Place counts

Practitioners in all branches of mental health quickly become aware of the importance of time and place as they develop their craft. The value of guidelines for good practice can be considerable, but if they become echoing mantras that are applied indiscriminately, they can impair good care. Psychiatry does not lend itself easily to sound-bites because so much of what we say and do has to be put into a context that is more complex than most other parts of clinical medicine. When giving a talk about the fundamentals of community psychiatry recently in a country with limited resources I was simply told after I had finished with a flourish, ‘this wouldn't work here’, and, after first feeling affronted, had to acknowledge in discussion that what I took for granted as universal, was clearly wrong.

One example is intervention through work. Although work has long been recognised as an important therapeutic asset in mental health, the best way of enabling this in practice is far from clear. The individual placement and support (IPS) model has attracted strong supportive evidence for its efficacy, first in the USA and Canada but now in many other countries ^{Reference Lehman, Goldberg, Dixon, McNary, Postrado and Hackman1–Reference Killackey, Jackson and McGorry3} and is becoming part of standard practice. So when we have a negative trial, as in this issue from Howard et al (pp. ), it is tempting for an editor to ignore it as an outlying study that does not demand close attention. But it looked to me and to our reviewers that Howard and colleagues had done all they could to replicate the methodology of earlier studies and so we moved ahead to publication. Readers now have to decide whether IPS does not work as well in the UK as elsewhere (the place issue) or whether it was implemented wrongly (the delivery issue). You will find arguments on both sides. Work is more difficult to obtain with the IPS model when local unemployment rates are high ^{Reference Burns, Catty, Becker, Drake, Fioritti and Knapp2} and the two London boroughs studied by Howard et al had higher rates than the UK average. The patients in the UK may be less motivated to gain work because of the alleged ‘dependency culture’ created by our benefits system, illustrated by one of my patients who responded with incredulity to my suggestion that employment might become a long-term goal with the remark, ‘Haven't you got it yet? Work is for suckers’. But Latimer (pp. ) argues that the input given to the IPS in the London trial was less than optimal, even if technically given with fidelity, and we have separate evidence that following the rules of the IPS model improves outcome. ^{Reference Catty, Lissouba, White, Becker, Drake and Fioritti4,Reference Campbell, Bond and Drake5} Place and delivery may also interact as there is evidence that we mental health professionals in the UK have somewhat odd ideas about supported employment and could definitely improve our encouragement and our performance. ^{Reference Howard, de Salis, Tomlin, Thornicroft and Donovan6–Reference Warner8} The same query about the influence of place comes from Doering et al (pp. ), whose report of the efficacy of transference-focused psychotherapy for borderline personality disorder in Austria and Germany comes hot on the heels of a negative study in which transference-focused psychotherapy in The Netherlands was less cost-effective than schema-focused therapy. ^{Reference Giesen-Bloo, van Dyck, Spinhoven, van Tilburg, Dirksen and van Asselt9,Reference van Asselt, Dirksen, Arntz, Giesen-Bloo, van Dyck and Spinhoven10} Is Austria/Germany very different from The Netherlands or does the control group of ‘experienced psychotherapists’, with its very high drop-out rate, support the general guideline view that individual unidisciplinary therapy is not recommended for this condition? ^{Reference Kendall, Pilling, Tyrer, Duggan, Burbeck and Meader11} Soeteman et al (pp. ) put all these questions into context – but of course their study is from The Netherlands only.

On the face of it no one could argue about the importance of place after reading the report by Claasen et al (pp. ), which shows the continuing influence of Emil Durkheim on social psychiatry. Distance from Ground Zero in New York predicted that in the area close to the twin towers attack where social cohesion was enhanced suicide rates went down, but they did not change elsewhere, and the figures seem to shine forth without any fear of contradiction. Now, they have a saying in the East End of London – ‘cut one and we all bleed’ – but I hope we do not have to test the ‘place hypothesis’ to see whether Newham shows the same findings as New York.

Succour, support and transference cures

Friis (pp. ) nicely brings together two other contradictory papers that suggest that gains of early intervention over standard care in first-episode psychosis are good at the beginning but not after 5 years (McCrone et al, pp. ; Gafoor et al, pp. ), and this is reinforced by other evidence. ^{Reference Bertelsen, Jeppesen, Petersen, Thorup, Øhlenschlaeger and le Quach12} Read one of his sentences again. ‘One lesson seems quite clear: specialised treatment for people with first-episode psychosis is effective as long as the treatment continues’. We should not be surprised at this as almost all our ‘cures’ in psychiatry have to be qualified. Was this the expectation when clinicians, researchers and planners set up these services? Before long I expect we will be told, but in the meantime, check the dates whenever you look at the outcomes of this bold experiment.