Is it fair to expect the NHS alone to stop health inequalities widening when they are shaped by so many factors in society - income, housing, education - or is the health service failing in its basic responsibilities?

The NHS enters its 61st year in pretty good shape. Among the public and politicians there seems to be consensus that a publicly funded health system should endure, that healthcare should be provided according to need and financial settlements remain relatively generous. However, despite improvements in health, there are enduring and widening health inequalities which should dampen the birthday celebrations.

The government has made reducing health inequalities a priority, although there have been complaints from some experts and practitioners about a lack of focus and investment. In 2002, the government set targets to narrow inequalities by 10 per cent. For men, the relative gap in life expectancy is now 2 per cent wider than in 1997-99, for women it is 11 per cent wider and the infant mortality gap is 4 per cent wider than it was in 1997-99. Neither of the targets is likely to be met by 2010.

Health secretary Alan Johnson has signalled his determination to reverse these trends, calling health inequalities an indictment on society. The government's recent Health Inequalities: progress and next steps report has many sensible propositions aimed at improving health for the most disadvantaged people and tackling some of the unhealthiest behaviours. The programme promises momentum, but there are of course no easy answers.

Charged with finding new ways that the NHS can help people to stay healthy, the next stage review outlines how primary care trusts and local authorities can improve health by commissioning well-being and prevention services in coalition with local authorities. While the importance of the issues has not escaped the government's notice, funding and delivery remain significant challenges.

Much of the evidence and analysis on health inequalities shows that health is so closely related to socio-economic status that the most significant contribution to reducing them would come from reducing income inequalities - and they are widening too. Recent years have seen a more sophisticated understanding of the links between health and social status. We know that health inequalities are closely bound with other social divisions. Income, lower levels of education, poor housing and unhealthy environments are likely to lead to worse health and shorter lives.

The government recognises that inequalities in health are unlikely to be tackled without broader changes in other policy areas - for instance ending child poverty, improving housing and cutting unemployment.

One commentator at the launch of the Health Inequalities report suggested that the money spent by the health service on reducing health inequalities might have more health impact if it had been spent on education.

Is it fair or realistic to expect the NHS and the Department of Health to deliver significant reductions in health inequalities? Although there is clearly a limit to how much the NHS can do to reduce inequalities when health is shaped by factors well beyond its remit and reach, the NHS does have a vital role in tackling inequalities by identifying and targeting those most at risk of poor health. It should also be providing an equally good service to everyone, regardless of who they are or where they live. A significant, but often unrecognised role is how the NHS may itself contribute to health inequalities.

Inverse care law

Inequalities in access to healthcare services first came to prominence in the 1970s, when GP and researcher Julian Tudor Hart pointed to the existence of an "inverse care law" in which those who needed care most had least access. The fact that, for instance, there are fewer GPs per head in deprived areas than richer areas with lower needs is now widely recognised and still persists. Efforts are being made to encourage more GPs to work in deprived areas, but there has been limited impact.

But inequities in healthcare go beyond GP access. There is less research into other aspects of inequity in the NHS and, as a result, there is insufficient understanding of how healthcare inequities relate to wider health inequalities. And less idea about what the NHS can do to reduce them.

Healthcare can be unequal in different ways. The inverse care law operates in other arenas of the NHS - for instance in access to treatments and mental health services, some secondary care and even preventive services, such as vaccinations and screening.

From 1998-2002, 19 survey studies of cancer screening were published which appear to have confirmed that cancer screening services are underused by minority ethnic women.

Moreover, several studies found that low income, less educated and older women are less likely to attend for screening. A 2005 DH report on inequalities in vaccination uptake shows that babies from the most deprived groups were less likely to receive childhood immunisations than babies from wealthier groups. Take-up of flu jabs too shows a slight socio-economic gradient.

It is not just socio-economic status which is important in shaping access to healthcare. Ethnicity has an impact on access. Research reported by the BMJ in 2002 shows that although Afro-Caribbean children had worse reported health than the general population and although they were more likely to consult GPs, they are less likely to be referred to secondary care. A survey for the DH shows that appointment requests for a specific GP vary too, with Bangladeshi patients experiencing the worst access to this service, followed by Indian patients.

In the UK it has been found that the uptake of breast cancer screening among minority ethnic groups is persistently low. Perhaps unsurprisingly, surveys of patient opinion show some minority ethnic groups are least satisfied with health services. On average, scores were lowest among South Asian respondents. In all the surveys, Bangladeshi respondents had the lowest score.

Evident differences

Age is important in shaping access to services and therapies and treatment by staff. Although people aged 65-85 report a relatively favourable impression of the NHS, there are variations in access to treatments by age. For instance, a 2005 study on differences in access to preventive therapies found that older men were less likely to be prescribed aspirin, statins, ACE inhibitors and beta blockers following a myocardial infarction than younger patients.

Studies reviewed by Institute for Public Policy Research in 2006 also suggest that informed middle class patients are better able to enter and navigate the system. They are also more likely to discuss and debate their treatment options with doctors and push for their preference more forcefully. There is also some evidence that middle class patients may spend longer with their GP and are referred for treatment more often.

Research reported by the Race Equality Foundation suggests other sources of inequity may also be rooted in the attitudes of some doctors and other staff towards patients. Evidence from the US would support this, suggesting some doctors' attitudes also shape differential levels of care. For instance Hispanics and Asians reported more difficulty communicating with their doctors than both white and black people. Adults whose primary language is not English are more likely to report that their providers sometimes or never listened carefully, explained things clearly, respected what they said, and spent enough time with them.

Questions remain as to how choice of treatment and provider will relate to existing health inequalities. Will access to best quality services and treatments be shaped by socio-economic status, ethnicity and age? Existing evidence suggests that it will.

Part of the contribution from the NHS in tackling health inequalities must involve a long, hard look at itself. Weakening the inverse care law is about more than improving access to GPs in deprived areas. There are widespread socio-economic, ethnic and age-related differences in access to best quality services, differences in access to information, differences in treatments and referrals offered, differences in levels of care and differences in satisfaction with the NHS. These differences add up, they build on other disadvantages and exclusions and they relate to and reproduce wider health inequalities. The cause of many of these inequities lies with the NHS, and the remedy often lies there too.

The new NHS promised by Darzi and based on a vision of demanding and well-informed patients holds much promise. But there are risks that new gaps will open and existing inequalities and inequities will be compounded. Part of the drive to ensuring an equitable NHS is to make certain that it is not just demanding and well-informed patients who receive the benefits from a 21st century NHS.