TABLE 3.
Analysis of Effect Modification by Study Characteristics: Perinatal and Neonatal Risk Factors With Heterogeneity, P < .10
| Perinatal and Neonatal Risk Factors with Heterogeneity, P < .10 | Significant Sources of Between-Study Heterogeneity: Study Characteristics, P < .10a | Summary Effect Estimate (95% CI) |
|---|---|---|
| Prolonged labor | None | 1.77 (0.76–4.14) |
| Induced/augmented labor | 1.21 (0.90–1.62) | |
| Population based (P = .03) | ||
| 4 Studies: population based | 0.99 (0.72–1.35) | |
| 3 Studies: clinic based | 1.72 (1.17–2.54) | |
| Precipitous labor | 1.52 (0.59–3.90) | |
| Population based (P < .001) | ||
| 2 Studies: population based | 0.71 (0.45–1.11) | |
| 3 Studies: clinic based | 4.10 (1.95–8.62) | |
| Exposure data collection (P < .001) | ||
| 4 Studies: prospective | 0.78 (0.52–1.16) | |
| 1 Study: mix of retrospective and prospective | 5.33 (2.12–13.40) | |
| Cesarean delivery | None | 1.26 (0.99–1.60) |
| Abnormal presentation | 1.44 (1.07–1.94) | |
| Exposure data collection (P = .03) | ||
| 9 Studies: prospective | 1.56 (1.17–2.08) | |
| 3 Studies: retrospective and mixed | 0.75 (0.32–1.76) | |
| Amniotic fluid not clear or meconium staining | 0.82 (0.25–2.69) | |
| Abnormal and multivariate versus univariate analysis (P = .03) | ||
| 3 Studies: normal control subjects and multivariate analysis or sibling control subjects | 1.10 (0.50–2.39) | |
| 1 Study: abnormal comparison group and univariate analysis | 0.08 (0.01–0.72) | |
| Twins or multiple birth | None | 1.77 (1.23–2.55) |
| April–June birth | 1.14 (0.80–1.63) | |
| Population based (P < .001) | ||
| 2 Studies: population based | 0.86 (0.77–0.97) | |
| 2 Studies: clinic based | 1.69 (1.20–2.38) | |
| October–December birth | None | 0.80 (0.58–1.10) |
| Fall birth | 0.84 (0.62–1.14) | |
| Population based or multivariate versus univariate analysis (P = .03) | ||
| 2 Studies: population based and univariate analysis | 0.99 (0.88–1.11) | |
| 1 Study: clinic based and sibling control subjects | 0.52 (0.30–0.91) | |
| Winter birth | None | 0.92 (0.69–1.24) |
| Spring birth | None | 1.14 (0.71–1.83) |
| Postterm birth | 1.14 (0.58–2.24) | |
| Abnormal (P = .02) | ||
| 9 Studies: normal control subjects | 1.40 (0.77–2.56) | |
| 1 Study: abnormal control subjects | 0.25 (0.14–0.44) | |
| Preterm birth | 1.16 (0.83–1.62) | |
| Abnormal (P = .003) | ||
| 14 Studies: normal control subjects | 1.30 (0.99–1.71) | |
| 1 Study: abnormal control subjects | 0.31 (0.17–0.57) | |
| Exposure data collection (P = .02) | ||
| 10 Studies: prospective | 1.54 (1.14–2.06) | |
| 5 Studies: retrospective and mixed | 0.74 (0.41–1.35) | |
| Diagnostic criteria (P < .001) | ||
| 10 Studies: narrow diagnostic criteria | 0.81 (0.55–1.20) | |
| 5 Studies: broad diagnostic criteria | 1.76 (1.44–2.16) | |
| Low birth weight (<2500 g) | 1.61 (1.18–2.19) | |
| Exposure data collection (P = .01) | ||
| 10 Studies: prospective | 2.03 (1.39–2.96) | |
| 4 Studies: retrospective and mixed | 0.86 (0.58–1.29) | |
| Birth weight < 2000 g | None | 2.20 (0.53–9.12) |
| Low 1–minute Apgar score | 1.08 (0.71–1.64) | |
| Abnormal (P = .003) | ||
| 5 Studies: normal control subjects | 1.25 (0.93–1.67) | |
| 1 Study: abnormal control subjects | 0.21 (0.07–0.62) | |
| Weak or no crying after birth or slow to cry | 1.84 (0.19–17.58) | |
| Abnormal (P < .001) | ||
| 2 Studies: normal control subjects | 5.93 (2.75–12.80) | |
| 1 Study: abnormal control subjects | 0.34 (0.17–0.68) | |
| Respiratory distress or no breathing after birth | 1.84 (0.84–4.21) | |
| Abnormal (P = .04) | ||
| 7 Studies: normal control subjects | 2.95 (1.14–7.61) | |
| 2 Studies: abnormal control subjects | 0.48 (0.24–0.94) | |
| Hyperbilirubinemia or jaundice | 1.39 (0.97–1.97) | |
| Diagnostic criteria (P = .05) | ||
| 8 Studies: narrow diagnostic criteria | 1.63 (1.12–2.37) | |
| 2 Studies: broad diagnostic criteria | 0.86 (0.53–1.40) | |
| Medical intervention in the first month | 1.34 (0.81–2.21) | |
| Abnormal (P = .01) | ||
| 5 Studies: normal control subjects | 1.58 (1.09–2.30) | |
| 1 Study: abnormal control subjects | 0.18 (0.05–0.67) | |
| Oxygen treatment or resuscitation | 1.38 (0.65–2.94) | |
| Abnormal (P = .001) | ||
| 6 Studies: normal control subjects | 1.89 (1.06–3.37) | |
| 1 Study: abnormal control subjects | 0.22 (0.11–0.46) |
a
Exposure data collection refers to the effect modification by exposure measurement (prospective versus retrospective). Diagnostic criteria refers to the effect modification by diagnostic criteria (narrow versus broad). Multivariate versus univariate analysis refers to effect modification by the degree of control for covariates. Population based refers to the effect modification by population-based versus clinic-based samples. Abnormal refers to the effect modification by use of a normal comparison group versus an abnormal comparison group. None refers to no effect modification (P < .10) by any of the above study characteristics.