Pulmonary surfactant has been used in the PARDS treatment since many years ago[10], it is not recommended as a routine drug in treating PARDS partly because its association with mortality, mechanical ventilation duration, and other direct prognostic indicators for PARDS is uncertain. Factors such as the inconclusive timing and dosage, the different components among different products, and the high cost also limit the routine use of PS[11]. The dosage of PS is related to the area of the alveolar-capillary membrane; however, an accurate evaluation is difficult to achieve. Patients aged ≤ 3 years were chosen as the sample population to reduce the bias and cost in this study.
The early administration of a certain dose of the PS was found to accelerate the improvement of the OI among patients within 72 h, particularly within 24 h of the moderate or severe PARDS onset. A theoretical basis for exogenous PS in improving OI in ARDS is sufficient[2]. Although there are numerous causes of PARDS, its common pathophysiological change is the decrease in the amount of endogenous PS and the inhibition of its activity caused by the damage of type II alveolar epithelial cells and capillary endothelial cells[12]. Endogenous PS deficiency leads to alveolar collapse and intrapulmonary shunt, thus reducing the lung volume and lung compliance, thereby causing unimprovable hypoxemia, and pulmonary hypertension, and may be further followed by right heart failure, shock, multiple organ failure, and even death. Exogenous PS can play an alternative role in this stage. Most studies on the effectiveness of the PS on PARDS describe an improvement in pulmonary oxygenation[3, 4, 6, 12].
In our study, the mortality in the PS group is significantly lower than that of the control group. Based on the results of the subgroup analysis, we consider that the PS may reduce the mortality by accelerating OI improvement in acute phase of PARDS especially in children with "the most severe oxygenation deterioration"[13]. Patients in the PS group were given PS once to thrice under their doctors’ recommendation according to the OI. Thus, patients in this group who used PS three times had the worst oxygenation level (as shown in Fig. 2). Some patients in the control group were not given PS because it was not recommended by their doctors as the oxygenation deterioration was relatively mild or could be improved rapidly after receiving common treatment with mechanical ventilation. Other patients in the control group had severe oxygenation deterioration and thus the use of PS was recommended, however, their guardians’ refusal of PS treatment causes this group of patients to have the highest mortality. In our study, children in the end-stage of diseases or with chronic cardiac/pulmonary diseases were excluded. Therefore, the dead patients in the control group were also considered to also have the worst oxygenation [baseline OI of the dead and surviving patients in the control group, 18.6 (11.6, 26.4) vs. 13.4 (9.95, 18.2), Z = 1.899, P = 0.058].
Although, most of the current studies confirm the effectiveness of the PS on improving oxygenation or decreasing the mechanical ventilation time, it is controversial whether PS can improve the mortality of PARDS[3, 5, 14–16]. However, few studies suggest that PS can improve the mortality of PARDS[15, 16]. We considered that is partly related to the different groups of patients included in different studies, and also the different timing and dosage of the PS used. The etiology of PARDS is very complicated, some patients with combined neuromuscular system diseases or have been in the end-stage of underlying diseases before the onset of PARDS. After the 2015 PALICC guidelines were published, some researchers also included children with chronic cardiac/pulmonary diseases. All the above-mentioned can contribute to the mortality and other primary outcomes of PARDS[7]. Such kinds of patients were excluded from our study. It has been reported that the OI in the early stage of PARDS can predict its outcomes, which partly supports the results of this study[17].
Oxygenation deterioration in some patients who have moderate/severe PARDS is less serious or can be well managed with common treatment such as mechanical ventilation. No evidence supports PS will be beneficial to these patients but rather increase medical expenses. Referring to the 25th percentile of baseline OI in subgroup analysis, the OI threshold of the PS treatment may be set to higher than 8 to avoid unnecessary waste of medical resources, except for some patients with rapid OI deterioration. Further studies are required to identify a new threshold. The reported duration of the endogenous PS damage is 4–8 days or longer, however, it is not supported with strong evidence therefore it is unclear[18]. In our study, a dosage of 140 mg of PS, given to the patients up to thrice with an interval of 8 h during the acute phase of PARDS can significantly decrease the mortality, therefore, prolonged treatment till the endogenous PS recovery is not recommended. Pulmonary surfactant is delivered via a thin catheter combined with postural changes, this process may not ensure the equal distribution of PS. It is a sample and available method for drug administration and has been used in some research[3, 19].The effectiveness of this method is supported in our study. We did not find any pneumothorax, pulmonary hemorrhage, mediastinal emphysema, or other complications that are directly caused by exogenous PS delivery, the safety usage of the PS is similar to other studies[5, 15, 16].
Standard PS dose has not been established and remains exploration. In several published studies, the dose range is 20–175 mg/kg [3, 15, 19]. The subjects in our research are between 29d to 3ys old, and majority of them weigh between 4–13 kilograms. The PS product we use is 70 mg per servings and is relatively expensive. So, we used a fix dose of 140 mg per time, and prepared to repeat 1–3 times according to pulmonary oxygenation, which can ensure the total dose within the range of published studies and control the cost. Several studies evaluated the effectiveness of PS within 12–24 h after the initial dose, the extent of pulmonary oxygenation function improvement was individualized, while some patients suffered oxygenation deterioration after initial improvement [3, 4, 15]. We repeated blood gas analysis, calculated OI and reviewed the clinical status of patients at an interval of 8 h to assess the patients more carefully due to dose bias existed. We repeated doses of PS when pulmonary oxygenation or relative clinical status deteriorated or not improved. We described an association between PS administration and improvement of OI and mortality under our dosage regimen, although we must admit the dose bias does exist. Further studies should increase the sample size to explore the exact dose of PS through stratified analysis by weight or body length. we will also consider further study on elder children when the effectiveness and applicable dosage of PS is clearer.