Medical writing often asks us to compress: a study into an abstract, a programme into a slide, a body of evidence into a recommendation. Compression is useful, but it creates a responsibility. The shorter the format, the easier it is to make an uncertain conclusion sound settled.
Uncertainty is part of the result
It can arise from sampling variation, missing data, measurement limitations, study design, indirectness, inconsistency, or simply a small evidence base. These are not editorial inconveniences. They determine how much weight a conclusion should carry and where it can be applied.
Regulatory benefit–risk assessment explicitly has to contend with uncertainty around the probability and size of desirable and undesirable effects. Scientific writing should not hide that complexity; it should organise it so a reader can make a more informed judgement.
Three habits that keep claims proportionate
Name the evidence. “The trial found,” “the analysis suggests,” and “the mechanism supports a hypothesis” are not interchangeable. The verb should reveal the kind of support behind the sentence.
Keep the denominator and comparator visible. Risk, response, and change become easier to misread when separated from the population, timeframe, baseline, or alternative against which they were measured.
Qualify precisely. Vague caution can be as unhelpful as overstatement. Instead of adding “may” everywhere, state the actual boundary: short follow-up, wide interval, exploratory analysis, surrogate endpoint, or limited generalisability.
The goal is not cautious-sounding prose. The goal is a conclusion calibrated to the evidence.
Clear and uncertain can coexist
A useful structure is to lead with the most decision-relevant finding, then place its magnitude, precision, and important limitation close beside it. A reader should not have to search a footnote to discover that the endpoint was exploratory or the estimate imprecise.
This also makes writing more durable. When new evidence appears, a bounded claim is easier to update than an absolute one. Responsible version control and visible corrections matter because scientific understanding evolves; the ICMJE treats correction as part of maintaining an accurate published record.
Confidence comes from traceability
The most trustworthy sentence is not necessarily the boldest. It is the one whose source, reasoning, and limits can be followed. In medical communication, confidence should come from that traceability—not from tone alone.