Previous studies have used multivariate regression analysis to determine the risk of extrapulmonary complications in MPP patients, such as pulmonary necrosis [16] and intrabronchial mucus embolism [17]. However, studies on the risk of BO coused by RMPP patients have not been reported yet. In the present study, we determined seven factors related to the risk of BO in children caused by RMPP through single factor and multivariate regression analysis, including WBC count, ALB level, consolidation range exceeding 2/3 of the lung lobes, the use of macrolides within 5 days, the use of glucocorticoid or fiber bronchoscopy within 2 weeks of disease course, and plastic bronchitis, all of which were used to establish a nomogram of the risk of BO. As far as we know, this is the first nomogram study of the risk of BO caused by RMPP. Based on AUC and calibration curve evaluation, this novel nomogram showed satisfactory performance in the research cohort. Therefore, the nomogram can be effectively used in clinical practice, which helps predict the development of BO in children due to RMPP, and emphasizes the importance of the timing of treatment with macrolides, glucocorticoids, and bronchoscopy after MPP.
Previous studies have shown that there is no difference in clinical characteristics and disease severity between MPP patients co-infected with viral and/or bacterial pathogens and MPP patients without any respiratory virus co-infection [18]. However, Lee et al. [19] have found that respiratory virus co-infection is a risk factor for BO after infection in children with MPP. In the current study, we found that although the proportion of children with RMPP pneumonia complicated with influenza infection was higher than that of non-BO children, multivariate regression analysis found that influenza infection was not a risk factor for the development of BO after RMPP. It is widely known that adenovirus infection is one of the risk factors for the development of BO in children [20]. In the current study, we did not find an association between adenovirus co-infection and the development of BO after RMPP, which may be related to the small sample size.
Inflammatory mediators are involved in the immune pathogenesis of MP infection [21]. As a negative acute-phase protein synthesized by the liver, serum ALB can rapidly decrease in acute infection [22]. The reduction of serum albumin level is considered to be a biomarker of local and systemic inflammation, and has important clinical value in predicting the severity and prognosis of pneumonia [22–24]. As a specific marker of fibrinolysis, D-dimer reflects the ability to dissolve fibrin [25]. Studies have shown that the clinical symptoms and chest imaging findings of MPP patients with elevated D-dimer levels are more severe than those with normal D-dimer levels [26, 27]. The increase of serum D-dimer level in patients with RMPP suggests excessive inflammation and prolonged vascular endothelial injury [28]. In the current study, compared with the non-BO group, the WBC count, CD8 + ratio and serum D-dimer level of the BO group were significantly increased, while the serum ALB level was significantly reduced. In addition, high WBC counts and low ALB level are independent risk factors for the occurrence of BO, suggesting that patients with RMPP complicated by BO have stronger systemic inflammatory responses. Persistent MP antigen stimulation and/or invasion greatly increases the occurrence of severe lung disease and pulmonary and extrapulmonary complications [29]. In the present study, the severity of pneumonia is an independent risk factor for the development of BO after RMPP, which suggests the need to monitor the lesions of RMPP patients in time.
Macrolides are the first choice of drug for MP infection in children [30]. However, many MP isolates in clinical samples showed resistance to macrolides [31]. Macrolide resistance in MP may weaken the response to drug therapy [32–34]. The presence of macrolide-resistant MP has been reported to be mainly related to persistent clinical symptoms such as fever, prolonged hospital stay, and increased antibiotic replacement rate [35]. Therefore, macrolide resistance may be the cause of RMPP [36]. However, some studies have shown that there is no significant difference in the detection rate of drug resistance genes between RMPP patients and children with ordinary MPP [37], revealing that macrolide resistance may not be the main cause of RMPP [38]. In the current study, we found that compared with the BO group, the non-BO group used a higher proportion of macrolides within 5 days of the disease course. Besides, the use of macrolides within 5 days of the disease course may be a protective factor for the risk of BO in RMPP patients, suggesting that the use of macrolides as soon as possible can reduce the occurrence of BO. But in the current study, we found that patients with BO showed longer treatment time for macrolides and glucocorticoids. This may be due to the late treatment time leading to excessive inflammation caused by infection, leading to prolonged clinical course. In addition, for RMPP patients with poor effect in the treatment of antibiotics and corticosteroids, fiberoptic bronchoscopy within 2 weeks of the disease course may effectively prevent the occurrence of BO.
Plastic bronchitis is a rare and underdiagnosed disease [39]. Influenza virusand mycoplasma pneumoniae infection are usually common causes of plastic bronchitis [40]. Plastic casts are caused by a variety of inflammatory cell infiltration and inflammatory mediators, leading to congestion, edema, necrosis, and lumen obstruction of the tracheal mucosa [41]. In the current study, we found through fiberoptic bronchoscopy that bronchial phlegm plugs and plastic bronchitis found in the BO group during the first bronchoscopy in the acute phase was higher than that of the non-BO group, which may be related to the excessive inflammatory reaction and bronchial endometrial damage in RMPP leads to the occurrence of BO. Moreover, the occurrence of plastic bronchitis was an independent risk factor for the occurrence of BO in RMPP patients, suggesting that the plastic under fiberoptic bronchoscopy may indicate the occurrence of BO.
The nomogram is very useful in using individual variables to predict the probability of a clinical event occurring in an individual. As an alternative, unlike multivariate regression analysis, a nomogram is a graphical description of a statistical model that calculates the probability of an individual patient’s specific outcome with satisfactory accuracy [42]. The nomogram has its own limitations. For example, when the covariate measurement changes, the performance of the nomogram may change greatly, and the influence of the nomogram-assisted decision-making on patient outcomes still needs to be studied [43]. Nonetheless, nomograms have been widely used to predict the short-term and long-term outcomes of various diseases (such as cancer and diabetes) [44]. In the current research, we provided a simple and easy-to-use risk prediction nomogram for the first time, which contains five factors that affect the occurrence of BO. The satisfactory performance of this model is reflected in its high predictive ability, and its AUC for the study cohort is greater than 0.8. The nomogram may improve and provide new ideas for the early identification and intervention of BO in RMPP patients.
There are still some limitations in this study. First of all, the study population was small, and only patients from a single center were included. Secondly, this is a retrospective study based on reviewing medical records, and patients with incomplete medical records were excluded. Thirdly, considering that macrolide resistance may be related to the occurrence of RMPP, antimicrobial susceptibility tests are still needed to classify patients with BO. In addition, multiple samples are still required to be listed to verify the feasibility of the nomogram model in this study. In order to further confirm the results of this study, it is necessary to conduct a multi-center in-sample study.