Patients and methods

Between 1 January 1993 and 31 December 1995, the aetiology of community acquired pneumonia was studied in 254 hospitalised children at the Department of Paediatrics, Turku University Hospital.2 The bacteria implicated wereS pneumoniae, M pneumoniae, Moraxella catarrhalis,Haemophilus influenzae,Streptococcus pyogenes, and Chlamydia pneumoniae. The viruses implicated were respiratory syncytial virus, rhinovirus, parainfluenza virus types 1, 2, and 3, adenovirus, human herpesvirus 6, influenza A and B virus, and coronavirus. Informed consent was obtained from the parents or guardians of children serving as study subjects.

M pneumoniae infection was identified by studying IgM and IgG antibodies in acute and convalescent phase serum samples and/or positive nasopharyngeal aspirate culture.2In some cases, the complement fixation (CF) test and/or cold haemagglutinin tests were carried out according to standard methods. For detection of S pneumoniae infection, pneumolysin IgG antibodies and pneumolysin immune complexes as well as C polysaccharide IgG antibodies and immune complexes were measured in acute phase and convalescent phase serum samples.2 Blood cultures were obtained from two patients. The methods have been described previously.2

Results

Pneumonia caused by M pneumoniae was diagnosed in 17 patients, and pneumonia caused by S pneumoniae in 93 patients. Of these, evidence of coinfection ofM pneumoniae and S pneumoniae was found in nine patients (table 1). Three of the nine children with coinfection also had evidence of viral infection (rhinovirus, influenza A virus, and human herpes virus 6). In addition, one child had evidence of H influenzaeinfection and one had evidence of M catarrhalis infection as a third possible causative agent. All nine patients were febrile before admission (>37.5°C). Seven patients had symptoms of respiratory tract infection. Five patients appeared ill. One patient had otitis media and one had maxillary sinusitis as well as tonsillitis. All patients had alveolar infiltrations in their chest radiographs: four solely and five with interstitial infiltrations. Five children had a C reactive protein concentration greater than 80 mg/l or a white blood cell count greater than 15 × 10⁹/l. In the hospital, seven patients were initially treated with β lactam antibiotics (table 2), and two patients received macrolide treatment. Finally, five of the nine patients received macrolide treatment either before hospitalisation, in the hospital, or after discharge. The mean duration of fever (>37.5°C) after onset of antibiotic therapy was 24.4 (SD 14.5) hours, ranging from 10 to 48 hours in eight patients who were febrile in the hospital. Four of six patients not initially treated with a macrolide had respiratory symptoms or fever for up to seven days after discharge. In two of them, the symptoms disappeared after onset of macrolide treatment. At the follow up visit three to four weeks after discharge, all patients showed clinical recovery from pneumonia, but four (of the eight) patients still had minor infiltrations on chest radiograph. One patient treated with penicillin and cefadroxil developed otitis media during follow up.

Discussion

S pneumoniae and M pneumoniae are the major bacterial causes of community acquired pneumonia in children, together accounting for up to 60% of cases.1-3 In the present study, half of the hospitalised children with M pneumoniae pneumonia had evidence of S pneumoniae coinfection. On the other hand, 10% of the children with pneumococcal pneumonia had evidence of M pneumoniae infection. Earlier aetiological studies have reported evidence of S pneumoniae and M pneumoniaecoinfection in 29 ambulatory or hospital treated children with pneumonia.3-7 In a recent study,3 patients with the coinfection accounted for 30% of cases ofM pneumoniae infection and for 23% of cases of pneumococcal infection.

Pneumonia in children may often be caused by multiple microbial agents.1 Intercurrent or preceding viral upper respiratory infections are believed to be risk factors for secondary bacterial disease. M pneumoniae infection can also precede viral or other bacterial infections by several days or weeks.8 Staugas and Martin9 reported five cases of M pneumoniae infection with probable secondary infection from H influenzae. Recently, Cimolai and coworkers8reported four patients with severe bacterial or viral infections either following or coinciding with M pneumoniaerespiratory infection. All these studies suggest thatM pneumoniae, like respiratory viruses, may predispose to secondary bacterial infection.

In this study, seven patients (78%) had been ill for a week or longer before admission, supporting the view that M pneumoniae infection probably precedes S pneumoniae infection. In our earlier study, only 25% of 85 children with bacteraemic pneumococcal pneumonia had had symptoms for seven or more days at the time of diagnosis.10 Patients with M pneumoniae pneumonia are often treated as outpatients.7 It is possible that an additional pneumococcal infection, which cannot usually be confirmed by standard laboratory methods, increases symptoms leading to hospitalisation in a patient with M pneumoniae pneumonia.