Mycoplasma pneumoniae is a common cause of community-acquired pneumonia (CAP) in children [1,2]. There is a scarcity of studies investigating co-infections of M. pneumoniae pneumonia (MPP) in children. The purpose of this study was to investigate the frequency and impact of bacterial co-infection in hospitalized children with MPP. Bacterial co-infection occur in respiratory MPP infections, but the attack rates and the clinical profile are not clear. The purpose of this study was to investigate the impact of bacterial co-infection in hospitalized children with MPP.

Methods

Medical records of all patients with MPP who were admitted to Beijing Children's Hospital from June 2006 to June 2014 were reviewed. The Pediatric Internal Medicine Department had 10039 MPP admissions during this time. Cases were eligible for enrolment if complete data were available. Pneumonia was diagnosed according to standard guidelines [3-5].

Patients were excluded if they had chronic pneumonia [6], tuberculosis (TB), fungal, Epstein-Barr virus (EBV) or Cytomegalovirus (CMV) infection, congenital immuno-deficiency, malignancy, or were receiving immuno-suppressant agents. A total of 8,612 children aged 0-17 years old were included in this analysis (Fig. 1).

Fig. 1 Study flow chart.

The acute and convalescent serum were obtained and measured for antibody response to M. pneumoniae by enzyme-linked immunosorbent assay methods (Serodia-mycoii, Japan) [7]. An acute infection was indicated by a 1:160 antibody titres [8]. Patients were also evaluated for viral, bacterial, tubercular or fungal infections.

All patients were screened for pulmonary tuber-culosis by the Purified protein derivative skin test with 5TU purified protein derivative. Blood, pleural effusion and bronchoalveolar lavage fluid (BAL) were sent for slide review and bacterial, M. tuberculosis and fungal culture.

A case with a co-infection was defined as any bacterial pathogen except M. pneumoniae detected in any specimen. A patient was considered to have a single infection if M. pneumoniae was the only pathogen detected.

The severity of pneumonia was assessed by scores from 0 to 5 according to the number of following clinical find-ings observed in the patients during admission (Table I): fever (>38.5º), rapid breathing (and/or lower chest wall indrawing), decreased oxygen saturation breathing room air (<92%), more than 7 days of hospital stay, more than 2 affected pulmonary lobes on chest X-rays. The patients with severity score ³3 were defined as severe pneumonia group and £2 as non-severe pneumonia group [5].

TABLE I	Severity Assessment of Pneumonia in Included Children

The management of CAP in infants and children was done as per standard guidelines [3-5]. Patients with severe M. pneumoniae pneumonia who required intensive care unit (ICU) admission were defined as per Infectious Diseases Society of America/American Thoracic Society criteria for severe CAP [9]. The symptoms mentioned above were typical of severe M. pneumoniae pneumonia.

Statistical analyses: Analyses were performed using SPSS 17.0 (SPSS Inc, Chicago, IL). The differences in age distributions among patients with various pathogens identified were tested by an independent sample t-test. P<0.05 was considered statistically significant. Para-metric data were compared with independent sample t-tests. Categorical data were analyzed by using the chi-square test. We have performed a univariate and multivariate Cox's regression analysis for various factors affecting hospital stay more than 7 days.

A total of 8612 children (age 2 m - 17 y; 51.6% males) hospitalized with MPP were included in the study. Characteristics of the 8,612 children hospitalized with MPP are shown in Table II. There were 1012 children with severe pneumonia; and their hospital stay was longer.

TABLE II	Clinical Characteristics of Children Hospitalized with  M. pneumoniae pneumonia 

Tests for bacterial, acid-fast bacilli and fungal infections were performed in all patients, and 2% (173/8612) of cases were positive for at least one bacterial pathogen in addition to M. pneumoniae. Bacterial isolates in these 173 cases are listed in Table III. One bacterial pathogen was identified in 93.1% (173/185), and two bacterial pathogens were identified in 6.9%(12/173). S. pneumonia, Haemophilus influenzae (H. influenzae) and Staphylococcus aureus (S.aureus) were the most common source of infection (Table III).

TABLE III Pathogens Indentified in Children With Mycoplasmal Pneumonia

Significant differences were observed in course of diseases, leukocyte count, and C-reactive protein between single and co-infections (Table II). There was no significant difference in Neutrophil, Lymphocyte, Platelet, Serum lactate dehydrogenase (LDH), Serum Creatine kinase (CK) and Serum Alanine amino-transferase (ALT) between patients with single infections and those who with co-infection (Table II). Hospital stay of children with single infections was shorter as compared to those with than bacterial co-infections (Table II).

Web Table I presents the results of univariate and multivariate Cox's regression analysis for various factors affecting hospital stay more than 7 days. Age was an important factor affecting hospital stay. Unilobar or Multilobar pneumonia was another important factors. Mixed infections and severe pneumonia also contributed to prolonged hospital stay

In this retrospective study from China, 2% of children with MPP were infected with another bacterial pathogen. S. pneumoniae was the leading cause of bacterial co-infection. Co-infections led to more disease severity in children with MPP compared with single infections.

There were several limitations to our study. First, nearly all children in our study received antibiotic treatment. This may have affected the results of bacterial culture. Second, we did not study co-infection with viruses.

Frequency of co-infections in our study was lesser than that seen in few other reports from China [10,11]. This could be related to inclusion of viruses as cause of co-infectioin in these studies. The distribution and age categorization of various bactria isolated in our study is in general similar to other reports from developing countries [12,13].

We conclude that bacterial co-infections are relatively uncommon in M. pneumoniae pneumonia. S. pneumoniae is the most common cause of bacterial infection in M. pneumoniae pneumonia.