The low availability of metadata elements for evaluating the quality of medical information on the World Wide Web.

Abstract

A great barrier to the use of Internet resources for patient education is the concern over the quality of information available. We conducted a study to determine what information was available in Web pages, both within text and metadata source code, that could be used in the assessment of information quality. Analysis of pages retrieved from 97 unique sites using a simple keyword search for "breast cancer treatment" on a generic and a health-specific search engine revealed that basic publishing elements were present in low frequency: authorship (20%), attribution/references (32%), disclosure (41%), and currency (35%). Only one page retrieved contained all four elements. Automated extraction of metadata elements from the source code of 822 pages retrieved from five popular generic search engines revealed even less information. We discuss the design of a metadata-based system for the evaluation of quality of medical content on the World Wide Web that addresses current limitations in ensuring quality.