Abstract
OBJECTIVE: To test the hypotheses that (1) infection increases ductal dilatory prostaglandins and inflammatory mediators that may influence the closure of a patent ductus arteriosus (PDA), increasing the incidence of late episodes of PDA (after 7 days) and the rate of closure failures, and (2) the concurrence of PDA and infection increases the risk of chronic lung disease (CLD).
METHODS: One hundred fourteen premature infants (birth weight, 500 to 1000 gm) were prospectively assessed for PDA and infection. Serum levels of 6-ketoprostaglandin F
1α and tumor necrosis factor alpha were measured routinely in all infants and when PDA or infection was present. Multivariate assessment of risk factors for PDA closure failure and for CLD was done by logistic regression, and expressed as an odds ratio and as 95% confidence intervals.
RESULTS: Late PDA episodes were more frequent in infants with infection than in those without infection. A temporally related infection (<5 days between both diagnoses) was associated with an increased risk of PDA closure failure (odds ratio, 19.1 [confidence interval, 4 to 90]). In addition to birth weight and the severity of initial respiratory failure, PDA and infection increased the risk of CLD (odds ratio, 11.7 [confidence interval, 1.7 to 81] for PDA; odds ratio, 3.1 [confidence interval, 1 to 11] for infection). Furthermore, when both factors were temporally related, they further increased the risk of CLD (odds ratio, 29.6 [confidence interval, 4.5 to >100]). Infants with infection and those with PDA had higher levels of 6-ketoprostaglandin F
1α than did control subjects. Levels of tumor necrosis factor alpha were also elevated in infants with infection and in those with late PDA.
CONCLUSIONS: Infection adversely influences PDA outcome by increasing the risk of late ductal reopening and PDA closure failures. Increased levels of prostaglandins and tumor necrosis factor alpha in infants with infection may explain the poor PDA outcome. The concurrence of PDA and infection potentiates their negative effects on the risk of CLD. (J P
EDIATR 1996;128:470-8)
