One must consider:
Validity: There are numerous journal articles and reports published on a daily basis, and not all are of equal value. If the article is a research study, is it published in a peer-reviewed journal? Is the study group large enough? The smaller the study group the more anecdotal the findings. Is the study biased in some way (eg, targets only a certain demographic, or is vendor/manufacturer sponsored) Consider the medium in which it is published. Is it a reputable website or company that is reporting the information?
Relevance: Determine if the trial or experiment is relevant to your question (eg, Do they use the same assay and do they have the same sample type?). Research the author(s). Do they hold the credentials needed to write an article in that field or specialty? Check the date and make sure the source you found is current-- preferably within the last three years. See if the author has published other related studies or is working at a reputable laboratory or university. Look at the disclosure statement in the article (all good journals now have them), and see if the author(s) are financially affiliated with the company/product they are reporting on.
Study design: Is the trial or test a blind comparison, or is there a current "gold" standard that it is compared to? Are test samples randomized (preferable)? Are outliers inappropriately removed from the study? Does the study compare 'apples to apples' when evaluating different tests or treatments? Do the author's conclusions make sense? If you think they are reaching, then they probably are!
The image on the right shows the number of publications in the laboratory field per year compared to other medical publications. With this drastically larger number of publications, it is important that you assess the validity and relevance of the publication.
(Note: we will revisit the search and assessment steps once we are finished outlining the entire process).