Itzhak Brook. Microbiology of Nosocomial Sinusitis in Mechanically Ventilated Children. Arch Otolaryngol Head Neck Surg. 1998;124(1):35–38. doi:10.1001/archotol.124.1.35
To assess the bacteriology of nosocomial sinusitis in mechanically ventilated children.
Retrospective review of sinus aspirate specimens obtained from 20 children with nosocomial sinusitis. The specimens were processed for aerobic and anaerobic bacteria.
A total of 58 isolates (2.9 per specimen), 30 aerobic or facultative (1.5 per specimen) and 28 anaerobic (1.4 per specimen), were recovered. Aerobes only were present in 8 patients (40%), anaerobes only in 5 (25%), and mixed aerobic and anaerobic flora in 7 (35%). The predominant aerobes were Pseudomonas aeruginosa (6 isolates), Staphylococcus aureus (5 isolates), Escherichia coli (3 isolates), and Klebsiella pneumoniae (3 isolates). The predominant anaerobes were Peptostreptococcus species (8 isolates), Prevotella species (6 isolates), and Fusobacterium species (4 isolates). Forty-one β-lactamase bacteria were recovered from 14 specimens (70%). Thirty isolates similar to the sinus isolates were also recovered from the trachea, 6 from blood culture specimens, and 6 from other sites. Anaerobes were more commonly isolated from sinus aspirate samples obtained after 18 days of mechanical ventilation (21 vs 7, P<.05 by χ2).
This study demonstrates the polymicrobial aerobic-anaerobic flora of nosocomial sinusitis in mechanically ventilated children.
NOSOCOMIAL sinusitis (NS) is often a complication of endotracheal intubation and mechanical ventilation. The lack of clinical signs often leads to underestimation of the incidence of this infection. The diagnosis of NS is based on findings of the radiological examination and microbiological culture of sinus fluid.1
Patients with NS generally require extended periods of intensive care involving prolonged nasotracheal or nasogastric intubation. Approximately one fourth of patients requiring nasotracheal intubation for more than 5 days develop NS.2 In contrast to community-acquired sinusitis, in which Streptococcus pneumoniae and Haemophilus influenzae are the predominant isolates, the usual pathogens recovered from adults with NS are gram-negative enteric bacteria (such as Pseudomonas aeruginosa, Klebsiella pneumoniae, Enterobacteriaceae species, Proteus mirabilis, and Serratia marcescens) and gram-positive cocci (such as streptococci and staphylococci).2- 4 However, no previous studies investigated the microbiology of NS in children.
This study describes the author's experience in evaluating the aerobic and anaerobic microbiology of NS of the maxillary sinus in mechanically ventilated children.
Twenty children who developed NS and whose sinus aspirate specimens were studied for aerobic and anaerobic bacteria were included in the study. The patients were examined and treated between May 1985 and August 1994 at area hospitals in Washington, DC.
The patients' ages ranged from 7 to 16 years (average, 11 years), and 6 patients were boys. Each patient received mechanical ventilation via a nasotracheal (13 patients) or tracheotomy (7 patients) tube connected to a ventilator. The patients included had been hospitalized for more than 72 hours and had a tracheal tube in place for more than 48 hours.
Each patient's medical chart was reviewed for the following factors: age, sex, date of intubation or tracheotomy, duration of mechanical ventilation, previous pulmonary infections, and antibiotics received before the day of sampling. Antimicrobial therapy was given to 14 patients before sinus sample collection: 7 patients received a penicillin, 5 a cephalosporin, 4 an aminoglycoside, and 2 erythromycin estolate.
Maxillary sinus aspirate specimens were obtained through inferior meatal antrostomies (after moving the tube to the other naris or to an orotracheal route when indicated). The site of aspiration was prepared by cleansing with povidone-iodine. Samples were also collected from other sites, including tracheal culture obtained through endotracheal suctioning in all patients, blood culture in 14, and other sites (ie, wounds and catheter sites) in 10. Sinusitis was judged to be present if the radiographic studies showed mucosal thickening and either an air-fluid level or complete opacification of the maxillary sinus. Radiological studies were performed because of the clinical suspicion of NS (ie, fever, pain, and postnasal drip). These studies included radiographs (in all patients) and computed tomographic scans (in 12 patients). Concordance between these radiological techniques was found in 10 patients.
Specimens were transported to the laboratory in a syringe sealed with a rubber stopper after evacuation of the air in the syringe. The time between the collection of materials and the inoculation of these specimens did not exceed 30 minutes.
Specimens were inoculated onto 5% sheep's blood, chocolate, and MacConkey agar plates for aerobic and facultative organisms. The plates were incubated at 37°C aerobically (MacConkey agar) or under 5% carbon dioxide (5% sheep's blood and chocolate agars) and examined at 24 and 48 hours. For anaerobes, the material was plated onto pre-reduced vitamin K1–enriched Brucella blood agar, an anaerobic blood agar plate containing kanamycin sulfate and vancomycin hydrochloride, an anaerobic blood plate containing colistin and nalidixic acid, and an enriched thioglycolate broth containing hemin and vitamin K1.5 The anaerobic plates and thioglycolate broth were incubated in jars (GasPak, Baltimore Biological Laboratories, Cockeysville, Md) and examined at 48 and 96 hours.
Anaerobes were identified by techniques described previously.5 Aerobic bacteria were identified by conventional methods.6 β-lactamase activity was determined for all isolates using the chromogenic cephalosporin analogue 87/312 method.7
Bacterial growth was present in specimens from all patients. A total of 58 isolates (2.9 per specimen), 30 aerobic or facultative (1.5 per specimen) and 28 anaerobic (1.4 per specimen), were recovered. Aerobic bacteria only were present in 8 patients (40%), anaerobic only in 5 (25%), and mixed aerobic and anaerobic flora in 7 (35%) (Table 1). The predominant aerobic bacteria were P aeruginosa (6 isolates), Staphylococcus aureus (5 isolates), Escherichia coli (3 isolates), and K pneumoniae (3 isolates). The predominant anaerobes were Peptostreptococcus species (8 isolates, including 3 Peptostreptococcus prevotii and 2 Peptostreptococcus micros), Prevotella species (6 isolates, including 3 Prevotella intermedia, 2 Prevotella melaninogenica, and 1 Prevotella oralis), and Fusobacterium species (4 isolates, including 2 Fusobacterium nucleatum and 1 Fusobacterium necrophorum).
A single isolate was recovered in 6 patients (30%): P aeruginosa (3 patients), S aureus (2 patients), and F nucleatum (1 patient). Polymicrobial infections, where the number of isolates varied between 2 and 6, were recovered in 14 patients (70%). Forty-one β-lactamase–producing bacteria were recovered from 14 sinus specimens (70%) (Table 1).
Thirty isolates similar to those recovered in the maxillary sinus were isolated in tracheal specimens (2 patients had concomitant pneumonia), 6 in blood culture specimens, and 6 from other sites (4 in wounds, 1 in urine, and 1 in catheter).
The underlying conditions leading to mechanical ventilation were status after surgery (6 patients), drowning (3 patients), head trauma (4 patients), respiratory distress syndrome (2 patients), brain abscess (2 patients), brain tumor (2 patients), and drug overdose (1 patient).
The duration of mechanical ventilation before sinus aspiration varied from 6 to 63 days (average, 14.2 days). Correlation between the bacteria isolated and the duration of mechanical ventilation showed a higher recovery of anaerobic bacteria in patients who had been receiving mechanical ventilation for longer than 18 days (21 isolates from specimens obtained after 18 days of intubation compared with 7 recovered from specimens obtained for <18 days, P<.05 by χ2). The 2 isolates of H influenzae and 1 isolate of S pneumoniae were recovered from patients who were intubated for less than 14 days. No other correlation was found between the microbiological findings and the clinical features.
Nasal decongestants were given to 17 patients for 5 to 9 days, and antimicrobial therapy was administered to all the patients for 18 to 43 days. Eleven patients received clindamycin, 6 amoxicillin-clavulanate potassium, 5 ceftazidime, 5 gentamicin sulfate, 4 oxacillin sodium, 3 amoxicillin, and 2 metronidazole.
Fourteen patients had complete resolution of infection. Nine of these discontinued mechanical ventilation. Two children required surgical drainage and 4 had at least 1 recurrence within 1-year follow-up.
This study demonstrates the uniqueness of the microbiology of NS in mechanically ventilated and intubated children. Similar to findings in adults with NS,1,8- 12P aeruginosa, Enterobacteriaceae species, and S aureus predominated. However, this study demonstrates for the first time the importance of anaerobic bacteria in NS in intubated children. Although these organisms were also recovered in a similar infection in adults,8- 12 they were generally not completely speciated and were recovered in fewer than 22% of the patients.4,8- 12 The higher rate of recovery of organisms reported in this study may be due to the strict methods used for transportation of specimens and cultivation of bacteria.
The isolates Wald et al13 recovered from patients with NS that developed after less than 14 days of intubation, H influenzae and S pneumoniae, are similar to those recovered from patients with acutely inflamed sinuses in commonly acquired NS. Similar to that found in patients with chronic sinusitis,14,15 anaerobic bacteria predominated in patients with NS that occurred after 18 days of intubation. However, in contrast to findings in patients with community-acquired sinusitis, gram-negative aerobic and facultative bacteria (P aeruginosa, K pneumoniae, E coli, and P mirabilis) predominated in patients with NS regardless of the length of previous intubation. These organisms were also isolated from tracheal aspirate samples and other sites, suggesting their potential source or dissemination from the infected sinus. Although these bacteria did not cause active lung infection in most patients at the time of the NS and probably only colonized the trachea, similar bacteria were recovered from the tracheal aspirate specimens of 2 of the patients who also had pneumonia, and previous studies identified similar isolates from pneumonia in intubated patients.16
The recovery of organisms similar to those isolated from the infected sinus from other body sites can assist in the empirical selection of antimicrobial therapy. However, since not all of the organisms isolated from sinuses were also recovered at other sites, aspiration of the sinus cavity may be useful to establish the exact bacterial etiology and to assist in the proper selection of antimicrobial therapy. Removal of the endotracheal, nasotracheal, or nasogastric tube is the mainstay of therapy. Surgical drainage is of paramount importance, especially if no clinical improvement has occurred.17
The recovery of β-lactamase–producing bacteria in 70% of patients with NS is not surprising since most had received multiple courses of antimicrobial agents, including β-lactams, during hospitalization. Such therapy can select β-lactamase–producing bacteria.18 The choice of antimicrobial agents effective against the organism causing NS should be individualized because the microbiology varies among patients. Empirical broad-spectrum coverage with a combination of a penicillin and a β-lactamase inhibitor, plus an antimicrobial effective against aerobic gram-negative rods (ie, an aminoglycoside, ceftazidime, or a quinolone in adults), or a carbapenem may be warranted initially. Specific antimicrobial therapy can be chosen after culture specimens are obtained. Endoscopic techniques can assist in obtaining such culture specimens noninvasively. Studies that include larger numbers of children with NS are needed to further elucidate the microbiology and proper therapy of sinus infection in these patients.
Accepted for publication September 22, 1997.
I acknowledge the efforts of the staffs of the Departments of Otolaryngology and the Clinical Microbiology Laboratory Services at the participating hospitals.
Corresponding author: Itzhak Brook, MD, MSc, PO Box 70412, Chevy Chase, MD 20813-0412.