AI Summary & Key Takeaways
AI InsightBecause classification is sensitive to the assumed day, prevalence from week-only registries should be reported with that assumption.
📌 3 Key Takeaways
Abstract
Aims: To assess the fit of the INTERGROWTH-21st newborn standards to the birth weight distribution of a Turkish tertiary cohort. Methods: Retrospective analysis of 41 967 live singleton births at 24-42 gestational weeks (2022-2026), classified as small (SGA) or large for gestational age (LGA) with the INTERGROWTH-21st standard (≥33 weeks) and very preterm reference (<33 weeks). Gestational age was recorded in completed weeks; the mid-week (week+3) value was used and all seven day assumptions tested. Proportions at seven percentiles were compared with nominal values by exact binomial tests; a probit fit gave location and scale estimates. Results: SGA prevalence was 8.3% and LGA 10.3%; the z-score distribution was estimated as mean +0.07, standard deviation 0.97. After correction for multiple comparisons, departures were confined to the lower half (below the median 45.9% vs 50%; all p<0.001); upper-tail proportions stayed close to nominal. Under the mid-week assumption 3479 newborns were SGA, 718 fewer than the 4197 expected at 10%. Across the seven assumptions SGA prevalence ranged from 5.8% to 11.0%, reclassifying up to 4547 newborns (10.8%). Adolescent maternal age [adjusted odds ratio (aOR) 1.49; 95% CI 1.31-1.71] and foreign nationality (aOR 1.28; 1.17-1.40) remained associated with SGA. Conclusion: Under the prespecified mid-week assumption, the standards fit this cohort closely but classify 718 fewer newborns as SGA than the nominal 10% implies, leaving them outside the threshold that prompts further assessment. Because classification is sensitive to the assumed day, prevalence from week-only registries should be reported with that assumption.