Association of Chorioamnionitis With Bronchopulmonary Dysplasia Among Preterm Infants: A Systematic Review, Meta-analysis, and Metaregression

IMPORTANCE: Bronchopulmonary dysplasia (BPD), a chronic lung disease of prematurity, remains one of the major and most common complications of very preterm birth. Insight into factors associated with the pathogenesis of BPD is key to improving its prevention and treatment. OBJECTIVE: To perform a systematic review, meta-analysis, and metaregression of clinical studies exploring the association between chorioamnionitis (CA) and BPD in preterm infants. DATA SOURCES: PubMed and Embase were searched without language restriction (last search, October 1, 2018). Key search terms included bronchopulmonary dysplasia, chorioamnionitis, and risk factors. STUDY SELECTION: Included studies were peer-reviewed studies examining preterm (<37 weeks’ gestation) or very low-birth-weight (<1500 g) infants and reporting primary data that could be used to measure the association between exposure to CA and the development of BPD. DATA EXTRACTION AND SYNTHESIS: The Meta-analysis of Observational Studies in Epidemiology (MOOSE) guideline was followed. Data were independently extracted by 2 researchers. A random-effects model was used to calculate odds ratios (ORs) and 95% CIs. Heterogeneity in effect size across studies was studied using multivariate, random-effects metaregression analysis. MAIN OUTCOMES AND MEASURES: The primary outcome was BPD, defined as supplemental oxygen requirement on postnatal day 28 (BPD28) or at the postmenstrual age of 36 weeks (BPD36). Covariates considered as potential confounders included differences between CA-exposed and CA-unexposed infants in gestational age, rates of respiratory distress syndrome (RDS), exposure to antenatal corticosteroids, and rates of early- and late-onset sepsis. RESULTS: A total of 3170 potentially relevant studies were found, of which 158 met the inclusion criteria (244 096 preterm infants, 20 971 CA cases, and 24 335 BPD cases). Meta-analysis showed that CA exposure was significantly associated with BPD28 (65 studies; OR, 2.32; 95% CI, 1.88-2.86; P < .001; heterogeneity: I(2) = 84%; P < .001) and BPD36 (108 studies; OR, 1.29; 95% CI, 1.17-1.42; P < .001; heterogeneity: I(2) = 63%; P < .001). The association between CA and BPD remained significant for both clinical and histologic CA. In addition, significant differences were found between CA-exposed and CA-unexposed infants in gestational age, birth weight, odds of being small for gestational age, exposure to antenatal corticosteroids, and early- and late-onset sepsis. Chorioamnionitis was not significantly associated with RDS (48 studies; OR, 1.10; 95% CI, 0.92-1.34; P = .24; heterogeneity: I(2) = 90%; P < .001), but multivariate metaregression analysis with backward elimination revealed that a model combining the difference in gestational age and the odds of RDS was associated with 64% of the variance in the association between CA and BPD36 across studies. CONCLUSIONS AND RELEVANCE: The results of this study confirm that among preterm infants, exposure to CA is associated with a higher risk of developing BPD, but this association may be modulated by gestational age and risk of RDS.