«Edited by ANNE MASON Research Fellow, Centre for Health Economics University of York and ADRIAN TOWSE Director, Ofﬁce of Health Economics Radcliffe ...»
Of course these authors are coming from an egalitarian perspective. User charges for libertarians empower consumers to express their freedom of choice, and they have less concern about the equity consequences of ‘taxing’ the ill, who are also often poor and elderly. As Stoddart notes, such instruments may also increase expenditure, with the asymmetry of market information facilitating supplier-induced demand.
The cycle of advocacy of changes in healthcare funding in socialised systems represents a continuing attempt by libertarians to undermine the egalitarian structures of healthcare ﬁnance. Often this advocacy of funding reform is accompanied by arguments that demand cannot be funded alone from tax/social insurance ﬁnance, even though such ‘single-pipe’ funding is a necessary, if not sufﬁcient condition for healthcare expenditure control, and that the instruments advocated by libertarians are unlikely to produce efﬁciency gains (see Tables 4.2 and 4.3). Nevertheless much sound and fury is spent on what is, in effect, an irrelevant debate for egalitarians.
However the mutual posturing of both sides of such debates appears to serve a purpose which is utility generating for each participant, i.e. with policy focused on the largely irrelevant demand/funding side, everyone can ignore supply-side inefﬁciencies which for decades have been well evidenced and maintained by a conspiracy of silence.
IDEOLOGY AND THE SUPPLY SIDEWhatever the nature of the organisation of a healthcare system, all appear to exhibit a common problem: a reluctance to deal with well-evidenced supplyside failures in terms of healthcare process and outcome. Let us brieﬂy review these problems: the application of the evidence base, the ubiquitous nature of clinical practice variation, and the absence of outcome measurement and management.
The application of the evidence base Firstly it is evident that despite the work of the Cochrane Collaboration, a considerable portion of medicine has no evidence base. The James Lind Alliance aims to establish partnerships between patients and clinicians to identify and prioritise current uncertainties about medical practice and to
THE PUBLIC–PRIVATE CHALLENGE IN HEALTH CARE 37ensure that they inform future research agendas. One product of the Lind alliance is a Database of Uncertainties about the Effectiveness of Treatments (DUETs, accessible at www.duets.nhs.uk). The extent of the uncertainty about clinical effectiveness is considerable (see Figure 4.1).
FIGURE 4.1 Uncertainty about clinical effectiveness.
Source: Figure taken from the BMJ Clinical Evidence website ‘About us’ section, www.clinicalevidence.com [accessed 5 September 2007]. Used with permission from the BMJ Publishing Group Ltd.
Such information is pertinent when investing in medical care. However, the important issue is that evidence-based treatments are not delivered to patients, thereby increasing morbidity and mortality. This is particularly the case for the chronically ill, where there are established ways of identifying patients in need, and relatively cheap and cost-effective treatments, although side-effects may affect compliance. Universally we can see evidence of this. For example, in the USA where 16% of a large GDP is spent on health care, Americans allegedly get only 55% of appropriate care (Kerr et al., 2004). Similar problems can be seen in most developed countries, middle-income countries such as China and Egypt, and low-income countries such as Kyrgyzstan.
In Britain, one policy response to this problem is the QOF in the 2004 contract for primary care (see Box 4.3). Despite the defects of the QOF, it is a notable attempt to remedy a long-term problem of failure to implement inter ventions of proven efﬁcacy and it can be improved over time. Parallel attempts in the USA have generally been feeble, with report systems being used on a voluntary basis by insurers and managed care companies, with some participants insisting that their responses are conﬁdential. ‘Competitive’ markets and commercial incentives do not improve compliance in the absence of regulation!
38 THE IDEAS AND INFLUENCE OF ALAN WILLIAMS
BOX 4.3 PRIMARY CARE INNOVATIONS IN BRITAINThe Quality and Outcomes Framework (QOF), introduced to NHS general practice in 2004, is based on ‘points’ collected by practices for meeting various clinical and organisational targets. This largely rewards primary care practices for doing what they should have been doing anyway and is deemed a ‘success’ by government, as achievement levels of 91% have exceeded planned targets of 75%. This success is ambiguous as there are no ‘before’ data to compare with the ‘after’ achievement levels. Further, the evidence base for the 10 clinical activities targeted by generous fee-for-service incentives has been criticised (Fleetcroft and Cookson, 2006), and the weighting between the categories appears to reﬂect not their relative clinical importance in terms of improving population health but the estimated workload for GP practices.
Variations in clinical practice With so little in medicine certain and supported by the evidence base, it is unsurprising that different clinicians treat similar patients in very different ways! Clinical practice variation has been the subject of considerable analysis for decades, and remarkably this evidence has had very little impact on practice and policy: the inefﬁciencies are well recognised by libertarians and egalitarians, both of whom have failed to mitigate them with effective supplyside reforms.
For example, in the US, Jack Wennberg and his colleagues at the Dartmouth Medical School have carried out the most sustained and insightful of this work. For over two decades this group has analysed Medicare data, initially comparing similar areas (e.g. New Haven and Boston) and more recently charting the geographical variations across the country (Wennberg et al., 1987, 1989). For instance US Medicare spending per capita in 2000 was $10 550 per enrolee in Manhattan and $4823 in Portland, Oregon. The differences were ascribed to volume effects rather than illness differences, socio-economic
status or the price of services. Fisher summarised this work as follows:
He went on to suggest that potential savings of 30% of the Medicare budget were possible if high spenders reduced expenditure and provided the safe practices of conservative treatment areas.
Recently in England as the government became frustrated by the reluctance of the NHS to follow the Prime Minister’s dictum and ‘act smarter’, academic
THE PUBLIC–PRIVATE CHALLENGE IN HEALTH CARE 39research and government agencies, in particular the Modernisation Agency and its successor the NHS Institute for Innovation have again emphasised the scope for ‘efﬁciency savings’. For instance the work of Bloor, who used routine NHS Hospital Episode Statistics to identify signiﬁcant differences in consultant activity measured in terms of volume (ﬁnished consultant episodes or FCEs) and in terms of ‘value’ (i.e. volume multiplied by Healthcare Resource Group (HRG) price), was an input into debates within the Department of Health about the adoption of fee-for-service payments to consultants as a means of shifting the modest mean of this distribution and reducing dispersion.
However, these data had little impact on an NHS that continues to use its own routine data to inform management by clinicians and non-clinicians (Bloor and Maynard, 2002, 2006).
One current focus of UK NHS reform is improving knowledge and management of variations in the adoption of proven technologies and reducing variations in clinical practice. This is remarkably reminiscent of the efforts of the Wilson Labour Government 30 years ago. The then Department of Health and Social Security (DHSS) sought to get the NHS to focus on the evidence base for day case surgery and the reduction of variations in clinical practice (DHSS, 1976). The failure of 30 years of NHS ‘redisorganisations’ of structures to alter clinical processes is noticeable!
The absence of outcome measurement and management Williams dedicated his life to persuading reformers, clinical and political, to the cause of patient reported outcomes measures (PROM). Whilst he had some success in getting such quantiﬁcation into health technology assessment, in particular the use of QALYs in the National Institute for Health and Clinical Excellence (NICE), one of his ﬁnal papers (Kind and Williams, 2004) again emphasised the absence of routine PROM measurement and management in the NHS. There are now signs that policy makers may invest in PROM and the measurement of, for instance, physical and mental functioning before and after elective procedures in the NHS. This could revolutionise the management of medical practice by making clinical activity more transparent as well as facilitating systematic appraisal of the success of the NHS in improving health.
IDEOLOGY AND ADJUSTMENT MECHANISMS: HEALTHCARE REFORMThe libertarian and egalitarian protagonists engaged in ideological battles for healthcare reform focus on the ‘ideal’ characteristics of the adjustment mechanism (see Table 4.2), even though the ‘actual’ nature of their systems are well deﬁned (see Table 4.3). The failure of both models to produce efﬁcient outcomes is a product of the incentives facing both consumers and providers.
In egalitarian systems, and ignoring the largely irrelevant arguments for funding changes proffered by libertarians, the usual focus of reform is the use of ‘competition’ in either the management of funds and/or the delivery of
40 THE IDEAS AND INFLUENCE OF ALAN WILLIAMSservices. Thus, in countries such as the Netherlands and Switzerland, ‘choice’ of insuring sickness fund is seen as an essential part of using competition to improve resource allocation. Whether these measures improve choice and the efﬁciency of service delivery for patients is unclear (Herzlinger and ParsaParsi, 2004; Reinhardt, 2004; Schut and Van de Ven, 2005).
One lesson to be learnt from this literature is the need for careful regulation of competition and the integration of each component part of the reform package so that they are complementary rather than conﬂicting. This is nicely epitomised by the Dutch approach to competition, and in Enthoven’s advocacy of regulated competition. In the latter’s Jackson Hole reform proposals for the USA (Ellwood et al., 1992), there was to be regulation of premium setting for competing insurers, deﬁnition of a basic package, which was to be guaranteed to the less afﬂuent through tax subsidies, and regulation to ensure both health technology appraisal and the national measurement of outcomes. These structures aimed to ensure risk spreading in the insurance system and to control the supplier-induced demand excesses of doctors. Regulation in the Dutch system includes risk-related premiums so that competing insurers are not disadvantaged by having disproportionate numbers of poor risks, and with detailed regulation of the supply side including anti-trust regulation to break up provider cartels.
The careful articulation of a regulatory framework in these schemes contrasts with the Blair Government’s conversion to ‘constructive discomfort’ (Stevens, 2004) and contestability in market structures in the English NHS.
Here a series of radical innovations has preceded the construction of a regulatory framework. This ‘cart before the horse’ approach has created ﬁnancial instability and associated media pressure about the ‘failure’ of the Blair reforms. Underlying this controversy is a fundamental issue related again to the choice of ideology, i.e. is the goal of policy to create a demand-led NHS or one based as in the past, on need?
This clash of objectives is epitomised by the tariffs policy (known in Britain as ‘payment by results’ (PbR), but similar to diagnostic-related groups (DRGs) in other systems). The issue here is the role of PbR in a cash-limited, public insurance/NHS system. In 2006–2007 tariffs are determining 80% of hospital income. For elective procedures there is a set of tariffs with no cap. For emergency procedures the full tariff is paid only for volume up to 2004–2005 levels plus 3%, after which the tariff is reduced by 50%. However again there is no cap and a 50% tariff can be attractive for hospitals operating at marginal cost. Those hospitals with capacity and costs below the PbR average have a clear incentive to trade up, while those with higher costs have to meet access targets and are also incentivised to increase throughput and maximise income.
Unsurprisingly, meeting local demand in this way contributes to expenditure exceeding budgets.
However, local budgets are determined by a weighted capitation formula that reﬂects local need through crude proxies such as mortality and pertinent,