DISCUSSION
The use of LUS in studying respiratory function of preterm infants has been well established. It has been used to predict the need for surfactant administration and/or respiratory support.17-20 Using LUS to diagnose neonates with pulmonary edema has been shown to be more sensitive and with greater negative predictive value than X-rays.21 Similarly, LUS showed good positive predictive value in the assessment of the early risk for BPD in ELGANs.22 Surprisingly, despite the controversy, LUS has not been systematically used in the evaluation of diuretic effect in premature infants diagnosed with BPD. Recently, a very small study used LUS to evaluate a heterogeneous group (4/18 neonates had PDA) of preterm neonates with pulmonary edema treated with diuretics. The study showed lower LUS scores in the group of diuretic responders compared to non-responders. However, this pilot study enrolled neonates quite early in life, with diuretics being administered at a mean of 31 days of life. The study concluded that responders were treated with diuretics for much shorter periods and ultimately developed less severe BPD.23
We found that diuretic therapy was associated with significant decrease in pulmonary edema, FiO2 requirement, and respiratory flow support compared to the control group by day 3. FiO2 requirements were not different between the diuretic and control groups on day 6, either because they were both low and approaching the room air target on day 6, or because clinicians preferentially weaned respiratory flow support before FiO2 in the diuretic group. It is also possible that infants developed tolerance to diuretics after prolonged use.24,25 In the diuretic group, 46% of infants were treated with diuretics for a duration of 7-10 days, and 3.6% of them had a rebound increase in need for FiO2 or respiratory flow support upon cessation of diuretic therapy. Our findings are consistent with previous studies showing that diuretics can improve O2 requirements and respiratory mechanics in infants.26 Although, there is only limited previous data on the use of pulmonary edema scores in neonates, the correlation of decreasing scores with decreasing respiratory support parameters in our subjects supports their validity.
A recent retrospective longitudinal cohort study confirmed the long-established association between chronic diuretic use and electrolyte disturbances requiring electrolyte supplementation, especially in infants treated with thiazides.27 In our study group, 25% of infants had to be supplemented with either NaCl and/or KCl. This common side effect together with other concerns related to diuretic use such as ototoxicity, metabolic bone disease, and nephrocalcinosis underline the importance of careful evaluation of the need for diuretics in infants with CLD.
Our study has several limitations. The specific indications for diuretics in BPD are unclear, and thresholds for treatment were heterogeneous in our cohort, based on clinical experience and preference. In addition, infants in the diuretic group received several distinct regimens based on clinician preference, and the study was not adequately powered to compare them. It is also important to note that, while diuretics have short-term beneficial effects on respiratory function, they have not been associated with a decrease in long-term outcomes or length of hospital stay.28 We could not evaluate the length of the hospital stay in our study because several infants from the control group eventually were later treated with diuretics.