Full-Text Review:Check What the Abstract Didn't Tell You

The question in full-text review is not "is this study interesting?" or "is it high quality?" It is: does this paper meet the predefined eligibility criteria, at the level of the full text?

Quality appraisal — risk of bias — is the next step. Mixing the two doubles your time per paper, so keep this pass strictly about eligibility.

These papers survived because the abstract couldn't settle them, so what you're checking is precisely the information the abstract left out. The usual suspects:

Detailed inclusion and exclusion criteria. The exact content and duration of the intervention or exposure. The nature of the comparison group. Outcome definitions and measurement timing. Follow-up length. All of it lives in the Methods section — so where to read is decided before you start.

The basic route: the population, intervention, and outcome descriptions in the Methods → Table 1 (the actual characteristics of the sample) → the main results figures and tables.

For an eligibility check, you can usually skip the Introduction and Discussion entirely. Feel no guilt about not reading start to finish. For this task, that is the correct way to read.

Two things are easy to miss. First, some papers put outcome definitions and protocol details only in the supplementary materials. If you can't find something in the main text, check the supplements before concluding it isn't reported.

Second, duplicate reporting of the same study. If several papers come from the same cohort, including more than one counts the same patients twice. Be suspicious when authors, institutions, registration numbers, or study periods overlap.

Exclusions at this stage go into the PRISMA flow diagram broken down by reason — so the moment you exclude a paper, jot one short line that maps to your criteria.

Included papers need records too. Note where you confirmed each criterion ("Methods p.4, Table 1") so you don't hunt for the same passage again during data extraction and risk-of-bias assessment.

Before diving into each paper, an AI summary of the study as a whole makes it easier to aim straight at the right part of the Methods.

But never use the summary itself to decide eligibility. This step exists to verify information the abstract compressed away — basing it on a summary, which compresses even further, defeats the purpose. The basis for the decision is always the main text and its figures and tables.

Import a full-text candidate into Paperfy and the AI summary, figures, tables, and original PDF sit side by side on one page per paper. The whole loop — glance at the summary for orientation, check the Methods and tables, add your decision and reasons to the summary — happens on a single screen.

Months later, when you return for write-up and risk-of-bias assessment, the paper is still there with every verification mark intact.