The point of our study was to see whether a very simple, cheap feedback intervention might work to improve quality of website information. Clearly, it either did not work or the effect was small. It is quite possible that more elaborate feedback interventions might work. This needs to be tested. However, if these were to work, would they be of any practical use? Is anyone going to go to the trouble of routinely monitoring website quality and personally contacting website developers to give them feedback? Who would fund this sort of work? There is also the related issue of who would resource website owners to carry out substantial revisions. In this regard, it is interesting that after our trial was over, one website administrator wrote to us saying that they had now revised their website in response to our feedback. The reason they cited for the delay is the limited resources they had as a non-government organisation.

Readers of our article may be interested in another study on feedback which only came to our attention after our trial was completed. This was a much larger randomised controlled trial (n = 299 URLs) from the field of pharmacology and gave feedback on quality of information on the drug sildenafil. Like our trial, this one found no effect of emailed feedback letters.