Kavita Bhola, Jann P Foster, David A Osborn

Patent ductus arteriosus (PDA) is associated with mortality and morbidity in preterm infants. Phototherapy is a common treatment for jaundice in preterm infants. However, phototherapy has been associated with failure of closure of the ductus arteriosus in preterm infants. To determine if chest shielding of preterm infants receiving phototherapy reduces the incidence of clinically and/or haemodynamically significant PDA and reduces morbidity secondary to PDA. We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (the Cochrane Library; 2015, Issue 3), MEDLINE, EMBASE, CINAHL, previous reviews, cross-references, abstracts, proceedings of scientific meetings, and trial registries through March 2015. Randomised controlled trials (RCTs), cluster-RCTs, or quasi-RCTs of chest shielding during phototherapy compared to sham shielding or no shielding for the prevention of a haemodynamically or clinically significant PDA in preterm infants. Three review authors independently assessed studies for eligibility and quality and extracted data. We defined a clinically significant PDA as the presence of a PDA with clinical signs of an effect on organ function attributable to the ductus arteriosus. We defined a haemodynamically significant PDA as clinical and/or echocardiographic signs of a significant ductus arteriosus effect on blood flow. We included two small trials enrolling very preterm infants (Rosenfeld 1986; Travadi 2006). We assessed both as at high risk of bias. No study reported clinically significant PDA, defined as the presence of a PDA with clinical symptoms or signs attributable to the effect of a ductus arteriosus on organ function. Rosenfeld 1986 reported a non-significant reduction in haemodynamically significant PDA with left atrial to aortic root ratio greater than 1.2 (risk ratio (RR) 0.23, 95% confidence interval (CI) 0.05 to 1.01; 74 infants) but a statistically significant risk difference (RD -0.18, 95% CI -0.34 to -0.03; number needed to treat for an additional beneficial outcome (NNTB) 5, 95% CI 3 to 33). Rosenfeld 1986 reported a significant reduction in PDA detected by murmur (RR 0.50, 95% CI 0.29 to 0.88; RD -0.30, 95% CI -0.52 to -0.08; NNTB 3, 95% CI 2 to 12; 74 infants). Rosenfeld 1986 reported a significant reduction in treatment with indomethacin (RR 0.12, 95% CI 0.02 to 0.88; RD -0.21, 95% CI -0.35 to -0.06; NNTB 5, 95% CI 3 to 17; 74 infants), and only one infant had a ductal ligation in the no-shield group. There were no other significant outcomes, including mortality to discharge or 28 days, days in oxygen, days on mechanical ventilation, days in hospital, intraventricular haemorrhage, retinopathy of prematurity, or exchange transfusion. The available evidence is very low quality and insufficient to assess the safety or efficacy of chest shield during phototherapy for prevention of PDA in preterm infants. Further trials of chest shielding are warranted, particularly in settings where infants are not receiving prophylactic or early echocardiographic targeted cyclo-oxygenase inhibitors for PDA.