Ann Dix's insightful report on the Castlefields experience of effective case management ('Happy Ever After', pages 28-31, 9 December 2004) contrasts markedly with the latest King's Fund document which could find only weak evidence to support the approach.

To my mind, this raises some key issues as the Department of Health rightly directs policy towards improving NHS services for people with longer-term conditions.

First, cost-saving should not be our sole objective. Case management as adapted at Castlefields has reduced hospital admissions and bed-day usage in a range of highrisk groups, but the objectives are best summarised as value for money. This encompasses quality measured by improved patient experience, quality of life for patients and clinical outcomes. Indeed these factors are the most powerful arguments for the approach.

There is also extensive evidence from the US of the positive impact of managed care, namely the audited surveys reported by the US National Council for Quality Assurance. The absence from the King's Fund paper of any reference either to this or our most successful UK pilot to date rather diminishes its findings.

Second, case management is incumbent on both project design and evaluation exercises to be precise about objectives, definitions and the content of the models. Otherwise the outcomes and findings will inevitably be inconclusive. This risks undermining the more progressive clinicians and managers striving to implement this policy.

Third, and a further lesson from Castlefields, it is not simply good project design which will deliver effective case management. The key critical success factors are the quality and commitment devoted to the clinical process. We do a disservice to the people delivering this if we fall short of a comprehensive approach when we presume to evaluate their work.

Dr David Cochrane, director, Conrane Consulting, and visiting senior fellow in health management, Keele University