Conventionally, a chest tube drainage is placed following PDA ligation to monitor possible bleeding and drain air or effusion postoperatively. The chest tube drain is typically inserted even in the absence of lung injury or bleeding. In most cases, after PDA ligation via thoracotomy, a chest tube drain is inserted for the evacuation of possible trapped air inside the thoracic cavity after skin closure and not for the drain of air leakage from potential lung injury as a surgical complication. Therefore, most patients were not found to have tangible pleural effusion or air leakage after PDA ligation. Instead, chest tube drainage is associated with significant thoracotomy pain, risk of infection, excessive pleural reactions, or impaired pulmonary function [9, 14]. Particularly, owing to their prematurity, preterm neonates are known to have many factors that predispose them to an increased risk of infection or sepsis [15]. Moreover, preterm neonates’ lung can be easily wounded by a relatively rigid chest tube because the lung parenchyma is very fragile.

In our experience, we evacuate the air by water-sealing with a 5-Fr silicon feeding tube during chest wall closure and fully expand the left lung. Neonates with chest tube drainage placed after PDA ligation had no air leak or effusion drained during the 24 h after surgery. One of the major concerns for doctors was the management of chest bottles. As there were no customized closed water-sealed chest bottles suitable for 5-Fr silicon feeding tubes, we assembled an open water-sealed chest bottle, which was unsanitary and easily spillable, causing the possibility of infection in the neonates. We always considered the necessity of a chest tube drainage, and a few studies indicated no risk of complications without routine chest tube drainage after PDA ligation [12, 16,17,18]. Moreover, previous reports have addressed the feasibility and safety of omitting chest tube drainage after thoracoscopic lung surgeries [4,5,6,7,8,9]. Therefore, we began omitting chest tube drainage in selected cases from 2013; since 2014, our institution has not routinely employed chest tube drainage in the pleural cavity after PDA surgical ligation in preterm neonates. In our study, no patients had significant postoperative complications, such as pneumothorax or hemothorax. However, 19 patients had minor complications of subcutaneous emphysema around the thoracotomy site. No patients required additional chest tube drainage for postoperative bleeding, pleural effusion, or progressive subcutaneous emphysema.

We also reviewed the mortality cases. As mentioned in the Results, all seven patients were extremely low birth weight and extremely preterm neonates. Four of the seven patients who underwent PDA ligation within two weeks preoperatively had severe generalized edema due to congestive heart failure, multi-organ dysfunction, and septic shock, barely maintained vital signs, and required careful consideration for PDA ligation. Neonatologists and pediatric cardiologists felt that PDA ligation might relieve congestive heart failure and eventually help overcome the sepsis state; however, unfortunately, the patients died within 30 days postoperatively.

Currently, we do not employ a chest tube drain in the pleural cavity after PDA ligation even in full-term born neonates. As previously mentioned, uncomplicated PDA ligation is a relatively short-time procedure with minimal or no damage to intrathoracic structures, causes no blood loss, does not involve lung resection, and the pleura absorbs a small amount of inflammatory pleural effusion or blood [12, 16]. Therefore, we also found no fluid and air with or without a chest tube drainage in the pleural cavity, corroborating the unnecessity of routine chest tube drainage after PDA ligation in preterm neonates.

Nevertheless, there are also several limitations to our study. Firstly, it is a single-center retrospective study with a small sample size, and the experimental results need to be further confirmed by a larger sample size study. Secondly, we had to use a non-commercialized product (5-Fr feeding tube with assembled open water-sealed chest bottle) since the patients’ pleural cavity was relatively too small for a 10-Fr trocar catheter or 16-Fr chest tube. Finally, we did not use the lung ultrasound to check for air leak, pneumothorax, or effusion. However, lung ultrasound would be a good option instead of chest radiographs as it reduces patients’ exposure to X-rays.