When The New NHS white paper was published six months ago, its focus on quality surprised many people. But while it made all the right noises, it also skilfully left much unsaid. The specifics would, we were told, be set out in a consultation paper.

Last week, that consultation paper finally arrived, after three months' delay caused, it seems, by vigorous debate at the Department of Health.

Ministers had pressed for an ambitious and centrally driven strategy for quality improvement in the NHS, something we have never had before. Others had advised caution, emphasising the need to secure the support of bodies like the royal colleges and the professions more widely.

The General Medical Council's inquiry into events in Bristol raised the stakes - making radical change easier, but also raising public expectations that whatever was proposed would make such tragic failures in the quality of care unlikely in future.

From reading the consultation paper, A First Class Service: quality in the new NHS, it seems that the case for radical change has won the day. The paper sets out an ambitious agenda at both a national and a local level. It proposes much greater managerial involvement in clinical quality and performance, and stronger central direction, control and oversight over clinical standards and practice.

The National Institute for Clinical Excellence is to be established as a special health authority, charged with providing national guidance to the NHS on effectiveness and quality. It will bring together responsibility for much current work on health technology assessment, national service frameworks, clinical guidelines and national clinical audits. Crucially, its advice will carry semi-statutory force, and HAs, trusts and clinicians will have to think very carefully before following a different course of action.

The consultation paper sets out what clinical governance will mean for trusts and primary care groups. It places clear statutory responsibility for the quality of care on chief executives and NHS boards. They will have to understand what their organisations are doing to improve quality, and direct and lead quality improvement in a way that most have never done before.

A 20-point checklist for clinical governance stresses the need for clear lines of responsibility and accountability, comprehensive participation in audit and quality improvement, and systems for identifying and dealing with clinical risk and performance problems.

The consultation paper also emphasises the need to integrate quality improvement into organisational structures and systems, and to focus not on the process of clinical governance but on its outcomes. We are promised more guidance in the autumn from the NHS Executive on local arrangements for clinical governance. Trusts will be required to have their systems in place by next April, and to produce their first annual reports on clinical governance at the end of 1999.

But it is the proposals for the Commission for Health Improvement that make the most interesting reading. CHI will be a statutory body, responsible directly to the health secretary.

It will undertake regular inspections of all trusts to assess their local arrangements for clinical governance. It will also lead a programme of service reviews to monitor local progress in implementing national service frameworks and guidelines. When there are serious clinical problems, CHI will be called in to undertake a rapid investigation. Increasingly, it will take on responsibility for organising inquiries into clinical failures.

There are striking parallels between CHI's remit and that of the recently abolished Health Advisory Service. It will have to tread a delicate path, balancing loyalties to the DoH and the NHS with its need to be independent. The relationship between CHI and some existing organisations is far from clear. What future is there for accreditation programmes, such as the King's Fund's Organisational Audit (now the Health Quality Service) if CHI sets up a national, statutory system of inspection and accreditation? The establishment of CHI can also be seen as an implicit criticism of the Audit Commission for failing to tackle issues of clinical performance and sticking to less controversial territory in its national and local studies.

Of course, there are caveats. Most obviously, the question of funding for NICE and CHI is fudged. We are told that they will be financed from existing resources. But even rough calculations of the costs would suggest that this just won't work. CHI alone will cost upwards of pounds 5m a year if it is set up and run properly. It seems curious, given the huge injection of funds that the NHS is to receive, that the DoH is unwilling to commit itself to providing the resources that its proposals on quality will need.

The extent to which these new arrangements will apply to primary care is also left open. While the concept of clinical governance will extend to primary care groups, at least in name, it seems as if they will be excused the practical implications - like published annual reports on clinical governance, and visits by CHI - unless and until they become primary care trusts.

There is a new requirement for all hospital doctors to take part in a programme of national audits, but nothing about doctors in primary care. Similarly, it is not clear how recommendations and guidelines from NICE which relate to primary care will be put into practice or monitored.

The proposals are out to consultation until mid-September, though it is clear that their main thrust is not up for negotiation. If, as seems likely, the ideas in the consultation paper become reality, the NHS could be a very different place to work in in three or four years' time. For the first time ever, the NHS will be taking quality seriously.