An individual-participant-data review reanalyzes records, not just published averages

Identify the material behind the synthesis

A review can combine estimates from published tables, obtain original participant records, or use both. Locate the actual approach for the result being quoted. A review having many participants does not establish which approach it used. Nor does the phrase individual data mean that identifying information is public or that readers should seek private records.

Cochrane’s individual-participant-data chapter describes obtaining, checking and reanalyzing the original data from included studies. That methods distinction is used here, without treating the approach as a quality badge or importing a treatment finding. Our reading task concerns data coverage and the analysis reported, not entitlement to participants’ information.

Cochrane: Cochrane Handbook Chapter26: Individual participant data

Make an eligibility-to-analysis coverage map

List the eligible studies named for the relevant review question. Add separate columns for data requested, usable data obtained and included in this analysis. Keep unreported statuses unknown. Some material may be available but unusable for a particular outcome, so receipt and analytical inclusion are different stages. A combined participant total does not explain each stage by itself.

In a hypothetical review, three eligible studies supply records and a fourth supplies only its published summary. The reader checks whether the final result uses three studies, combines all four through stated methods, or describes another choice. That is a coverage question, not evidence that the unavailable study had a favorable or unfavorable result. No actual review is described.

Find what the reanalysis added

Original records may allow consistent definitions, checks or participant-level analyses that published summaries could not support. Read which of those tasks the authors actually carried out. Having data available creates possibilities; it does not show that every useful analysis was done or that missing measurements were recovered. Keep the outcome and assessment time attached to each finding.

Also check the analytical plan and whether participants remain grouped within their original studies appropriately. An introductory reader need not fit a model to ask whether the explanation is present. If the analysis is too technical, preserve its stated design and ask a qualified methodologist about it rather than interpreting access alone as proof of a trustworthy result.

Describe the benefit and the remaining boundary

A bounded summary names the participant-data coverage, the actual reanalysis and any combined aggregate evidence. It can acknowledge useful additional detail without claiming that the entire literature became complete. If the review could not obtain some data, retain its explanation and limitations rather than inventing the missing results or condemning the whole project.

For a health decision, seek professional interpretation of the actual evidence and your circumstances. A personal tracker record is not a research dataset that should be sent to a review team without an authorized study process. Finish with a transparent coverage map showing what supports the quoted analysis, while keeping your ordinary quit planning independent of a methodological label.