Dear Editor,
We write in response to the recent short report on the ‘Attitudes and practices in the management of ADHD among healthcare professionals who responded to a European study’ (Fitzgerald & McNicholas Reference Fitzgerald and McNicholas2014). The objective of the report and the clearly exhaustive collaborative process are both commendable. It is difficult, however, to agree with the finding that ‘intriguing country- and profession-specific differences emerged’, due primarily to a lack of generalisability of their results.
Web-based surveys have a number of attractions; large number of subjects can be contacted, they are inexpensive and they reduce the time needed for data collection (Heiervang & Goodman Reference Heiervang and Goodman2011). Caveats include that the subjects are self-selecting rather than random or necessarily representative (Braithwaite et al. Reference Braithwaite, Emery, de Lusignan and Sutton2003). In Fitzgerald and McNicholas’ study, the authors acknowledge that their decision to issue the survey in only one language led to potential for further bias.
While we acknowledge that it is well-recognised that response rates are lower for online surveys (Manfreda et al. Reference Manfreda, Bosnjak, Berzelak, Haas and Vehovar2008), the response rate here was extraordinarily low at 0.6%. That this was reported is in some ways to be applauded, as the non-reporting of response rates is occurring. It clearly, however, falls short of the widely accepted 50%–60% response rate for meaningful conclusions to be drawn (Fincham et al. Reference Fincham2008). Means of increasing response rates, with modest evidence, include personalising the contact details on the emails (Cook et al. Reference Cook, Heath and Thompson2000) or telling the respondent that they have been selected as part of a small group to participate (Porter et al. Reference Porter and Whitcomb2003), but one could not expect for any intervention to increase a response rate from 0.6% to 60%.
Absolute numbers are important, and little could be drawn from a survey of a non-self-selected group of 134 clinicians across Europe. We disagree with the authors that any meaning can be inferred from the fact that four health professionals in France expressed a different view from 11 health professionals in Italy. We would be of the same view were these to be 15 non-self-selected participants.
