Now that you have determined the type of question you are trying to answer, choose the study type that is best for that type of question.
If your question is:
Therapy: Use Randomized Controlled Trials, Systematic Reviews, or Meta-Analysis
Diagnosis: Use Prospective, bind comparison to gold standard (compares the old standard with the new item to be tried: Example: Rapid HIV Test versus the standard ELISA test are the results just as reliable?)
Etiology/Harm: Use Cohort or case study
Prognosis: Use Cohort
Prevention: Use Randomized Controlled Trial, Systematic Reviews, or Meta-Analysis
Clinical Exam: Use Prospective, blind comparison to a gold standard
Cost: Use Economic Analysis
All 3 of the examples below are Therapy questions, so we would search for Randomized Controlled Trials, Systematic Reviews, or Meta Analysis to answer these questions.
Example: In a child under 2 years of age with otitis media, is watchful waiting versus amoxicillin x 10 days as effective in reducing symptoms.
Example: In a patient with Bell's Palsy, is valacyclovir versus prednisolone more effective in improving facial paralysis?
Example: For all persons entering the critical care unit, is the addition of hand rubbing with a waterless, alcohol based solution more effective than just hand washing with soap and water in reducing nosocomial infections?