Abstract
In this issue Elizabeth Penn and Derek Tracy review in great detail the effects of antidepressants in an article headed ‘The drugs don’t work?’. The question mark is important: Penn and Tracy conclude that drugs do work but only after a delay and not in everyone. Much of their discussion centres on the now famous analysis of Irving Kirsch who postulated that antidepressants were only really more effective than placebo in the most severe depression [Kirsch et al. 2008]. Of course what is often forgotten is that placebo is itself a potent antidepressant with an effect size (0.92, according to Kirsch) greater than most medical treatments. In addition to this, Kirsch’s definition of a clinically significant difference is three points on the Hamilton Depression Rating Scale (HDRS). This makes no sense at all because the HDRS is an ordinal scale: someone with a score of 20 is not twice as depressed as someone with a score of 10. Moreover, a three-point difference on the suicide item of the HRDS is a very different thing from a difference of three points on sleep items. It is also worth reflecting on the peculiarity of placebo-controlled studies in depression and other mental illnesses: placebo actually works. This is some contrast with most medical illnesses (think diabetic ketoacidosis, myocardial infarction, most neoplastic disease) where placebo is completely inactive. Add to that the fact that we cannot actually measure the severity of depression with any confidence or any accuracy (unlike, say, blood pressure or plasma glucose) and the surprise is that we have managed to find anything better than placebo for depression. Nonetheless, as Penn and Tracy note, hard endpoints in depression are hard to come by, but where they do exist they support the efficacy of antidepressants: the more use they get, the lower the rate of suicide, for example.
Few would argue about the efficacy of antipsychotics in schizophrenia, particularly when considering olanzapine which is often shown to be somewhat more efficacious than your average antipsychotic. You might think a depot formulation of olanzapine would go off like a rocket but in many countries it has seen only very limited use. This situation has come about largely because of the regulatory demand for a 3-hour postinjection monitoring period owing to the very small risk of postinjection sedation syndrome. Deirdre McGlennon and Chris Bushe describe their experience of using this new formulation in a UK NHS setting. To nobody’s surprise, I suspect, olanzapine long-acting injection was effective in the three patients in whom it was used. What is surprising, however, is the ease with which the 3-hour monitoring schedule was accommodated in normal practice. The injections were given at a day care centre where observation by nursing staff was almost unavoidable and where numerous activities and interventions were available to occupy (in a productive way) those being monitored. There were essentially no cost implications for giving olanzapine pamoate in this environment.
Outside the developed nations, many countries are still making do with conventional depot injections. James Bawo and colleagues report from Nigeria on their survey of psychiatrists’ attitudes to depot medication. Their findings are of interest not only because of the relatively lower use of atypical long-acting preparations but also because there is a view amongst patients in Nigeria that injections are more effective than tablets and because paternalistic attitudes persist in psychiatrists, making patient choice a rarity. These competing concepts combine to produce a picture of widespread use of depots in Nigeria, particularly among those psychiatrists with a patient-centred attitude. Psychiatrists who viewed depots as coercive were less likely to prescribe them.
Each of the three papers in this issue discusses published data relevant to an evidence-based analysis of the utility of different drugs and formulations. What is behind each of the reports, however, is clinical experience of patient benefit: experience of the clear effectiveness of antidepressants and of the relatively greater effectiveness of olanzapine and long-acting antipsychotic injections. It is worth remembering that antipsychotics and antidepressants were shown to be effective (to just about everyone’s satisfaction) without the ‘advantage’ of randomized, controlled trials. Each paper, in its own way, calls for better integration of evidence-based ideals with clinical observation.