Participant-data bias, robustness and certainty

Assess availability and analytical assumptions

Assess unavailable studies and harmonization choices alongside study-level bias. IPD access does not make the evidence base complete or unbiased.

Flow diagram

Participant-data bias, robustness and certaintyF1.S1Assess study and result biasF1.S2Assess data availability biasF1.S3Evaluate harmonization assumptionsF1.S4Specify sensitivity analysesF1.S5Assess certainty and limitationsF1.N1IPD sensitivity registerConcernMainspecificationAlternativeEvidenceInterpretationUnavailable IPDHarmonizationMissingoutcomesStudy influenceF1.N2Assumptions and rationaleF1.N3Method and protocol evidenceF1.S0 Start flowParticipant-data bias, robustness andcertaintyF1.S999 End flowEnd flow

Steps

Assess study and result bias

Record design-specific domains and evidence for judgments.

Assess data availability bias

Compare supplied and unavailable studies using accessible characteristics.

Evaluate harmonization assumptions

Identify outcome mapping, missingness and follow-up differences.

Specify sensitivity analyses

Compare modeling choices, influential studies and aggregate-only inclusion.

Assess certainty and limitations

Link outcome-specific conclusions to study and availability limitations.

Participant-data bias, robustness and certainty

End flow

Tables and supporting information

IPD sensitivity register

IPD sensitivity registerConcernMainspecificationAlternativeEvidenceInterpretationUnavailable IPDHarmonizationMissingoutcomesStudy influence

Document prospective alternatives rather than choosing favorable results.

Assumptions and rationale

Participant-level access improves flexibility but does not remove study bias or selection into the available dataset.

Method and protocol evidence

Attach the protocol, methodological sources and versioned software specification supporting these decisions. Record citations and provenance; do not enter participant identifiers.

Sources