Timing of bronchoscopy and application of scoring tools in children with severe pneumonia

Fibreoptic bronchoscopy (FOB) and bronchoalveolar lavage (BAL) help diagnose many pulmonary diseases in children [1, 2] and are often used in paediatric patients with lower respiratory tract lesions and suspected pulmonary infections [3], and even in the assistance of cardiopulmonary surgery [4]. Overall, BAL is considered to provide valuable diagnostic information, and the safety of FOB in paediatric populations and even neonates has been demonstrated [5]. In the paediatric intensive care unit (PICU), FOB is generally used to identify the aetiology of acute respiratory failure in children, such as laryngeal obstruction and foreign bodies [1], and the study of bronchoalveolar lavage fluid (BALF) can assist in early identification of lung disease, pathogens and causes of partial infiltration [5]. Rejection in children who have undergone cardiopulmonary transplantation can be assessed by cellular changes in BALF [6]. In recent years, the use of FOB and BAL in paediatric patients has increased significantly, and the use of FOB and BAL in the PICU has received increasing attention; however, the impact of decisions about their management has not been fully evaluated in previous studies [7] on the diagnosis and treatment of pulmonary diseases. The timing of FOB and BAL use in the treatment of mechanically ventilated patients remains controversial, and the rate of diagnosis in critically ill patients has varied across studies [4, 8]. There is still a lack of effective scoring criteria for assessing the severity of pulmonary infection associated with changes in the endobronchial lining of the bronchus in children.

This study retrospectively analysed the clinical data of 229 paediatric patients with severe pneumonia treated with FOB combined with BAL to investigate the value of endoscopic bronchitis scores and assess the timing of BAL.

The overall beneficial effect of microbiological diagnostic bronchoscopy on clinical outcomes in pneumonia is well known. In previous studies of patients without human immunodeficiency virus (the same as the current study’s population), the aetiological diagnosis of FOB and BAL was 52–84% [12]. The present study found that the pathogen detection rate reached 78.6%, which is higher than that found by Panse et al. (67.2%) [13] and Ma et al. (75%) [14]. This was probably because of the development of diagnostic techniques such as pathogenic metagenomic technology in recent years [15]. However, these techniques have limitations in clinical application due to various factors, such as cost [16].

In this study, the pathogen with the highest detection rate was mycoplasma, and it was the highest in each age group. The 13–36 month group had the highest incidence, and the difference was statistically significant. The main pathogens detected were Mp, RSV, Staphylococcus aureus, Haemophilus influenza (Hi) B virus, and Streptococcus pneumoniae (Sp). Ma et al. [14] found that 53.7% of the respiratory pathogens detected in patients were viruses, 32.9% were bacteria, and 13.4% were Mp; the reason for this difference in the results of this study and that of Ma et al. may be related to the different sample sizes. The sample size of this study is relatively larger than that of Ma et al.’s study. Slightly different from the findings of the study conducted by Ma et al., the top five pathogens in this study were RSV, Sp, Hi, Mp, and adenovirus (AdV). It is worth noting that in this study, RSV was a common virus in critically ill patients under 1 year old, followed by the influenza virus. No AdV was detected, however, and the fungal population was surprisingly low. Conversely, this virus was found to be more common in the research conducted by Ma et al. and Huang et al. [14, 17, 18], especially with the development of pathogenic metagenomic technology. The reason for this difference may be related to disease severity and regional factors in the included patients.

There is still no well-recognised endoscopic scoring tool for assessing the severity of pneumonia in children. The most commonly used method is to predict the severity of pulmonary infection by clinical indicators rather than intrapulmonary manifestations. Zhao et al. [19] constructed a nomogram to predict the development of plastic bronchitis in children with refractory mycoplasma pneumoniae pneumonia (RMPP) by studying six variables: peak body temperature, neutrophil ratio, PLT, IL-6, LDH and atelectasis. In Cheng’s study [20], indicators such as lactate dehydrogenase, albumin, neutrophil ratio and high fever were used. None of these indicators can objectively reflect the degree of inflammatory changes in the lung, and most of the previous studies were single-centre studies [21].

As early as 1989, Thompson developed a semi-quantitative bronchoscopy scoring system based on mucosal fragility, oedema and erythema to describe secretions of intraluminal airway inflammation [22]; however, it is not widely used. It was not until 2018 that researchers, such as Thomas, developed a bronchial inflammation scoring tool [11]. In 2020, a team led by Thompson validated six visual characteristics – secretions (number and colour) and mucosal appearance (erythema, pallor, elevation and oedema) – in a prospective study and concluded that secretion volume and colour, and mucosal oedema and erythema were significantly associated with BAL neutrophils [23]. Lung inflammation can be reflected by the endobronchial lining and secretions. However, a large number of studies failed to verify this, and there are few reports on the severity of pneumonia and the results of altered bronchoscopy.

In studies conducted by other researchers, BALF microbial cell counts increased with increasing bronchoscopy grades of secretions and were associated with bronchial inflammation [11]. Microbial growth in BALF was greater in patients with bronchoscopy grade IV secretions and above [24]. Nursoy et al. found that the incidence of culture-detected microorganisms in BALF of patients with bronchoscopy grade IV secretions and above increased. With an increasing bronchoscopy secretion score, the first second of expiratory volume and forced vital capacity values gradually decreased, and the bronchial Bhalla score and bronchial mucosal inflammation increased [25]. However, no study has linked scoring tools to disease severity. In this study, the endoscopic mucosal oedema, erythema and pallor scores and the total score in the invasive ventilation group were higher than those in the non-invasive ventilation group. In the FOB and BAL groups, within 1 day of admission to the ICU, the colour score of the secretions was lower than that in the operation group after 1 day. When investigating the effect of endoscopic scores under different ventilation modes, mucosal erythema, pallor, oedema and the total score had better aROCs, which was basically consistent with the performance of the SOFA score aROCs. However, most of the aROCs in this study were no higher than 0.8, and the predictive value was not perfect, which may be related to the small sample size, therefore further prospective studies are still needed. Moreover, as this study is a retrospective study, the percentage of neutrophils in BAL could not be collected, so the validity of this tool needs to be further verified.

Bronchoscopy procedures are often avoided in critically ill patients, especially those on mechanical ventilation because of the risk of complications. However, it has also been reported that bronchoscopy can reduce the incidence of pneumonia, length of hospital stay and mortality in patients who have undergone tracheostomy [26]. Other studies have shown that bronchoscopy for aspiration injury pneumonia can help to reduce ICU length of stay, hospital costs and even mortality by addressing airway problems [27]. The present study also showed that patients in the invasive ventilation group had higher APACHE II scores, SOFA scores, lengths of ICU stay, total lengths of stay and ICU mortality than patients in the non-invasive ventilation group, and early FOB could reduce the length of ICU stay and total length of stay; the lavage treatment was completed within 24 h in 60 cases (26.20%), multiple organs were involved in 128 patients (55.90%), and 22 patients (9.61%) died.

Studies have shown that bronchoscopy within 24 h is beneficial for clinical outcomes in mechanically ventilated patients with aspiration pneumonia [28]. These studies suggest that therapeutic bronchoscopy may benefit patients with severe pneumonia by reducing airway inflammation. The results of the current study showed that early FOB can shorten the length of ICU stay; however, there was no difference in mortality. The survival analysis indicated that the survival rate of the early FOB group was lower than that of the late FOB group. This indicates that bronchoscopy may not improve the survival rate of ICU patients with severe pneumonia. A retrospective analysis of the clinical data of the death cases revealed that they were not related to bronchoscopy. Two cases underwent ECMO treatment for severe ARDS, one had intracranial haemorrhage, and four had multiple malformations, which may be related to the severity of the disease and insufficient sample size. Related studies have also speculated that necrotizing bronchitis and diffuse alveolar damage of bronchioles may occur early in critically ill patients [13]. These factors contribute to poor tolerance of BAL and increase the risk of complications. A recent study of AdV pneumonia also showed that there was no statistically significant difference in the duration of fever and hospital stay between patients in the severe group who received BAL in the early stages and those who received BAL in the later stages [18]. It is even speculated that receiving BAL in the early stages may increase the chance of mechanical ventilation and the occurrence of surgical complications in severe cases. As this study shows, adverse events during or after BAL in ICU patients are not uncommon. According to this study, the occurrence of serious complications, such as pneumothorax, may be related to the inexperience of the performing physician and the longer operation time. Studies have concluded that the inexperience of physicians performing BAL may be a major predictor of clinically significant adverse events in non-intubated patients [8].

Fibreoptic bronchoscopy combined with BAL appears to be an effective treatment for severe pneumonia and also increases its detection rate. In addition, early BAL can reduce hospitalisation and time spent in ICU; however, it cannot reduce mortality, which may be related to the severity of the disease. This study also demonstrated that the endoscopic score can reflect the severity of lung inflammation. However, the sample in the current study was small, so the authors will conduct a multi-centre study with a larger sample in the future.