Use a cohort-specific appraisal tool to judge selection, measurement, and confounding, recording a judgement and reason for each item instead of one global score. The NHLBI Study Quality Assessment Tools are organized by study design and direct reviewers to internal validity and potential flaws in methods or implementation; the same source notes that users must define their own judgement parameters. STROBE is a reporting checklist for cohort, case-control, and cross-sectional studies and assesses reporting completeness, so use it to locate what the paper reports and the appraisal tool to judge those methods.
- Confirm the design and pick the cohort tool. Check that the study follows a group over time and reports the exposure and outcome as a cohort; if the paper is cross-sectional or case-control, switch to the matching tool. Expected result: one appraisal form whose items fit the design.
- Set your judgement parameters before reading the results. Decide what counts as adequate for each item, such as how the cohort was selected, how winter illness was measured, and which confounders were addressed. Expected result: a consistent scale you apply to every item.
- Appraise selection. Record how participants entered the cohort, who was excluded, and whether the compared groups came from the same source population. Expected result: a judgement with the reason and the page or table supporting it.
- Appraise measurement. Record how exposure and winter illness outcomes were defined and ascertained, and whether the same method applied to all participants. Expected result: an item-level judgement on measurement quality.
- Appraise confounding. List the confounders the authors identified and the design or analysis steps used to address them. Expected result: a judgement on whether confounding was handled, with the specific confounder named.
- Complete the form. Enter a judgement and reason for every item, then summarize the pattern of weaknesses. Expected result: a completed appraisal form with item-level justifications that a reader can trace back to the paper.
Use STROBE to check whether the cohort's methods and results are reported completely enough to appraise; a missing detail leaves that item unresolved, and the next step is to check the paper's supplementary material or contact the authors.