NICE as an agent of inflation

B&G tell us that by 2002 NICE’s guidance had added an estimated £575 million to the NHS budget. The news, if it is indeed bad, is worse than that – an updated number from 2005 now brings that estimate to £836 million (www.publications.parliament.uk) of a £80 billion annual NHS net expenditure (Department of Health, 2006). In an earlier review of the Ontario Drug Quality and Therapeutics Committee, B&G argued that agreeing to fund new interventions with incremental cost-effectiveness ratios (ICERs) seen as reasonable by historical standards will predictably generate new costs to a health care system (Gafni and Birch, Reference Gafni and Birch2003). They advocate using cost-effectiveness methods instead to explore which substitution of new services for the status quo will improve outcomes within the current budget. Their starting point, then, appears to be the preservation of budget neutrality.

How troubled by health spending growth should we be? That depends on ones view of public spending on health care, given other competing social needs. Certainly in the US, where 45.8 million persons are uninsured (Holahan and Cook, Reference Holahan and Cook2005) in the face of 16% of GDP spent on health care (Smith et al., Reference Smith, Cowan, Sensenig and Catlin2006), there are persuasive reasons to worry about new expenditures. As Fleck has eloquently pointed out, many newly introduced and expensive interventions have little effect on overall population health; but, in adding to the already significant cost burden, they build a system that further diminishes equity in access and outcome (Fleck, Reference Fleck2006).

Concerns regarding growth in health spending need to be seen in the context of the health care system by which one is surrounded. In a system such as the United Kingdom’s National Health Service, a public system that serves all citizens, but has been historically under-financed by industrialized nation standards, budget neutrality may not be the objective to strive for. Value for money, with assurances of comprehensive care provision for all, seems more the target. Were the bloated US health care system to begin to use cost-effectiveness analysis within a federally funded program such as Medicare, wringing out budget neutrality should certainly be on the table. Sadly, this is a far-fetched idea at this time in US health policy. The US tolerates a system where no public entity has responsibility for assessing the value for money of new interventions and, further, the government is legislatively barred from negotiating Medicare drug prices with industry. Unsurprisingly, per capita costs of pharmaceuticals in the US are two fold those in the UK and the gap has been widening. A recent Canadian review of the international experience in drug pricing shows that between 2002 and 2005 the average rate of price increase for non-patented, brand-name drugs in the US was 6.4% per year; during this same period the UK averaged a decline of nearly 8% (PMPRB, 2006).

Health policy in the UK appears to recognize that new treatments are coming along in response to the populace’s ever-growing desires for health and longevity, that they need to be evaluated for worth and value; and that the government needs to find different levers for expanding what is available while controlling costs. We believe that the US’s failure to implement a NICE-like appraisal process in a rapidly expanding private medical sector allows costs to grow far more rapidly than they would with formal economic evaluations in place.

Opportunity costs: what is value for money?

On the opportunity cost issue, B&G return to concerns flagged previously by themselves (Birch and Gafni, Reference Birch and Gafni2002) and others (Sculpher et al., Reference Sculpher, Drummond and O’Brien2001 and Cookson et al., Reference Cookson, McDaid and Maynard2001). Their central worry is that given the fixed budget scenario facing local health care economies in the NHS, making mandatory the uptake of favorable NICE guidance requires local trusts to forego other possibly higher priority investments. They note that implementation of NICE guidance has not resulted in ‘proof’ of net health gains.

We note first that absence of proof of net gains in population health should not be mistaken for lack of effect. Evidence of this sort is not trivial to demonstrate in the short period of time in which new interventions are added to a medical armamentarium. In any event, the problem of opportunity costs is not news to NICE. A 2001 NICE-initiated survey of Directors of Public Health (Davies and Littlejohns, Reference Davies and Littlejohns2002) provided pointed feedback with respect to NICE’s failure to usefully address the prioritization debate as it played out locally. Recent attention on the part of NICE to implementation issues at the local levels, for example, through the creation of a Director for Implementation, together with an NHS budget that has grown by 9% a year for the last several years has forestalled some of these problems, but we agree that much needs to be done.

Opportunity costs remain a vexing methodological challenge that lies at the heart of many of the equity issues that B&G discuss later. A real look at opportunity costs from the government’s perspective would require looking beyond the boundaries of the medical care system to see what gains in health and longevity would be associated with investments in other areas of the public sector. There are resource issues – intellectual capital as well as pounds sterling – in developing an economic evidence base that adequately maps out what is gained and what is lost when every possible health care intervention is compared, and there are therefore likely to be constraints as to what NICE can accomplish. Previously, B&G have laid out approaches to tackling this issue, which they have suggested are not overly ‘data-hungry’ (Gafni and Birch, Reference Gafni and Birch2003); we believe they bear further examination.

We accept that under the current arrangement of centralized decision making the real opportunity cost of NICE’s guidance to the NHS will vary from one locality to another. An inevitable consequence of a national process is variation in opportunity costs (and some inefficiency) across regions. We understand B&G’s fundamental challenge here as questioning whether centralized analysis and decision making are to be preferred over local evaluation and decision making within the NHS. We believe – with a caveat – that centralized analyses and judgments are for the moment preferable. Were it possible to develop centrally a sophisticated set of mathematical programming techniques that permitted local health authorities to model the implications of different investment strategies, the argument for ‘self-rule’ with respect to prioritization of new technologies would grow much stronger.

From a pragmatic perspective however, finding the technological and economic capital to support evidenced-based decision making locally is difficult at present. Local trusts are known to vary enormously with respect to analytic capacity (Williams and Bryan, submitted). In the US it is organizations such as Kaiser-Permanente and Blue Cross Blue Shield Association, with large in-house research and analytic resources that have moved fastest to implement evidence-based practice (Garber, Reference Garber2004). Lastly, from a philosophical perspective, we worry that leaving all decision making to local discretion simply returns the NHS to the difficulty that NICE was intended to solve: inequity of access by reason of geography.

Equity and opportunity costs

Is equity about equivalence in access to services on the basis of need, or is it about fairness in health outcomes? If it is simply the former, there is every reason to believe that the latter will remain elusive for those who are materially disadvantaged (Braveman and Gruskin, Reference Braveman and Gruskin2003). Investing in those whose baseline risk of poor health is higher is almost always inefficient compared with investing in those whose wealth or physical health is better. We do not believe that an equitable national health system can, at a policy level, justify withholding interventions of known overall cost-effectiveness on the basis of underlying health, demography, or geography.

B&G tell us that NICE 2 deemphasizes equity considerations (which, given their concerns with NICE 1, one would have guessed would please them, but no such luck!). They argue that because the new NICE reference case treats all QALYs as being of equal value, regardless of the characteristics of the population to which they accrue, equity – which they seem to understand as fairness in outcomes – is, by definition, unattainable. This is because attaining the same outcome for a disadvantaged group is likely to require greater investment per outcome, hence coming in at a less favorable CE ratio. Following on the arguments advanced by the Panel on Cost-Effectiveness in Health and Medicine, we do not believe that sub-group analysis, here on the basis of personal characteristics such as social class, is an ethically warranted approach to coverage determinations on the basis of cost-effectiveness at the policy-making level (Gold et al., Reference Gold, Siegel, Russell and Weinstein1996). We cannot find an acceptable philosophical perch from which to comfortably bar citizens living in one locality (as a surrogate for class) from access to effective treatment available elsewhere, so that others in that locality can benefit first from the more cost-effective treatments. Such a policy would by region systematically discriminate against people with particular illnesses.

We think that the practice of using a CE ratio as a first piece of information, then to be debated and understood by stakeholders and decision makers as part of a judgment process, remains the more appropriate use of economic analyses. Toward that end, NICE has been proactive in examining views of equity by members of its Citizen’s Council. The June 2006 meeting of the Council took on directly the question of whether NICE, in issuing its guidance for public health and medical care interventions, should take account of health gaps between social classes. Asking citizen representatives whether or not equity of outcome should be explicitly aimed for through NICE guidance is quite a bold question to put to people. Certainly it is not the sort of question that gets asked of the U.S. public by its government. We see this effort as providing an important starting point for broader social discourse as to what the function of a health care program is, and we applaud it.

The other key development concerning equity and NICE is the research programme, sponsored in part by NICE and currently ongoing, to explore public attitudes towards equity issues (http://www.pcpoh.bham.ac.uk/publichealth/nccrm). The funded projects (led by researchers at the Universities of Sheffield and Newcastle-upon-Tyne) are seeking to explore the extent to which the public supports equity weighting as part of technology coverage decision making in the NHS, and the objective is to deliver information that can be used by NICE as part of its appraisal process. We are surprised that there is not fuller discussion of this research in the B&G paper.

Conflicts of interest

Before we discuss this issue we report our conflicts. MG had dinner in New York with Sir Michael Rawlins, Chair of NICE, to discuss replicating US-based work in public deliberation regarding CEA in the U.K. environment. She has since attended a Citizen’s Council Meeting with expenses covered by Vision 21, a consulting group paid by NICE. SB has recently been re-appointed as a member of the advisory Appraisal Committee of the National Institute for Health and Clinical Excellence (NICE), having served on the same Committee during 2004/05. Further, SB is a member of University of Birmingham team that conducts CEAs for NICE, and was principal investigator on a research project (funded by the UK Department of Health) looking at the use of CEA by the Institute in its technology appraisals process (Williams et al., 2007; Bryan et al., 2007).

That said, we share B&G’s concern in relation to conflicts of interest and agree that it is a first-order issue in the context of any public organization that has such a central role in influencing government expenditures. The suggestion made by B&G, that expert advisors to regulatory agencies be obliged to suspend all conflicts during their time in office is one we join in endorsing and extending through to future posts when advisors could receive substantial benefit from policies that were effected while they were in office. We were not, however, able to discern from B&Gs innuendoes the extent to which such departures from this guidance are present (not to mention widespread) within NICE’s management or advisory groups. For example, we do not know what to make of the notation that ‘only 3 of the 14 members of the original NICE Guidance Steering Committee and none of the Methodology Working Party for NICE 2 listed any relationships with the pharmaceutical industry in the published guidelines documentation’. Is the implication that the remaining 11 members and all of the Methodology Working Party are in bed with the pharmaceutical industry? If so, this needs to be said straight out and documentation furnished, with the intent to explicitly correct wrong-doing. Otherwise, it seems to us that a little circumspection would go further in providing credibility to what is, until its final two paragraphs, an academic critique of NICE.

Conclusions

The policy world, by its nature, is a messier one than is academe. What is elegant in theory often breaks down a bit in the rough and tumble world of implementation. While we recognize that there are aspects of NICE’s methods and resources that could stand building, on balance we think that U.K. health economists should sleep more soundly at night for its presence, and U.S. economists should advocate more vociferously for its analogue to be created in the States. Our view remains that the creation of NICE represented a bold step by the UK Department of Health to begin the process of incorporating economic considerations into the setting of priorities. Because of NICE, national health policy decisions are increasingly evidence-based, with explicit consideration given to cost and cost-effectiveness information (Devlin et al., Reference Devlin and Parkin2004, Bryan et al., 2007). We see this development as positive and believe that the Institute deserves the support of the academic community. Some of that support will come in the form of thoughtful critique, and that sort of criticism should be welcomed by NICE as it seeks to bring about improvements in its methods and operations. Thus, we endorse B&G’s efforts to continue to raise important issues and challenges for NICE (many of which we see good sense in) at the same time that we urge them to do so accurately and constructively.