Daily PubMed evidence board
Comparative Diagnostic Performance of Early and Term MRI in Preterm Infants: a Diagnostic Test Accuracy Systematic review and Bayesian Meta-analysis.
This systematic review and Bayesian meta-analysis compared MRI scans done early (before 36 weeks) versus at term-equivalent age (36-44 weeks) in babies bo…
Signal score64Research triage score
CertaintyModerateVerify in full text
PMID42247320Source identifier
Research triage, not medical advice
Do not use this summary, score, or benefit-cost estimate to diagnose, treat, prescribe, or change care without reviewing the full study and consulting qualified professionals.
Verify before acting
Check full-text methods, eligibility, outcomes, risk of bias, harms, conflicts, funding, replication, and applicability.
Plain-English signal
This systematic review and Bayesian meta-analysis compared MRI scans done early (before 36 weeks) versus at term-equivalent age (36-44 weeks) in babies born preterm to see which timing better predicts later developmental problems. Across 30 included studies, MRI done at term-equivalent age was overall more predictive of later abnormalities (higher sensitivity and specificity for lesions and white matter injury) than early MRI. White matter injury seen at term best predicted later motor problems. Using advanced MRI measures (volumetric and diffusion imaging) improved prediction for cognitive outcomes. The authors note variability in MRI methods and outcome definitions across studies, and call for standardised scoring and combining multiple MRI measures to improve early risk identification.
Why it matters
- This analysis directly compares the predictive performance of early (<36 weeks) versus term-equivalent (36-44 weeks) MRI for later neurodevelopment in preterm infants, addressing a clinically relevant question about optimal timing for prognostic imaging.
- Findings suggest term-equivalent MRI has higher sensitivity and specificity for lesion detection and white matter injury, which matter for counseling families, planning early intervention, and resource allocation in neonatal follow-up services.
- Volumetric and diffusion MRI metrics may improve prediction of cognitive outcomes, indicating a potential role for advanced imaging biomarkers in early risk stratification and research prioritization.
Primary outcomes
- Motor and non-motor neurodevelopmental outcomes assessed between 12-36 months
- Presence of MRI-detected lesions (including intraventricular haemorrhage, white matter injury, cerebellar haemorrhage)
- Volumetric and diffusion MRI biomarkers predictive of cognitive outcomes
Effect summary
Abstract-reported pooled diagnostic metrics: term-equivalent MRI showed higher diagnostic odds ratios, sensitivity, and specificity than early MRI for lesion presence and white matter injury (e.g., lesion presence DOR 14.17 vs 2.67; sensitivity 84% vs 76%; specificity 73% vs 45%). White matter injury at term best predicted motor outcomes (DOR 18.43). Volumetric and diffusion measures improved prediction for cognitive outcomes. Heterogeneity in protocols and modest study numbers limited precision.
Benefit-cost lens
| Quick take | Term-equivalent MRI showed better diagnostic predictivity than early MRI in pooled studies, but implementation or policy change requires absolute predictive values, local baseline risk, costs, and operational feasibility. |
|---|---|
| BCR anchor | 2 |
| Time horizon | 3 |
| Discount rate | 0.03 |
| Assumptions | Assessment based on abstract-level metrics (DOR, sensitivity, specificity) rather than absolute predictive probabilities; implementation requires full-text verification of pooled metrics, participant characteristics, and applicability to local settings. |
Benefit-cost fields are assumptions-based unless explicitly source-derived. Treat them as prompts for deeper economic review.
Risk of bias
| Tool | rapid-abstract-screen |
|---|---|
| Verdict | Some concerns |
| Notes | Assessment based only on abstract information. The study is a systematic review/meta-analysis which used PRISMA-DTA and PROSPERO registration is reported, but abstract notes heterogeneity in MRI protocols, biomarker definitions, and outcomes and a modest number of studies - all factors that can increase risk of bias or reduce applicability. Full-text risk-of-bias assessments (study-level QUADAS-2 or similar) and sensitivity analyses are needed. |
Harms, equity, conflicts & implementation
| Implementation | Full-text review to extract absolute predictive values and subgroup applicability; local capacity assessment for MRI timing and advanced sequences; cost and workflow analysis (scan slots, neonatal transport, staffing); stakeholder (neonatology, radiology, follow-up services) engagement; and validation in local cohorts before practice change. |
|---|---|
| Equity impact | Unclear from the abstract; equity implications depend on access to MRI (including advanced sequences), regional differences in neonatal follow-up services, and whether certain subgroups were underrepresented - full-text subgroup data needed. |
| Harms | Harms are not described in the abstract. Potential tradeoffs include resource diversion, need for transport/immobilization/sedation in some infants, and false reassurance or overdiagnosis leading to unnecessary interventions; these require full-text verification and contextual assessment. |
| Registration | CRD42024528207 |
| Replication | Unknown from automated PubMed triage; replication would require review of included studies and independent re-analysis. |
Source links — verify original
Use PubMed and full-text links to confirm methods, population, outcomes, harms, conflicts, and applicability.