Beantwortbare Fragen - Step 3

Translate the identified keywords and key-concepts into answerable questions - use the PICO scheme as help:

• Patient - Is it necessary to specify patient or disease characteristics inside the search question?
• Intervention - Which is the intervention of interest?
• Comparison - What comparators shall be included in the search?
• Outcome - How can patient-relevant outcomes be specified inside the search question?

Looking for Things that exist inside the Database

All of the above identified concepts can be sought for inside a database if they are applied to existing, real facts.

Thus, you can now build a search strategy with the identified components.
In cases where patient characteristics are coupled to the disease or symptom in question, you can just search for this disease or symptom (like prostate cancer - only if you want to specifically know something about the unusual young patient you need to include this patient characteristic into the search formulation).
Usually you want to know about some intervention - whether it is useful, i.e. better than another intervention targeted at the same problem, or better than the "natural course". Thus, you have to find information about that intervention - preferably as "head-to-head"-comparison between this intervention and the chosen comparator, and preferably on evidence level "1", i.e. you are looking for randomized controlled trials (RCTs) as primary sources of your evidence.
Sometimes, no "head-to-head" comparison can be found. Then one needs a common comparator - like, one study comparing strategy "A" with the natural course, another study comparing strategy "B" with the natural course. Or, there could be a comparison between "B" and "C", giving some - hopefully - absolute outcome parameters, like intervention effect on absolute life-span plus disease related mortality or some other absolute parameter. Then, if you are interested in therapy "A", and you find any other study giving information regarding the same absolute parameters in connection with this therapy, you can indirectly compare therapy "A" with therapy "B" and "C".

Sometimes, also no RCTs can be found - in which case you will be looking for "the next best thing", which could be a well designed cohort trial, or an equally well-designed non-randomized controlled trial. Even data from disease registers can be reliable sources of evidence - but you have to make sure about the quality of the data inside the register.
Thus, the design of a trial gives you a good hint at the reliability of the evidence elicited from that study - but you have to be wary, since there are always such things as applicability of the results due to patient selection, like the artificial circumstances inside the trial frame, which make an exact replication of results within the context of "normal" primary care unlikely, etc., etc. ... Do not be over simplistic and do not take the results of any RCT, however well designed overall, as some absolute truth.