Community‐acquired pneumonia in pediatric patients with acute neuromuscular respiratory failure: A microbiologic perspective

We are grateful to read the comments sent to the journal by Dr. Massimiliano Don regarding our article “Combined noninvasive ventilation and mechanical in-exsufflator in the treatment of pediatric acute neuromuscular respiratory failure”. These valuable comments bring our attention to the microbiologic aspect of pneumonia in pediatric patients with neuromuscular diseases (NMD), an important issue that has not been well addressed. To date, early or prophylactic use of antibiotics as clinical management plans for community-acquired pneumonia (CAP) in NMD children are suggested. However, the literature is sparse regarding the etiology of CAP in these children and beneficial evidence of empiric antibiotics in these populations is still lacking. Thus, we are in line with Dr. Massimiliano Don that investigations elucidating the etiology of CAPmay improve outcome of these patients and avoid adverse effects of unnecessary antibiotics. To this end, we re-analyzed our data in our article, and tried to explore possible microbial factors associated with the outcome of these patients receiving noninvasive ventilation (NIV) and mechanical in-exsufflator (MIE) due to pneumonia. To compare the differences between success group (n1⁄4 12) and failure group (n1⁄4 4) in our patients, we analyzed infection-related variables including presence of fever at admission to our pediatric intensive care unit, pathogen survey by sputum culture, serum mycoplasma antibody or nasopharyngeal antigen detection of respiratory syncytial virus (RSV), and inflammatory markers including white blood cell count and serum C-reactive protein (CRP). However,we found that therewere no significant differences of these variables. The presence of fever at admission was comparable, 42% (5/12) versus 50% (2/4) between success group and failure group. In the success group, there are only four events with identified pathogens, including two with RSV and two with mycoplasma pneumonia, while in the failure group there is one with mycoplasma pneumonia infection and one with streptococcus infection. Comparing the success group with the failure group, the initial WBC countswere 9,263 2,618ml 1 versus 10,225 2,646ml 1 (P1⁄4 0.54) and the initial CRP levels were 17.6 14.6mg/L versus 21.7 22.8mg/L (P1⁄4 0.68, normal range1⁄4 0–6mg/L). Take together, we did not find any prognostic factor associated with outcomes of NIV/MIE use in our study. In addition, we speculate that in those events without identifiable pathogen (n1⁄4 10, 67%), virus infection may be the most probable etiology given with the low averageWBC counts and CRP levels. Of note, the small sample sizes in our study lacked the statistical power to rigorously conclude the validity of these lacks of difference. Future studies, properly powered, will determine with more certainty whether these preliminary observations are correct. Despite advances of bacteriological identification procedures in the pathogen survey of pneumonia, the frequency of microbiologically documented CAP is only around 25% among in-patients. Furthermore, in those patients with pneumonia not receiving endotracheal intubation, the diagnostic reliability of deep cough-produced sputum and nasopharyngeal aspirates remains uncertain because of the problem of contamination with upper airway flora. Accordingly, in children with severe CAP without intubation, their inability to produce effective cough often preclude the reliability of pathogen investigations in pneumonia.