I congratulate Lieb et al on their excellent systematic review. ^{Reference Lieb, Völlm, Rücker, Timmer and Stoffers1} However, it is interesting that studies until June 2008 were included in this review; moreover, that in January 2009 the National Institute for Health and Clinical Excellence (NICE) guidelines advised that ‘drug treatment should not be used specifically for borderline personality disorder or for the individual symptoms or behaviour associated with the disorder’. ²

I am surprised that there were no randomised controlled trials (RCTs) available at the time of study on the usefulness of quetiapine, although some RCTs of aripiprazole and olanzapine were. A few open-label studies have been done highlighting the usefulness of quetiapine in reducing impulsivity and affective symptoms, ^{Reference Villeneuve and Lemelin3–Reference Adityanjee, Romine, Brown, Thuras, Lee and Schulz7} and it is evident in clinical practice that it does have some beneficial effects on mood instability and aggression.

It is a pity that forest plotting could not be done, which would have shown how much variation existed among studies and the degree of precision of each study, although one can understand the various difficulties faced by the authors.

Lastly, I would like to seek clarification regarding somewhat conflicting statements in the paragraph ‘Implications for practice and research’; it initially states ‘nor can low-dose antipsychotics be advised for cognitive–perceptual symptoms as earlier recommended by the American Psychiatric Association Practice Guidelines’, but later states ‘the SGAs (aripiprazole, olanzapine) should be the first choice for treating cognitive–perceptual symptoms’.