Learn R Programming

jarbes (version 2.4.2)

preterm: Meta-analysis of Cognitive Outcomes in Very Preterm Children

Description

This dataset contains study-level data extracted from the systematic review and meta-analysis Trends Over Time in Cognitive Outcomes of Children Born Very Preterm by Behboodi et al. (JAMA Pediatrics, 2025). The review synthesizes evidence from observational cohort and matched studies comparing cognitive outcomes of children born very preterm (VPT) or extremely preterm (EPT) with children born at full term (FT).

Arguments

Format

A data frame with 61 rows and 18 columns. Each row represents a single study included in the meta-analysis. The columns include:

study

First author and publication year.

n_EPT.VPT

Number of children born extremely or very preterm.

mean_EPT.VPT

Mean cognitive score in the EPT/VPT group.

sd_EPT.VPT

Standard deviation of cognitive scores in the EPT/VPT group.

n_FT

Number of children born full term.

mean_FT

Mean cognitive score in the full-term group.

sd_FT

Standard deviation of cognitive scores in the full-term group.

SMD

Standardized mean difference (Hedges’ g) comparing EPT/VPT with FT.

seSMD

Standard error of the standardized mean difference.

mean_age

Mean age at cognitive assessment.

mean_ga

Mean gestational age of the study population.

mean_bw

Mean birth weight of the study population.

median_birth_year

Median year of birth of the cohort included in the study.

quality

Methodological quality score based on a modified Newcastle–Ottawa Scale, ranging from 0 to 7, with higher values indicating better study quality.

matched

Indicator of whether the study used a matched design to compare preterm and full-term participants ("yes" or "no").

birth_level

Classification of the preterm subgroup included in the comparison: VPT for very preterm only, EPT for extremely preterm only, and VPT_and_EPT for combined very and extremely preterm groups.

country

Country where the study was conducted.

analysis

Indicator of whether the study contributed data to the primary analysis, secondary analysis, or both, as defined in Behboodi et al.

Details

The standardized mean difference (SMD) is a measure of the difference between two group means, expressed in units of their pooled standard deviation. In this dataset it quantifies the difference in mean cognitive test scores between children born preterm (EPT/VPT) and those born full term (FT), with negative values indicating lower scores in the preterm group.

The SMD is calculated as Hedges’ g. First, Cohen’s d is computed as

$$d = \frac{\bar{X}_{EPT.VPT} - \bar{X}_{FT}}{s_p},$$

where \(\bar{X}_{EPT.VPT}\) and \(\bar{X}_{FT}\) are the mean cognitive scores in the EPT/VPT and FT groups, respectively, and \(s_p\) is the pooled standard deviation.

The pooled standard deviation is given by

$$s_p = \sqrt{\frac{(n_{EPT.VPT}-1)\, s_{EPT.VPT}^2 + (n_{FT}-1)\, s_{FT}^2} {n_{EPT.VPT} + n_{FT} - 2}}.$$

Here, \(n_{EPT.VPT}\) and \(n_{FT}\) are the group sample sizes, and \(s_{EPT.VPT}\) and \(s_{FT}\) are the corresponding standard deviations.

Cohen’s d is multiplied by the small-sample correction factor

$$J = 1 - \frac{3}{4\,(n_{EPT.VPT} + n_{FT} - 2) - 1}$$

to obtain Hedges’ g:

$$SMD = g = J \times d.$$

The standard error of the standardized mean difference (seSMD) measures the uncertainty of the estimated SMD. Smaller values indicate more precise estimates of the difference between the preterm (EPT/VPT) and full-term (FT) groups.

The seSMD is calculated as

$$seSMD = \sqrt{ J^2 \left( \frac{n_{EPT.VPT} + n_{FT}}{n_{EPT.VPT}\, n_{FT}} + \frac{d^2}{2\,(n_{EPT.VPT} + n_{FT} - 2)} \right) }.$$