Deciding to integrate evidence from non-randomised studies of interventions (NRSI) in an evidence synthesis has methodological and practical implications. New GRADE guidance supports transparent judgements by providing clear criteria for when NRSIs can complement, replace, or follow RCT evidence, and when integration should be avoided.
Session 7 of the GRADE Learning Live series introduced the structured and practical GRADE approach for integrating (or not integrating) RCTs and NRSIs in evidence syntheses.
Topics included:
- Assessing the certainty of evidence (CoE) from non-randomised studies: bias and upgrading factors
- Non-randomised studies as complementary, sequential or replacement evidence
- Four steps for deciding when and how NRSIs can meaningfully complement or replace RCT evidence:
1. Assessing the CoE of both RCTs and NRSI
2. Assessing congruence of effect estimates from RCTs and NRSI
3. Identifying and comparing GRADE domains that are affected by certainty ratings
4. Deciding whether and how to use both types of studies and how to present the information in GRADE tables
This webinar was suitable for those wanting to use GRADE to interpret and summarise findings in a systematic review. An understanding of systematic review and GRADE methods for assessing certainty of evidence is assumed.
The session was delivered in October 2026 and below you will find the recording with the accompanying slides to download [PDF].
Presenter Bio
Carlos A. Cuello-Garcia, MD, PhD is a clinical epidemiologist at McMaster University whose work focuses on evidence synthesis, GRADE methodology, and guideline development. He is a member of the GRADE Working Group and works closely with health decision-makers in Canada and internationally.
