Pneumonia in children: Inpatient treatment (Update)
William J Barson, MD
Morven S Edwards, MD
George B Mallory, MD
Mary M Torchia, MD
All topics are updated as new evidence becomes available and our peer review process is complete.
Literature review current through: Nov 2017. | This topic last updated: Dec 01, 2017.
Community-acquired pneumonia (CAP) is defined as an acute infection of the pulmonary parenchyma in a patient who has acquired the infection in the community, as distinguished from hospital-acquired (nosocomial) pneumonia. CAP is a common and potentially serious illness with considerable morbidity.
The inpatient treatment of CAP and hospital-acquired pneumonia in children will be reviewed here. The outpatient treatment of CAP is discussed separately, as are the epidemiology, etiology, clinical features, and diagnosis. (See “Community-acquired pneumonia in children: Outpatient treatment” and “Pneumonia in children: Epidemiology, pathogenesis, and etiology” and “Community-acquired pneumonia in children: Clinical features and diagnosis”.)
The recommendations provided below are largely consistent with practice guidelines provided by The Pediatric Infectious Diseases Society/Infectious Diseases Society of America and the British Thoracic Society [1,2].
Indications — The decision to hospitalize a child with community-acquired pneumonia (CAP) is individualized based upon age, underlying medical problems, and clinical factors including severity of illness (table 1) [1-3]. Hospitalization generally is warranted for infants younger than three to six months of age, unless a viral etiology or Chlamydia trachomatis is suspected and they are not hypoxemic and relatively asymptomatic. Hospitalization is also warranted for a child of any age whose family cannot provide appropriate care and assure compliance with the management plan. Additional indications for hospitalization include [1,2]:
- Hypoxemia (oxygen saturation [SpO2] <90 percent in room air at sea level)
- Dehydration, or inability to maintain hydration orally; inability to feed in an infant
- Moderate to severe respiratory distress: Respiratory rate >70 breaths/minute for infants <12 months of age and >50 breaths per minute for older children; retractions; nasal flaring; difficulty breathing; apnea; grunting
- Toxic appearance (more common in bacterial pneumonia and may suggest a more severe course) 
- Underlying conditions that may predispose to a more serious course of pneumonia (eg, cardiopulmonary disease, genetic syndromes, neurocognitive disorders), may be worsened by pneumonia (eg, metabolic disorder) or may adversely affect response to treatment (eg, immunocompromised host)
- Complications (eg, effusion/empyema, abscess)
- Suspicion or confirmation that CAP is due to a pathogen with increased virulence, such as Staphylococcus aureusor group A Streptococcus
- Failure of outpatient therapy (worsening or no response in 48 to 72 hours)
Indications for intensive care — The decision to treat a child with pneumonia in an intensive care setting is individualized, based upon clinical, laboratory, and radiologic findings. Treatment in an intensive care setting generally is warranted for children who manifest [1,2]:
- The need for ventilatory support beyond that which can be provided outside the intensive care unit (eg, mechanical ventilation, noninvasive positive pressure ventilation, failure to maintain oxygen saturation [SpO2] >92 percent in FiO2>0.5)
- Signs of impending respiratory failure (lethargy, increasing work of breathing, and/or exhaustion with or without hypercarbia)
- Recurrent apnea or slow irregular respirations
- Cardiovascular compromise with progressive tachycardia and/or hypotension that requires or is refractory to fluid management
Care in the intensive care unit also may be warranted for children with two or more of the following :
- Respiratory rate >70 breaths/minute for infants <12 months of age and >50 breaths/minute for older children
- Increased work of breathing (retractions, dyspnea, nasal flaring, grunting)
- PaO2/FiO2ratio <250
- Multilobar infiltrates
- Altered mental status
- Pleural effusion
- Comorbid condition (eg, sickle cell disease, immune deficiency, immunosuppression)
- Unexplained metabolic acidosis
- Pediatric Early Warning Score >6 
Infection control — CAP can be caused by a variety of microbial agents requiring a variety of infection-control measures . If possible, rapid diagnostic tests should be performed at the time of admission, to facilitate decisions regarding appropriate precautions. (See “Community-acquired pneumonia in children: Clinical features and diagnosis”, section on ‘Rapid diagnostic tests’.)
Hand washing is the single most important procedure to prevent the spread of infection. Additional infection control measures depend upon the likely pathogen(s), as follows [6,7]:
- Respiratory syncytial and parainfluenza viruses – Gown and gloves (ie, contact precautions)
- Influenza virus, group A Streptococcus(for the first 24 hours of treatment), methicillin-susceptible S. aureus, Bordetella pertussis(until patient has received five days of effective therapy), and Mycoplasma pneumoniae – Mask within 3 feet (ie, droplet precautions)
- Adenovirus – Contact and droplet precautions
- Methicillin-resistant S. aureus and other multidrug resistant organisms – Special organism precautions; contact and droplet precautions and dedicated patient equipment
These precautions are discussed separately (see “Infection prevention: Precautions for preventing transmission of infection”). Guidelines for hand hygiene in healthcare settings can be accessed through the Centers for Disease Control and Prevention.
Supportive care includes ensuring adequate antipyresis, analgesia, respiratory support, and hydration.
Antipyresis and analgesia — Children hospitalized with pneumonia usually have fever and may have pleuritic chest pain, which can lead to shallow breathing and impaired ability to cough. Administration of antipyretics and/or analgesics (eg, acetaminophen, ibuprofen) can be used to keep the child comfortable; opioid analgesia is rarely necessary in children without a chest tube in place. Adequate pain control may promote coughing, which facilitates airway clearance. Antitussives should be avoided as none have been found to be effective in pneumonia . Symptomatic treatment of cough is discussed separately. (See “The common cold in children: Management and prevention”, section on ‘Cough’.)
Respiratory support — Children hospitalized with pneumonia should receive ventilatory support as indicated by their clinical condition [1,2]. A supported sitting position may help to expand the lungs and improve respiratory symptoms .
We suggest that children with oxygen saturation [SpO2] <95 percent in room air be treated with supplemental oxygen to maintain oxygen saturation ≥95 percent while they are in respiratory distress. Different thresholds for supplemental oxygen are suggested by other experts (eg, the British Thoracic Society guidelines suggest supplemental oxygenation to maintain oxygenation saturation >92 percent) . Gentle bulb suction of the nares may be helpful in infants and children whose nares are blocked with secretions. Minimal handling seems to reduce oxygen requirements. (See “Continuous oxygen delivery systems for infants, children, and adults”.)
In children who are severely ill, it may be necessary to monitor carbon dioxide tension via blood gas analysis in addition to oxygen saturation (SpO2) by oximetry. Hypercarbia is an important sign of impending respiratory failure, particularly in the young infant who is tiring but may have preserved oxygenation.
Fluid management — Children who cannot maintain adequate fluid intake because of breathlessness, fatigue, or risk of aspiration  may require intravenous fluid therapy. Nasogastric (NG) tubes should be avoided if possible because they may compromise breathing; if necessary, the smallest NG tube possible should be used . (See “Maintenance fluid therapy in children”.)
Children with pneumonia are at risk for inappropriate secretion of antidiuretic hormone (SIADH) [10,11]. Serum electrolytes, fluid balance, and urine specific gravity should be monitored if there is clinical suspicion of SIADH . Confirmation of SIADH is discussed separately. Isotonic, rather than hypotonic, intravenous fluids should be provided if SIADH is suspected. (See “Pathophysiology and etiology of the syndrome of inappropriate antidiuretic hormone secretion (SIADH)”, section on ‘Pulmonary disease’ and “Maintenance fluid therapy in children”, section on ‘Hospitalized children’.)
Chest physiotherapy — Chest physiotherapy is not beneficial for children with uncomplicated community-acquired pneumonia (CAP) . In randomized and observational studies in children and adults, chest physiotherapy had no conclusive effect on length of hospital stay, duration of fever, or radiographic resolution [12-17].
Adjunctive glucocorticoid therapy — We do not routinely provide adjunctive glucocorticoid therapy to children hospitalized with pneumonia. Although a systematic review and meta-analysis of randomized trials in adult patients hospitalized with CAP found that corticosteroid therapy may be beneficial in reducing the development of acute respiratory distress syndrome, need for mechanical ventilation, and the duration of hospitalization , additional studies in children are necessary. A retrospective study evaluating adjunctive glucocorticoid therapy for children being treated for CAP in the outpatient setting found an association between adjunctive glucocorticoid therapy and treatment failure in children without underlying asthma .
Overview — Prompt initiation of antimicrobial therapy is crucial in children with community-acquired pneumonia (CAP). The initial treatment of children who are hospitalized with pneumonia is empiric (table 2). Factors that must be considered include the spectrum of likely pathogens, antimicrobial susceptibility, simplicity, tolerability, palatability, safety, and cost .
The recommendations of most guidelines are based on in vitro susceptibilities of the most likely pathogen or pathogens, rather than evidence of the superiority of one antibiotic over another. Clinical response to empiric therapy and results of microbiologic studies, when available, help to determine whether additional evaluation or changes in therapy are necessary [1,2]. (See “Community-acquired pneumonia in children: Clinical features and diagnosis”, section on ‘Microbiology’ and ‘Specific therapy’ below and ‘Response to therapy’ below.)
There are few randomized controlled trials to guide the choice of empiric antibiotics in children with CAP. Decisions regarding empiric therapy are complicated by the substantial overlap in the clinical presentation of bacterial and nonbacterial pneumonias [21-23]. Treatment decisions usually are based upon algorithms that include patient age, epidemiologic and clinical information, and diagnostic laboratory and imaging studies (table 2) . The scope of empiric therapy (ie, narrow or broad) depends upon the severity of illness and presence of complications. Agents other than those suggested in the table may be more appropriate if there are clinical or epidemiologic features strongly suggestive of a specific cause (eg, mediastinal or hilar lymphadenopathy, residence in the central United States, and exposure to caves and/or bat guano suggestive of pulmonary histoplasmosis) .
Consultation with a specialist in infectious disease may be helpful in children with medication allergies, comorbid conditions, failure of outpatient therapy, or multiple-drug-resistant organisms. Consultation with a pediatric pulmonologist may be helpful in children with recurrent pneumonia. (See “Community-acquired pneumonia in children: Clinical features and diagnosis” and “Community-acquired pneumonia in children: Outpatient treatment”, section on ‘Treatment failure’.)
Etiologic clues — Certain clinical and epidemiologic features can be used to determine the most likely pathogen(s) to aid in decisions regarding empiric therapy. Because these features often overlap, they cannot be used with complete confidence, but are helpful in guiding empiric therapy until results of microbiologic tests are available (table 3). These features are discussed in greater detail separately. (See “Community-acquired pneumonia in children: Clinical features and diagnosis”, section on ‘Clues to etiology’ and “Community-acquired pneumonia in children: Clinical features and diagnosis”, section on ‘Etiologic clues’.)
Neonates — The treatment of neonatal pneumonia is discussed separately. (See “Neonatal pneumonia”.)
Viral pneumonia — Most children younger than three to five years of age who are admitted to the hospital with pneumonia have viral pneumonia (eg, respiratory syncytial virus) . This is particularly true in the absence of lobar (or lobular) infiltrate and pleural effusion . Viral pneumonia does not require antibiotic therapy, unless a mixed infection or secondary bacterial infection is suspected. (See “Respiratory syncytial virus infection: Treatment”, section on ‘Overview’ and “Respiratory syncytial virus infection: Clinical features and diagnosis”, section on ‘Clinical manifestations’.)
No effective antivirals are available for most viral pneumonias, with a few important exceptions, described below.
Influenza pneumonia — Initiation of antiviral treatment for influenza (eg, oseltamivir) as soon as possible is recommended for children hospitalized with presumed influenza pneumonia; laboratory confirmation should not delay initiation of antiviral therapy. The diagnosis and treatment of influenza in children are discussed separately. (See “Seasonal influenza in children: Prevention and treatment with antiviral drugs”, section on ‘Antiviral therapy’ and “Seasonal influenza in children: Clinical features and diagnosis”, section on ‘Diagnosis’.)
For children with influenza pneumonia in whom secondary bacterial pneumonia is suspected, empiric antibiotic therapy should include coverage for S. aureus, including methicillin-resistant S. aureus (MRSA). Coinfection with S. aureus may be particularly severe and rapidly fatal.
Other viral pneumonias — Acyclovir can be used in the treatment of pneumonia due to herpes simplex virus (HSV) or varicella zoster virus (VZV). Ganciclovir be used in the treatment of pneumonia due to cytomegalovirus (CMV). (See “Treatment of varicella (chickenpox) infection”, section on ‘Individuals with complications’.)
Common respiratory viruses may cause serious infections in immunocompromised children and require consideration of antiviral therapy: ribavirin for respiratory syncytial virus (RSV) or parainfluenza and cidofovir for adenovirus. Concomitant immunoglobulin therapy is an additional consideration: palivizumab for RSV, CMV immune globulin for CMV, and intravenous immunoglobulin for the other viral etiologies. (See “Respiratory syncytial virus infection: Treatment”, section on ‘Pharmacotherapy’ and “Diagnosis, treatment, and prevention of adenovirus infection”, section on ‘Treatment’.)
Uncomplicated bacterial pneumonia — Streptococcus pneumoniae is the most common bacterial cause of pneumonia in children of all ages [4,26]. Other potential bacterial pathogens that may need to be included in empiric therapy for hospitalized children include S. aureus, including MRSA, S. pyogenes (group A Streptococcus), Haemophilus influenzae type b (Hib) (if unimmunized), nontypeable H. influenzae, and Moraxella catarrhalis [2,4,26-31].
The table provides several suggested parenteral empiric antibiotic regimens for uncomplicated bacterial pneumonia in hospitalized children when S. aureus is not a consideration (table 2) [4,32,33]. The treatment of complicated CAP and severe CAP (particularly when S. aureus is a consideration) are discussed below. (See ‘Complicated CAP’ below and ‘Severe CAP requiring ICU admission’ below.)
Ampicillin or penicillin G generally provides adequate coverage for the fully immunized child (table 4) in communities without substantial prevalence of penicillin-resistant S. pneumoniae [1,34,35]. We suggest a third-generation cephalosporin (eg, cefotaxime, ceftriaxone) for children younger than 12 months and those who are not fully immunized because third-generation cephalosporins provide coverage for the beta-lactamase producing pathogens (eg, H. influenzae and M. catarrhalis) that may occur in these children. We also suggest third-generation cephalosporins for children with more severe illness (table 1) because third-generation cephalosporins provide coverage for a broader range of pathogens, including penicillin-resistant S. pneumoniae, than ampicillin [1,36,37]. The fifth-generation parenteral cephalosporin, ceftaroline, is approved by the US Food and Drug Administration (FDA) for treatment of community-acquired bacterial pneumonia due to S. pneumoniae, methicillin-susceptible S. aureus (MSSA), and H. influenzae in children ≥2 months of age. Although ceftaroline exhibits in vitro activity against MRSA , clinical experience is insufficient to suggest its use when MRSA is a consideration. In a randomized trial in children between 2 months and <18 years who were hospitalized with CAP, ceftaroline and ceftriaxone had similar cure rates . Three children with S. aureus infection (two with MSSA recovered from sputum and one with MRSA recovered from blood) were successfully treated with ceftaroline. However, patients considered at risk for MRSA infection or those with sputum demonstrating a predominance of gram-positive cocci in clusters were excluded from the trial, precluding conclusions about efficacy in this population.
A macrolide may be added (table 2) if M. pneumoniae, C. pneumoniae, or legionellosis is suspected, although the benefits of combination therapy are uncertain. In a prospective population-based study of 1418 children hospitalized with radiographically confirmed CAP, the addition of a macrolide to beta-lactam antimicrobial therapy was not associated with decreased length of stay, intensive care admission, rehospitalization, or self-reported recovery . In subgroup analysis, combination therapy was not associated with decreased length of stay in children in whom atypical bacteria were detected, children older than five years, children admitted to the intensive care unit, or children with wheezing. (See ‘Atypical pneumonia’ below.)
We suggest that children who require hospitalization for treatment of CAP be treated initially with parenteral antibiotics. However, oral amoxicillin may be an alternative for infants and children fully immunized against Hib and S. pneumoniae with uncomplicated pneumonia that is not thought to be due to S. aureus. In a multicenter randomized trial, treatment with amoxicillin was equivalent to treatment with penicillin G in children with CAP who required hospital admission but did not have wheezing, hypotension, chronic pulmonary conditions (other than asthma), immunodeficiency, pleural effusion requiring drainage, or oxygen saturations <85 percent in room air . The British Thoracic Society guidelines suggest that oral antibiotics are safe and effective even for children with severe pneumonia as long as they are able to tolerate oral fluids, are not vomiting, and do not have signs of septicemia or complicated pneumonia .
Atypical pneumonia — Atypical bacterial pathogens include C. trachomatis in afebrile infants, and M. pneumoniae and C. pneumoniae in older children and adolescents. The table provides several suggested empiric regimens for atypical bacterial pneumonia in hospitalized children (table 2) [4,32].
For children older than four years, coverage for typical bacterial pathogens (eg, ampicillin or a third-generation cephalosporin) may be added to empiric coverage for atypical pathogens if there is strong evidence of a bacterial cause. Strong evidence of a bacterial cause includes white blood cell count >15,000/microL, C-reactive protein (CRP) >35 to 60 mg/L (3.5 to 6 mg/dL), chills, or no response to outpatient therapy with a macrolide or doxycycline [4,42].
Fluoroquinolones (eg, levofloxacin, moxifloxacin) may be reasonable empiric therapy for the older child and adolescent with suspected atypical pneumonia who could actually have pneumococcal pneumonia. The fluoroquinolones also may be used in the older child or adolescent who has a type 1 hypersensitivity (table 5) to beta-lactam antibiotics. In addition to their excellent gram-negative spectrum, the fluoroquinolones are active against a number of the pathogens responsible for CAP, including beta-lactam-susceptible and nonsusceptible S. pneumoniae, M. pneumoniae (including macrolide-resistant M. pneumoniae), and C. pneumoniae . However, S. pneumoniae resistant to levofloxacin have been identified .
Topic 6054 Version 50.0