Children diagnosed with lobar pneumonia caused by MP in the Department of Respiratory Medicine of our hospital from February 1, 2023 to January 31, 2024, who had undergone bronchoscopy were selected as the subjects of this study. The PB group consisted of all 44 children from whom plasticized materials could be removed during bronchoscopy. The non-PB group was composed of 88 children who were matched by gender structure to the PB group through 1:2 propensity score matching from over 1,000 children who met the established diagnostic criteria (i.e., “lobar pneumonia caused by MP infection and had undergone bronchoscopy”) but did not have plasticized materials. Our study has received approval from the Ethics Committee of our hospital, and all children’s guardians were required to assist in completing the informed consent form. Inclusion criteria for the children were: ① Those who met the diagnostic criteria for Mycoplasma pneumoniae pneumonia (MPP) as outlined in the Expert Consensus on the Diagnosis and Treatment of Pediatric MPP (2015 edition); ② Those who met the diagnostic criteria for lobar pneumonia; ③ Bronchoscopy procedures that complied with the regulations in the Chinese Pediatric Flexible Bronchoscopy Guidelines (2018 edition); ④ Parents of the children signed the informed consent for bronchoscopy. Exclusion criteria were: ① Children with severe arrhythmia, severe congenital heart disease, bleeding disorders, or insufficient cardiopulmonary or liver and kidney function, or congenital immunodeficiency; ② Children allergic to treatment drugs or with poor compliance; ③ Children with contraindications to bronchoscopy. A total of 132 children were included in the study, with 44 in the PB group and 88 in the non-PB group. The flowchart for the methods was shown in Figure 1.

The indicators included in our statistical analysis are as follows: ① Duration of fever before this hospital admission; ② Duration of cough; ③ Length of preoperative illness; ④ Duration of macrolide antibiotic use before completing fiberoptic bronchoscopy; ⑤ Duration of other antibiotic use before completing fiberoptic bronchoscopy; ⑥ Pulmonary signs; ⑦ Laboratory test indicators: white blood cell count, neutrophil percentage, lymphocyte percentage, platelet count, C-reactive protein, procalcitonin, neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), systemic immune inflammation index (SII), systemic inflammation response index (SIRI), lactate dehydrogenase, D-dimer, and fibrinogen content; ⑧ Co-infections with other pathogens: bacterial, viral, or mixed; ⑨ Chest imaging results.

Our statistical analysis was performed using SPSS 26.0 software. Measurement data that conform to a normal distribution are represented as X ± S, and comparisons between groups are made using the t-test; measurement data that do not conform to a normal distribution are represented as M (P25, P75), and comparisons between groups are made using non-parametric tests. Count data are expressed as rates, and comparisons between groups are made using the Pearson χ² test, continuity correction χ² test, and Fisher’s exact probability test. Multivariate logistic regression analysis is conducted for indicators with significant differences between the two groups. A p-value of less than 0.05 is considered statistically significant.

The basic demographic, general medical history, and laboratory parameters of the participants are shown in Table 1.

The age difference between the two groups was statistically significant (t = −2.201, p = 0.029). The test results indicated that the children in the PB group were elder than those in the non-PB group, with an average age of 7.48 years in the PB group and 6.51 years in the non-PB group.

The duration of fever, duration of cough, preoperative course of disease, and pulmonary signs showed no significant statistical difference between the two groups of children (p>0.05).

There was no significant difference in the total number of white blood cells in the blood of the two groups of children. However, the ratio of neutrophils to lymphocytes in white blood cells showed a significant statistical difference. In the PB group, the proportion of neutrophils was significantly higher, while the proportion of lymphocytes was significantly lower (p < 0.001). The levels of CRP, PCT, NLR, PLR, SII, LDH, and D-dimer were all significantly higher in the PB group than in the non-PB group (p-values were <0.001, <0.001, <0.001, <0.001, 0.001, <0.001, <0.001 respectively). There was no significant difference in the comparison of other indicators (p > 0.05).

In addition, there was no significant difference in the proportion of patients with stenosis, occlusion, or mucus plugs in the chest imaging findings between the PB group and the non-PB group. Meanwhile, there was also no significant difference in the composition ratio of other pathogens apart from Mycoplasma pneumoniae between the two groups of children (p > 0.05).

The preoperative use of antibiotics in the two groups of children showed different results depending on the type of antibiotic. There was a significant difference in the composition ratio of children using macrolide antibiotics between the two groups (p = 0.031). In the category of using macrolide antibiotics for ≥2 courses, the PB group had a higher composition ratio than the non-PB group, while in the category of <2 courses, the result was the opposite. However, there was no significant difference between the two groups of children in the use of other antibiotics, such as β-lactam antibiotics (p > 0.05).

From the aforementioned study, we extracted several indicators with statistical differences among blood cell-related parameters and examined their value in predicting the occurrence of PB, and the final curves were shown in Figure 2. We found that D-dimer had the highest sensitivity (81.82%, AUC = 0.7771), followed by SII (77.27%, AUC = 0.6761). PCT had the highest specificity (92.05%, AUC = 0.6761), followed by LDH (87.5%, AUC = 0.8035).

Subsequently, we conducted a multiple regression analysis on easily quantifiable laboratory indicators, including age, CRP, PCT, NLR, PLR, SII, LDH levels, and the ratio of D-dimer as independent variables, with concurrent PB as the dependent variable (no = 0, yes = 1), and performed a binary multiple logistic regression analysis, as they were shown in Table 2. Since some significant results may be collinear with new indicators, we adjusted for neutrophils and lymphocyte subgroups. The results showed that PLR (OR = 1.011, 95%CI = 1.002–1.020, p = 0.013), SII (OR = 0.999, 95%CI = 0.998–1.000, p = 0.044), and D-dimer (>normal value but <10 times the maximum normal value OR = 5.460, 95%CI = 2.000–14.905, p<0.001; >10 times the maximum normal value OR = 33.450, 95%CI = 2.879–388.676, p = 0.005) still had significance in the multiple analysis and were jointly related to the risk factors of PB occurrence.

Since there were no indicators with both high sensitivity and high specificity in the aforementioned independent predictive indicators for PB diagnosis, we conducted another logistic regression analysis. By establishing a triad model of PLR, SII, and D-dimer for prediction, we found that this model had good diagnostic value for the formation of PB (AUC = 0.831), with a sensitivity of 84.09% and a specificity of 72.73% as shown in Figure 3. The formula for this triad model is Logit(P) = −3.251 + 0.011PLR − 0.001SII + 1.901D-dimer (> normal value but <10 times the normal value) + 4.130D-dimer (>10 times).

This study has several limitations. First, this is a single-center study. Although Chengdu is the largest and most populous city in Southwest China, its radiation scope is inevitably limited. As a city located in a basin, this study did not include participants from northern China, plateau areas, or different ethnic groups, which may lead to certain limitations in the generalizability of the study results. Second, plastic bronchitis is a relatively rare disease, resulting in a relatively small sample size in this study. In addition, due to the limited age range for eligible pulmonary function tests (e.g., forced expiratory volume in 1 s, lung diffusing capacity), inconsistent compliance of children’s caregivers, and incomplete pulmonary function data attributable to the retrospective design of this study, we did not investigate the impacts of plastic bronchitis on children’s pulmonary function, quality of life, exercise capacity, and family care-related indicators. Thus, the dimensions of outcome assessment in this study are not sufficiently comprehensive. Furthermore, conventional chest imaging examinations have no definitive diagnostic value for plastic bronchitis in Mycoplasma pneumoniae-associated lobar pneumonia, and the search for alternative visual imaging modalities remains warranted—this also constitutes a limitation of the present study.

Based on the limitations of this study and the clinical and research needs in the field of pediatric respirology, we put forward the following recommendations for future research. First, multicenter, large-sample prospective cohort studies should be conducted to enroll pediatric participants from different geographical regions and hospitals at all levels, expand the research coverage of plastic bronchitis (PB), and verify the generalizability of the findings of this study. Integrating artificial intelligence and machine learning technologies, a large-scale dataset should be established to further refine predictive models, which can then be formulated as expert consensus and promoted nationwide. Second, with an expanded sample size, analyses of indicators related to children’s pulmonary function, quality of life, exercise capacity and family care should be carried out, together with targeted clinical intervention studies; meanwhile, long-term follow-up should be conducted to explore whether early intervention alters clinical outcomes. A multidimensional evaluation system should be constructed to more comprehensively investigate the clinical and molecular mechanisms underlying Mycoplasma pneumoniae infection, lobar pneumonia and plastic bronchitis, thereby providing more reliable clinical and theoretical support for the precise diagnosis and treatment of pediatric respiratory diseases. Third, in view of the limitation that conventional chest imaging examinations have poor diagnostic performance for PB complicated with lobar pneumonia, more intuitive alternative imaging methods should be explored. For example, recent studies have shown that emerging pulmonary ventilation imaging techniques represented by hyperpolarized gas magnetic resonance imaging (MRI) (e.g., propane, butane, diethyl ether) may make up for the shortcomings of conventional chest imaging (50–53)—this technique enables direct visualization of pulmonary ventilation function with excellent spatial resolution, which may help improve the specificity of differentiating PB from other pulmonary diseases and integrate the assessment of pulmonary ventilation function with imaging findings, achieving two objectives with one measure.

In summary, neutrophil ratios, CRP, PCT, NLR, PLR, SII, LDH, and D-dimer are independent risk factors distinguishing PB from non-PB cases. Among these, the three-marker model combining PLR, SII, and D-dimer offers the strongest predictive capability for PB in children with lobar consolidation or atelectasis. This study also highlights the need for timely fiberoptic bronchoscopy when extensive consolidation or atelectasis persists beyond two macrolide treatment courses. Our findings provide robust support for clinical decision-making and enhance patient–doctor communication in managing MPP-associated PB.