Write down the actual question
Who was studied, what was compared and what outcome was measured? Use the paper’s own wording in your private notes, then write a plain-language version. If you cannot identify the comparison, leave that field open rather than filling it with an assumption.
Keep different evidence separate
A laboratory measurement, an account of one patient and a comparison between groups are different kinds of report. Ask what the study can actually establish. A finding about a material or an experimental model does not by itself establish a benefit for people using a product.
Make uncertainty visible
Record the limitations beside the result. Look for a registered protocol, missing data, funding information and whether independent work reaches a similar conclusion. A useful reading note can end with an unanswered question; it does not need to produce a treatment recommendation.
