In response to the article on NHS Direct ('Dial M for. . . medical advice, pages 24-26, 24 May), our research raises further issues. It is based on three years' data for the unique integrated site for the Harmoni GP co-operative - serving 900,000 people - and NHS Direct West London, serving 1. 3 million. Through Harmoni, many patients have been familiar with nurse-led computer triage since 1996, and it was the first site to incorporate NHS Direct within an existing GP co-op.

Our study indicates at least double the activity of any of its comparators: more than 5,500 in 1998, 10,000 in 1999 and nearly 10,000 in 2000 (when networking of calls accounted for a slight fall). Changes to organisations' strategic purpose and intent cannot be underestimated: NHS Direct started as a supplementary service but is now becoming the primary point of access.

This affects professional roles and patient attitudes.

NHS Direct as a national networked service could run contrary to the concept of locality-based primary care teams. Further research is needed on the impact on demand for in-hours care in general practice, but NHS Direct may account for the problems facing some GPs.

Statements about workload can disguise substantial differences: total volume changes in GP out-of-hours referrals may hide practice- or GP-dependent reductions and may be affected by GPs' inhours availability.

NHS Direct has the potential to deliver and reveal healthcare activity. But concerns remain about the difference between what we may want to know and what we can afford to find out.

Annabelle Mark Reader, organisational behaviour and health management Middlesex University Business School