In a paper entitled ‘The dementia of dementia praecox’, published in 1978, Johnstone et al refocused attention on cognitive impairment and negative symptoms in schizophrenia. This refocusing was timely because many patients with schizophrenia remained seriously disabled despite the encouraging efficacy of antipsychotic medication in treating positive symptoms. Clearly, therapy focused on positive symptoms was inadequate. This book examines the situation 23 years later. In those 23 years a huge amount of evidence regarding the nature of negative symptoms and cognitive impairment has accumulated. In particular, the link between cognitive impairment and poor functional outcome has been confirmed. Tantalisingly, we have a new generation of antipsychotic medication that offers the prospect of greater efficacy in treating both negative symptoms and cognitive impairment.

However, as I read the book, I experienced growing disappointment. In large measure, the reason for this was the limited utility of the assembled evidence. We still cannot define the criteria for a meaningful trial of treatment for negative symptoms in a way that would satisfy a rigorous regulatory body such as the US Food and Drug Administration. Concepts such as the distinction between primary and secondary negative symptoms appear useful until one attempts to define them precisely. Existing definitions of primary negative symptoms are too strongly tied to treatment non-responsiveness, which creates a bias against the older treatment in any trial comparing a new with an older drug. It might not invalidate a comparison of a new adjunctive treatment with adjunctive placebo, but the pharmaceutical industry has shown little enthusiasm for trials of adjunctive treatment.

Our knowledge of the pathophysiology of negative symptoms has advanced only a little since 1978, when the evidence pointed towards an association with enlarged cerebral ventricles. The more recent evidence indicating the involvement of potentially remediable disorders of neurotransmission offers some hope. However, until our understanding of mechanisms improves, rational therapeutic strategies directed at specific symptoms will be difficult.

In the case of cognitive impairments, the task of defining criteria for a meaningful treatment trial should be easier. The three main conceptual issues in trial design are the multiplicity of different aspects of cognition, which generates the problem of multiple outcome measures; the phase of illness in which the trial might be most meaningful; and the appropriate duration of the trial. This book touches on these issues but does not resolve them. Furthermore, the data offered on treatment of cognitive impairment was mainly derived from open-label trials.

One should not shoot the messenger just because the message is disappointing. None the less, the fault lies not only in the limitations of the evidence. There is a lack of detailed guidance for the future. For example, the discussion of cognitive tasks that might be useful in future trials of new therapies, or in clinical practice, is disappointing. We now know a lot about which aspects of cognition are most relevant. However, issues such as suitability of tests for repeated administration and the time-scale of testing are addressed scantily.

The timing of publication of the book is another problem. Many of us found Keefe et al's (Reference Keefe, Silva and Perkins1999) review of (mainly) open-label trials of the effects of atypical antipsychotics on cognition tantalising, but our real interest was directed towards the outcome of the larger, randomised double-blind trials in progress. The results of those trials have yet to be published fully, and this book provides little advance on the state of knowledge in 1999.

This book provides much information about one of the most important current issues in the treatment of schizophrenia and it covers both pharmaceutical and psychosocial aspects of treatment. It does not, however, provide enough useful answers.