For webpages there are at least three levels of quality assessment that may be desirable. The first would be provenance, the second content coverage and the third the quality of that coverage. The "provenance" quality is most easily assessed. Silberg et al. (1997)
proposed the four criteria of authorship, disclosure, attribution and currency. While these are important, it remains true that a site could satisfy all these criteria and still be misleading or inadequate. Clough & Oliver (2002)
proposed that orthopaedic patient information pages on the Internet should cover a set of topics relating to the condition, nonoperative treatment, and operative management. However, even if those topics are covered, there is no certainty that they are covered well. Informatics faces a huge task defining and assessing quality. It will involve the creation of standards, the assessment of pages against those standards and some form of "enforcement" of those standards. An intriguing finding comes from a study of patient information on carpal tunnel syndrome
. Fricke et al found that "indicators that make use of the invisible link structure of the Internet are more reliable guides to accurate information". This suggestion that better pages received more links should be further studied and implies an inherent quality mechanism in the Internet.
Some form of central directory of orthopaedic information would seem to be essential to any quality control mechanism. If a listing on the directory became a desirable promotion of traffic to the site (i.e. readership), then the directory itself would be in a position to suggest standards.
