Much has been written about the motivation of perpetrators in Munchausen by proxy syndrome (MBPS), but little, as far as I know, about the motivation of doctors who "collude" with parents in these behaviours. In general, I have no quibble with the analysis by Jureidini and colleagues.1 It is interesting, as they have done, to put forward a theory as to why doctors who are motivated to make "the diagnosis" might fail to recognise invented illness; however, the motivation of the "relentless investigator" is not the only one that should be examined.

Indeed, doctors do not practise in a vacuum and are subject to pressures, societal among others, which may change with the clinical setting, affecting their practice. As a result, we may make mistakes in aspects of the clinical encounter — whether in our assessment of the process, the history, the examination or our interpretation of tests. Thus, when dealing with children, persistent parents may force the hand of even-thinking doctors because of current medicolegal and societal pressures.

I do agree that children are more at risk from doctors who are enthusiastic about making a medical diagnosis and who ignore the social aspects of the consultation. However, concentrating on the motivation of the parent in MBPS cases has not proven useful in practice because of the complex psychopathology underlying such behaviour,2-4 and I doubt that focusing on the motivation of doctors will improve practice for similar reasons.

While the relentless investigator is a particular problem in these cases, factors such as being busy, stressed or forgetful may also account for failure to identify the problem sooner. My concern is that, for an understanding of motivation to be useful in preventing the phenomenon of MBPS, it would be necessary to both reliably identify doctors' motivations and intervene to change these, and I do not believe that either of these steps is feasible.

As a practising paediatrician, I suggest that a more useful approach to this problem of paediatric diagnosis would be to establish guidelines for the problem of the unresolved clinical problem or "difficult diagnosis" (see Box).