Chapter 56. Infections in Children with Tracheostomy

Tracheotomy is one of the most common surgical airway procedures performed in children. Approximately 5000 children undergo tracheotomy each year.1 In the past, tracheotomy was predominately performed in older children with acute upper airway compromise secondary to infection such as epiglottitis and croup.2 Currently, these acute upper airway infections represent less than 5% of tracheotomy performed in children.1,3 Tracheotomy is now more commonly being performed in children who require prolonged mechanical ventilation or who have significant anatomic rather than infectious causes of upper airway obstruction. The changes in the indications for tracheotomy reflect the extended survival of children born prematurely as well as those with chronic underlying illnesses such as neuromuscular disease and congenital anomalies.2–4 These underlying illnesses influence the etiology of tracheotomy-related infections encountered in children. This chapter focuses on the childhood tracheotomy-related infections of stoma cellulitis, tracheitis, and bacterial pneumonia.

Definition and Epidemiology

Tracheostomy stoma cellulitis is a bacterial infection of the epidermis lining of the tracheostomy opening. It is a dangerous infection that, if untreated, can spread contiguously through the tracheostomy site into the trachea or into the deep tissues of the neck and mediastinum. In the past, stoma infections occurred approximately in one-third of patients undergoing surgical tracheotomy placement.5 In more recent studies the rates of stoma infection have been significantly lower, ranging from 0% to 3%.6,7

Pathogenesis

There is limited information on the pathogens responsible for tracheostomy stoma cellulitis. It appears that the bacteria colonizing the granulation tissue surrounding the tracheostomy cause most cases of stoma cellulitis. The epithelial granulation tissue lining the tracheostomy site is colonized with bacteria by direct contact with the bacterial flora of the surrounding skin. Polymicrobial colonization has been reported in the majority of tracheostomy granulation tissue specimens, with a mixture of gram-positive, gram-negative, and anaerobic bacteria. Approximately 6 bacteria are detected in each granulation tissue specimen sent for culture.8,9 The bacteria most frequently recovered from granulation tissue culture include alpha-hemolytic streptococci, Staphylococcus aureus, Peptostreptococcus species, Bacteroides species, Fusobacterium species, and Pseudomonas aeruginosa.9,10 Most isolates produced beta-lactamase, including all isolates of S. aureus and Bacteroides species.9

The tracheostomy tube itself may also become colonized as it provides a portal of entry from the outside directly into the normally sterile trachea. Tracheostomy tubes are typically composed of plastic, with differing degrees of flexibility depending on the specific type of material (i.e., silicone or polyvinyl chloride). Bacteria with a polysaccharide shell can bond to the surface of a tracheostomy tube lumen, forming organized matrices of bacterial colonies. The colonies are often referred to as biofilm.11 The longer that a tracheostomy tube is used, the more likely that it will develop biofilm.

The bacteria that most frequently colonize the tracheostomy tube include S. aureus and P. aeruginosa.12 Exposure to these pathogens likely occurs in the hospital. Most children undergo tracheotomy in an intensive care setting. Pretracheotomy intubation times may exceed more than 1 month4,13,14 while overall hospital lengths of stay associated with tracheotomy placement may exceed 50 days.1 This significant nosocomial exposure provides ample opportunity for colonization with flora endemic to the intensive care unit setting.

Although many of the bacterial species that colonize granulation tissue and tracheostomy tubes are considered normal skin flora and are typically nonpathogenic, others can cause clinically significant infection. The most common bacterial pathogens believe to be responsible for tracheostomy stoma cellulitis include group A β-hemolytic streptococci, S. aureus, and P. aeruginosa in addition to various anaerobic bacteria. Additionally, candida species may also be responsible for tracheostomy stoma cellulitis.

Clinical Presentation

On history, parents of children with tracheotomy who have stoma cellulitis will commonly report a change in color of the granulation tissue of the tracheostomy stoma from pink to red. This redness may spread circumferentially throughout the skin surrounding the tracheostomy. Foul-smelling, yellow/green drainage and bleeding at the site have also been reported. The infection may be accompanied by a history of fever, malaise, overall discomfort, and localized pain.

On physical examination, children may be febrile, tachycardic, and ill-appearing. Frank erythema, edema, and/or induration of the skin surrounding the tracheostomy site may be observed. Bleeding may be present. The skin may also feel warm to the touch. Purulent exudate may be visualized emerging from the stoma tissue. Supraclavicular and cervical lymphadenopathy may be palpable. Confluent papular erythema with erythematous satellite papules may be present with candidal tracheostomy cellulitis.

Differential Diagnosis

Noninfected, healthy granulation tissue or fibrous scar tissue is commonly mistaken for cellulitis. Most children with tracheotomy will likely develop pink granulation stoma epithelium to some degree (Figure 56–1). Characteristics that distinguish health stoma granulation tissue from stoma cellulitis are shown in Table 56–1. Stoma cellulitis is unlikely in the presence of exudate without other clinical features of infection.

Superficial stoma cellulitis should also be differentiated from the extremely rare but life-threatening deep tissue infections such as cervical necrotizing fasciitis and mediastinitis.15 In these deep tissue infections, the erythema may progress rapidly; the patient is often ill-appearing with systemic signs of toxicity.16 In cases with cervical necrotizing fasciitis, the skin may appear gray, green, or purple. Subcutaneous emphysema may also be observed.16 Mediastinitis is difficult to recognize early; however, significant respiratory distress or even respiratory failure will likely be present.17 Patients who have undergone tracheotomy in relation to cardiac surgery are at increased risk for deep-tissue infection.18

Diagnosis

Tracheostomy stoma cellulitis is a clinical diagnosis. Laboratory and radiographic diagnostic tests are not helpful in making the diagnosis but may help guide empiric antibiotic therapy and facilitate early detection of complications such as deep-tissue infection. Gram stain and culture from the tracheotomy stoma tissue or exudates may help identify potential causative pathogens. However, since polymicrobial colonization of healthy granulation tissue is common, the presence of bacteria in a culture alone does not necessarily indicate cellulitis. We recommend that Gram stain and culture be obtained prior to treatment as a guide to antibiotic therapy, rather than as a diagnostic indicator of stoma cellulitis.

The diagnostic role of blood inflammatory markers (white blood cell [WBC] count, differential, C-reactive protein, etc.) has not been evaluated in children with tracheostomy stoma bacterial infections. WBC counts less than 15,400 cells/mm3 have a negative predictive value of 99% excluding necrotizing fasciitis in adult patients with stoma cellulitis.19 Chest radiography or magnetic resonance imaging provide evidence of deep-tissue infection such as subcutaneous emphysema (on radiograph or magnetic resonance imaging) and fascia inflammation or abscess formation (on magnetic resonance imaging).

Treatment

Evidence-based guidelines for the treatment of tracheostomy stoma cellulitis do not currently exist. Clinical trials of antibiotic therapy for this infection have not been performed. However, since polymicrobial infections are common, initial antibiotic therapy should be broad-spectrum. Therapy should also be directed toward the most common pathogens (Table 56–2).

Intravenous antibiotics are recommended for the ill-appearing patients and those with significant or rapidly progressing cellulitis. Close follow-up is recommended, especially given the possible direct route of bacterial spread into the trachea and into the deep tissues of the neck and chest if the infection progresses. For this reason, hospitalization is strongly encouraged for the initial portion of therapy for ill-appearing patients with stoma cellulitis. Intravenous options for single initial therapy may include ampicillin-sulbactam or, if Pseudomonas species are strongly suspected, piperacillin-tazobactam or imipenem. Children with severe penicillin allergy may receive imipenem alone or a combination of clindamycin and an aminoglycoside (i.e., tobramycin, amikacin, or gentamicin).20 Additional broad-spectrum coverage may be provided with the addition of either ciprofloxacin or cephalosporins with activity against P. aeruginosa, such as ceftazidime or cefepime. The stoma culture antibiotic susceptibilities may be used to adjust or narrow antibiotic therapy. Vancomycin may be required if methicillin-resistant S. aureus (MRSA) is suspected.21

Transition from intravenously to enterally administered antibiotics is reasonable for patients who have initial treatment as indicated by decreased stoma site erythema, exudate, and induration. Reasons for failure in treatment response may include resistant organisms and extension of infection into the trachea or soft tissues. Surgical debridement may be warranted for deep-tissue infections.18

Initial enteral antibiotic therapy may be appropriate for the nontoxic appearing patient with mild cellulitis, assuming that antibiotic absorption is not compromised by conditions such as diarrhea or short gut syndrome. In previous studies, bacteria present in granulation tissue at the tracheostomy site have been susceptible to amoxicillin-clavulanate or ciprofloxacin, suggesting that these antibiotics may be options for enteral therapy.8 The American Academy of Pediatrics does not endorse the use of fluoroquinolones for skin or soft-tissue infections in children younger than 18 years, unless there is no other available alternative therapy.22 Other enteral choices include clindamycin or trimethoprim-sulfamethoxazole.21 If candidal cellulitis is suspected, therapy with a topical antifungal medication (e.g., nystatin) may be appropriate.

Most clinical trials of skin infections in children have used 10 days of antibiotic therapy.21 Current data suggests that 10–14 days of antibiotic therapy should be considered for children with tracheostomy stoma cellulitis. It is important to dispose or thoroughly clean home tracheostomy equipment at risk for bacterial contamination when a child experiences a stoma infection. This equipment may include trach ties, suction catheters, suction tubing, humidification nose, oxygenation and ventilation tubing.23 Consultation with an otolaryngologist and intensivist should be undertaken in situations where a tracheostomy tube must be replaced secondary to stoma cellulitis.

Definition and Epidemiology

Bacterial tracheitis is a serious infection of the tracheal epithelium. The prevalence rate in children with tracheostomy has not been reported, although it is believed to be very rare. Bacterial tracheitis is associated with significant morbidity. Respiratory failure requiring intubation and mechanical ventilation is commonly reported.24 Mortality rates have been reported as high as 18–40%.25

Pathogenesis

Children with tracheostomy are at risk for bacterial infection in the trachea as a result of direct airway exposure to exogenous pathogens, colonized bacteria on the tracheostomy tube, and tracheal injury. Over time, the presence of the tracheostomy tube within the trachea may result in epithelial ulceration and denudation, which may predispose the trachea to infection.26 Some children with tracheostomy may have a history of gastroesophageal reflux disease that, with chronic tracheal aspiration, may result in tracheal injury as a predicate to tracheitis.27S. aureus, group A Streptococcus, and Haemophilus influenzae have been reported as the leading bacterial causes of tracheitis in children, with S. aureus isolated in the majority of cases.24,27 Although gram-negative organisms are rarely isolated in tracheitis, P. aeruginosa should be considered given the high rates of tracheostomy colonization with this organism.12

Clinical Presentation

On history, children with bacterial tracheitis have a 2–4 day prodrome of upper respiratory tract infection symptoms, including rhinorrhea, low-grade fever, cough, and sore throat. The hoarseness and “barky” cough frequently encountered in otherwise healthy children with tracheitis may be more difficult to recognize in children with tracheostomy.

Typically, the child with tracheitis will rapidly develop signs of respiratory distress following this prodrome. Increased work of breathing, airway compromise, higher fever, and toxic appearance are commonly observed.27 Parents may also report changes in the quantity and quality of tracheal secretions. With bacterial tracheitis, the amount of secretions may increase dramatically and change from thin and clear to thick and yellow or green in color.

On physical examination, the child with bacterial tracheitis may appear ill. Biphasic (inspiratory and expiratory) stridor may be audible over the tracheostomy if the infection is severely compromising the airway. Tracheostomy suctioning may reveal continuous, thick, purulent secretions. Auscultation of the lower airways may reveal gurgling sounds transmitted from the upper airways. If pneumonia is absent, the lower airways may be clear, without focal findings. Signs of respiratory distress, including tachypnea and retractions, are common.

Differential Diagnosis

The most common condition confused with bacterial tracheitis in children with tracheostomy is viral laryngotracheobronchitis (croup). Differentiating croup from early bacterial tracheitis in children with tracheostomy can be challenging. Typically, children with bacterial tracheitis have progressive respiratory distress despite the use of racemic epinephrine and corticosteroids. Pneumonia should also be considered in the differential diagnosis, as up to 25% of children with bacterial tracheitis may have concurrent pneumonia.28

Diagnosis

If bacterial tracheitis is suspected in the child with tracheostomy, maintaining patency of the upper airway requires immediate medical attention. Delaying attention to the airway secondary to obtaining diagnostic tests could place the child in immediate danger of respiratory collapse.

The diagnosis of bacterial tracheitis is usually made by direct visualization with laryngoscopy or tracheoscopy.29 Direct visual examination of the trachea can help distinguish trachieitis from croup and epiglottitis. A normal-appearing epiglottis and larynx with visual evidence of tracheal inflammation, the presence of purulent tracheal secretions, and a sloughing tracheal pseudomembrane is diagnostic for tracheitis.27,30

A Gram stain of tracheal secretions to evaluate for the presence of polymorphonuclear leukocytes and bacteria may be a helpful initial screen. This Gram stain should be compared with the child's previous tracheal aspirate results. If the previous aspirate revealed gram-negative rods and few white cells, but now demonstrates gram-positive organisms and sheets of white blood cells, the index of suspicion must be significantly higher for a tracheitis, especially if there is no evidence of pneumonia.27 Bacterial culture of the tracheal secretions and a blood culture are recommended with suspected tracheitis in children with tracheostomy. These cultures may be used to determine bacterial sensitivities and guide antibiotic therapy.

Radiographs of the upper airway may be performed in stable children without signs of impending respiratory collapse. An anterior–posterior neck film may demonstrate a “steeple sign” that is also commonly seen in children with croup. A lateral neck film may demonstrate a tracheal air column that appears diffusely hazy, with multiple luminal soft tissue irregularities. These irregularities may be indicative of pseudomembrane detachment.29

Treatment

Children with tracheostomy and suspected bacterial tracheitis should be considered to have a critical airway. At least half of all children with bacterial tracheitis require endotracheal intubation although some clinicians recommend routine intubation for airway protection.31,32 Prompt airway stabilization with transfer to an intensive care unit with a team of experienced nurses, respiratory therapists, and practitioners skilled in advanced airway management in children should be considered.

Once the airway is stabilized, antibiotic therapy should be initiated. Empiric broad-spectrum antibiotic administration provides coverage for S. aureus, Streptococcus spp., and H. influenzae. Parental antibiotics are strongly recommended. Vancomycin may be appropriate for initial therapy if MRSA is suspected.27 Combination therapy with a third-generation cephalosporin, such as cefotaxime, and/or clindamycin may also be considered. If MRSA is not suspected, therapy may be narrowed from vancomycin to a penicillinase-resistant penicillin pending bacterial sensitivities. Ten to fourteen days of intravenous antibiotics is recommended given the high morbidity and mortality associated with bacterial tracheitis.

Tracheoscopy may be therapeutic for bacterial tracheitis in children with tracheostomy. Frequently, a thick inflammatory exudate with sloughed mucosa that obstructs the lumen of the trachea and main bronchi can be visualized. This thick material may be removed with suction and foreign body forceps if necessary.29

Definition and Epidemiology

Pneumonia is defined as an infection of the lower airways. In children with tracheostomy, there is a high prevalence of pneumonia. Eighty-eight percent of patients with tracheotomy will develop a lower airway infection within one year of tracheotomy placement. Children with tracheotomy have an average of 2.8 episodes of pneumonia annually.33

Pathogenesis

Children with tracheotomy are at increased risk for developing both community- and hospital-acquired pneumonia caused by impaired respiratory immunity, enhanced exposure to pathogens during routine tracheotomy care, bacterial colonization of the lower airways, and impaired cough reflex which compounds the difficulties with bacterial clearance from the airway.

The lower airways of healthy children are sterile. Healthy lungs have integral defense mechanisms designed to maintain sterility. These mechanisms, often referred to as the “mucociliary escalator,” include mucus secretion, upward beating cilia and secretory immunologlobulin A.12 In children with tracheotomy, this defense mechanism is impaired. The tracheotomy itself provides direct exposure of these pathogens to the lower airway, bypassing the humidification and warming from the nose and oropharynx. This decreased humidification and warming impairs tracheal cilliary activity, resulting in thicker secretions and decreased mucous clearance.34 For those children with tracheotomy who have coexisting neurologic-impairment, a weakened or absent cough reflex further prevents the expulsion of respiratory secretions and infectious pathogens.

Impaired respiratory immunity is largely responsible for bacterial colonization of the airways of children with tracheotomy. Colonization refers to the presence of bacteria in the lower airways without evidence of infection. Up to 95% of children with tracheotomy are colonized with bacterial pathogens in their lower airways.33S. aureus and P. aeruginosa have been reported as the most common exogenous bacteria responsible for colonization.12,35 Over time, colonized bacteria in the lower airways may replicate, infecting the lower airways.35

Routine tracheostomy care provides a portal for exogenous pathogens to enter the lower airways of children. Daily respiratory care maintenance for a child with a tracheostomy requires suctioning and tube tie changes. Routine suctioning is necessary to ensure a patent airway. This involves the placement of a suction catheter through the tracheostomy into the trachea to remove mucus and other debris. Improper cleaning of the suction catheter or poor hand hygiene of the caregiver may lead to direct contamination of the trachea with exogenous pathogens.

Clinical Presentation

Caregivers of children with a tracheotomy-related pneumonia will often report an increase in the quantity of secretions suctioned. Secretions that become thicker and change color from clear to yellow, brown, or green may be associated with pneumonia. Overall, caregivers often report that the child sounds “more junky.” Children with tracheotomy who require oxygen may demonstrate desaturations and increased oxygen requirement. Constitutional signs of fever, malaise, and increased somnolence may be present. Caregivers may also report signs of respiratory distress, including tachypnea and retractions.

On physical examination, children with tracheotomy who have pneumonia may be ill-appearing, febrile, and tachycardic. Lung auscultation may reveal crackles, rhonchi, or decreased breath sounds over the affected area. Additionally, percussion of the affected lung field(s) may reveal muffled tones. It is important to auscultate the child immediately following suctioning or a solid cough. Upper airway mucous plugging can be mistaken for lower airway rhonchi. Tachypnea and retractions may be noted in children who are experiencing respiratory distress.

Differential Diagnosis

The most common condition confused with bacterial pneumonia in children with tracheotomy is an upper respiratory infection in a child with “noninfectious” bacterial colonization of the lower airways. Almost all children with tracheotomy will eventually become colonized with bacteria such as S. aureus and P. aeruginosa. These bacteria may be present in the child's lower airway without inducing an inflammatory response or pneumonia. When these bacteria are isolated in a patient without fever, cough, sputum changes, lung auscultation findings, or respiratory distress, the child likely does not have pneumonia. The distinction between bacterial colonization and pneumonia becomes less clear when children exhibit one or two of the above symptoms, such as sputum change and cough. These symptoms are not specific for pneumonia. They can be associated with a self-limiting upper respiratory infection.

Viral pneumonia and aspiration pneumonia should also be considered in the differential diagnosis of pneumonia in a child with trachesotomy and respiratory distress. Viral pneumonia, such as influenza or RSV infection, will be more likely to occur during seasonal peaks of incidence in the late fall and winter. Aspiration pneumonia may be associated with a temporal history of a choking, gagging, or vomiting event 1–2 days prior to the development of respiratory symptoms.

Diagnosis

Features that may help distinguish bacterial pneumonia from bacterial colonization are summarized in Table 56–3. The use of sputum culture or tracheal aspirate to diagnose bacterial pneumonia in children with tracheotomy is complicated because of the difficulty of differentiating bacterial colonization of the trachea from bacterial infection of the lower airways. The presence of a high bacterial load and leukocytes are also not specific for lower-airway infection.23 Bacteria isolated from the trachea may not correlate with bacteria found in the lower airways. Previous studies support the use of bronchoalveolar lavage (BAL) or quantitative deep endotracheal aspirate (QDEA) for diagnosing ventilator-assisted pneumonia in adult patients.36,37 Clinical studies using these techniques to diagnose ventilator-associated pneumonia in children are in progress. It is unlikely that sputum or tracheal aspirate will assist greatly in the diagnosis of pneumonia in children with tracheostomy. However, these specimens may be useful to determine antibiotic sensitivities of the bacteria present in the airway and direct antibiotic therapy.

Chest radiograph is the gold-standard for diagnosing pneumonia, if the pneumonia is visible on chest radiograph. A substantial number of children suspected of pneumonia will not have radiographic evidence of an infiltrate on initial chest X-ray.38 Comparing prior chest radiographs to the current film is extremely important in assessing pneumonia in children with tracheostomy. Children with neurologic impairment who have a tracheostomy may exhibit chronic lung findings on radiographs, including chronic atelectasis at the lung bases. These markings may be confused with a pneumonia infiltrate (Figure 56–2).

The utility of laboratory diagnostic tests to assist in the diagnosis of bacterial pneumonia is variable. Evaluation of these tests to diagnose pneumonia in children with tracheotomy has not been performed. Increased rates of pneumonia have been reported in children with fever and WBC counts greater than 15,000 cells/mm3.39,40 High procalcitonin (≥2 ng/mL) and C-reactive protein (≥65 mg/L) have exhibited 90% or greater sensitivity in diagnosing bacterial compared with viral lower-respiratory infection in children.41

The combination of symptoms, laboratory and radiographic findings has been the most useful in diagnosing pneumonia in adult, ventilated patients. A new radiographic infiltrate with at least two of the following: fever, leukocytosis, or purulent sputum has been shown to increase the likelihood of pneumonia.42 The absence of a new infiltrate on a plain chest radiograph and fewer than 50% neutrophils on cell count analysis lowered the likelihood of pneumonia.42

Blood culture has not proved useful in the diagnosis or initial treatment of pneumonia in hospitalized adult patients.43 Following the introduction of H. influenzae type B vaccine, rates of bacteremia in healthy children with pneumonia have been reported as less than 2%.44 Rates of bacteremia associated with pneumonia in children with tracheostomy have not been described. The rates may be higher in children with tracheostomy, given their impaired respiratory immunity. Until further data is available, we recommend that blood culture be performed in children with tracheostomy and suspected pneumonia.

Treatment

Evidence-based guidelines for the treatment of pneumonia in children with tracheotomy do not exist. The recommendations described below are based on studies of patients with ventilator-assisted pneumonia or chronic bacterial colonization. The antibiotic options discussed are intended for the treatment of bacterial pneumonia without effusion or empyema.

Early broad-spectrum antibiotic administration has been associated with decreased morbidity and mortality in adult patients with ventilator-assisted pneumonia.45 Initial intravenous antibiotic therapy for tracheostomy-related pneumonia should be based on the recent bacterial colonization pattern of the patient (Table 56–4). Pipercillin-tazobactam, ticarcillin-clavulanate,46 or ceftazidime may be necessary for children previously-colonized with antibiotic-susceptible P. aeruginosa. Therapy with multiple antibiotics is often necessary. Clindamycin may be added for children colonized with methicillin-sensitive S. aureus or if aspiration pneumonia is suspected.47 Vancomycin may be recommended as antibiotic therapy for MRSA pneumonia depending on individual institutional guidelines.48 For children not colonized with P. aeruginosa or S. aureus, a third-generation cephalosporin is suggested for coverage of S. pneumoniae, group A β-hemolytic Streptococcus, and typeable or nontypeable H. influenzae.49,50

Narrowing the spectrum of therapy, once final sputum, tracheal or blood culture results are available, is recommended if the child has clinically responded to initial therapy. Once the child's respiratory status has returned close to baseline, a transition to enteral antibiotics is suggested. Options for enteral therapy include amoxicillin-clavulanate, clindamycin, or cefdinir.51,52 Fluoroquinolones have been endorsed for the exacerbation of pulmonary disease in patients with cystic fibrosis who have colonization with P. aeruginosa, when no alternative therapies are available, and the child can be treated in an ambulatory setting.22 Fluoroquinolones may be an option for enteral therapy in the ambulatory setting for children with tracheostomy who are colonized with P. aeruginosa, when no other alternative is available. Consultation with a pulmonary specialist is recommended under these circumstances.

Previous studies evaluating duration of therapy suggest that at least 14 days of therapy for patients with P. aeruginosa pneumonia is necessary to avoid recurrence.53 For other pathogens, 10–14 days of therapy may be generally appropriate. Children with tracheostomy and pneumonia who are ill-appearing, dehydrated, in respiratory distress, or in need of enhanced oxygenation or ventilatory support should be cared for initially in a hospital setting. Children who respond to initial therapy may be eligible for treatment completion as an outpatient.