It is ungenerous and unjustified for Helen Killaspy ^{Reference Killaspy1} to accuse George Lodge ^{Reference Lodge2} of nostalgia and wearing rose-tinted spectacles just because she disagrees with him. Newer is not automatically better. We have had altogether too much frenetic reorganisation of mental health services where every change, no matter how hare-brained, is hailed ‘an innovation’. Nobody waits to see whether it makes any difference, never mind delivers an improvement. It is whether an idea is right or not that matters, not how long it has been around. Similarly, it is disingenuous of her to claim that the service changes she describes were ‘informed by research’.

New developments arise from a mixture of creative thinking and professional ambition, and there is nothing wrong with that. By the very nature of the beast, evidence comes later. We need the new services in place to research them rigorously or make judgements from mature experience. The National Service Framework is a case in point. Only one of the new teams imposed had any evidence for it at the time, and assertive outreach teams’ international evidence was unravelling in the UK context as they were being rolled out. ^{Reference Burns, Creed, Fahy, Thompson, Tyrer and White3}

There was not a single randomised controlled trial of crisis teams until Johnson's excellent, but still unrepeated, 2005 study. ^{Reference Johnson, Nolan, Pilling, Sandor, Hoult and McKenzie4} The only two randomised controlled trials of early intervention teams also came later, and neither found a significant advantage in their declared primary outcomes. A more measured position is probably justified.

Continuity of care can be a complex concept to define ^{Reference Burns, Catty, White, Clement, Ellis and Jones5} but it is not that difficult to recognise. We can all grasp the importance of being treated by familiar individuals who know our situation and illness, of not being passed on, and not having to repeat our history to an endless stream of new staff whom we then have to learn to trust. Everybody who is asked, patients, staff or families, insists that they value continuity. I know I do.

Whatever else mental illnesses are, they are experienced, expressed and treated in relationships. George Lodge is right that these relationships have been given altogether too low a priority in recent planning and strategy. Our decade of fragmentation may have contributed some improved understanding of process, but undoubtedly at a cost of simple humanity and attention to the unique individuals for whom the whole edifice exists. Helen Killaspy is right that we have a progressive discipline, responsive to an expanding evidence base. That does not mean that every change is improvement, nor that more specialised services (with their inevitable fragmentation of care) are necessarily better for patients.