Unfortunately, there are no features on history or physical examination that are highly specific for either viral or bacterial disease. Although bacterial disease is more often associated with consolidations, validated radiographic criteria for reliably distinguishing between viral and bacterial pneumonia do not exist. Furthermore, children, unlike adults, commonly develop viral pneumonia or bacterial-viral coinfections; one large prospective study of children hospitalized with CAP in the USA reported that 66% had a respiratory virus detected and 26% had proven coinfections. Consequently, though physicians are aware that antibiotics do not improve the course of children with viral illnesses, the inability to reliably discriminate between viral and bacterial disease results in the vast majority of children with even non-severe CAP (ie. that can be managed as an outpatient) receiving antibiotics. This may be because clinicians are not aware how much paediatric CAP is viral, but is probably also related to physician beliefs/attitudes. These include the desire to be 'safe', the belief that families want antibiotics, and the perception that it is easier and quicker to prescribe than to explain why antibiotics are not required.
Minimizing inappropriate antimicrobial prescribing is an important strategy to prevent further emergence of circulating antimicrobial resistance. The WHO's Global Action Plan on Antimicrobial Resistance states that 'evidence-based prescribing should be the standard of care'. Optimizing antimicrobial prescribing, also known as 'antimicrobial stewardship', is the main strategy to deal with escalating antimicrobial resistance and has been called 'a fiduciary responsibility for all healthcare institutions across the continuum of care'. Antibiotic stewardship is critical to reducing the dramatic increase in antimicrobial resistance seen in Canada and worldwide, which is why CIHR continues to encourage antimicrobial stewardship research. The Public Health Agency of Canada's 2023 Pan-Canadian Action Plan on Antimicrobial Resistance explicitly calls for research to 'support the development and uptake of diagnostics…that support antimicrobial stewardship, including point-of-care tools'. Antimicrobials routinely prescribed to children can produce problematic side effects and impact normal microbiota, which can influence the development of obesity, atopy, and other disorders.
It has now been definitively established that short course (3-5 day) antibiotic treatment is equally effective as longer courses of antibiotics for non-severe paediatric CAP. The investigators conducted the first of these trials to be published, a multi-year RCT at McMaster Children's Hospital (MCH) and the Children's Hospital of Eastern Ontario (CHEO), enrolling children with non-severe CAP, randomizing to either 5 or 10 days of amoxicillin (duration blinded using placebos), and demonstrated that clinical response at 14-21 days was comparable between groups. Shortening durations further would not seem advisable, given that one trial in Israel found more failures in children given 3 days of antibiotics and that it has been shown that 3 days of antibiotics is superior to placebo for the treatment of paediatric CAP. The investigators believe a better strategy would be to determine how to better identify children with purely viral disease, given that these cases represent the majority of paediatric CAP, and/or enable front-line clinicians to more effectively withhold antibiotics entirely from children at lower risk of bacterial disease. (The investigators note that in these referenced randomized trials and systematic reviews, antibiotic treatment was not associated with large improvements in cure rates; this probably speaks to the fact that only a small proportion of paediatric CAP is bacterial, as reliable discrimination between viral and bacterial disease is not possible on clinical grounds.)
There have been efforts to use biomarkers alone to improve prescribing for respiratory infections. One systematic review found that point-of-care C-reactive protein (POC CRP) testing was associated with significantly decreased antibiotic prescribing for those with acute respiratory infections (not confirmed bacterial disease) presenting to primary care (RR 0.77, 95%CI 0.69-0.86, 12 trials [10218 participants], moderate-certainty evidence). Although the performance characteristics of this test are insufficiently good to recommend its use in isolation, this evidence would suggest that this assay could play a useful role when integrated into a care pathway. Other biomarkers have not been proven to improve prescribing; one Swiss study using procalcitonin to guide antibiotic treatment for paediatric CAP did not find any benefit. However, this may well have been due to the fact that standard prescribing practices at these Swiss centres were already extremely good, as only 56% of children in the control group were given antibiotics. North American clinicians are much more liberal with antibiotic prescribing for CAP; unpublished MCH and CHEO data indicate that 99% of children diagnosed with CAP in the paediatric ED receive antibiotics, and American data show that most children diagnosed with CAP at outpatient clinics and EDs are also treated.
There have been very few trials investigating care pathways to improve antibiotic treatment of children with non-severe CAP in the ED. One cluster-randomized trial enrolled children with respiratory infections to assess the utility of an online digital calculator (developed to predict the likelihood of bacterial infections using many clinical and laboratory variables), where those randomized to the intervention arm were only prescribed antibiotics if the probability of bacterial disease was >10%; unfortunately, only 315/995 participants had CAP and there was no effect observed of the intervention on antibiotic use. Another study done in Italy using a different paediatric CAP care pathway did not even attempt to decrease unnecessary prescribing.
Overall, the effectiveness of care pathways will vary widely depending on how they are constructed and by what constitutes local standard of care. In the PIONEER pilot study that just finished enrolling at MCH, 100% of participants with CAP who received standard care received antibiotics, whereas only 53% of participants with CAP whose care was guided by the novel care pathway (the intervention of interest) were prescribed antibiotics within 14 days. The absolute benefit of our intervention may prove to be substantially larger than that reported for other stewardship interventions aiming to improve antimicrobial use for respiratory infections.
