Most respiratory tract infections in children are viral, and antibiotics have minimal effects

A prospective study in children presenting to their GPs with 'more than a simple cold' found that a demonstrable viral cause could be shown in over three-quarters: antibiotic treatment had minimal effects except in those shown to have influenza. For many years, GPs have been encouraged to avoid prescribing antibiotics for 'simple coughs and colds' to reduce the spread of resistance; prescribing has been reduced, but has not declined further of late. Much of the evidence supporting these recommendations has weaknesses, such as excluding younger children and those with more severe symptoms.

This study aimed to address some of these issues. GPs were asked to identify children presenting with cough and fever and considered to have more severe symptoms, for whom they would consider prescribing an antibiotic: the study looked at the aetiology and times course of the infection in these children.

Participants were children aged 6 months to 12 years with a cough, fever reported within past 72 hours, symptom severity suggesting a respiratory tract infection more than a simple cold, and for whom the GP considered prescribing an antibiotic. Those actually studied had a nasopharyngeal aspirate taken for viral identification (by PCR) and both GP and parents were asked to rate the severity of illness on a 0 to 10 scale. The child's illness was tracked using a parental symptom diary. Outcomes included cause, severity and time-course of illness, and the effects of antibiotics where prescribed.

A total of 425 children were initially selected, of whom 408 were analysed (most common reason for exclusion was lack of parental consent to sampling). In these, a probable viral cause of infection was identified in 77% (316/408), with the main virus identified being influenza (33%), respiratory syncytial virus (RSV, 14%), parainfluenza (14%), and human metapneumovirus (8%). Those with RSV infection were assessed as most severely ill by GPs (mean score 5), however there was considerable overlap between scores. Clinical symptoms identified the virus in 45% of cases.

Duration and severity scores of illness were very similar for all viruses, with the median symptom scores substantially the same on each day, and scores falling to a median of 0 by nine days from presentation. About a third of children (34.1%) were prescribed an antibiotic, most often in human metapneumovirus and RSV infections, however antibiotic prescription had no significant impact on rate of recovery. There was no effect on duration of fever overall with antibiotics, however there was an indication that antibiotics may reduce duration of fever in those with influenza virus infection.

The authors conclude that the use of molecular diagnostic technology and nasopharyngeal aspirates allowed identification of a probable cause for infection in a much higher proportion of cases than in previous studies. It shows that in the majority of children with 'more than a simple cold' for whom an antibiotic would be considered, the infection has a viral aetiology. In this study, prescribing an antibiotic made no difference to the overall time course of the illness. Their results, therefore, underpin existing guidance that these infections in the community are predominantly viral: in the absence of any significant respiratory difficulty they are self-limiting and do not require antibiotics. The consistent time course of illness found with all infections allows a reliable prognosis to be given. They discuss the specific observation of a reduction in duration of fever in influenza infections, and suggest that this is consistent with reports of 10-17% secondary bacterial infection rate in children with 'flu: this observation would support a trial of early use of antibiotics in children with flu.

Arch Dis Child 2007; 92: 594-7 (link to abstract)