Kieran Walshe is right to welcome the new emphasis on quality in The New NHS: modern, dependable (Open Space, 18 December). But his suggestion that the recommendations on the use of new medicines and technologies from the new National Institute for Clinical Excellence (NICE) should have statutory force sits uncomfortably with that.

Assessment of the clinical and cost-effectiveness of new technologies, including medicines, is both necessary and desirable. Those assessments need to be authoritative, and to be disseminated effectively. And there need to be structures and procedures in place to assess performance, which can take account of those recommendations. NICE has the potential to co- ordinate the considerable amount of work that is already being done in this field and add real value to it. But to give its 'recommendations' statutory force is another matter.

Central to the notion of quality is effectively meeting patients' needs. One of the distinctive characteristics of patients, as any clinician will tell you, is their infinite variety. Guidance on the clinical or cost- effectiveness of a new technology is inevitably based on averages. It may be that for the 'average' patient (or perhaps more accurately, the majority of patients) a particular technology delivers little or no benefit. But there are few technologies which never provide benefits. A blanket ban based on averages will inevitably mean that those who could benefit will be denied that opportunity. For them it will be a reduction, not an improvement, in quality.

The right approach is guidance which is just that - which suggests what will normally be the most appropriate response but allows for clinicians to reach a different decision where the circumstances of an individual patient justify. That will deliver quality for everybody. A 'fourth hurdle' approach which prohibits use of a medicine or technology based on averages is no better than the 'nonsense' - which Kieran Walshe rightly castigates - of each health authority making its own decision.

Jim Furniss,

Managing consultant,

Cambridge Pharma Consultancy,

Cambridge.

David Hunter's review of Geoffrey Rivett's history of the NHS's first 50 years (Book Reviews, 8 January) is right to hope that this is not the end of history for the NHS: the half-century must not mark its cradle to grave.

He notes the government's New NHS focus, using health action zones and primary care groups (or trusts). But how will these new bodies ensure they serve all residents (as envisaged by the NHS founding principles), including those whose care may be difficult or expensive?

The question of accountability is surely key. People must not be left out just because they don't fit neatly into the new arrangements. People such as those with HIV, who are not spread evenly throughout the population and who are sadly still subject to widespread discrimination (and are thus not likely to identify themselves publicly at local consultation meetings), have not always received fair treatment from GPs or local community services, and have missed out between the sharpest lines of the 'health' and 'social' interface. Research suggests that 50 per cent of people with HIV change their GP after diagnosis, and that 'family doctors' do not always have experience of 'minority' conditions.

So there must be strategic planning and real decision making, covering regional or similar geographical areas, rather than delegating to smaller and smaller units within the NHS. Local may be better, and the corner shop could be preferable to the supermarket, but our health services need to be national - as befits the 50th anniversary - comprehensive, and accessible to everyone in need.

John Nicholson,

Director,

George House Trust,

Manchester.