Community-acquired pneumonia (CAP) is a common infectious disease that is sometimes fatal and seriously threatens the health of children. CAP remains a significant cause of morbidity and death worldwide. Mycoplasma pneumoniae pneumonia (MPP) is one of the most frequent causes of respiratory tract infection, accounting for almost 30% of all cases of CAP (1, 2). Under normal circumstances, MPP is mainly a benign, self-limiting disease; however, the number of children with severe Mycoplasma pneumoniae pneumonia (SMPP) is increasing every year due to drug abuse, drug resistance, clinical misdiagnosis and missed diagnosis (3). Moreover, children with SMPP also have high rates of extrapulmonary organ dysfunction and macrolide resistance (4). In addition, Streptococcus pneumoniae (S. pneumoniae) is one of the most common leading pathogens of CAP, causing mild to life-threatening invasive diseases. Because of their few side effects and lower toxicity, β-lactam and macrolide antibiotics are used as first-line therapies in children (5). In recent years, the treatment of Streptococcus pneumoniae pneumonia (SPP) has become increasingly difficult due to the high levels of macrolide-resistant S. pneumoniae. Reports have shown that approximately 95% of Streptococcus strains show high resistance to erythromycin and clindamycin (6, 7). Therefore, it is difficult to manage SPP by medication with macrolide antibiotics as the first-choice drug. On the other hand, severe pneumonia associated with S. pneumoniae and Mycoplasma pneumoniae (M. pneumoniae) remains an important reason for ICU admission, and more than 50% of children with pneumonia in China die of S. pneumoniae (7). If acute MPP in children is not controlled in time, it may progress to bronchiolitis obliterans, which can greatly affect the quality of life of children (8). These facts indicate that the identification of MPP and SPP in the early stage could be important. To our knowledge, there are already some scoring systems that propose a differential diagnosis between MPP and some bacterial pneumonias for selection of the appropriate antibiotics for the treatment of CAP (9). In fact, there are many situations in which we cannot accurately distinguish between the two groups of patients. When MPP presents as lobar pneumonia, the patient has similar clinical manifestations and imaging features to SPP, even those experienced radiologists and pediatricians are difficult to distinguish between the two diseases.

Thus, the purpose of the present study was to determine whether we could differentiate between pneumonia caused by M. pneumoniae and S. pneumoniae on the basis of epidemiological and clinical manifestations and laboratory results before infection was serologically confirmed. The findings from our study are anticipated to provide evidence-based medical guidelines for early diagnosis, enabling the treatment of CAP caused by M. pneumoniae and S. pneumoniae, thereby leading to improved prognosis.

In total, 506 patients with MPP and 311 patients with SPP who had been hospitalized at the Children’s Hospital of Chongqing Medical University were included. Previous studies have shown that M. pneumoniae and S. pneumoniae are the major causative agents of CAP in children (11). According to the WHO, pneumonia accounts for 16% of all deaths in children under 5 years old, with S. pneumoniae being the most common cause of bacterial pneumonia (5).

In our study, a higher incidence of MPP was recorded in summer and autumn, which was the same as a previous study in Nanjing, China (12). However, S. pneumoniae CAP was recorded as common in spring and autumn, which differed from some reports. Previous studies have shown that M. pneumoniae infections are uncommon in children under the age of 5, with the maximum frequency occurring between the ages of 5 and 15 (2, 13), and other studies have shown that children aged 2–5 years are more susceptible to M. pneumoniae infections (14). Consistent with some other studies (11), our results show that the highest incidence of MPP was among preschool children aged around 5 years old, while the highest incidence of SPP was in infants and young children aged around 2 years old.

In both groups of patients, males were affected more often than females, which correspond to the findings of some previous studies (6, 13). We observed that 93.9% of MPP and 86.5% of SPP patients received antibiotics before admission, while there was no significant difference in the proportion of severe pneumonia and the average duration of hospitalization in both groups.

We observed several differences between the symptoms and signs produced by MPP and SPP. It has been reported that MPP usually presents with fever, cough, diarrhea, and some other nonspecific symptoms and that the clinical symptoms, laboratory data and radiologic findings of MPP are different from those of other bacterial pneumonias (15–17). Previous studies have demonstrated that both M. pneumoniae and S. pneumoniae are pathogens that cause persistent wheezing, which causes an inflammatory stimulus in children with persistent wheezing (18, 19). Acute M. pneumoniae infection can cause not only acute attacks in children with asthma but also wheezing in nonasthmatic children (20). In our study, the patients with MPP were more likely to have a fever, dry cough and diarrhea than patients with SPP but less likely to have wheezing, cyanosis and irritability.

Both groups of patients were observed as being more likely to have crackles and less likely to have rhonchi in the lungs. A higher proportion of SPP patients had signs, either crackles or rhonchi. In addition, 73.95% of SPP and 44.66% of MPP patients had bilateral signs, while 20.26% of SPP and 37.15% of MPP patients had no signs in the lungs. In our study, 15.2% of MPP patients and 12.9% of SPP patients were diagnosed with severe pneumonia. These data may suggest that some children with SMPP had few signs in the lungs, which led to misdiagnosis and missed diagnosis.

In recent years, hematological markers have gradually begun to be used in the clinical diagnosis and prognosis of infectious diseases, and the neutrophil to lymphocyte ratio (NLR) may reflect the degree of inflammation and the state of immunity (21). A previous study reported that the NLR may have clinical value in the differential diagnosis of bacterial pneumonia, and the NLR value was significantly higher in children with bacterial pneumonia than in children with MPP (22). Analysis of the laboratory findings in our study showed that there were significant differences between the two groups in mean leucocyte count, N%, L%, ALT levels, AST levels, LDH levels and incidence of accelerated PCT. Similar to a previous study, the perpheral leucocyte count of the patients with MPP is usually normal at less than 10,000/UL (17). The higher leucocyte counts and PCT levels may suggest that the inflammation of Streptococcus pneumoniae pneumonia is more serious than that of Mycoplasma pneumoniae pneumonia. Although significant differences in inflammatory markers were observed between the two groups, these markers were not found to be adequate for predicting SPP. This may be due to the large overlap of serum inflammatory markers between groups or the inherent characteristics that these markers are nonspecific and reflect the level of whole-body inflammation (23).

In our study, age, dry cough, polypnea, LDH level and PCT level were independent predictive factors associated with MPP and SPP. Younger age, no dry cough, no polypnea, lower levels of LDH, and higher PCT might lead to a diagnosis of SPP. MPP can lead to liver damage in patients, which may be directly caused by M. pneumoniae or may result from the liver’s immune response to M. pneumoniae. LDH can be released into the blood when hepatolysis or cell membrane damage of hepatocytes, resulting in significantly elevated LDH levels in children with SMPP. Unlike previous studies have shown that serum LDH levels may indicate the severity and prognosis of pneumonia (24–26), our results showed that LDH was associated with M. pneumoniae infection. However, the area under the ROC curve of serum LDH levels in our study was smaller, and the prediction value was smaller. Previous studies suggest that age may be a main factor associated with M. pneumoniae infection, but the underlying mechanisms remain unclear (6, 27). In our study, age > 48.5 months was an independent predictive factor for early evaluation and identification of MPP. This study provides a possible theoretical basis for distinguishing between MPP and SPP.

In summary, there are many differences between Mycoplasma pneumoniae pneumonia and Streptococcus pneumoniae pneumonia. Age, dry cough, polypnea, LDH level and PCT level were independent predictive factors associated with MPP and SPP. An age > 48.5 months could be a possible predictor for the early evaluation and identification of MPP. This is the first study to elucidate the differences between MPP and SPP, providing a theoretical basis for early identification. There are several strengths of our study. First, to the best of our knowledge, this is the first study to analyze the differences in clinical features between children with MPP and those with SPP. Second, our study identified predictive factors to distinguish between children with MPP and those with SPP. However, the study still has some limitations. First, it is a retrospective study, which appears to have a selection bias. Second, the presence of other pathogens could not be detected due to shortcomings in the existing detection technology. Third, the distribution of the number of patients among the two groups could not be balanced, which could have influenced the statistical analysis. Fourth, the cases in our study were all children from southwestern China, and multicenter study with a large sample size is required in future studies.