Abstract
From a population perspective, the first stage in optimising value is the resource allocation process. Allocative value is optimised when it is not possible to switch resources from one budget to another and get more health for the population as a whole. As emphasised in a previous article, resources are traditionally allocated to institutions, to health centres and hospitals for example, but increasingly resource allocation to different subgroups of the population is coming up the agenda, driven in no small part by the Commissioning for Value Packs of NHS RightCare. 1 Allocating resource to programmes allows a much clearer understanding of what happens when resources are switched from one programme to another, using the method called marginal analysis the origin of which is entertainingly described in the free RAND book called How much is enough. 2
Furthermore, allocating resources to programmes such as the programme for people with visual problems, or the programme for people with mental health problems engages the clinicians involved in thinking hard about choices, as they decide how much to allocate to the various subgroups within that programme. Within the eyes and vision programme for a population, for example, there are six principal subgroups set out below:
People with glaucoma People with age-related macular degeneration People with cataract People with diabetic retinopathy People with low vision or blindness Children with eye problems
Each of these six subgroups needs a system of care designed using the principles described in the preceding article but how is value to be optimised within each system?
Value-based payment
Numerous papers have been published in the last year describing the development of value-based payment in the United States, in particular, within Accountable Care Organisations. The value-based payment relates the outcomes of the patients treated to the resources used in their treatment. However, the meaning of the term value is different in the United States than in the United Kingdom.3–5
The patients’ outcomes and the costs of patient treatment are the two elements in the American definition of value, but in the United Kingdom this would count as efficiency because when looking at the resources that have been used, it is essential for any jurisdiction responsible for a population to identify people who have not been treated and to identify people who may have been treated but have got less value from the service than would have been realised if those resources had been used for the people in greater need who missed out on treatment. The relationship between productivity, the original economists’ term, efficiency and value is shown in the diagram below.
Relationship between Productivity, Efficiency and Value.
Once a system has been designed and set up it is essential to ensure that value is optimised for the population in need. First, it is important to emphasise that the four activities that have dominated debate in the last 20 years are still of vital importance and need to continue and these are set out below:
Preventing disease, disability, dementia and frailty to reduce need Improving outcome by providing effective, evidence-based interventions Improving outcome by increasing quality and safety of process Increasing productivity by reducing cost Ensuring that every individual is helped to make a decision that is in line with their values – personalised decision-making. This will be covered in a separate article. Ensuring the right patients are seen, namely the people in the population who benefit most. Optimising the allocation of resources for the whole care pathway. Introducing high value innovation quickly and comprehensively funded by reducing expenditure of lower value interventions.
These are necessary but not sufficient, and four new activities are required in addition and these are as follows:
Seeing the right people and providing best current knowledge to everyone in need
The system of care for people with broken legs sees all the right people without delay. The system for people with cancer sees all the people with cancer, although there may be inappropriate delays. However, for conditions of which there is no clear-cut diagnosis or dramatic symptom, the set of people with the condition is perfectly distributed between those who reach the services and those who do not. The relationship between the two subsets is shown in Figure 2.
Seeing patients who will get most value.
There is a considerable variation in the rate of presentation to a generalist, in the UK a GP, and then there is variation in the rate of referral from GPs to specialists. This rate of variation can be classified as unwarranted, namely it is rarely justified by variation in need or by the explicit preferences of patients. One of the first, and most elegant, depictions of this phenomenon was with respect to a set of conditions where diagnosis is imprecise and where a degree of variation is not surprising, namely mental health problems. In their classic book called Pathways to community care published in 1980, Goldberg and Huxley 6 used Venn diagrams to illustrate the different subsets of the population in need and developed the concept of filters through which patients have to pass to reach a new level of care.
In developing the concept of population-based medicine or population-based clinical practice to use a more generic term because the same principles apply to other professions such as nursing, pharmacy or physiotherapy, it is important to ensure that every specialist team has at least one person, even one day a week, who has a map on the wall and monitors the source of the people who are referred to the specialist service. Simple interventions can reduce the impact of this problem, not only the publication of clear guidelines but also measures such as the identification of new GPs who arrive in the population and ensuring that those new arrivals are briefed on the indications of referral.
In addition, the specialist services have a role to play in education and support for both patients and generalists so developing population-based healthcare. The specialist service should not only be considered as a piece of real estate concerned only with the people who cross the threshold but as a part of the clinical network with specialist knowledge that needs to be made available to all the people in need and all the professionals whether generalist or specialist.
Optimising the distribution of resources along the care pathway
A system is a set of activities with a common set of objectives, a network is a set of organisations and individuals that deliver the system locally, often called the local service. A pathway is the route that patients follow through the network and there are usually several very well defined pathways that can be described as part of the range of resources for ensuring that the service delivering the system to the defined population does so effectively.
Often, however, different paths of the pathway are funded from different sources and budgets. Prevention may come from one budget, generalist care from another and specialist care from a third and in some places super specialist care comes from a fourth budget. Furthermore, the balance of preventive spend to generalist spend to specialist spend varies from one part of the country to another because the particular balance of resources at any point of time is a reflection of trends that have taken place in the preceding decades. Using a tool called the Socio Technical Allocation of Resources (STAR) 7 the people providing the service, always including patients and carers, can reflect on the balance of resources that they have inherited and make decisions to improve value by moving resources from one part of the pathway to another, from treatment to prevention, or vice versa if that makes it better value.
Obviously there are bureaucratic problems but Directors of Finance are often in a position to be able to manage these shifts if they are involved in the process from the start. When the STAR tool was used by IMPRESS (a collaboration between primary and secondary care experts in respiratory disease) as a means of analysing the relative value of interventions for chronic obstructive pulmonary disease. Their deliberations led to recommendations to shift resources from triple drug therapy to rehabilitation and smoking cessation.
8
It is only possible to do this when there is a clear system with a budget and even though the budget cannot be defined precisely people are able to put estimates on the level of resources being used. Similarly people are able to put estimates on the benefit that would occur if resources were shifted from one part of the pathway to another. As always in decision-making, judgement is required and judgement is partly influenced by the evidence but also requires ethical awareness as Herbert Simon
9
emphasised. “In making administrative decisions it is continually necessary to choose factual premises whose truth or falsehood is not definitely known and cannot be determined with certainty with the information and time available for reaching the decision.”… In ordinary speech there is often confusion between the element of judgment in decision and the ethical element. This confusion is enhanced by the fact that the further the means-end chain is followed, i.e. the greater the ethical element, the more doubtful are the steps in the chain, and the greater is the element of judgment involved in determining what means will contribute to what ends.”
Increasingly clinicians and patients need to work within a defined envelope of resources, still able to bid for more resources but held responsible for reflecting on ways in which resources have been used and taking responsibility for optimising the use of those resources.
Ensuring the introduction of high value innovations
When healthcare budgets are organised by primary and secondary care in huge horizontal slabs there is always reluctance to consider innovation particularly when finances are limited. However, the development of population-based systems and budgets offers another opportunity.
In the old world, an enthusiastic clinician would simply be told ‘there is no more money’ but with the system budget the clinician can ask, after an enthusiastic reception about the importance of the proposed innovation, ‘what do you know is happening at present that is of lower value than this and how can we shift the resources?’ People who pay for healthcare know a little about a lot. For example, a new service for rheumatoid arthritis to improve grip strength through specialist physiotherapy will be understood by the payer but the payer will not know enough about the treatment of rheumatoid arthritis in depth to identify its relative value. Furthermore, the balance of the service for rheumatoid arthritis in one part of the country may be different from another so even if the payer has had experience of dealing with rheumatoid arthritis elsewhere the opportunities for value upshift may be different. This is where the clinician comes in. It is also important to emphasise that this is another reason for programme budgeting because if the rheumatologists requesting resources to invest in a high value intervention can genuinely say that there is no low value intervention that can be cut or reduced but the rheumatologist can then be asked for other activities within the musculoskeletal programme budget to which she might quickly reply ‘injections of the spine for backache’ or some other activity she observes happening which she knows is not based on strong evidence of high value.
We are now moving beyond evidence-based decision-making to evidence and value-based decision-making. Evidence of effectiveness and cost-effectiveness is still necessary but it is not sufficient in the new world of limited resources. Furthermore, there may be a need to support innovation even when there is no evidence of high value. This particularly applies to developments in surgery and the use of technology and it was for this reason that the IDEAL method was developed in Oxford. 10
IDEAL stands for Innovation, Development, Exploration, Assessment and Long term study. It recognises that when a new surgical operation comes along or the innovative use of a piece of complex technology, there may not be strong evidence because there is insufficient activity to organise a randomised controlled trial. Furthermore, there may be a learning curve which means that surgeons will have to start doing the new operation and acquiring the skill to do it well before it can be assessed. The IDEAL method accepts this constraint but requires that all the patients treated with the new technology should be entered in a register so that evidence can start to be collected about the effectiveness of the intervention, and as least as important, the harm.
Health services around the world have been too slow to adopt innovation of high value but the development of systems and programmes a method for funding these innovations is available even when there is not strong evidence from systematic reviews or randomised controlled trials.
Moving to a hybrid organisation
It is vitally important to have well-run primary and secondary care services but there is a need to move from this as the only dimension of healthcare to what Andy Grove called a hybrid organisation, namely an organisation … in which functional units and mission orientated units work together and the accompanying principle of dual reporting, like a democracy, are not great in and of themselves. They just happen to be the best way for any business to be organised.