To the best of our knowledge, there is no data available comparing the risk factors, outcomes, and microbiological characteristics of severe CAP and VAP in the paediatric population. In this study, we have analysed data from the last five years on children admitted to the PICU with pneumonia, both CAP and VAP. It has revealed some data that is of interest to clinicians, since these are both associated with a high mortality rate, especially VAP, in which mortality can reach up to 20% 25.
The main objective of the present study was to analyse the differences between CAP and VAP in children. According to these results, there are important differences between both diseases, since patients with VAP had worse prognosis than patients with CAP, needing longer mechanical ventilation, more inotropic support and had higher mortality. Moreover, patients with VAP were mainly infected by Enterobacteriaceae and had more multidrug resistant microorganisms.
Patients that developed VAP were more prone to having comorbidities than patients with CAP. This is consistent with other studies in adults, in which underlying diseases and comorbidities have been described as risk factors for developing VAP 25. In addition, patients that developed VAP were more likely to have a tracheostomy than those with CAP. Since tracheostomised children are at an increased risk of developing pneumonia, it is important to implement preventative care bundle measures in these vulnerable patients when they’re admitted to the PICU 26.
Moreover, patients with VAP were admitted due to a cardiovascular disease in the 20% of the cases, which was higher than in patients with CAP. The vulnerability of patients with cardiovascular diseases has been reported in other studies 27, as they are more susceptible to developing VAP and, additionally, nosocomial infections in these patients are an important cause of morbidity and mortality. In some studies performed on adults, they even propose decontamination or pre-emptive antibiotic therapy in order to prevent the development of VAP 28,29.
While patients with CAP were diagnosed with infection at admission, patients that ended up with VAP received this diagnosis later (after 11 days). This fact highlights that even patients that do not develop pneumonia initially are at a high risk of developing it if they are on MV. For this reason, strategies to prevent VAP such as the elevation of the head of the bed, regular oral care, maintaining ventilator circuits, the use of cuffed endotracheal tubes, and minimising the duration of MV are highly recommended 30.
Patients with VAP had higher inotropic requirements, longer MV, a longer length of stay, and required more extracorporeal support than patients with CAP. Furthermore, the mortality was higher in patients with VAP than in patients with CAP. Kollef et al. described in a recent multicentre study that patients with VAP seem to have worse outcomes than patients with CAP, in terms of mortality rate and length of stay 31. Therefore, even if both types of patients are in a critical care unit and on MV, we should be especially concerned about the ones with VAP, since they are more likely to require more support and have a poorer outcome.
As for the microbiological data, remarkable differences were found between the two groups. The bacteria most frequently involved with CAP were gram negative species, especially Haemophilus spp, followed by gram positive species such as S. pneumoniae and S. aureus. In children, CAP is usually caused by a virus, followed by S. pneumoniae and others like Haemophilus spp and S. aureus 32. In recent years, the development of vaccines against S. pneumoniae (the 13-valent pneumococcal conjugate vaccine) and Haemophilus influenzae type B has helped to decrease the incidence of CAP related to these microbes, especially severe cases of CAP 21. Furthermore, after the implementation of these vaccines, other changes in the epidemiology have been revealed: non-vaccine serotypes have been isolated in very few cases but the other hand, an increasing prevalence of CAP with viral involvement has been described 33,34. Despite this, the cases of CAP related with S. pneumoniae and Haemophilus spp are still relevant, mainly due to persistent inequities in access to healthcare, especially in low and middle-income countries 35.
In contrast, VAP was caused mainly by Enterobacteriaceae. In adults, it has been widely described how Enterobacteriaceae are involved in a high percentage of VAP cases 36. In our study, the specific microorganisms most frequently involved with VAP were Pseudomonas aeruginosa, followed by Enterobacteriaceae. As previously reported, MV duration and length of PICU stay were significantly longer in the group with VAP. Chomton et al. found that the median MV duration at VAP diagnosis was longer for VAP due to nosocomial microorganisms such as P. aeruginosa or E. coli when compared with VAP due to community-acquired bacteria such as H. influenzae and S. pneumoniae 10. This fact was also explained by Kollef et al.; oropharyngeal and tracheal colonisation with Pseudomonas and enteric gram-negative bacilli increases in step with length of hospital stay and severity of illness 37.
One of the major concerns worldwide nowadays is the increasing prevalence of multidrug-resistant microorganisms and the lack of new antimicrobial agents for use in paediatric pneumonia 3. Patients with VAP required more days of antibiotic treatment than patients with CAP. Moreover, they were more likely to need their antibiotics to be switched. This is probably related to the differences in the type of microorganism the pneumonia involves. While CAP is usually caused by drug-sensitive bacteria and empirical treatment is normally sufficient, in VAP, as reported, there are more MDR and ESBL bacteria, and patients therefore more commonly need broad-spectrum antibiotic treatment. This observation is consistent with the results of other studies, in which ESBL has been related with VAP 38. A significant morbidity and mortality associated with inadequate or delayed antibiotic treatment is reported in adults, so in children it can be assumed that adequate antibiotic use is also a key prognostic factor 10.
We acknowledge several limitations in this study. Patients admitted to the Neonatal ICU (<1 month) were not included (because the neonatal and paediatric ICUs are two separate units with a high inflow of patients), and therefore we do not have results for the neonatal population. In addition, it is a single-centre study, so the results might be difficult to extrapolate to other populations. However, for a paediatric study, it has quite a large number of patients, so the results may be useful to other PICUs.
Despite these limitations, to our knowledge this is the first study comparing the risk factors and outcomes of severe CAP and VAP in children. Therefore, we believe that it provides valuable information on the paediatric population.
To sum up, children that develop VAP seem to be more vulnerable than those with CAP, because they had a higher proportion of comorbidities and they had an increased prevalence of cardiovascular diseases. Moreover, patients with VAP required more inotropic support, longer MV, and had longer hospitalisation times than patients with CAP, and they ended up having a higher mortality than patients with CAP. In light of this, strategies to prevent nosocomial infections should be carefully executed in order to avoid VAP, since it worsens patients’ prognosis. Finally, considering that VAP is usually related to Enterobacteriaceae, the correct antibiotic treatment should be implemented as soon as signs of infection appear so as to improve the outcome of the patient.