According to data from the World Health Organization, CAP accounted for 15% of deaths among children under 5 years of age in 2015, and more than 900,000 deaths among children of all ages worldwide.4 Pneumonia is the greatest single cause of infant mortality worldwide.1 In developed countries, however, this rate is virtually zero in children without previous comorbidities.

Regarding hospitalization for CAP, in the Etiology of Pneumonia in the Community (EPIC) study conducted in 2638 children in the US, a hospital admission rate of 15.7 per 10,000 was observed in children under 18 years of age, rising to 62.2/10,000 in children under 2 years of age.5 In European countries, the incidence of consultations in hospital emergency departments for CAP is 14.4/10,000 in children aged 0–16 years and 33.8/10,000 in children under 5 years.2 However, in recent years, since the introduction of pneumococcal vaccination, a significant reduction in hospitalization for this cause has been reported.6,7 A recent study8 conducted in Spain between 2001 and 2014 observed an annual reduction of 3.4% in the pneumonia admission rate in children under 2 years of age. In the same period, these and other authors describe a parallel increase in the frequency of empyema, which they attribute, among other causes, to an increase in the prevalence of different pneumococcal serotypes capable of invading the pleural space that were not included in the 7-valent pneumococcal conjugate vaccine (PCV7), primarily serotypes 19A, 1 and 7 F.9 This increase in the frequency of complications had already been observed in recent decades, even before the introduction of PCV7, so it is possible that other factors, such as the use of antibiotics, antimicrobial resistance, or the epidemiological trend itself, might also have played an important role.

The development of pneumonia in children and its clinical severity are the result of a complex interaction between host and environmental factors.10 In healthy children, especially boys, the greatest risk is to be younger than 5 years of age and, in particular, younger than 2.8 In terms of underlying diseases, asthma is the most common in children hospitalized for CAP, followed by neurological diseases,8 which are associated with higher morbidity and mortality. Other risk factors include: bronchopulmonary dysplasia, congenital heart disease, prematurity, immune deficiency, overcrowding, exposure to environmental pollution, etc.11,12

It has been suggested that certain polymorphisms in genes involved in the innate or specific immune response may be associated with increased susceptibility to developing CAP, as can be seen from the results of a recent meta-analysis in which adult carriers of the Toll-like receptor 4 (TLR4) A299 G polymorphism seemed to present a higher risk. However, these conclusions are not applicable to children, because only 1 pediatric study was included.13

Because of the efficacy of vaccines against pneumococcus and Haemophilus influenzae (H. influenzae) type b, failure to vaccinate should be considered one of the most important preventable risk factors for the development of CAP in childhood.

Host-related factors associated with more clinically severe CAP include: comorbidities, age under 2 years, absence of breastfeeding or duration less than 4 months, malnutrition, and passive smoking.8,14

The epidemiology of CAP is seasonal and closely related to the circulation of the major etiological agents (Table 1).15 In temperate countries, the highest incidence of CAP is recorded in the winter months, coinciding with the peak circulation of respiratory viruses, especially respiratory syncytial virus (RSV).16 It has been suggested that colonization by both pneumococcus and RSV triggers a synergistic effect that favors the pneumococcal infection and the pulmonary inflammatory response, increasing the incidence of CAP, with or without bacteremia.17 Moreover, co-infection with RSV appears to increase pneumococcal virulence and clinical severity.17,18

The association of influenza virus with bacterial pneumonia, particularly when caused by pneumococcus and Staphylococcus aureus (S. aureus), is also an important factor.19

Patients with underlying diseases have a higher incidence of pneumonia and a worse clinical course than healthy children. The etiological spectrum of CAP in this population varies according to the severity of anatomical and physiological alterations and the degree of immunosuppression. In general, the agents that usually cause CAP remain the most prevalent in these patients, but pulmonary infections with gram-negative bacilli, S. aureus and respiratory microorganisms of low virulence, such as non-typeable H. influenzae, are of greater importance.32

In children with more severe immunosuppression, opportunistic pathogens such as oral 〈-hemolytic streptococci, Pneumocystis jirovecii, Legionella pneumophila, cytomegalovirus and fungi should also be considered as possible etiological agents.33

In children with aspiration syndromes, pneumonia is usually caused by oral and oro/nasopharygeal aerobic and anaerobic flora, such as streptococci, Bacteroides and Fusobacterium species, and Prevotella melaninogenica.34

In patients with cystic fibrosis, S. aureus, Pseudomonas aeruginosa, non-typeable H. influenzae, and Burkholderia cepacia are common microorganisms.

A tuberculous etiology should be considered in children who have recently traveled to an endemic area or who have been in contact with a patient with active tuberculosis.35

In summary, age is the parameter that best predicts the etiology of CAP3 in childhood, although other factors, such as immune status, the presence of underlying disease, or the child's vaccination status, are also important (strong recommendation, high evidence).

Since the diagnosis is primarily clinical, an appropriate medical history and physical examination are essential to establish clinical suspicion and guide additional testing.36

Signs and symptoms related to the current disease should be recorded, and any possibly relevant factors (concomitant diseases, vaccination status, recent antibiotic use, day care, travel, exposure to infectious diseases, etc.) should be investigated.

The clinical presentation of CAP varies depending on age, causative agent, patient response, and extent of the disease, and both signs and symptoms are non-specific.37 Clinical manifestations are diverse and may occasionally be imperceptible, particularly in neonates and smaller infants.38

The presence of cough and fever, whether preceded or not by upper respiratory infection, is indicative of pneumonia, especially if associated with tachypnea and increased use of accessory muscles (chest indrawing), expiratory grunt, and nasal flaring.39 Other symptoms, such as chest or abdominal pain, diarrhea, vomiting, or headache due to meningism, sometimes occur, especially in the case of pneumonia located in the upper lobes.

Fever, the common sign of CAP in children, is a variable that has relatively low sensitivity and even less specificity.37 Although a pattern of lower, or absent, fever continues to be noted in pneumonia caused by viruses or atypical bacteria (C. pneumoniae, M. pneumoniae), compared to the high fever associated with a probable bacterial or mixed etiology (the latter being related with greater severity), the clinical value of this differentiation is not well established.37

Like hypoxemia or intercostal chest indrawing, the presence of tachypnea (Table 2)40 may be a predictor of pneumonia in children, mainly in children under 3 years of age. A recent systematic review of 23 studies in children under 5 years of age concluded that respiratory rates (RR) higher than 40 breaths per minute are associated with an increased likelihood of radiological pneumonia, although the ability of RR to accurately distinguish children with and without pneumonia is very limited (likelihood ratio [LR] 1.5; 95% confidence interval [CI], 1.3–1.7).37 The severity of pneumonia (reflected by the level of hypoxemia) also correlates with the grade of tachypnea,41 especially in infants under 1 year of age with an RR greater than 70 bpm.42 In any case, it should be remembered that RR depends not only on age, but also on other factors, such as temperature, stress and anxiety, sleep, wakefulness, etc., and that, in the absence of tachypnea, the probability of pneumonia is very low.37

Chest pain is a symptom usually reported by older children; in isolation it has little value in the diagnosis of pneumonia and is often associated with pleuritis or pleural effusion. The same applies to fever and cough as isolated parameters. Dry or productive cough, depending on the time since onset of the disease, may be absent at the beginning of pneumonia. When symptoms of upper respiratory tract infection associated with generalized wheezing and low fever predominate, bacterial pneumonia is unlikely.43

With regard to the physical examination, special attention should be paid to general condition, skin color (cyanosis, paleness), and respiratory distress.44

Bacterial pneumonia, as in the case of M. pneumoniae, may be accompanied by other symptoms,45 including cutaneous involvement and mucositis.46,47

Auscultation continues to be an essential component of the physical examination,48 although studies indicate that its findings do not correlate closely with radiological pneumonia, and may be normal in children. The subjectivity and technical difficulty involved in this procedure, especially in infants, probably contribute to its low diagnostic yield. The most common findings in pulmonary consolidations include rales, crackles, hypophonesis, and bronchial breath sounds. Wheezing, in addition to fine crackles in the bronchi, indicates bronchial involvement typical of bronchopneumonia, and points towards a viral origin or the presence of atypical bacteria such as M. pneumoniae or C. pneumoniae.44

Hypoxemia, often present without cyanosis, correlates positively with severity.37 Low levels of arterial oxygen saturation (SaO2) along with signs of increased work of breathing (chest retraction, nasal flaring, etc.) are the findings most robustly associated with the diagnosis of pneumonia.1 However, the diagnostic utility of these parameters is limited, as the LRs are less than 5 for both hypoxia and work of breathing: the LR for SaO2 ≤ 96% is 2.8 (95% CI, 2.1–3.6), with 64% sensitivity and 71% specificity,1 while the LR for SaO2 ≤ 95% is 3.5 (95% CI, 2.0–6.4) with lower sensitivity but higher specificity (96%).49 The LR for increased work of breathing is 2.1 (95% CI, 1.6–2.7).37

From a clinical point of view, CAP can be classified by severity (Table 3),2,50 and also as typical or atypical, albeit with exceptions, by combining clinical, radiological, and analytical characteristics (Table 4).51

The diagnosis of CAP and its possible etiology depend on the integral assessment of the child's age, clinical and radiological signs, and the sum of certain clinical laboratory findings.

Blood tests should not be performed routinely, as the white blood cell count and differential are not useful for determining the etiology of pneumonia.1 White blood cell counts greater than 15,000/mm3 with left shift are not clearly associated with a bacterial etiology, although C-reactive protein (CRP) values > 60−80 mg/mL do suggest this etiology.52

Procalcitonin (PCT) may be a potentially useful marker for making therapeutic decisions in emergency units2 or for guiding treatment.53 In fact, values < 0.25 ng/mL would rule out typical bacterial CAP (negative predictive value: 96%) and would help identify children who do not need antibiotic treatment.54 Some studies report that levels ≥ 1 ng/mL increase by up to 4-fold the probability of bacterial pneumonia55 and levels greater than 2 ng/mL have a specificity of 80% in predicting this etiology.56 However, there is no consensus on the usefulness of these biomarkers in determining microbial etiology in CAP,57 and there is no evidence that the use of a combination of these parameters (leukocytes, neutrophils, erythrocyte sedimentation rate [ESR], CRP and PCT) will increase sensitivity or specificity.58,59 PCT correlates positively with the severity of CAP2,60 and may be an indicator of the risk of bacteremia,60 but it is not useful in uncomplicated CAP.2

Other markers are being investigated61–64 and their clinical usefulness remains to be proven in coming years.

A definitive etiological diagnosis of CAP can only be established by isolating a pathogenic microorganism from a normally sterile fluid (blood, pleural fluid, lung biopsy). In children, this is only achieved in 30%-40% of cases and in less than 10% at disease onset, when this information would be useful for determining treatment. As a result of this low sensitivity, together with the poor cost-benefit ratio and the difficulty in obtaining adequate samples, microbiological testing not routinely recommended in previously healthy children with mild-moderate CAP who can be treated as outpatients.2

Rapid molecular diagnostic techniques (multiple PCR) are now available, significantly increasing diagnostic sensitivity in blood or pleural fluid samples,65 and, in the case of S. pneumoniae, distinguishing different serotypes involved in the disease. However, the results of multiplex panels should be interpreted with caution because they do not distinguish colonization from infection,66 nor can they be used alone to rule out a bacterial infection, given the possibility of concomitant viral-bacterial infection, particularly in more severe cases.67 Microbiological testing should be reserved for the following situations: pneumonia associated with bacteremia or pleural effusion; immunosuppression or immunosuppressive treatment; moderate-severe or slow-progressing pneumonia; and epidemic outbreaks.1 In these cases, the diagnostic tests to be used would be2: blood culture; Gram stain and sputum culture, PCR detection of virus or immunofluorescence in nasopharyngeal secretions or nasal swabs; serology, for respiratory viruses, M. pneumoniae, and C. pneumoniae, and, if a pleural fluid sample is available, direct microscopy, culture (low sensitivity), detection of pneumococcal antigen (sensitivity and specificity > 90%), and PCR for S. pneumoniae.

A predominant microorganism or intracellular organisms detected in an appropriate sputum sample (≤ 10 epithelial cells and ≥ 25 polymorphonuclear leukocytes [⋅100]) is indicative of the etiological agent.68 In these cases, Gram stain has a sensitivity and specificity for pneumococcus of 85% and 62%, respectively. However, the diagnostic yield of induced sputum is not sufficient to include this procedure in the routine diagnosis of CAP.69Nasopharyngeal aspirates or exudate are used for the detection of Bordetella pertussis and respiratory viral antigens (enzyme immunoassay and immunochromatography), but not for bacterial cultures, as the causative bacteria may be the same as those of the normal oropharyngeal flora.70 In hospitalized children, the study of respiratory viruses in nasopharyngeal secretions may be helpful (strong recommendation, high evidence).

The same is true of the detection of pneumococcal urinary antigens, which can be positive in children < 4−5 years old, pneumococcal carriers, or patients who have recently received a pneumococcal vaccine.1,2,71,72 After this age, pneumococcal urinary antigen has a similar value to adults72 and may be useful as a negative predictor of pneumococcal infection in older children.2 Detection of Legionella antigen in urine should be considered in cases with clinical and epidemiological suspicion of legionellosis.

Serological diagnosis should not be included as a routine test since it requires the comparison of 2 samples at an interval of 2–4 weeks (acute phase and convalescence), so is not helpful in decision-making, and the seroconversion of some pathogens is also difficult to assess. Even so, the British guidelines propose storing serum samples in the acute phase and taking a sample during convalescence in cases where a microbiological diagnosis was not reached during the acute stage, and recommend this procedure for the diagnosis of atypical and respiratory virus infections.2

Other tests, such as the tuberculin test or Interferon-Gamma Release Assays (IGRA), are not recommended in all pneumonias. They are justified only in the case of clinical or epidemiological suspicion of tuberculosis, refractory pneumonia, or in marginal settings, and travel and migration of the population from areas of high prevalence in tuberculosis. Table 4 shows etiological, clinical and laboratory correlations, and Table 5 shows the microbiological tests proposed at each level of care.

Studies to determine the etiology of community-acquired pneumonia in children only establish the cause in 20%-50% of cases2 (moderate recommendation, moderate evidence).

Chest X-ray is the gold standard for establishing a diagnosis of pneumonia, but since it does not modify therapeutic decisions or improve clinical outcomes,73 it is not mandatory in clinical guidelines for CAP,1,2 and may be dispensed with in previously healthy children with a first episode of pneumonia and no severity criteria, or in children with fever but no tachypnea, unless warranted by the patient data3 (strong recommendation, high evidence). Indications are given in Table 6.12

In most cases, frontal projection is usually sufficient for the diagnosis of CAP.2 Lateral projection increases the radiation dose and does not usually provide significant information. As such, it should be reserved for inconclusive cases, diagnostic doubts, or suspected lymphadenopathies1,2 and not routinely performed (strong recommendation, high evidence). The radiological expression of pneumonia is consolidation or pulmonary parenchymal infiltrate, with 2 characteristic radiologic patterns (alveolar and interstitial) that, while conventionally associated individually with a type of infection (bacterial/viral or M. pneumoniae), are not unique to any specific etiology.55 The alveolar pattern is characterized by lobar or segmental consolidation, with or without air bronchogram or alveologram. The interstitial pattern is caused by bilateral, diffuse and irregular parahiliar infiltrates, air trapping or segmental atelectasis due to mucosal plugs and peribronchial thickening, and can be observed in viral pneumonias or those caused by M. pneumoniae, C. pneumoniae and Legionella sp. There is also a mixed pattern, which combines the above characteristics and is another, not uncommon, form of presentation of CAP.74

An image of “round pneumonia” is indicative of pneumococcal infection73 and the presence of pneumatoceles, cavities, or large pleural effusion is associated with bacterial pneumonia. The first radiological sign of effusion is usually the occupation of the costophrenic angle, which, when it appears, already indicates the existence of a major effusion; therefore, if this complication is suspected an ultrasound should be performed to evaluate the amount of fluid present and decide on the action to take.2

In summary, analytical and radiological studies cannot be used to establish the viral or bacterial etiology of pneumonia in children with absolute certainty. In routine clinical practice, the results of microbiological and laboratory tests should always be interpreted together with the medical history, physical examination findings, and chest X-ray75 (moderate recommendation, moderate evidence).

Current evidence supports ultrasound as a useful imaging alternative for the diagnosis and follow-up of pneumonia in children76–79 (moderate recommendation, moderate evidence), because it is readily available, quick to learn, easy to perform, affordable, and radiation-free.80 Some authors believe that it should take precedence over radiography.81 Two recent meta-analyses80,82 confirm its high sensitivity (93%-94%), specificity (93%-96%), and area below the curve (0.98). Ultrasound provides data on the lung parenchyma: bronchogram (distorted or normal), homogeneity or heterogeneity of consolidation, areas with avascularity or low echogenicity due to necrosis, vascularized wall areas associated with the formation of abscesses, etc.77,83 It is much more sensitive than radiography for confirming minimal effusions84–86 and provides more information than computed tomography (CT) in terms of the amount, nature of the effusion (septated or not), and location of the puncture site, if necessary.

CT may be indicated if complications are suspected (severe or complex pneumonia, pneumonia in immunocompromised patients, antibiotic-refractory pneumonia, recurrent or unresolved pneumonia, patients with clinical suspicion of pneumonia but normal or questionable chest X-ray findings, and pneumonia with suspected underlying diseases87), or if there is difficulty in differentiating CAP from other diseases.86 It is very useful for assessing the pulmonary parenchyma, and defines necrotic lesions, cavities, pneumatoceles, abscesses or bronchopleural fistulas with great precision. It also complements the ultrasound evaluation of empyema, locating the drainage tube, and evaluating possible parenchymal reexpansion failures after tube placement.86,88

Although recent studies suggest that magnetic resonance imaging (MRI) is slightly more sensitive than X-ray in detecting pneumonia, its high cost, limited availability, and the meagre radiation dose that would be avoided if used instead of X-ray rule it out as a first-line diagnostic option.88,89 It may be considered in children with complicated pneumonia, replacing repeated CT scans as a follow-up test, or in patients with chronic diseases who will undergo multiple CT scans throughout their life, as a method for evaluating the lungs without increasing the radiation dose.88

Fiberoptic bronchoscopy (FB) is reserved for cases of severe or potentially severe CAP, slow progress, or persistent radiological abnormalities, and for children with recurrent pneumonia in the same site.90 It is also indicated in oncological or immunocompromised patients who do not respond adequately to initial treatment, and in whom the causative agent must be identified.91 In these patients, the yield is higher (80%) than in immunocompetent children, in whom the isolation rate may be higher if the procedure is performed early.92

FB should always be accompanied by bronchoalveolar lavage (BAL) at one or more sites for cell culture and analysis.92 The isolation of certain microorganisms, such as Mycobacterium tuberculosis (M. tuberculosis), RSV, influenza virus or M. pneumoniae, is sufficient for them to be considered the etiological cause of the pneumonia, although they may not be the only causative agents involved. The detection of other bacteria (S. pneumoniae, H. influenzae, etc.) that typically occur in the oropharyngeal flora could indicate contamination, and in these cases quantitative cultures should be used, with a colony-forming unit per milliliter (CFU/mL) ≥ 104 indicating infection. Bronchial brushing has a lower diagnostic yield than BAL, so its use in the pediatric setting has declined.93

Fig. 1 shows the diagnostic algorithm for suspected CAP.

Fig. 1.

Diagnostic algorithm for suspected community-acquired pneumonia.

(0.11MB).

After pneumonia has been diagnosed, its severity must be evaluated and decisions taken on whether hospitalization is needed. In uncomplicated cases, 90% of patients will become afebrile within 48 h of starting antibiotic treatment, and only a small proportion will require hospital admission (Table 7).50,94 In the community setting, the pediatrician should perform a clinical assessment 48 h after pneumonia has been diagnosed and treatment started.1,2

If, contrary to expectations, there is no significant improvement after 48−72 hours of treatment (persistence of fever, worsening general condition, dyspnea, etc.), the reasons for this should be considered, such as incorrect diagnosis, ineffective treatment (poor compliance or incorrect dose, presence of resistance), appearance of complications (necrotizing pneumonia, parapneumonic effusion, pulmonary abscess), viral etiology or a rarer causative microorganism (M. tuberculosis, Actinomyces sp., fungi, protozoa), previously undiagnosed immunodeficiency, non-infectious cause and associated bronchial obstruction, underlying process such as congenital lung malformation, or non-infectious disease: pulmonary hemorrhage, pulmonary edema, diaphragmatic hernia, eosinophilic pneumonia, organizational pneumonia, pulmonary thromboembolism. All these possible causes must be evaluated and explored (Table 8).50,94

In clearly deteriorating cases, hospitalization is necessary to administer intravenous (IV) antibiotic treatment and any necessary support, and to perform complementary examinations to evaluate the existence of potential complications and clarify the etiology (new chest X-ray, ultrasound, blood tests, ESR, PCR, blood culture, viral antigen detection, serology, tuberculin testing, FB, BAL, CT, etc.).50

If there is no clear deterioration, but no significant improvement is observed within 72−96 hours, antibiotic resistance, presence of another pathogen, or a non-infectious cause of pneumonia must be considered. If resistance to antibiotics or other pathogens is suspected, the administration of another antibiotic to expand coverage against S. pneumoniae and atypical bacteria must be considered. There is no specific waiting time, but if after 72 h there is no clear improvement, the above options should be considered.94

However, with proper treatment, clinical progress is favorable in most cases.95 Residual cough may persist for a few weeks, especially after viral or M. pneumoniae-associated CAP.96 Recovery is usually complete and without sequelae in previously healthy children so, in general, no clinical laboratory monitoring is necessary when the course is normal and the patient remains asymptomatic.97

It is important to remember that although clinical resolution is usually rapid in most cases, radiological changes usually take 3−7 weeks to normalize. The various guidelines and consensus documents specify that routine radiological follow-up is not necessary in patients who remain asymptomatic after uncomplicated CAP.97 In contrast, a follow-up X-ray is justified in the following cases: persistent symptoms, history of recurrent pneumonia, presence of atelectasis, round pneumonia, empyema, pneumatocele, pulmonary abscess, or other concomitant disease.1,2,50,94,98

If the pneumonia required hospital admission (Table 7), follow-up after discharge may be conducted by the primary care pediatrician or in the hospital, depending on the reason for admission, severity, complications, etc.

With regard to clinical course, 2 other possibilities in addition to rapid cure must be taken into account, namely slowly resolving pneumonia, in which the cure process is slower than usual, but clinical and radiological normalization is finally achieved, and unresolved or persistent pneumonia, defined as clinical or radiological symptoms persisting over a certain period (≥ 1 month), despite the administration of antibiotic treatment for 10 days.99,100 Few authors specify the time after which pneumonia should be considered persistent, but this ranges from 1 to 3 months. The real incidence is unknown since cases of persistent and recurrent pneumonia overlap in the published series.101,102

In the case of slowly resolving or persistent CAP, once normal prolonged radiological resolution has been ruled out, the possible causes of a hypothetical anti-infectious treatment failure must be raised, without forgetting the possibility of a non-infective etiology of a persistent radiological image103–106 (Table 8). In all these cases, FB, with or without BAL, combined with other studies (immunological, imaging, etc.), has been shown to help elucidate the underlying cause.106,107

Empirical treatment of CAP and its complications is established according to the common pathogens. However, one of the most important problems is the correct distinction between cases of probable viral etiology and cases of probable bacterial etiology. In general, antibiotic treatment is indicated in cases of typical CAP where bacterial etiology is suspected. In cases of atypical CAP, antibiotic therapy will only be indicated in patients older than 4−5 years, whereas in children under this age antibiotic treatment will not be prescribed unless they meet some of the criteria described below.115

If the decision is taken to start outpatient antibiotic treatment in typical CAP, taking into account that most are caused by S. pneumoniae, almost all of which are currently susceptible to penicillin and amoxicillin, the antibiotic of choice is amoxicillin PO, at a dose of 80 mg/kg/day, every 8 h (Table 9)114,116–119 (strong recommendation, high evidence). The use of amoxicillin-clavulanic acid PO (80 mg/kg/day) would only be justified in patients with underlying diseases predisposing them to a wider range of bacteria, pulmonary aspiration, or incomplete H. influenzae type b vaccination (strong recommendation, high evidence). The recommended duration of treatment in a patient with a typical uncomplicated CAP that does not require admission is a maximum of 7 days.119

In the case of atypical CAP in children under 4−5 years of age, etiology is usually viral, so in principle antibiotics will not be prescribed. In those older than 4−5 years, in whom an etiology of M. pneumoniae is rare and C. pneumoniae even rarer, the use of macrolides PO is recommended120 (Table 9) (strong recommendation, high evidence).

If a properly vaccinated patient needs to be admitted to hospital, and pneumococcal etiology is assumed or confirmed, the antibiotic of choice in typical CAP is currently high-dose penicillin G sodium or ampicillin IV, given the excellent tolerance (Table 9)112 (strong recommendation, high evidence).

In children under 6 months of age with typical CAP who will not yet have completed their initial Hib vaccination, admission is recommended depending on their general condition, degree of hypoxemia, etc., for IV treatment with amoxicillin-clavulanic acid or cefuroxime.114 In children under 3 months, empirical treatment consists of ampicillin and cefotaxime.114

In cases associated with influenza, the most frequent superinfection is S. pneumoniae and, to a lesser extent, S. aureus, S. pyogenes and H. influenzae, so empirical therapy with amoxicillin-clavulanic acid may be appropriate. MRSA co-infection has been found to be more severe, so when a patient with concomitant influenza and severe pneumonia is admitted to the pediatric ICU, empirical use of clindamycin (or vancomycin, depending on local susceptibility data) together with a cephalosporin (cefuroxime or cefotaxime) may be appropriate. In immunocompromised patients or patients with other risk factors who develop severe respiratory infection due to the influenza virus, treatment with an antiviral (usually oseltamivir) should be initiated early, given the greater severity of influenza virus infections in hosts this of type.120

If aspiration is suspected, amoxicillin-clavulanic acid is recommended.115

Complicated forms of CAP, such as necrotizing pneumonia and pulmonary abscess, are usually initially treated empirically with ampicillin or cefotaxime combined with clindamycin for 4 weeks, or for at least 2 weeks after fever has disappeared118 and, if the etiological agent is known, the antibiotic therapy may be subsequently adjusted.

In case of pleural empyema, a dose increase is recommended to reach higher concentrations in the pleural space (Table 9).114 When the patient has remained afebrile for 24−48 hours and once the pleural tube has been removed, treatment can be switched to the oral route.116 The total duration of treatment is usually 2–4 weeks.118

Video-assisted thoracoscopy (VATS) is useful for staging the effusion, rupturing septae, draining fibrinopurulent material, reducing the bacterial load in the early stages, and placing the drainage tube in the correct position.120 It is also useful for visualizing the appearance of the underlying lung, its expansion capacity, and the location of possible bronchopleural fistulas.120 The complication rate, including air leak or persistent pneumothorax, pneumatocele, or bleeding, is low (6%-7%).141

VATS is used in some centers as initial treatment of moderate or large PPEs,142 since, compared to drainage alone, it reduces the duration of fever, which usually resolves 24−72 h after the procedure, and shortens hospital stay,126,143 although it is more commonly used in the treatment of cases in which drainage with fibrinolytics has failed (moderate recommendation, moderate evidence) and in the treatment of complications, such as bronchopleural fistulas.

There is no universal consensus on the treatment of parapneumonic effusion, since in many respects the evidence is weak, partial, or non-existent.126 In our setting, mortality is minimal in children and the long-term prognosis is excellent. The different therapeutic approaches are proposed to optimize health resources, including shortening hospital stay.

For the management of effusions, it is useful to classify them as144:

• -

  Small: < 10 mm on lateral X-ray, or < 1/4 of hemitorax on anterior-posterior (AP) X-ray.

• -

  Moderate: > 1/4 and ≤ 1/2 on AP X-ray.

• -

  Large: > 1/2 on AP X-ray.

Small effusions respond well to antibiotic treatment without the need for drainage. If significant respiratory compromise or uncontrolled sepsis occurs despite antibiotic treatment, interventional therapy is indicated.140,144 Most guidelines1,120,129 recommend pleural drainage associated with fibrinolytics or VATS in all cases of moderate and large effusion, since the duration of treatment and hospital stay is considerably longer in cases treated with antibiotic alone or with drainage alone120,129(strong recommendation, high evidence). However, in recent series,131,144,145 a percentage of children with moderate (50%) or large (25%-30%) effusion showed good resolution with antibiotic treatment alone, without the need for drainage and without prolonging hospital stay,144 suggesting that there may be a subgroup of children with moderate or large effusions in whom interventional treatment may not be necessary. To our knowledge, these approaches − antibiotic-only treatment, initial antibiotic treatment and intervention at 2−3 days in case of lack of response, or initial interventional treatment − have not been compared in any clinical trial.

In a retrospective study, delays of 1−2 day in chest drainage placement were associated with lower mortality than delays of more than 7 days.145

There is good evidence from clinical trials in children to support interventional treatment where this approach is needed or selected. The largest study was a multicenter study conducted in Spain, with 50 patients in the VATS group and 53 in the fibrinolytic group,132 while another 2 studies have been performed, 1 in the United Kingdom (UK) and 1 in the US, with a maximum number of 30 children per arm.146,147 None of these 3 studies revealed any differences between the 2 groups in terms of hospital stay or complications (strong recommendation, high evidence). The cost of fibrinolysis was lower than that of VATS, 35% less in the US study and 20% less in the UK. Therefore, given the greater simplicity of fibrinolytic administration that does not require surgical intervention, this approach is proposed as the first option before the use of VATS,1,120,143 which should be reserved for non-responders.

A meta-analysis140 revealed no differences in perioperative complications between the use of fibrinolytics and VATS. The need for reintervention and the duration of postoperative hospital stay were both lower in the VATS group (–0.7 days). However, this meta-analysis includes a clinical trial that compared VATS with drainage alone, and 4 observational studies, which, along with the small differences in hospital stay, does not warrant modifying the previous recommendation.140

A recent Spanish multicenter study (Spanish Multicenter Study on the Use of Dexamethasone in Parapneumonic Pleural Effusion [CORTEEC])148 included 60 children who were randomized to receive antibiotic treatment at baseline, combined with either dexamethasone (0.25 mg/kg/6 h IV) or placebo for 2 days. In children who received dexamethasone, the time to recovery was shorter (median 2.8 days), but the risk of hyperglycemia was higher. In just over 90% of cases, effusions were small (<1/3 hemitorax), so the results cannot be extrapolated to patients with larger effusions.

In summary, the choice of treatment for parapneumonic effusions in hospitalized children appears to depend largely on the preferences of physicians based on their experience and the resources available.143 In the US, the use of VATS peaked about 10 years ago, and this technique is gradually being replaced by pleural drainage and fibrinolytics.143

In a survey conducted in Spain on the best approach to initial treatment in children with parapneumonic effusion or empyema,149 44% respondents preferred chest drainage with fibrinolytics; 18%, antibiotics only; 17%, simple drainage; and 17%, VATS. Overall, 54% of the centers said they followed the guidelines published in Spain. Another survey conducted in Central European countries also revealed a complete lack of agreement on the therapeutic approach to parapneumonic effusions in children.138 There is also a lack of consensus on the use of diagnostic thoracentesis in cases where no chest tube is placed; in the European survey, it was only used in 3% of cases138 and in the Spanish survey it was used in 46%.149

Multicenter clinical trials are thus needed to provide more evidence for the management of parapneumonic effusions in children and to develop evidence-based algorithms.137

Fig. 2 shows the algorithm recommended by this consensus document, following the best available evidence and expert agreement.

Fig. 2.

Algorithm for significant parapneumonic pleural effusion.

a Dexamethasone 0.25 mg/kg/6 h for 2 days at the beginning of treatment may reduce recovery time in small effusions. b For moderate parapneumonic effusion without severe symptoms, the options are to administer antibiotic treatment only, to which approximately 50% of patients will respond, or to place a chest tube and administer intrapleural fibrinolytics. If pleural drainage is not selected, diagnostic thoracentesis may be an option. c Worsening or treatment failure: development of respiratory failure or persistent tachypnea, fever, or deterioration of general condition within 48−72 hours of starting antibiotic therapy (duration of fever should not be the only criterion in cases of antibiotic treatment without drainage). dOther causes of poor progress may include the presence of necrotizing pneumonia or the development of a pulmonary abscess.

(0.16MB).

Vaccination against certain microorganisms has been shown to have an impact on the incidence and mortality of CAP. The etiologic agents for which vaccines are available are S. pneumoniae, H. influenzae type b and influenza virus.150

The inclusion of pneumococcal conjugated vaccines, both PCV13 and PCV10, in the official pediatric immunization schedule in more than 150 countries worldwide has shown a moderate reduction in cases and hospitalization for CAP in children, especially in children under 5 years of age150–158 (strong recommendation, high evidence), although the data on the impact of group immunity on the incidence of CAP in adults, including unvaccinated individuals over 65 years of age, is conflictin.153–156 In Spain, data have not been published since the introduction of routine vaccination with PCV13 in 2016, but experience in the Community of Madrid has shown that systematic vaccination of children reduces the incidence of CAP, including complicated forms, caused by serotypes contained in PCV13.158 In countries such as France, which has greater experience with routine PCV13 vaccination, pneumococcal pneumonia in children has reduced significantly, with S. pyogenes currently being the most frequent cause of parapneumonic pleural empyema (45%).159

The AEP Vaccine Advisory Committee recommends systematic H. influenzae type b and PCV13 pneumococcal vaccination in all children under 5 years of age.160–163 Moreover, children in risk groups should receive the influenza vaccine after 6 months of age and the 23-valent unconjugated vaccine after 24 months of age.160–163