3.1 General SVJs: as reported in general NICE guidance

The Secretary of State's Directions to NICE require that its evaluations of health care interventions should take into account, among other things, “the degree of clinical need of patients with the condition or disease under consideration” (SoS Directions). Rawlins et al. (Reference Rawlins, Barnett and Stevens2010) describe severity of illness as one of the ‘special circumstances’ to which NICE's advisory bodies have given ‘special weighting’, reflecting evidence that many people value the health gains accruing to severely ill individuals more highly than equal-sized health gains accruing to those whose illness is relatively mild (Shah, Reference Shah2009).

In January 2009, NICE issued supplementary advice to its TACs to be taken into account when appraising life-extending end of life treatments (EoL Guidance). This indicated that, if certain criteria are met, it may be appropriate to recommend the use of such treatments even if their ICERs exceed the upper end of the range normally considered acceptable. The TACs are instructed to consider “the impact of giving greater weight to QALYs achieved in the later stages of terminal diseases, using the assumption that the extended survival period is experienced at the full quality of life anticipated for a healthy individual of the same age” (Longson and Littlejohns, Reference Longson and Littlejohns2009). They are also instructed to consider the magnitude of additional weight that would need to be assigned to the health gains in order for the ICER to fall to an acceptable level. This has been argued to reflect the special value society places on prolonging the lives of terminally ill patients (Rawlins et al., Reference Rawlins, Barnett and Stevens2010). Some aspects of the supplementary advice were revised following a five-week public consultation exercise. The criteria that need to be satisfied in order for the advice to be applied are currently as follows (EoL Guidance):

• • “The treatment is indicated for patients with a short life expectancy, normally less than 24 months;

• • There is sufficient evidence to indicate that the treatment offers an extension to life, normally of at least an additional three months, compared to current NHS treatment;

• • The treatment is licensed or otherwise indicated, for small patient populations”.

The NICE SVJ guidance does not say anything about whether the unavailability of alternative effective treatments might affect the acceptability of a given technology's cost-effectiveness. The supplementary advice regarding end of life treatments proposed initially included, in addition to the above, the following criterion:

• • “No alternative treatment with comparable benefits is available through the NHS”.

This criterion was used in the end of life test cases. For example, in the appraisal of sunitinib for renal cell carcinoma, the Committee noted that the technology provided a ‘step change’ in the first-line treatment options for advanced and/or metastatic disease (NICE, 2009c). Eventually, however, the criterion was dropped from the supplementary advice following the public consultation. Among other things, comments made during the consultation exercise included pragmatic concerns about what types of treatment could be considered as an ‘alternative’ (e.g. any treatment theoretically available, or only those that have been approved by NICE, or only those that are used routinely). The consultation also revealed concerns that there is little scientific evidence to support the premise that society is prepared to fund life-extending end of life treatments that would not meet the cost-effectiveness criteria used for other treatments. NICE has sought to address this lack of evidence by making end of life a central topic in a recent Citizens’ Council meeting and by funding a study seeking to investigate the extent of public support for its supplementary policy (NICE Decision Support Unit website, http://www.nicedsu.org.uk/End-of-Life(2474843).htm, accessed 30 January 2012).

Finally, NICE SVJ guidance explicitly addresses the issue known in the literature as the ‘rule of rescue’ or the ‘identifiable victim effect’ (McKie and Richardson, Reference McKie and Richardson2003; Cookson et al., Reference Cookson, Tsuchiya and McCabe2008). The second edition of the Social Value Judgements: Principles for the development of NICE guidance document states: “There is a powerful human impulse, known as the ‘rule of rescue’, to attempt to help an identifiable person whose life is in danger, no matter how much it costs … NICE recognises that when it is making its decisions it should consider the needs of present and future patients of the NHS who are anonymous and who do not necessarily have people to argue their case on their behalf. NICE considers that the principles provided in this document are appropriate to resolve the tension between the needs of an individual patient and the needs of present and future users of the NHS. The Institute has not therefore adopted an additional ‘rule of rescue’ ” (SVJ Guidance, pp. 20–21). This guidance on ‘rule of rescue’ was issued in 2008, following deliberation and consultation about this specific issue including a Citizens’ Council meeting in 2006, and pre-dates the supplementary guidance on life-extending end of life treatments issued in 2009. However, as the SVJ guidance makes clear, NICE sees these as separate issues and does not seek to favour identifiable patients who need an intervention being appraised by NICE over anonymous patients who need other forms of care.

3.2 Case-specific SVJs: as reported in NICE guidance on health care technologies and public health interventions

As noted in section 3.1, since early 2009 NICE has instructed its advisory bodies to consider giving special weighting to the health gains offered by life-extending treatments for individuals in the later stages of terminal illness. However, there are cases of NICE giving special consideration to such treatments that pre-date this supplementary guidance. For example, in the appraisal of riluzole for motor neurone disease, the TAC considered the ‘severity and relatively short lifespan’ of affected patients, and recommended the use of this technology despite the fact that the ICER estimates ranged between £34,000 and £43,500 per QALY gained (NICE, 2001). Similarly, trastuzumab combination therapy was recommended for use in the treatment of advanced breast cancer even though the manufacturer provided an ICER estimate of £37,500 per QALY gained. This was due in part to the TAC's observation that “improvements in survival of this magnitude due to therapeutic intervention have rarely been recorded in women with metastatic breast cancer” (NICE, 2002).

Between January 2009 (when the supplementary guidance was issued) and April 2009, nine technologies were considered by NICE in the context of the end of life criteria, because they were associated with a life expectancy of less than 24 months and the estimated ICERs exceeded the range normally considered acceptable (Longson and Littlejohns, Reference Longson and Littlejohns2009). All nine related to treatments for cancer. Three of these technologies were considered not to fulfil all of the criteria – bevacizumab for renal cell carcinoma because it was licensed for other indications that have large eligible patient populations; lapatinib for breast cancer and cetuximab for head and neck cancer because the available evidence suggested that neither offered a life extension of at least three months. For the remaining six technologies, it was assumed that the extended survival period would be experienced at the full quality of life anticipated for a healthy individual of the same age, in accordance with the supplementary guidance. In addition, the TAC also considered the magnitude of implicit weightings that would need to be assigned to the health benefits offered by the technologies in order for the ICERs to fall within the normal threshold range. For example, in the case of sunitinib for renal cell carcinoma, it was noted that the health gains would need to be valued 70 percent higher than those achieved in other circumstances for the ICER to fall to £30,000 per QALY gained. Three of the six technologies were eventually recommended as treatment options: sunitinib for the first-line treatment of renal cell carcinoma (most plausible ICER without end of life weighting: ∼£50,000 per QALY gained), lenalidomide for multiple myeloma (more than £43,800 per QALY gained), and sunitinib for gastrointestinal stromal tumours (£31,800 per QALY gained).

We did not identify any clear-cut examples of TACs giving explicit additional weight to QALY gains from treatments for severe but non-terminal illnesses. However, there are cases where TACs have taken severity of illness into account in a rather different way: by making a special effort to identify a small sub-group of severely ill patients for whom the technology might be considered cost-effective. In the appraisal of omalizumab for severe persistent allergic asthma, for instance, the Committee was persuaded that the technology could be considered cost-effective for a small sub-group of severely ill patients, even though the ICER estimates for the broader patient group exceeded the range normally deemed acceptable (NICE, 2007). The guidance document explained this decision as follows: “Overall, therefore, the Committee concluded that there were a number of considerations which meant the ICER was higher than acceptable for patients with severe persistent allergic asthma. However, the Committee was persuaded that for a narrowly defined severely affected group of asthma patients, at an elevated risk of asthma-related mortality, cost-effective treatment with omalizumab was possible, if therapy was discontinued in non-responders at 16 weeks and if vial wastage could be minimised to reduce costs” (NICE, 2007).

4.1 General SVJs

NICE must comply with legislation on human rights, discrimination and equality, at both the scoping and appraisal stage (TA Guide, p. 10). In particular, the 2010 Equality Act specifies nine ‘protected characteristics’, which cannot be used as a reason to treat people unfairly: age;Footnote 3 disability; gender reassignment; marriage and civil partnership; pregnancy and maternity; race; religion or belief; sex and sexual orientation. The current version of the UK public sector equality duty, which came into force on 5 April 2011, states that public bodies must have due regard to the need to: eliminate unlawful discrimination; advance equality of opportunity between people who share a protected characteristic and people who do not share it; and foster good relations between people who share a protected characteristic and people who do not share it (http://www.homeoffice.gov.uk/equalities/equality-act/equality-duty/, accessed 22 September 2011).

As well as these nine “protected characteristics”, NICE's concern for non-discrimination extends to three further characteristics: conditions associated with stigma, behaviour-dependent conditions and socio-economic status (SVJ Guidance, pp. 23–25). NICE ensures that “patients are not denied access, or have different or restricted access, to NHS care because of their race, disability, age, sex/gender, sexual orientation, religion, beliefs, or socioeconomic or other status” (SVJ Guidance, p. 23).

NICE seeks to treat the same those who are the same in a relevant respect – a horizontal equity objective (Culyer, Reference Culyer1995). The Institute's general principle is that ethnicity, age, gender, sexuality and other social characteristics should not affect decisions about whether to recommend a given intervention. However, these characteristics may be taken into account if there is clear evidence that they are indicators for the risks or clinical effectiveness of that intervention. More specifically, principle 7 of Social Value Judgements: Principles for the development of NICE guidance states that: “NICE can recommend that use of an intervention is restricted to a particular group of people within the population (for example, people under or over a certain age, or women only), but only in certain circumstances. There must be clear evidence about the increased effectiveness of the intervention in this subgroup, or other reasons relating to fairness for society as a whole, or a legal requirement to act in this way” (SVJ Guidance, p. 25).

NICE also seeks to treat differently those with different needs – a vertical equity objective (Culyer, Reference Culyer1995). For example, it “should take special account of the needs of disabled people, which includes whether there are obstacles that might prevent them from benefitting from NICE guidance” (SVJ Guidance, p. 23). However, similar considerations have not been explicitly applied to other characteristics that may be indicative of obstacles to benefitting from NICE guidance (such as socio-economic status or religious beliefs).

NICE explicitly does not discriminate according to the cause of illness or how ‘deserving’ of treatment an individual may be. In particular, “NICE should not produce guidance that results in care being denied to patients with conditions that are, or may have been, dependent on their behaviour” (SVJ Guidance, p. 25). This would include conditions that may have been contracted or exacerbated due to the individual's lifestyle or occupation. In line with principle 7, however, “if the behaviour is likely to continue and can make a treatment less clinically effective or cost-effective, then it may be appropriate to take this into account” (SVJ Guidance, p. 25). In part, this reflects the general principle that the NHS ought to distribute health care according to need, rather than desert. However, NICE's position also reflects the fact that it is “often impossible, in an individual, to decide whether the condition was dependent on their own behaviour or not” (SVJ Guidance, p. 24). Further, taking into account the cause of illness may entail discrimination on grounds of gender (e.g. where a behavioural health risk is associated with male-dominated occupations) or socio-economic status (e.g. where a behavioural health risk is more prevalent among low income individuals).

NICE also advises that it should evaluate ‘orphan drugs’ (treatments for rare conditions) “in the same way as any other treatment”, although the Department of Health does not refer ultra-orphan drugs (treatments for very rare conditions) to NICE (SVJ Guidance, p. 20). It also considers that the standard approach to assessing cost-effectiveness is appropriate for the evaluation of conditions associated with stigma – for example, sexually transmitted diseases or growth failure in individuals with growth hormone deficiency (NICE, 2010) – although it expects its advisory bodies to consider the possibility that “stigma may affect people's behaviours in a way that changes the effectiveness of an intervention” (SVJ Guidance, p. 24).

4.2 Case-specific SVJs

There is rarely any explicit discussion of non-discrimination principles in the ‘considerations’ section of NICE appraisals. This is because non-discrimination principles are used mainly to rule out of consideration any decision options which might discriminate against certain patient groups, rather than to weigh up the pros and cons of the decision options that are actively considered. One exception to this is the appraisal of alendronate for the primary prevention of osteoporotic fragility fractures in post-menopausal women (NICE, 2008c), in which the TAC considered the special needs of patients who cannot take alendronate because of a contraindication or a disability that prevents them from complying with the instructions for its administration.

In this section, we describe a test case involving an appeal in which NICE was accused of violating non-discrimination principles by favouring a technology predominantly used to treat a disease linked to occupational exposure to asbestos. The appeal was submitted by a Primary Care Trust (PCT), doubtless concerned about the financial implications of the recommendation. The appeal was rejected, on the grounds that the technology was cost-effective and the TAC had not given any special priority to this patient group on the grounds of occupational exposure. This is a potentially important test case, as it suggests that non-discrimination principles do not just apply to ‘within-treatment’ decisions about restricting the use of the same treatment to one patient sub-group rather than another, but may also apply more generally to ‘between-treatment’ decisions about whether to recommend use of one treatment for one condition rather than a different treatment for a different condition.

Malignant pleural mesothelioma (MPM) is a type of cancer that occurs in the pleura, the membranous lining surrounding the lungs. Presentation and diagnosis often occur at an advanced stage, at which point the prognosis for most patients is extremely poor (median survival is 9–13 months). Approximately 90 percent of cases of MPM are linked to asbestos exposure. A wide range of occupations, notably those in the shipbuilding, railway engineering and asbestos product manufacture industries, are associated with an increased risk of MPM. A peak incidence of MPM is expected in men in the 1948–1953 birth cohort due to the high utilisation of asbestos in the 1970s. Family members of people whose work clothes were contaminated with asbestos fibres have also developed MPM. The use of asbestos was banned in the United Kingdom in 1999.

In 2006, NICE evaluated the use of pemetrexed with cisplatin for the treatment of MPM. The initial decision was negative, as the TAC judged that “there was insufficient evidence to demonstrate that pemetrexed plus cisplatin was superior to other, far less costly treatment regimens and was not persuaded that it was cost-effective in the treatment of MPM” (NICE, 2006a). The manufacturer submitted an appeal against this decision, claiming (among other things) that the use of QALYs may be flawed when assessing treatments for “patients who did not have long to live, who were elderly, or who had a rare disease” (NICE, 2006b). The appeal decision document noted that the TAC “considered pemetrexed to be innovative and of proven benefit in a grave condition” and that the Appeal Panel were satisfied that the TAC “had taken into account these relevant factors”. However, the Appeal Panel upheld the appeal on three of the points raised by the manufacturer.

Following the outcome of the appeal, the appraisal was sent back to the TAC for revaluation (NICE, 2008d). The results of the pivotal trial showed that the pemextrexed with cisplatin combination treatment confers a survival benefit of approximately three months compared with cisplatin alone, as well as offering advantages in terms of time to disease progression, tumour response rate and quality of life. The Assessment Group estimated an ICER of £34,500 per QALY gained for the patient sub-group comprising the majority of MPM patients seen in United Kingdom clinical practice, which was higher than the range normally considered acceptable.

The TAC considered that the economic analyses may have underestimated the overall quality of life benefits offered by pemetrexed combination therapy, while potentially overestimating the costs of the treatment in UK clinical practice. It also noted that MPM is “a rare and aggressive malignancy caused by occupational exposure to asbestos” and that it has a “very poor prognosis”. Considering these factors together with the available clinical and cost-effectiveness evidence, the TAC agreed that the ICER was likely to fall within acceptable levels, and therefore recommended the use of pemetrexed in combination with cisplatin as a treatment option for MPM for the defined sub-group.

However, Birmingham East and North PCT submitted an appeal against this decision, alleging that the TAC had failed to take account of the principles of NICE's own SVJ guidance and that the final recommendations were perverse in light of the evidence submitted (NICE, 2008e). The PCT alleged, inter alia, that the TAC appeared to have taken into account the fact that MPM is typically contracted via occupational exposure. This, in the view of the PCT, was an irrelevant consideration. The PCT argued that, in reaching decisions about treatment, NICE should assess patients as they present and irrespective of the circumstances that led to them contracting the condition. By taking into account occupational exposure, it risked discriminating against patient groups and/or on gender grounds.

In response to this point, the Appeal Panel heard that when reaching its conclusion the TAC had taken into account those matters relating to the likelihood of greater quality of life benefits and lower costs than assumed by the cost-effectiveness analyses. Matters relating to the cause of the disease had been noted by the TAC but were not actually factors in its recommendation. The Appeal Panel accepted these reassurances and dismissed the appeal on this point.

5.1 General SVJs

NICE distinguishes two approaches to allocating health care resources: the ‘utilitarian approach’ and the ‘egalitarian approach’. More specifically, “the utilitarian approach involves allocating resources to maximise the health of the community as a whole”, whereas “the egalitarian approach involves distributing healthcare resources to allow each individual to have a fair share of the opportunities available” (SVJ Guidance, p. 9). The Guide to the Methods of Technology Appraisal defines a reference case assumption that all equal-sized health improvements “should receive the same weight regardless of any other characteristics of the people receiving the health benefit” (TA Guide, p. 49) (often referred to as the ‘a QALY is a QALY is a QALY’ rule). However, the Institute recognises that this simple approach – which accords with the utilitarian approach – may, in some circumstances, not be appropriate: treating QALYs equally is not necessarily the same thing as treating individuals fairly.

One important dimension of the egalitarian approach defined above is the concern to reduce health inequalities. This is importantly different from concern to distribute health care according to need, as the focus is on health rather than health care. Equal health care is not the same thing as equal health, as health care is only one among many important determinants of health.

In general, and in accordance with the relevant equality legislation, NICE seeks to identify opportunities for promoting equality in all forms of guidance (SVJ Guidance, p. 12; PH Methods, p. 10) and expects its advisory bodies to ensure that implementing NICE guidance will not widen existing inequalities (SVJ Guidance, p. 28). Arguably, however, use of the term ‘inequalities’ in this context refers to non-discrimination principles and social inequalities in general rather than health inequalities in particular.

Specific concern for health inequality reduction is largely confined to NICE documentation relating to public health. The Methods for the development of NICE public health guidance document emphasises NICE's concern to reduce health inequalities, whereas the Guide to the Methods of Technology Appraisal makes little mention of this issue. This suggests that health inequality reduction principles have been considered to be more relevant to public health rather than health care decisions. However, this is an implicit distinction; there is no explicit ruling that health inequality reduction is less relevant to health care decisions. It may also be the case that in a health care context it is harder to navigate tensions between equity principles relating to health inequality reduction and equity principles of non-discrimination. This tension is clearly brought out in Principle 8 of Social Value Judgements: Principles for the development of NICE guidance, which states: “When choosing guidance topics, developing guidance and supporting those who put its guidance into practice, the Institute should actively consider reducing health inequalities including those associated with sex, age, race, disability and socioeconomic status”. However, it goes on to state: “NICE should not recommend interventions on the basis of individuals’ income, social class or position in life. Nor should individuals’ social roles at different ages affect decisions about cost-effectiveness” (SVJ Guidance, p. 25).

NICE distinguishes between inequity and inequality: “health inequity is concerned with social justice, value or politics, while inequalities in health are a matter of fact” (NICE glossary, http://www.nice.org.uk/website/glossary/, accessed 5 August 2011). However, NICE offers little specific guidance about the criteria for assessing how far particular health inequalities are unjust or unfair or inequitable, or how concern for reducing unfair health inequality is to be balanced against concern for improving population health.

One specific recommendation is that, although special efforts need to be made to improve the health of the most socially disadvantaged groups, efforts should “focus on all levels of the [health and socioeconomic] gradient” (PH Methods, p. 12).

A second specific recommendation is that it may sometimes be justified to recommend interventions that improve population health but slightly widen health inequalities. For instance, Methods for the development of NICE public health guidance states that, ideally, a health intervention should “benefit health in all groups, while reducing health inequities. However this intention may in practice be difficult to achieve. It may be justified in certain circumstances to increase overall population health while relative differences between groups remain the same (or get slightly worse)” (PH Methods, p. 11). This excerpt indicates that the utilitarian approach may be followed at the expense of increasing health inequalities in some situations, as the overall increase in health may compensate for the deterioration in the distribution. This is important because public health interventions that produce overall health improvement often exacerbate relative health differences between the most and least advantaged groups in society (PH Methods, p. 12). For example, policies aimed at reducing hazards in ‘white collar’ workplace environments will not benefit those who are unemployed or who work in (traditionally lower pay) manual occupations. Thus, there may be a trade-off between achieving overall health improvement and reducing health inequities (Graham and Kelly, Reference Graham and Kelly2004). Methods for the development of NICE public health guidance states that decisions need to be made on a case-by-case basis, and that it may be appropriate to apply hybrid actions, which contain elements of both universal approaches designed to produce overall population health improvement and a targeted focus on the most disadvantaged (PH Methods, p. 142).

Finally, NICE's adherence to the principle of non-maleficence (SVJ Guidance, pp. 8–9) suggests that it is not permissible to recommend an intervention that may reduce the health of advantaged individuals, even though this may reduce health inequality. For example, no-one would seriously recommend a policy of free vouchers for cigars and fast food for well-to-do individuals and communities, however effective this might be in ‘levelling down’ socio-economic inequalities in health.

5.2 Case-specific SVJs

We did not identify any cases to date in which NICE has explicitly departed from the ‘a QALY is a QALY is a QALY’ principle in order to pursue health inequality reduction objectives. It has, however, ‘taken into account’ health inequality reduction objectives. Furthermore, we shall show that, NICE has implicitly departed from this principle by recommending targeted rather than universal versions of an intervention.

It is often the case that public health programmes tend to benefit individuals who are already quite healthy, and therefore result in increased health inequalities. This is true of smoking cessation services, for example, which in the past have been found to be least likely to attract people from sectors of the population where smoking rates are high. Hence, when asked to evaluate workplace-based smoking cessation interventions, PHIAC considered not only the available evidence on the effectiveness and cost-effectiveness of the options, but also the fact that the workplace setting offered a means of targeting groups that traditionally have been difficult to reach (such as young men), and could help therefore to reduce health inequalities (NICE, 2008f). However, the cost per QALY gained estimate was well within NICE's normal threshold range anyway, so the intervention could be recommended as cost-effective without recourse to giving additional weight to health gains for hard-to-reach groups.

It is a common feature of PHIAC decision making that the relevant cost per QALY gained estimates are well within NICE's normal threshold range (in contrast to TAC decision making, where a sizeable proportion of the estimates tend to be either at the upper end of or beyond this range). Health inequality reduction concerns can therefore be addressed without needing explicitly to depart from the ‘a QALY is a QALY is a QALY’ principle. In some cases, however, there may be an implicit departure from this principle when PHIAC recommends an intervention targeting disadvantaged populations. This is because it may be possible to find a universal form of the same intervention with an even lower cost per QALY gained. For example, suppose the targeted programme has a cost per QALY gained of £15,000, but spending the same amount of money on a universal version of the programme (which mainly benefits more advantaged individuals) would be associated with a cost per QALY gained of £10,000. In this case, it would be possible to increase population health by approving a universal version of the programme rather than the targeted version. The decision to spend money on the targeted intervention is then an implicit decision to depart from ‘a QALY is a QALY is a QALY’ and to give greater weight to the health gains for the disadvantaged population. In such cases, PHIAC itself typically does not explicitly seek to calculate whether a more universal variant of an inequality-reducing intervention might be more cost-effective, so any potential departure remains implicit.