to do.
New Zealand's Health System is already very
The good news comes in the next slide [#3].
The New Zealand health system is already cost effective. The top left hand quadrant is the place to be, and moving further into that quadrant is the holy grail of cost-effectiveness. Unlike the previous slide, this slide demonstrates no strong clustering. Instead it demonstrates that the linkage between investment and health outcome isn’t strong – that influences outside the health system and cost-effectiveness within the health system matter a lot. It also demonstrates yet again that the US health system tells us very clearly where we should not be going.
You may wish to ponder why we are cost-effective. I certainly do. I wonder whether it is a combination of our size and intimacy, our seven decades of public health delivery, the medico-legal advantage of having a no-fault accident compensation system, our attention to the determinants of health or simply the influence of uncomplicated common sense.
New Zealand's Drive for Equity
Cost is an easy concept to grasp, effectiveness less so. Life expectancy is the obvious high-level measure of effectiveness, but so is health expectancy. Some might argue for other measures – access, quality, timeliness, priority – the list could go on. But sticking with life expectancy the next slide [#4] should give us cause for pause. Not all is rosy.
This slide shows two sobering gradients – socio-economic and ethnic. Rich people live longer than poor people and non-Maori live longer than Maori. Putting the two together we find that a poor non-Maori has a slightly better life expectancy than a reasonably well to do Maori.
These gradients are repeated throughout the New Zealand health system, almost irrespective of what is measured – whether it is by geography, disease incidence, timely access to treatment, predisposing conditions or whatever.
When all census data is combined with all mortality data over an extended period of time the result is a landmark study, the third and final chapter of which has just been released. It is called ‘Decades of Disparity’ and it quantifies somewhat the disparity that can be explained by ethnicity alone – that part of the disparity that has a cultural origin or which is the product of alienation or exclusion. Some call it institutional racism, but whatever its label it is solid reason to have more Maori in the health workforce and to pay more attention to the delivery of health services to Maori, particularly primary health services. Dealing with disparities improves effectiveness.
Of course other ethnic, gender, geographic and especially socio-economic gradients exist and they matter. Attending to all of them, especially through the Primary Health Care Strategy, is a significant feature of our evolving health system.
Now that the third chapter of this census-mortality study is out, the researchers have advised that they intend to change the name of the series from Decades of Disparity to some other title. The reason is that they have discovered that the disparity is narrowing and they anticipate the narrowing to persist or even accelerate. It is very early days yet, but it is a hopeful sign. As the browning of our population continues, addressing such disparities is not only socially just, it is straight forwardly necessary.
This Labour-led government believes cost-effectiveness isn’t [#5] something to be achieved by putting a competitive market model at the core of the system, or by shifting costs onto individuals through part charging, or by deflating salaries, or by removing democratic input, or by increasing private sector capacity for the sake of doing so, or by attacking universality.
I have, of course just described the nineties. More concerning, I have also just described the still current view of those politicians sitting opposite the Government. Labour and National both say cost-effectiveness is an important feature of health policy but Labour doesn’t use the term as code for slash and burn. We are committed to building New Zealand's health infrastructure and service, and we are committed in the long-term.
National's prescription is preoccupied with short-term gain. In the nineties salaries were subject to downward pressure. Over the short-term costs go down and cost-effectiveness goes up. The problem is that before long there are not enough nurses. Which is precisely where the need for last year’s pay-jolt for nurses came from.
Another short-term technique to improve cost-effectiveness is to increase the barriers to accessing health care [#6].
If the costs of accessing the health system go up then access reduces, costs go down and cost-effectiveness goes up. Soon after, however, cost-effectiveness will inevitably plummet when the patient does finally show up, in a sorrier and more expensive state than might otherwise have been the case. New Zealand knows this better than most countries. We experimented on ourselves in the nineties and through “Decades of Disparity” we measured the result. We need to learn from that history, avoid a re-run of it, and approach cost-effectiveness as a long-term and permanent idea.
Let's now move from the general to the particular, and look at some examples. We should start with two simple examples, then delve later into some more complex ones.
Simple Examples of
This is Pharmac’s view of how New Zealand’s drug costs would look if we did not deploy a central procurement system [#7].
The savings are very large about $900m per year, about 9 per cent of the total health budget, yet the volume and mix is comparable to similar countries. The buying techniques Pharmac deploys, especially reference pricing, come under understandable scrutiny from pharmaceutical companies. Less understandably they come under attack from the current crop of National MP’s, even though Pharmac began under National. Why any New Zealand politician would attack Pharmac's role in the pharmaceuticals market is beyond me. A new generation of drugs, especially cancer drugs, is the latest challenge Pharmac faces, Herceptin being just one. That challenge, like earlier ones, needs to be addressed. The Government has embarked on developing a long-term medicines strategy as part of its agreement with the United Party.
One of my jobs is to see that central procurement is extended beyond drugs, into consumables or communications or transport, or into insurance – which is where we have made a start. District Health Boards have recently decided to club together to buy their insurance as a collective and to save money by doing so.
A second simple example of cost-effectiveness seeks not to decrease costs, but to increase them a little, in order to increase effectiveness a lot.
Pharmac will do again, for this example. To an extent our society tends to overuse drugs or to overly medicalise conditions. But we under-use some drugs, such as statins and this under use follows the sadly familiar gradients of socio-economic and ethnic status. Notwithstanding ongoing criticism directed by some toward Pharmac and their alleged approach to statins, the most recent round of which was last weekend, they have actually been out and about seeking to increase the uptake of statins, with much success. Statin use has risen to almost an optimal rate of use and a reduction in cardiovascular disease can be anticipated.
It is at this point that cost-effectiveness in health becomes a little more complex, and one reason is that it is hard to work out where health starts and stops. Some cost effective measures are clearly part of the health system. In the personal health arena green prescriptions will do as an example and in public health we can point to many things, such as immunisation.
But other examples are more dependent on action outside the health system, such as the Healthy Eating Healthy Action strategy that seeks to address the obesity epidemic - the most important public health challenge facing New Zealand families. Fluoridation is another example.
As we move progressively outside the health system we soon arrive [#8] at The Ottawa Charter, which reminds us that the so called determinants of health lie substantially outside the health system – education, housing, employment, household wealth, a sense of inclusion, good air and water
The Minister of Health is therefore mighty pleased that we have very low unemployment, that we replaced market-based state housing rentals with income-related rents, that the minimum wage has gone up by about 50 per cent since the change of Government, that the Working for Families package has rolled out, that a high premium is placed on tolerance and inclusion, that there is a national environmental standard for air quality, or that we have a high participation rate in tertiary education. These determinants of health, as much as a good health system, contribute to good health outcomes. Failure to improve the determinants of health, as much as the era of health restructuring in the nineties, caused the findings in the Decades of Disparity study.
Here is an
Graph [#9] shows a startling health gain for which there may be any number of explanations. In an ideal world children shouldn’t be admitted to hospital for avoidable conditions. In South Auckland the numbers have dropped recently and significantly. Why is this?
It could be insulation retrofitting that I paid attention to as Ministry of Energy, or the Healthy Homes project that Mark Goshe got underway as Minster of Housing, or the Primary Health Care Strategy begun by Annette King, or the rise in employment, or vaccination, or cheaper housing rentals, or a milder winter – or all of the above. But it's probably more the case that families and communities have prospered because people have had greater continuity of income and public services, and that both individual responsibility and civic action have flourished in response.
So having acknowledged the determinants of health, and having acknowledged the power of a newly empowered community, let me return to the health system, to cost-effectiveness and lets begin to look at complex examples.
Let's start with size and let's start here in Canterbury. I can’t name an outfit in Canterbury that is bigger than the Canterbury DHB. Biggest budget, biggest employer, biggest funder of contracts. [#10]
Add to size, medicine. Humans are complex and different. And that’s just the patients. There is also an astonishing array of different professional and administrative expertise in this DHB as in others.
To this complexity we need to add the regulatory framework, industrial relations, budget constraints, patient expectations, fiefdom behaviour, tradition, new technology, public scrutiny and politics.
So health is both big and complex. Changes to health delivery are continuous and endless yet those changes are resisted by the inertia that resides, innately, in big and complex systems. Health professionals are always seeking improvement but often attempt it in isolation because it's easier to do that than to change the system as a whole. This leads to the interesting phenomenon, which someone has labelled 'drowning in good ideas'. Ideas, which in isolation, make a lot of sense, are often frustrated as the system pushes back, because that is what systems do.
So how do changes to the system occur? How can a big and complex system become more responsive? And how can cost-effectiveness in the system as a whole be achieved?
A good place to start is the primary health care system. By international standards we already have a good primary health care system. The Commonwealth Fund in Washington measures us against Australia, UK, US, Canada and Germany, we come out well or very well on the variety of measures and we rank second to Germany overall. Our system is easy to access, and patients report favourably on many aspects of care. However these same patients may not have gone to their doctor early enough, may not pick up their prescriptions or not complete their course of drugs, and some costs are higher than some other countries. In addition the system as a whole is not responsive enough to chronic disease.
Last week [#11] a Christchurch School of Medicine study hit the popular press showing that about a third of all hospital admissions were avoidable and that those admissions cost about $100m. Unsurprisingly chronic disease featured strongly.
Avoidable hospital admissions went up quite steeply in the late eighties and early-mid nineties, since then they have more or less plateaued. The trick now is to reduce them.
Both studies are a ringing endorsement of the government’s Primary Health Care Strategy. [#12]
This strategy seeks to move primary health care towards a population based model, to increase emphasis on the prevention, early detection and active management of chronic disease, to undertake outreach to people or communities who don’t use primary health care services adequately or early enough, to undertake health promotion and lastly to reduce doctor’s fees.
The strategy was strengthened considerably, and in effect future proofed, Tuesday before last when DHB’s and PHO’s reached agreement on various matters, particularly on a mechanism which stops any recalcitrant GP from hiking fees and in effect pocketing the government's substantial investment in fees reduction. It is not that we had by any means an epidemic of such behaviour. Rather it is that the strategy could not have survived such an epidemic.
From tomorrow 45-64 year olds will see fees reduce, and in one year’s time reduced fees for 25-44 year olds will complete the low fees roll out.
Also tomorrow the first phase of the performance management system kicks in for most practices. It allows GPs to record progress against various standards – blood sugar control or the level of child immunisation and to be rewarded for progress. The system will be progressively expanded as reliable parameters are agreed. The sector is widely supportive of this initiative. It is a very exciting development.
In all the primary health care strategy is a very significant change in New Zealand’s health system and certainly the most dramatic change in primary health care for decades. As it happens the market value of a GP has risen by over 30 per cent in recent years; something to be celebrated if primary health care itself is to be valued.
Much of the development of the strategy is however still in front of us, especially the development of a team-based approach to patient care. Nurses are under-utilised as are many other health professionals. Many remarkable team-based examples of primary care delivery exist, but they are still in the minority.
Into the midst of this sea-change in primary health care I should assert the importance of another cost-effective idea – the expert patient. Many patients are already experts – many diabetics for example. But the idea is far from fully exploited and a patient who receives early medical or nursing advice on the detail of their chronic condition will manage it better than one who rolls up to the clinic because they are in trouble and don’t know why.
If improving cost-effectiveness through better cheaper primary health care is complex then the dynamic interface between primary and secondary care is equally so.
This interface has many different facets so I shall limit myself to just one, access to a first specialist assessment. There are about 450,000 such assessments each year so there is a lot of traffic.
There is also a lot of politics. The government believes that someone who is accepted for a first specialist assessment should be seen be a specialist within six months. We believe that waiting longer than that is unreasonable. Most are seen well within six months. 450,000 are seen. But every year 10-15,000 are returned to their GP without being seen by a specialist, after they were told they would be, on the basis that it is far better to be reassessed by one’s GP than it is to be seen by no-one. [#13]
And that is where the politics comes in. It may be that only 3 per cent of people are returned to their GP, but those 3 per cent are entitled to be annoyed. For them the system has failed, demonstrably.
The solution lies in further improvements to the primary-secondary interface. For example a number of referrals, some say 25 per cent of them, need not be seen by the specialist. Instead advice to the GP can be given off the records or by a phone discussion or both. Or a person who needs a medical image of some sort may not always need to see a specialist first. Improving the primary-secondary interface is important, but a little complex. It is certainly cost-effective.
If primary health care and the primary-secondary interface are complex examples of improving cost effectiveness then hospitals themselves seem a lot more complex. Yet it is here that some really big gains have been made, by reducing average length of stay over the years, greater use of day surgery, out patient surgery, of new technology and on it goes. Change and improvement is ongoing.
More and more that change and improvement is secured by viewing processes through the eyes of a patient. This thinking has many names and variants. In Waikato they deploy a body of work and experience called practice management. Here in Canterbury attention is paid to patient flow. The Cancer Control Strategy has another lens through which to view similar thinking called the patient’s journey.
All of these ideas seek to design systems around patients. At Flinders in Adelaide for example a person’s discharge planning begins at the point of admission, even at a trivial level. For example if a person is likely to be a short stay patient do they have a set of house keys on them so they can get in the door when they go home? In some places in New Zealand, I'm told that this planning begins even before a patient is admitted to hospital.
This increasing tendency to design systems around patients is coincidental with another tendency, the move towards standardised care. As the body of medical knowledge deepens, more and more certainty arises about how to treat a patient with a given condition. More and more patient care is standardised care.
One of the more interesting papers I
have come across is called “Blending Custom and Standard
It is written by Richard Bohmer, a New Zealand doctor working at the Harvard Business School. The paper – published last year – draws in part on an earlier paper by John Wennberg, which shows wide differences in intervention rates for the same condition in different US cities. The proposition is that this is due to clinician proclivity or local fashion rather than a good evidence base.
Many writers respond to this sort of data with calls for standardised care in which medicine becomes something of a technology and where decisions are made according to a predetermined decision tree. Reductionism is the name of the game.
Bohmer doesn't do that. Instead he acknowledges that patients are different, that exceptions abound and that standardised care must be balanced with customised care. The technology of medicine must be balanced by the science and art of medicine.
For those who are interested the references are shown on this slide [#15]. The third reference is but one example in the literature where the authors tell a startling series of stories about the value of getting the little things done well. In this instance we can learn of the significant gains that can be made by ensuring that a patient's 'bloods' are done on time. We are told, in detail, of the many things that conspire to stop the 'bloods' being done on time and how our heroine, a nurse, overcame that conspiracy. The story sings the praises of a standardised approach, not to the patient's treatment, but to ensuring that 'bloods' are always ready and accurate. These sorts of stories are always a rollicking read because the forces of good prevail.
It is time to draw conclusions. Tomorrow the new year begins and I have listed cost-effectiveness as one of my six priorities. So the question that is begging is “what does a good year’s work on cost-effectiveness look like?” How will the health system be even more cost-effective one year from now?
Well, here is a list of things that I wish to see progressed:
- I want to see more central procurement. The central purchase of insurance is a good start, but what about self-insurance, including ACC. I would like to see plans well advanced, or implementation underway, for a range of other products – consumables or cars or communication products.
- I want to see simple highly cost-effective interventions happening more dependably. The statins story is a good one, but then there are immunisation rates, or retinal screening or the optimal use of aspirin.
- At the primary-secondary interface there is room for a lot of action. The sector has just concluded an analysis of itself and the result is a paper called “addressing disincentives” I am releasing today. That paper identifies and describes barriers in this complex thing called the health system. It reminds us that if a GP accesses hospital radiology or a sleep apnoea test directly, avoiding a specialist assessment that the access criteria have not been developed. It reminds us that if a GP and a specialist discuss a patient, avoiding a specialist assessment, then as far as the accountant is concerned the specialist has not done any work.
- Within hospitals themselves there remain many opportunities for cost-effectiveness and one of the really interesting drivers is that this government is building or refurbishing a lot of hospitals. I announced the 24th and 25th, at Blenheim and Wanganui, on Wednesday. In every case issues of design around patient flow feature strongly, especially in the bigger projects such as Wellington.
But something else happened recently too. Junior doctors had a five day strike and for five days the established culture of managing a hospital was suspended. New relationships were developed, but more importantly new insights were made. Around the country health professionals started thinking about how the system ticked and upon their return junior doctors no doubt joined the conversations. There is no such thing as a good strike and this is a quirky way to end a speech. But in health as elsewhere, necessity is mother of invention, in recent weeks some reinvention has definitely been underway.