Evaluating evidence found on the Internet

AORN Journal, Nov, 2000 by Suzanne C. Beyea

Searching the Internet for evidence to support practice decisions has become a common strategy for both practitioners and patients. Numerous data resources on the Internet provide up-to-date, reliable, valid, and relevant information. The challenge is differentiating quality sites from those that post unsubstantiated reports or articles, as well as information not based on research or clinical evidence. As clinicians and health care information consumers, nurses must develop the skills essential for evaluating web sites for trustworthiness and be skilled at helping patients critically examine information they find on the Internet.

LEVELS OF EVIDENCE

Keep in mind that in the published literature, many different types of evidence exist. Not all published material is based on fact, research, or scholarship. One author describes the "publishing wedge for therapy," which consists of seven levels of evidence within the health care literature, as an approach to illustrate different levels of evidence. Level one includes ideas, editorials, letters, and opinion papers. Level two information consists of case reports, case studies, and unusual happenings. Levels three through seven present research findings from laboratory studies (ie, level three), animal studies (ie, level four), and studies involving human subjects (ie, levels five, six, and seven). The various levels of human research reflect increases in complexity, scope, and rigor.(1) Some clinical problems have been studied through level seven or may not be amenable to clinical trials, so regardless of where information is published, always consider the level of the evidence.

When searching for evidence, one might identify systematic research reviews, research-based protocols, and clinical practice guidelines. These types of reports can be helpful because they provide a comprehensive literature review and synthesize research evidence. Regardless, the reader must evaluate the report or guideline critically to determine if the information is accurate and appropriate to the clinical situation. Most guidelines are developed using a framework to rate both the type and strength of the evidence and use a rating scale to help the reader understand the classification.(2)

Keeping these concepts and frameworks in mind, consider how one can approach evaluating evidence found on the Internet. Remember that information found on the Internet has to be appraised critically in a process similar to the approach used when evaluating other published materials. For example, one would expect the trustworthiness of information found in the AORN Journal to be much higher than that found in a tabloid newspaper. Within the AORN Journal, readers find level one articles (ie, editorials and opinion articles), as well as research studies with human subjects (ie, levels five, six, and seven), so the reader must always consider the strength of the evidence within a publication. Never be misled by the belief that, if a research study is published in a reputable journal, it guarantees high-quality research. Some of the most reputable medical journals have been criticized for the inclusion of research studies that lack appropriate attention to scientific rigor. The reader must evaluate critically every article he or she reviews, whether published in scientific journals or on the Internet.

The Internet poses some special challenges to identifying up-to-date, valid, reliable evidence-based references and materials. Consider the vastness of the resources on the Internet. For example, if one performs a search on the term diet using http:// gotonet.google.com/, more than 924,000 web sites are identified. Narrowing that search to the term low fat diet results in more than 11,300 possible hits. Trying to review and evaluate all of those web sites obviously is impossible. Both of these searches identified numerous web sites proposing strange, unusual diets and supplements that might suggest quackery to many professionals and patients. Users must identify and use sites they have evaluated as reliable and trustworthy and carefully critique any unfamiliar sites.

In the aforementioned situation, one might look at the American Heart Association's (AHA) web site (ie, http://www .americanheart.org) and examine its recommendations for a low fat diet. At such a site, one would expect to find the latest evidence-based recommendations about following a low fat diet. On this web site, in fact, there are links to scientific papers and statements for professionals, as well as consumers.

Web resources, such as the AHA, the National Library of Medicine (NLM), the Agency on Health Care Quality and Research, and the National Guideline Clearinghouse, are recognized for the timeliness, quality, and trustworthiness of information; however, one must evaluate all the information found on the Internet and consider the level and strength of the evidence, regardless of the source. Within the NLM databases, users may find references to editorials, opinion articles, and case studies, as well as controlled clinical trials with human subjects.