Slade et al (Reference Slade, Leese and Cahill2005) have published a potentially important study of the relationship between patient-rated unmet needs, quality of life and the effect of meeting those needs. They draw the conclusion that ‘meeting patient-rated unmet needs should be the starting point for mental healthcare’. Although much psychiatric care is indeed directed towards reducing unmet need, we believe that this research shows (over the time scale of the study) that reducing unmet need is actually largely ineffective. A longer study might confirm continuing incremental improvement but this would need to be demonstrated.

In the descriptive part of the study the authors show that low quality of life is associated with high unmet need. Figure 1 shows a clear gradient which can be estimated to be –0.2 by inspection (no summary statistics are given). By contrast, in the second part of the study, which looks at the effect of reducing unmet needs, Fig. 2 shows almost no relationship between change in unmet need and change in quality of life (summary statistics: B= –0.04, s.d.=1). Although B indicates high statistical significance it seems to be clinically irrelevant: one would have to meet 25 unmet needs to improve quality of life by one point; B is very small compared with the standard deviation and importantly is only one-fifth of the gradient in Fig. 1.

Thus quality of life and unmet need are associated (gradient= –0.2) but meeting unmet needs has a negligible effect (gradient B= –0.04) on quality of life. This suggests that unmet needs do not cause low quality of life and that the relationship between the two may be mediated by some third factor, such as psychiatric illness, that causes both. If this were the case, treating psychiatric illness should be the starting point for mental healthcare and not ‘meeting patient-rated unmet needs’.

Furthermore, if the justification for meeting unmet needs of psychiatric patients is to improve quality of life per se, then this research shows that in terms of size of effect (and over the period of the study), reducing unmet need is largely ineffective, and is therefore a questionable use of resources.