Our current study demonstrated that an independent significant factor associated with the deterioration in the clinical course was the consolidation pattern observed on initial chest radiograph in children with the A/H1N1 infection in Japan where NIs have been widely used. The percentage of patients who were admitted to the PICU and required MV was higher among those with consolidation on initial chest radiographs compared to those without. These findings support our hypothesis that the presence of consolidation pattern on chest radiographs representing alveolar damage is directly correlated with the disease severity in children with A/H1N1 infections, even in the area where NIs have been widely used [24–26].
The pattern and extent of consolidation on initial chest radiographs was a significant factor in the clinical course of patients with A/H1N1 infection. Additionally, the body temperature and oxygen demand on admission were independently associated with the clinical course of the patients who were administered in the PICU and who required MV, respectively. This may be related to the fact that the body temperature of patients hospitalized in the PICU might have already been controlled by their having received antipyretics or anesthetics in preparation for MV. Additionally, oxygen demand on admission may reflect the direct involvement of infection in the lung parenchyma, which is closely related to consolidation on chest radiographs. Meanwhile, oxygen demand might have been masked, because certain patients under serious condition had supplied oxygen at the time of SpO2 measurement.
Several previous studies of A/H1N1 reported that one of the most important risk factors for complications of severe illness with A/H1N1 infection were age [10, 27–29]. In the current study, however, age was not a significant factor associated with rapid disease progression of A/H1N1 infection in children. This discrepancy may be explained by the fact that age distribution of A/H1N1 infection in children in Japan was toward school-aged children compared to infants and younger children . Similar trends were observed in other Asian countries , suggesting that the influence of host genetic factors or the existence of maternal cross-reacting antibodies protected infants and younger children. Additionally, the use of NIs may contribute to the difference. In Japan, nearly all patients with laboratory confirmed influenza by the rapid antigen test are prescribed NIs . A further study is warranted to validate these findings.
According to a report, reviewing the chest radiographic findings in patients with A/H1N1 infection demonstrated that the predominant chest radiographic findings in children were interstitial pneumonia (34.3%), followed by consolidation and/or atelectasis (23.8%), normal radiographs (22.3%), lobar or segmental consolidation (19.4%,), and pleural effusion (14.9%) . In addition, several studies reported the findings of chest radiographs and computed tomography (CT), especially in terms of the patterns, distributions, and extent of lung involvement [10, 17, 32–36]. One study in adults revealed that normal chest radiographs were common among patients presenting to the hospital for A/H1N1 infection-associated symptoms without evidence of respiratory difficulties, but that the factors associated with an increased likelihood for abnormal chest radiographic findings were dyspnea, hypoxemia, and diabetes mellitus . The most common abnormal radiographic finding for lung involvement was a mixture of air-space consolidation and reticulonodularity or diffuse haziness of the lungs with a patchy pattern and lower/middle zone predominance [17, 33, 35, 36].
According to the report from Korea, abnormal chest radiographic findings were uncommon, but the major abnormal findings in chest radiographs and CT images were prominent peribronchial markings and patchy consolidation with mediastinal lymph node enlargement, pleural effusion, and pneumomediastinum in children with A/H1N1 infections . A few previous studies evaluated the relationship between disease progression of A/H1N1 and findings on chest radiographs. Lee et al. demonstrated that bilateral, symmetric, and multifocal areas of consolidation associated with reticulonodularity or diffuse haziness of the lungs were the predominant radiographic findings in pediatric patients with a more severe clinical course of A/H1N1 infection ; however, limited patient characteristics were reviewed in that study, and no multivariate analysis was performed to evaluate possible confounding factors. A similar trend was described in other two studies, and the findings of a multifocal patchy consolidation pattern with bilateral or diffuse lung involvement on admission should alert of the impending severity of disease [18, 19]. The clinical and pathophysiological differences between patients with consolidation and those with reticulonodularity or diffuse haziness of the lungs are likely due to the nature and extent of the inflammation. Consolidation on chest radiographs reflects the presence of alveolar spaces containing varying numbers of neutrophils and mononuclear cells admixed with fibrin and edema fluid, which may lead to airway obstruction, including atelectasis and plastic bronchitis. In addition to primary viral infection, secondary bacterial infection may be associated with consolidation on chest radiographs, because the normal epithelial cell barrier to infection and loss of mucociliary clearance enhance bacterial pathogenesis [23, 37–40]. In addition, reticulonodularity or diffuse haziness of the lungs on chest radiographs indicates that inflammation affects the interstitium, and pulmonary opacities appear as airspace, linear, or bandlike opacities, in a nonfocal, patchy, or mottled distribution of varying density. In addition, reticulonodularity or haziness of the lungs on chest radiographs is diffuse and is a radiographic pattern that reflects noncardiogenic pulmonary edema related to sepsis or diffuse pneumonias .
In the current study, most of patients received antivirals; those with consolidation on initial chest radiography (97%) and those with other findings (91-94%). Meanwhile, disease severity was significantly higher in those with consolidation than in those with other findings regardless of administration of antivirals. Although clear benefits of early use of NIs in adults have been described in the previous literatures , other data in Japan did not demonstrate the differences in timing of NIs treatment between fatal cases and non-fatal, severe cases . However, no data regarding the efficacy of early use of NIs were available in children with A/H1N1. Early administration of NIs might influence clinical course and outcome; however, in the current study, clinical course and outcome had already determined at the time of admission regardless of NIs administration in this population who received NIs.
We acknowledge that there were several limitations to our study. First, this study was a retrospective study, and we only included patients who required admission and whose initial chest radiographs were available. Some patients in the study may have undergone initial chest radiography at outside hospitals, but their chest radiographs were not available. Second, early medical check-ups were available in Japan, and the majority of patients in this study were prescribed NIs during the early stage of the illness. These clinical interventions could have modified radiological findings and clinical course of the disease compared to the data from other countries, where not all patients may have been able to visit the hospital as soon after the onset of illness. Third, secondary bacterial infection may affect clinical course and radiological findings; however, most of the present patients with a consolidation pattern on initial chest radiographs received antibiotics without microbiological evaluation. Fourth, A/H1N1 infection was diagnosed with a rapid test or medical history, without subsequent confirmation from tests such as polymerase chain reaction. Finally, it was impossible to investigate the relationship between radiographic findings and histopathologic findings , because none of the patients underwent a lung biopsy.