- A systematic approach should be developed for reading and evaluating scientific papers.
- First, the decision whether to read a particular paper should be made and based on the study relevance to the subject of interest. (title, abstract, considering the site and facilities).
- Evaluate the validity of the study e.g.:
- A study of a new diagnostic test requires independent blind comparison with a gold standard of diagnosis
- In a study of the clinical course of a disease, a control group should be included.
- Comparison of two therapies requires randomization and blinding.
- Accepting the results should be biased on the validity of the study by reviewing the methodology (blinding, randomization, control group) and the quality of evidence.
- A statistically significant difference may not have clinical relevance.