Stillbirths: Contribution of preterm birth and size-for-gestational age for 125.4 million total births from nationwide records in 13 countries, 2000-2020

Abstract

Objective: To examine the contribution of preterm birth and size-for-gestational age in stillbirths using six 'newborn types'.

Design: Population-based multi-country analyses.

Setting: Births collected through routine data systems in 13 countries.

Sample: 125 419 255 total births from 22⁺⁰ to 44⁺⁶ weeks' gestation identified from 2000 to 2020.

Methods: We included 635 107 stillbirths from 22⁺⁰ weeks' gestation from 13 countries. We classified all births, including stillbirths, into six 'newborn types' based on gestational age information (preterm, PT, <37⁺⁰ weeks versus term, T, ≥37⁺⁰ weeks) and size-for-gestational age defined as small (SGA, <10th centile), appropriate (AGA, 10th-90th centiles) or large (LGA, >90th centile) for gestational age, according to the international newborn size for gestational age and sex INTERGROWTH-21st standards.

Main outcome measures: Distribution of stillbirths, stillbirth rates and rate ratios according to six newborn types.

Results: 635 107 (0.5%) of the 125 419 255 total births resulted in stillbirth after 22⁺⁰ weeks. Most stillbirths (74.3%) were preterm. Around 21.2% were SGA types (PT + SGA [16.2%], PT + AGA [48.3%], T + SGA [5.0%]) and 14.1% were LGA types (PT + LGA [9.9%], T + LGA [4.2%]). The median rate ratio (RR) for stillbirth was highest in PT + SGA babies (RR 81.1, interquartile range [IQR], 68.8-118.8) followed by PT + AGA (RR 25.0, IQR, 20.0-34.3), PT + LGA (RR 25.9, IQR, 13.8-28.7) and T + SGA (RR 5.6, IQR, 5.1-6.0) compared with T + AGA. Stillbirth rate ratios were similar for T + LGA versus T + AGA (RR 0.7, IQR, 0.7-1.1). At the population level, 25% of stillbirths were attributable to small-for-gestational-age.

Conclusions: In these high-quality data from high/middle income countries, almost three-quarters of stillbirths were born preterm and a fifth small-for-gestational age, with the highest stillbirth rates associated with the coexistence of preterm and SGA. Further analyses are needed to better understand patterns of gestation-specific risk in these populations, as well as patterns in lower-income contexts, especially those with higher rates of intrapartum stillbirth and SGA.

Keywords: gestational age; newborn; pregnancy; premature birth; preterm; stillbirths.

FIGURE 1

Input data for stillbirth analyses, 2000–2020. (A) Flowchart of data inclusions and exclusions. *For the sensitivity analysis: 232 488 stillbirths and 612 436 live births at 22–27 weeks’ gestational age were excluded. Total number of births at ≥28 weeks: 402 619 late gestation stillbirths, 124 171 712 live births. **Due to overlaps of missing and implausible data, the total excluded values do not add up to the difference between box 3 and box 4 and between box 3 and box 5. (B) Number of stillbirths (635 107) and total births (125.4 million) by country. *UK: England & Wales: stillbirths = 13 831, total births = 3.2 million, SBR = 4.3; Scotland: stillbirths = 5127, total births = 1.1 million, SBR = 4.5. Map legend shows the distribution of 125.4 million total births (124.7 million live births and 635 107 stillbirths at ≥22⁺⁰ weeks) with information to classify by ‘newborn types’ included in this study.

FIGURE 2

Distribution of ‘newborn types’ among: (A) all stillbirths (see Table S4a for country‐level data), (B) late gestation stillbirths (see Table S4b for country‐level data), (C) live births, (D) neonatal deaths, 2000–2020. *Data not available from Argentina, Iran or Malaysia.

FIGURE 3

(A) Stillbirth rate ratio by ‘newborn types’ among all stillbirths (≥22⁺⁰ weeks), 2000–2020. (B) Stillbirth rate ratio by ‘newborn types’ among late gestation stillbirth (≥28⁺⁰ weeks), 2000–2020. Each point represents the stillbirth rate ratio, box plots summarise median values and IQR (25th and 75th percentiles) (A: countries = 13, n = 635 107; see Table S4a for country‐level data. B: countries = 13, n = 402 619; see Table S4b for country level data).