Patient choice has become central to the UK government’s vision for reform of the National Health Service. Subsequently there has emerged a novel claim for the benefits of choice. According to the former secretary of state for health, Patricia Hewitt, “choice is important … because—far from entrenching inequality—it will help us create a more equal society.”1 Choice is associated with individualism and autonomy; conversely, equity is associated with collectivism and justice. As such, choice and equity have been regarded as fundamentally competing ideals.2 It is therefore worth examining where Hewitt’s novel assertion came from, and whether it is based on sound analysis or evidence. If it is not, then the implication is that the “new Labour” party’s attitude towards equity in healthcare is disingenuous.

Initial calls for increasing patient choice as a central policy objective came not from within government but from its critics. As recently as December 1999, new Labour opposed such calls. For example, John Denham, a junior minister, thought it “much better to plan for the vast majority of patients to receive high-quality services in their most local hospital” and objected to the notion that patients should have to “chase all over the place to try and find out where the best treatment is available”.3 In June 2000, former secretary of state for health Alan Milburn accepted that greater choice was sacrificed in favour of greater equity in the NHS, stating “we are not prepared to trade off being free and fair for efficiency and responsiveness to demands of patients”.4 Yet, in 2001, Milburn felt the NHS “must offer patients not just fairness but more choice”, and the challenge was to “demonstrate that we can expand choice without compromising equity”.5 By 2003, when John Reid had been appointed secretary of state for health, “one of the main reasons for increasing choice in the NHS” was to “increase the fair distribution of access to health services”.6

The calls for patient choice that Denham and Milburn were dismissing 6 years ago came from commentators outwith new Labour and across the political spectrum. These calls can be caricatured as belonging to two main camps.

First, there were those who appealed for patient choice as a component of the expansion of market mechanisms in the NHS. This was patient choice as a step towards competition, the rationale being that this would force providers to raise their game. According to a pamphlet authored by Conservative member of parliament Peter Lilley:

A central aim of introducing more patient choice is to drive up the quality of all aspects of care … hospitals will only be able to expand if they are both more efficient and offer higher quality of treatment …7

Lilley outlines three key steps towards achieving this situation: choice of hospitals at the point of referral by a general practitioner, publishing information on the performance of hospitals and specialties, and “let taxpayers’ money follow patients’ choice”.7

Second, there are those, such as Charles Leadbeater of the think tank Demos, who have argued for choice as an essential component of greater personalisation in public services. In healthcare this might include greater self-care, “intimate consultation” with “clients” to “unlock their needs, preferences and aspirations”, with greater responsibility for patients in the design of their own personal services, rather than the involvement of patient representatives in the design of services generally.8

It can be difficult to pin down what is meant in practice by the rhetoric of personalisation. However, it is clear that the choices Leadbeater advocates are qualitatively different from those a simple consumer model implies—he is not concerned with competition:

… users of public services want to be treated well, as customers, but that does not necessarily mean they want to become consumers, shopping around for the best deal or even threatening to do so.8

An important point is that neither of these disparate rationales for increasing patient choice—competition or personalisation—maintains that increased choice would lead to increased equity in healthcare. Indeed, Leadbeater cites potential inequality as the biggest challenge to personalised public services. Therefore, while the call for “more patient choice” was conceived outside of government, the claim for choice advancing equity came only from new Labour and its immediate policy advisors.

So, when the government made its volte-face over patient choice, which rationale was it following? The 2004 NHS Improvement Plan stated the aim that by 2008 every patient in England referred for outpatient consultation would have the choice of any NHS or private provider that meets a standard charge. Information to support this choice—for example, hospital healthcare commission ratings—is available online.9 It is intended that by 2008, 90 per cent of inpatient, day-case and outpatient activity will be paid for through the “payment by results” system, which reimburses hospitals for each episode of care based on a set, diagnosis-related tariff.10 All of Peter Lilley’s key steps towards competition have been taken.

Meanwhile, some ideas fitting the personalisation mould of patient choice have been suggested—examples include easier access to repeat prescriptions, direct access to a midwife of the patient’s choice and space on electronic records for registering treatment preferences. These micro-policies do not seem to add up to a systemic application of the macro-rhetoric of personalisation. Some commentators have, rightly, concluded that “much of the debate and the political rhetoric about choice has been narrowly focused; couched in terms of competition or the threat of competition (contestability), markets (of a sort), and supply side incentives ... What is missing is a broader view of choice.”11

This, then, is the recent historical and policy context in which government ministers have made the novel assertion that choice, as a prerequisite for competition, is good for equity in healthcare. It is instructive to detail where this claim has been made, but also, first, where it has not been made.

Consider some of the government’s key strategic documents on health inequalities. The Acheson report on the independent inquiry into inequalities in health, commissioned by new Labour within months of gaining office, makes 14 recommendations relating to the National Health Service; patient choice in any of its guises is not mentioned.12 Similarly, the 2003 Tackling health inequalities: a programme for action set out plans to reduce health inequalities over a period of 3 years and extensively addressed the role of healthcare services but did not refer to the choice agenda at any stage. Instead these documents focused on steps to increase uniformity, such as national service frameworks. In addition they advocate measures to match increased funding to areas of increased need, rather than funding following consumer demand—for example by “identifying geographical areas or groups not receiving care according to need, such as [CABG] referrals or statin prescribing”.13

The case for choice increasing equity in healthcare has instead been made in speeches and pamphlets by successive secretaries of state for health and in articles published by a handful of new Labour’s closest health policy advisors and have been echoed in the Department of Health documents Building on the best: choice, responsiveness and equity in the NHS and the English public health white paper Choosing health.

Simon Stevens is a former Number 10 Policy Directorate chief health policy advisor. He addressed the issue of choice and healthcare equity in the published proceedings of a meeting of the Health Equity Network in 2003.14 Stevens suggests that while it is “frequently asserted that more patient choice would mean less equity … this popular nostrum could be wrong”. Having pointed out that the current “limited choice” NHS has not successfully eliminated inequities, Stevens then presents three strands to his argument:

1. Richer people currently have the choice to switch from under-performing public providers, as they can gain access to private healthcare. Extending patient choice would give everyone this option. This argument is in turn echoed in Building on the best: choice, responsiveness and equity in the NHS.15

2. Choice, through market incentives, can improve equity by putting pressure on low-quality providers. Or as Alan Milburn, when secretary of state for health, asserted: “poorer performance is often concentrated in poorer areas … the power to choose between services will drive standards up”.16

3. If the NHS is not, through greater choice, made more responsive to the consumerist demands of better-off patients, these patients will leave the NHS, thus fatally weakening the risk-pooling principle the NHS is built on. Again, according to Milburn, “greater choice can enhance social cohesion not diminish it”.16

This third point is curious, as it does not make any positive analytical argument as to why choice would increase equity in the NHS. In addition, the assumption that choice is the key to stemming the drift of patients to the private sector might be contested. A study by the Institute for Fiscal Studies suggests that support for universal public services does not fall even among those who have opted to pay for private services; instead they express support for the ideal of public services.17 If there needs to be a new bargain struck to secure continued middle-class support of the NHS, then is plurality of providers the right price? Might reduced waiting list times be a simpler and just as potent incentive to stay with NHS procurement, for example?18

In contrast, Stevens’ first point above makes a more positive argument for choice increasing equity in healthcare. However, in a system with little excess capacity, how can everyone “escape to better services”? Either we must have a substantial (and inefficient) excess of capacity, or we must rely on a universal improvement through market pressure. This implies poor hospitals improving services or good hospitals expanding capacity in the medium to long term—Stevens’ second point above. Still, in a health system with finite resources, not all demand can be met; in a market where individual demand is not tethered to individual ability to pay, which consumer is king? Some commentators have expressed concerns:

With greater choice, people may demand longer consultation times, and/or greater use of the most expensive (although not always more incrementally effective) pharmaceuticals and medical technologies, and providers may be under increasing pressure to provide these services. If the better educated and wealthier sub-groups in society are more aware of the choices they are being offered, it is possible that—in a resource constrained system—resources will be increasingly redirected to the provider units that tend to serve these particular patients. Of course, regulation can limit or even eliminate these events occurring, but what if we find that extensive regulation to guard against the ill effects of greater choice is necessary, is it really worth introducing more choice in the first place?”19

If “poorer performance” hospitals in “poorer areas” are to go to the wall, that is not an equitable outcome. To draw an analogy, out-of-town supermarkets may offer a greater choice of goods, but it is arguable whether they are offering a better service to the residents of inner city areas where corner shop grocers have gone out of business.

It has been argued that outcome is not important. Whether patients accept an offer of better treatment or not, if all patients are offered the choice, then equity concerns are resolved. This relies on a libertarian definition of equity; individuals are best placed to decide what maximises their utility, provided they are making informed choices. Another policy advisor, Julian Le Grand—who succeeded Simon Stevens at the Number 10 Policy Directorate—has been a proponent of this definition of equity. In an essay in Lancet, Oliver, Healey and Le Grand stated that “many differences in health across socioeconomic groups would not always be seen as inequitable if we remember that people have informed choice over diet, alcohol consumption, levels of exercise, etc.”20 John Reid, while secretary of state for health, was keen to apply Le Grand’s ideas about choice and equity in health to the patient-choice debate and equity in healthcare. Le Grand’s book Equity and choice: an essay in applied philosophy is quoted in a pamphlet authored by Reid in 2005:

Against this [complaint of inequity] we could point out, however, that both patients were at least offered the same choice. As Le Grand has argued: “Our judgements concerning the degree of inequity inherent in a given distribution depend on the extent to which we see that distribution as the outcome of individual choice …”21 (Italics in original.)

So Reid has applied this libertarian approach to the debate about patient choice, but this is by no means a widely accepted definition of equity. The problem of determining which choices are free and informed complicates this characterisation. Consequently, others use quite different definitions. Barbara Starfield, for example, sees equity as the eradication of “systematic and potentially remediable differences in health across populations defined geographically, demographically, and socially”.22 We argue that the existence of systematic differences in health-related choices between groups implies imperfect freedom of choice for the individuals within these groups. People may have different views on how to live their lives, but why should these individuals be grouped in a non-random way into particular geographic or social groups? Similarly, we would argue that offering choice to all patients does not resolve concerns about equity if there are systematic differences in the uptake of this choice.

The second condition, that differences are remediable, also seems relevant to the choice and equity debate. Starfield sees equal opportunity of access to a set of services as insufficient:

[H]orizontal equity will not resolve the problem of unequal need … even if access were vertically equitable, however, there would be no assurance of approaching equity in health, because access alone will achieve little in the absence of appropriate services … Equality of opportunity to interventions, then, is not equivalent to opportunity for equity in health, nor is it an adequate substitute for it.22

Everyone having the choice of using a better hospital clearly falls short of Starfield’s definition, if those utilising the better hospital are more likely to come from a defined social or geographical group, or if the set of services on offer is more appropriate to the needs/demands of an advantaged rather than a disadvantaged group. More accurately, if services on offer are more appropriate to the needs or demands of the advantaged, then the principle of equal access for equal need has not been met.

The aim here is not to champion one definition over another. An important point can, however, be made. No clear definition of equity in healthcare is being used by Stevens in his arguments above, or in the documents and ministers’ speeches that echo those points. Nor, indeed, is there a clear theoretical framework of equity in healthcare underpinning these arguments, in spite of an extensive literature which does attempt to provide such a context; see, for example, Goddard and Smith.23 So, while the white paper Building on the best might avow to ensure that the NHS does not discriminate against disadvantaged groups by placing the “power of making real choices about healthcare and exerting real influence over those choices … in the hands of all the users of services”, this purported solution is not “framed in causal explanations of the problem being addressed”.24 Any justification for the claim that “choice” will increase equity in healthcare must be grounded in an analysis of the evidence that has been collected on the causes of healthcare inequalities.

And such a claim, framed in causal explanations, has been made—in a review of equity in the NHS by Dixon and colleagues (of whom Le Grand was one), published in 2003.25 This review defines an equitable service as one that “offers equality of access to healthcare to individuals in equal need”—that is, horizontal equity. The authors accept the logic of Goddard and Smith in their earlier review, that evidence of “variations in access to care can only be considered useful for policy purposes if … presented in conjunction with the likely causes of such variation”, and so consider the evidence for a number of explanations. They conclude that there is evidence for several factors contributing to healthcare inequalities in the UK:

• Travel and time costs (rather than distance), with factors such as car ownership and loss of earnings in low-autonomy jobs being important.

• Unequal capacities, in terms of confidence and ability to articulate demand successfully—“voice” and “sharper elbows”—result in unequal ability to “promote referral” and navigate the healthcare system.

• Differences in health beliefs, health-seeking behaviour and health literacy across socioeconomic groups—for example fatalistic attitudes, normalisation of symptoms and lower expectations of healthcare—all combining to reduce access to, and therefore utilisation of, services.

The review goes on to suggest that increased patient choice can diminish the bearing of one of these factors on equity in healthcare:

Empowering all patients to make informed choices about their care (together with clinicians) could equalize the advantage that middle class patients currently exercise through their use of voice and connections … Together with the right to exit (choice to move to another provider), this should improve the responsiveness of services to tackle negative perceptions of the NHS.25

In short, capacity for “choice” is better than capacity for “voice” in disadvantaged socioeconomic groups—market mechanisms can empower those less able to navigate the current system, particularly if they are supported in making these choices. It is this argument, and a focus on unequal voice or ability to “work” the system, that is found in many of the government’s speeches and publications. On the day when Dixon and colleagues’ review was published, John Reid (the then secretary of state for health), in a speech entitled “Equity, choice, capacity and culture”, stated that inequality in the NHS goes beyond distribution of resources: “middle class people may get more out of the health service because they are more articulate, usually more confident and more persistent”.26 Reid believes disadvantaged groups are not as experienced at making choices and the rich are “better at choosing”, but that the less well off can be empowered by making choices explicit.21

This is an assumption on Reid’s part. Other commentators might, conversely, assume that explicit choice might further empower those with greater voice.27 What has Reid based this assumption on? Dixon and colleagues refer to a “preliminary analysis” of the London Patient Choice Pilot (LPCP) and report that it showed that uptake of choice was the same across primary trusts of varying levels of deprivation. However, two comprehensive reviews of the evidence relating to patient choice have been compiled in the UK, and they are both less sanguine about the equity performance of the UK choice pilots. A “scoping review” by Fotaki and colleagues, commissioned by the National Co-ordinating Centre for NHS Service Delivery and Organisation (NCCSDO), states the following:

An important experiment, which is often quoted to support the … argument that choice may increase equity, is the London Patient Choice Pilot. However, conclusions from this study need to be tempered by the fact that several disadvantaged groups were excluded from participation.28

The other review of the evidence on choice in public services—authored by Williams and Rossiter and published by the Social Market Foundation—held that there is a “lack of definitive evidence to indicate to what extent socio-economic factors influence take up of choice” in the context of the LPCP.29 Both the reviews highlighted a study by Burge and colleagues, who surveyed patients eligible for inclusion in the LPCP to elicit their attitudes towards choice of hospital.30 This study showed that patients who are older, female, looking after children or have a lower education level are less likely to choose an alternative hospital for faster treatment.

Both the Williams and Rossiter review and the NCCSDO review examined a range of other sources of indirect evidence—for example, international experience and other sectors such as education. While both emphasised the need for further research, Fotaki and colleagues concluded that

The empirical evidence on this question is fairly clear cut. Research from several countries shows that the impact of choice on equity is consistently negative, although the effects are sometimes small. Providing more choice increases inequity. This is partly because the better off are more able to exercise choice when it is offered … At the very least, choice policies have the potential to increase inequity. Therefore, if choice is actually to be used to address inequity, the policy needs to include targeting of specific groups that are likely to be disadvantaged … This does not appear to be part of current Government policy.28

So the assumption that explicit choice will empower those with relatively limited voice is based on quite selective attention to the evidence base. It should also be noted that the other two causal factors in healthcare inequities identified by the review of Dixon and coworkers—health beliefs/behaviour and costs of attending healthcare—are apparently not to be addressed by the patient-choice agenda.

CONCLUSIONS

The UK government has adopted, from its critics, “patient choice” as a policy program in healthcare reform. This patient-choice agenda has been developed in terms of a prerequisite for competition and marketisation. While adopting this policy program, new Labour has appended the claim that choice—and the market mechanisms this will facilitate—will make the NHS fairer. This claim has not developed prospectively from an analysis of the causes of healthcare inequity, or even with a consistent normative definition of equity. The limited justification that is framed in causal explanations of inequity has suffered from an apparent disregard for the available evidence.

The UK government should take seriously the possible negative impacts of choice on equity in the NHS and consider the policy design features that might minimise these costs—as detailed in recent literature reviews. As patient choice is rolled out in England, the equity impacts should be monitored by an independent body, so that the government may be held to account for its novel claim.