I wish to clarify some of the points raised by Dr Darely. First, it is important to remember that both groups in our study received ‘treatment as usual’, while those subjects randomised to receive a green card also had the facility to access emergency telephone consultation. We know from our data that there were no significant differences between groups with respect to their management following deliberate self-harm (DSH) assessment, and it is important to remember that the management plan was presented to patients before randomisation to avoid any subsequent bias in the treatment offered. We did not document other advice given over and above the main management plan but this is likely in both groups to have included advice to consult with the patient's general practitioner or psychiatric keyworker (if applicable).

Details of how the telephone support system was used, together with its effects on patient's uptake of other routine medical and psychiatric services, are soon to be published in a separate paper (further details available upon request). Speculation about why the green card appears to have a detrimental effect in DSH patients with a previous history of self-harm and a positive effect in ‘first-timers’ must remain tentative as these were secondary subgroup findings. Further research, in the form of a large multicentre trial, is needed to clarify the effects of the green card on patients presenting with DSH for the first time.

The mechanisms for such effects of the green card are even more speculative at this stage. For first-timers (only a minority of whom will use the card) it is not clear whether knowledge that the card is there to be used should a crisis ensue (the ‘safety net’ hypothesis) is the most important ingredient or whether the consultation itself makes the difference. Further qualitative work in this area utilising patient interviews would be welcome. For some patients with a history of repeated DSH, it could be suggested that the green card may heighten their experience of gratification resulting from a use of self-harm as an albeit distorted form of communication. Another hypothesis might be that a brief, focused telephone consultation from a psychiatrist in training is not containing enough and actually increases self-harming behaviour in such individuals. It seems unlikely that this apparent paradox will be explored further in view of ethical considerations when proposing research methodology. However, it may well be a pertinent issue given the projected prominence of ‘NHS Direct’.