Study design and participants
Our center is a referral university-based hospital in southern Thailand for all newborns presenting with cyanosis and diagnosed with congenital heart disease. Neonatologists support our team in treating them. This retrospective study was conducted with consecutive TGA newborn patients between January 2005 and December 2020 in Songklanagarind Hospital. TGA cases are diagnosed by a pediatric cardiologist on echocardiogram. If the patient is still hypoxemic after optimal treatment, we consider complications such as PPHN. The diagnosis of PPHN was made on physical examination, showing reverse differential cyanosis, and confirmed by echocardiogram. The exclusion criteria in this study were TGA with right or left ventricular obstruction.
We usually prescribe prostaglandin E1 (PGE1) infusion at a dose of 0.01–0.2 mcg/kg/min to maintain oxygen saturation of more than 70% or PaO2 more than 30 mmHg in suspected TGA cases before transport to our center. Mechanical ventilation is started if oxygen saturation is less than 70%. Moreover, we recorded oxygen saturation in the upper and lower extremities first, for early detection of PPHN. If the O2 saturation indicates reverse differential cyanosis between lower and upper extremities of more than 5–10%, PPHN may develop later. The diagnosis was confirmed by echocardiogram, which also illustrated possible risk factors for PPHN. For instance, restrictive atrial septal defect (ASD), D shape left ventricle (LV), and PDA flow direction may indicate a risk of PPHN in TGA cases. All TGA cases diagnosed with PPHN were given conventional treatment that included optimal ventilation, such as High-frequency oscillatory ventilation (HFOV), sedative medication (morphine or midazolam), volume support with normal saline solution(NSS), or blood transfusion if the hematocrit was less than 45%. Inotropic drugs, such as milrinone, were administered if blood pressure was normal but there was ventricular dysfunction. However, hypotension in critically ill infants with TGA and PPHN requires immediate treatment. Systemic blood pressure was supported with vasoactive drugs such as epinephrine, norepinephrine, dobutamine, and dopamine. This helped increase pulmonary blood flow due to increased systemic vascular resistance and raised the left to right shunt bypassing the PDA (systemic to pulmonary flow), resulting in improved cardiac output as well. The vasoactive-inotropic score (VIS) was calculated for some inotropic drugs. Emergency BAS was performed in cases with a restrictive atrial septum and clinical signs of hypoxia, in the cardiac catheterization laboratory. After BAS, the patient is re-evaluated for increased oxygen saturation, no reverse differential in oxygenation, and arterial blood gas (ABG) and undergoes an echocardiogram demonstrating PDA flow and cardiac function. In addition, iNO was started, at 20 parts per million (ppm), if patients with TGA and PPHN developed severe respiratory failure (OI > 25). Therefore, iNO doses could be decreased if there is no reverse differential and the OI level is less than 15, before arterial switch operation (ASO). Other pulmonary vasodilator drugs, such as bosentan, sildenafil, and iloprost, have also been used [8, 9]. In cases of TGA with PPHN diagnosis was confirmed by echocardiogram to identify the PDA flow direction. We suggest that monitoring oxygen saturation in the upper and lower extremities are a crucial examination.
The clinical data from the electronic medical records were reviewed. We collected the demographic and clinical information of newborn patients, including gender, gestational age, birth weight, type of TGA, age at presentation of cyanosis, initial oxygen saturation, oxygen index (OI), age, PEG1 drugs doses, arterial blood gas, echocardiogram result, severity of PPHN, inotropic drugs, VIS score, shock, medication for the treatment of PPHN, age at surgery, postoperative complications, and mortality. PPHN mortality was defined as patients dying before and after surgery.
Defining PPHN by severity 
TGA patients presenting with mild to moderate PPHN can be diagnosed by echocardiography showing bidirectional flow through the PDA that reverses differential saturation from pre- to post-ductal between 5% and 15%. Patients with severe PPHN presented with profound reverse differential cyanosis, with the echocardiogram revealing continuous right to left flow through a PDA and reverse differential saturation of 15% or higher.
Definition of the value of VIS [10, 11]
The value of VIS was calculated by a standard formula where the amount of inotrope used was measured by VIS, which was defined as dopamine dose (ug. kg_1. min_1) + dobutamine dose (ug. kg_1 min_1) + 100 x epinephrine dose (ug. kg_1 min_1) + 10 x milrinone dose (ug. kg_1 min_1) + 10,000 x vasopressin dose (U. kg_1. min_1) + 100 x norepinephrine dose (ug .kg_1. min_1).
Definition of the OI formula [12, 13]
OI was calculated using the standard formula (FiO2 x mean airway pressure x 100) / PaO2. Previously, the OI was widely used to initiate management in neonatal patients with PPHN, with OI > 25 being treated with iNO. However, recently, OI > 20–25 is the criterion for the use of iNO, and OI > 40 identifies a candidate for extracorporeal membrane oxygenation (ECMO).
The R program version R4.2.0 (R Core Team (2022). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria) was used to analyze the data. Descriptive data are presented as numbers and percentages. Continuous data are described as means ± standard deviations (SD) and medians (interquartile range [IQR]) depending on the distribution of data. The data were initially analyzed by univariate analysis. Factors with a P-value < 0.05 in univariate analysis were further analyzed by multivariate logistic regression analysis. The Mann–Whitney U or Fisher’s exact tests compared data between the two groups. Accuracy measures, including sensitivity, specificity, positive predictive value, negative predictive value, and area under the curve (AUC) were analyzed and used to determine the optimal cutoff values of peak VIS and OI for PPHN in TGA patients. Differences were considered statistically significant at a P-value < 0.05.