The study of Mackin et al was interesting but so much highly relevant information is missing that it is hard to determine whether the findings have validity. The clinicians are effectively trial participants, yet we are not told the method of selection for doctors in each country. Training and employment structures are so different in the three countries that the clinicians are likely to have had very different degrees of experience and specialisation (the American system in particular favouring greater sub-specialisation). We are required to make the assumption that the groups are similar in all respects except the culture of the country of practice, yet there is no way to tell this without a socio-demographic profile of the participants from each country. There should be an attempt to make them representative of the total population of psychiatrists in their country in terms of ethnicity, gender and other factors which have a strong subcultural influence. There is no unifying ‘culture’ for psychiatrists in the UK, where at least one-third are trained outside the UK, and in some areas of the country the significant majority of doctors are non-UK-trained. Sampling such a small group from the UK (n=20) would be most unlikely to give a representative picture of British psychiatry as a whole. Similarly, India and the USA are also among the most multicultural countries in the world, and the same issues of systematic sampling bias apply.

Furthermore, we do not actually know the ethnic and cultural background of the two videotaped individuals with mania. They are described only as ‘American’ – but can this be a meaningful term when describing an individual's culture in such a varied society? The authors minimise the implications of these difficulties by stating that ‘similar variability is likely to be present when ranking patients in routine clinical practice’. Few would debate the existence of inter-observer variability, but the core issue here is whether the authors’ data support culture as being a central factor in this phenomenon. The design of the study simply does not permit this conclusion.