Critical Appraisal of the Role ofUreaplasma in the Development of Bronchopulmonary Dysplasia With Metaanalytic Techniques

Schelonka, Robert L. MD^*; Katz, Brenda MD^†; Waites, Ken B. MD^†; Benjamin, Daniel K. Jr MD, MPH, PhD^‡

From the Departments of *Pediatrics and †Pathology, University of Alabama at Birmingham, Birmingham, AL; and the ‡Department of Pediatrics and Duke Clinical Research Institute, Duke University, Durham, NC

Accepted for publication June 9, 2005.

Supported by National Institute of Child Health and Human Development, grant HD043327 (to Dr Schelonka) and grant HD044799 (to Dr Benjamin).

E-mail[email protected]. Reprints not available.

The Pediatric Infectious Disease Journal24(12):p 1033-1039, December 2005. |DOI:10.1097/01.inf.0000190632.31565.83

Abstract

Background:

Controversy exists over whether or notUreaplasma colonization or infection of the respiratory tract contributes to the development of bronchopulmonary dysplasia (BPD). Because BPD is a major cause of morbidity and mortality in preterm infants and a potential therapeutic intervention with antimicrobials is possible, we sought to evaluate and critique the current medical literature and to document the reported association betweenUreaplasma and BPD.

Methods:

We analyzed all peer-reviewed articles and previous reviews including cross-references that reportedUreaplasma respiratory tract colonization or infection and development of BPD in neonates published from January 1966 to December 2004. Inclusion criteria included a cohort limited to all neonatal intensive care unit admissions or all colonized infants, articles that did not define a numerator and a denominator for BPD andUreaplasma or that included patients from other reports were excluded from the analysis. We evaluated BPD at 28 postnatal days (BPD28) or 36 weeks post-menstrual age (BPD36).

Results:

Twenty-three studies with an aggregate of 2216 infants reported BPD28, and 8 studies with 751 infants reported BPD36. Although there was significant association betweenUreaplasma colonization and both BPD28 and BPD36, there was substantial heterogeneity (Q test statistic,P < 0.01). We therefore focused on describing the study characteristics associated with an increased relative proportion of BPD. The greatest contribution to effect was from the studies enrolling fewer than 100 infants.

Conclusion:

Ureaplasma colonization is associated with higher reported rates of BPD, but the greatest reported effect is seen in small studies; reporting bias may be partially responsible for this effect.