This statement provides guidelines for therapy of children with serious infections possibly caused by Streptococcus pneumoniae. Resistance of invasive pneumococcal strains to penicillin, cefotaxime, and ceftriaxone has increased over the past few years. Reports of failures of cefotaxime or ceftriaxone in the treatment of children with meningitis caused by resistant S pneumoniae necessitates a revision of Academy recommendations. For nonmeningeal infections, modifications of the initial therapy need to be considered only for patients who are critically ill and those who have a severe underlying or potentially immunocompromising condition or patients from whom a highly resistant strain is isolated. Because vancomycin is the only antibiotic to which all S pneumoniaestrains are susceptible, its use should be restricted to minimize the emergence of vancomycin-resistant organisms. Patients with probable aseptic (viral) meningitis should not be treated with vancomycin. These recommendations are subject to change as new information becomes available.
- CDC =
- Centers for Disease Control and Prevention •
- MIC =
- minimum inhibitory concentration •
- CSF =
- cerebrospinal fluid •
- IV =
- intravenously •
- MBC =
- minimal bactericidal concentration •
- NCCLS =
- National Committee for Clinical Laboratory Standards
ANTIMICROBIAL RESISTANCE IN STREPTOCOCCUS PNEUMONIAE
Epidemiology of Antimicrobial Resistance
The Centers for Disease Control and Prevention (CDC) suggest that the same nomenclature used to define resistance for other bacteria, including methicillin-resistant Staphylococcus aureus and vancomycin-resistant enterococci, be used for S pneumoniae. All organisms with a minimum inhibitory concentration (MIC) equal to or greater than that defined for the intermediate category of resistance will be classified as nonsusceptible (Table 1).Resistant organisms will have an MIC equal to or greater than that defined for the resistant category. These latter organisms were formerly known as highly resistant. Strains in the resistant category for the penicillins, cephalosporins, and other β-lactam antibiotics are of particular concern, because they are generally also resistant to most other antibiotics, including the cephalosporins, trimethoprim-sulfamethoxazole, erythromycin, chloramphenicol, and tetracycline.1 In children older than 2 months, Streptococcus pneumoniae is the most common cause of invasive bacterial infections. The prevalence of infections caused by penicillin-nonsusceptible S pneumoniae for all ages has increased in the United States during the last 5 years (Figure).2-6 In recent years, in children, the proportion of penicillin-nonsusceptible invasive isolates has varied from 0% to 41%. Of thesenonsusceptible isolates, 5% to 21% were resistant (Table 1). The incidence of cefotaxime- and ceftriaxone-nonsusceptible isolates has increased to 20% in some areas (Table 2).7-25
Mechanism of Antimicrobial Resistance
Penicillins, cephalosporins, and other β-lactam antibiotics killS pneumoniae by binding irreversibly to high molecular weight enzymes located in the bacterial cell wall. These enzymes, also known as penicillin-binding proteins, are responsible for synthesizing peptidoglycan for new cell wall formation. Chromosomal gene changes can alter the structure of these enzymes, thereby decreasing the binding affinity for penicillin and the cephalosporins and resulting in resistance.26 Resistance of S pneumoniae to penicillin and cephalosporins is not mediated by β-lactamase enzymes. Consequently, treatment of pneumococcal infections resistant to these antibiotics with β-lactamase-resistant drugs, such as extended spectrum cephalosporins or combinations of broad spectrum penicillins plus clavulanate or sulbactam, offers no advantage.
The highest levels of resistance reported have been MICs of 8 to 16 μg/mL for penicillin27 and 16 to 32 μg/mL for cefotaxime and ceftriaxone.28-30 Although uncommon, resistance also has been described for clindamycin, rifampin, and imipenem.
In developing this statement, the Committee reviewed all available information relative to several areas: (1) prevalence of β-lactam-resistant pneumococcal disease in children in the United States; (2) optimal laboratory diagnosis of drug-resistant pneumococci; (3) in vitro antimicrobial susceptibilities ofnonsusceptible pneumococci, particularly those using time-kill methods; (4) pharmacokinetic and efficacy studies of antimicrobials for the treatment of meningitis in animal models; (5) pharmacokinetic studies of antimicrobials alone and in combination in children with meningitis; (6) case reports; (7) retrospective and prospective studies of nonsusceptible pneumococcal infections in children examining both bacteriologic and clinical outcome variables; and, (8) the effect of dexamethasone on cerebrospinal fluid (CSF) antimicrobial concentrations in animal models and in children.
A med-line search initially identified 250 articles for review. Additional articles published during the time of statement development as well as lectures and abstracts presented at regional and national meetings during this time were included in the Committee's deliberations. Additional articles were identified from personal files of Committee members, CDC reports, and bibliographies of articles. After evaluating all the articles for relevance and validity, 160 were selected for complete review.
The recommendations were drawn from an analysis of this literature and were augmented by expert consensus opinion. In areas for which the literature did not provide strong scientific evidence to support specific recommendations, the Committee used nominal group process to achieve consensus.
Expert consultants, including all of the Committee liaison representatives, reviewed the recommendations on several occasions. They were also reviewed by the AAP Committee on Practice and Ambulatory Medicine and The CDC Drug Resistant Streptococcus pneumoniaeTherapeutic Working Group.
Antimicrobial Susceptibility Testing
Pneumococci, which are nonsusceptible to penicillin, cefotaxime, ceftriaxone, and other antimicrobial agents, are divided into the categories of intermediate and resistantbased on the predicted ability of these drugs to treat pneumococcal infections effectively (Table 1).31 β-Lactam antibiotic concentrations in CSF adequate to treat nonsusceptibleorganisms may not be achieved consistently.32 Serum concentrations of antibiotics after parenteral administration are much higher. Therefore, for nonmeningeal infections, organisms usually may be treated successfully with those β-lactam antibiotics to which they are nonsusceptible.33 For someresistant organisms, however, a change of antimicrobial may be necessary depending on the clinical course.
All pneumococcal isolates from normally sterile body sites should be tested for penicillin, cefotaxime, and ceftriaxone susceptibility using the guidelines of the National Committee for Clinical Laboratory Standards (NCCLS).31
Manual agar or broth dilution techniques provide the most accurate quantitative results and should be used whenever possible.34 Many hospital laboratories are not equipped to perform these techniques and will depend on the E test (Epsilometric test, AB Biodisk North America Inc., Culver City, CA). The E test is a plastic strip containing an antimicrobial concentration gradient. It is simple to use and reasonably reliable for determining MIC values.25,34 However, the separation between susceptible, intermediate, and resistant categories may not always be distinct, and interpretation may vary with the observer.34,35 If the patient is not responding as well as anticipated, the possibility that the E test could have been misread by one MIC value should be considered. Other test methods approved by the Food and Drug Administration for the quantitative susceptibility testing of pneumococci include Microtech, Pasco, or Sensititre. Other automated test methods are not reliable and should not be used.36
Most laboratories using NCCLS guidelines for susceptibility testing will need 48 to 72 hours to provide results for pneumococcal antibiotic susceptibilities.31 To abbreviate this process, when the CSF smear shows characteristic gram-positive diplococci, some laboratories place the penicillin and cefotaxime (or ceftriaxone) E test strips directly on the agar plates at the time of CSF inoculation. This procedure is not standardized, and the results should be confirmed using the standard NCCLS protocols.
When quantitative testing methods are not available, the qualitative screening test using a 1-μg oxacillin disk reliably identifies all penicillin-susceptible pneumococci that have a disk zone diameter of 20 mm or greater. For organisms with an oxacillin disk zone size less than 20 mm, indicating potential nonsusceptibility to penicillin, additional quantitative susceptibility testing must be performed. Up to 40% of these organisms will be susceptible to penicillin (although resistant to oxacillin), and the size of the zone around the disk does not accurately distinguish between organisms with an MIC defined asintermediate or resistant (Table1).14,37
If the organism is penicillin-nonsusceptible by any form of testing, additional quantitative susceptibility testing should be performed for cefotaxime or ceftriaxone. The MIC results for either drug adequately parallel each other; thus, only one needs to be tested. Strains resistant to penicillin are more likely to beresistant to cefotaxime and ceftriaxone. Furthermore, clones of pneumococci have been isolated that are much more resistant to cefotaxime and ceftriaxone than to penicillin.23,29,30Susceptibility testing also should be determined for other clinically relevant drugs for meningitis treatment, which should include vancomycin, meropenem, and rifampin. For nonmeningeal invasive infections, susceptibility testing should be performed for erythromycin, trimethoprim-sulfamethoxazole, clindamycin, cefuroxime and potentially imipenem, meropenem and chloramphenicol (Table 1). Strains resistant to erythromycin are also resistant to the related antibiotics clarithromycin and azithromycin.
Isolation of S pneumoniae with a vancomycin MIC greater than 1 μg/mL should be reported immediately to the state health department. Because no isolates of S pneumoniae with an MIC this high have been identified, the state health department will arrange for confirmation testing. Identification of such isolates is critical for both local and national surveillance programs and for providing feedback to area physicians. For the same reasons, organismsresistant to penicillin and cefotaxime, or ceftriaxone should also be reported to the state health department.
Distinguishing Aseptic (Viral) From Bacterial Meningitis
Many more cases of aseptic meningitis occur than pneumococcal meningitis. In 1993, 12 848 cases of aseptic meningitis were reported to the CDC.38 Because most physicians do not report such cases, it is probable that a more accurate annual incidence figure would be at least threefold higher, or approximately 30 000 cases per year. The estimated incidence of pneumococcal meningitis for all age groups is 1 to 2 cases per 100 000 population.39 For the population of the United States, this represents approximately 3000 to 5000 cases per year. Thus, there will be 6 to 10 cases of aseptic meningitis for every case of pneumococcal meningitis, particularly during the summer and fall.
To prevent the development of vancomycin-resistant S pneumoniae, it is imperative that use of vancomycin be minimized. Children with suspected or proved aseptic meningitis should not receive vancomycin. For children pretreated with antibiotics or those for whom the clinical and CSF findings are equivocal, the decision of whether to begin antibiotic therapy is more difficult.40Several indices may be helpful in reaching a decision. For younger children with pneumococcal meningitis, 90% to 100% of the time, the Gram-stained smear of the CSF is positive.41 In older children and adults, only 50% to 68% of Gram-stained smears of CSF are positive.42,43 The latex agglutination test using CSF is more rapid and more sensitive than Gram-staining,44 but a negative test does not rule out a diagnosis of pneumococcal meningitis. Preceding antibiotic therapy is unlikely to change CSF parameters sufficiently to interfere with the ability to distinguish viral from suspected bacterial meningitis.45-47 For the young infant, the distinction between aseptic and bacterial meningitis may be particularly difficult.48 A total protein concentration of 100 mg/dL or greater is rare in infants with aseptic meningitis and may be helpful in separating bacterial from viral meningitis.
In some circumstances, distinguishing between viral and early bacterial meningitis in children older than 1 month continues to be difficult. In most settings, a bacterial cause will be responsible in less than half of these children. The likelihood of children having bacterial meningitis caused by S pneumoniae resistant to cefotaxime and ceftriaxone is less than 20% in most areas. If previously administered oral or intramuscular antibiotics suppress the presence of the organism on a Gram-stained smear, resistance of the causative organism to the high doses of cefotaxime or ceftriaxone administered for meningitis will be unlikely.
Initial Therapy for Children With Bacterial Meningitis
Optimal management of bacterial meningitis requires treatment with antimicrobial agents at dosages that achieve concentrations providing bactericidal activity in CSF diluted 1:8.49
Antimicrobials of Potential Use for the Treatment of Pneumococcal Meningitis
The recommended dosage of penicillin G for pneumococcal meningitis is 250 000 to 400 000 U/kg/d (150 to 240 mg), given intravenously (IV) in four to six divided doses (Table 3). A dose of 250 000 U/kg/d (150 mg) given in six divided doses results in mean CSF concentrations of 0.8 μg/mL sustained throughout the 4 hours between infusions.50 For penicillin-susceptible pneumococci (MIC ≤0.06 μg/mL), this concentration of penicillin approaches the desired bactericidal concentration in the CSF. However, for penicillin-nonsusceptible pneumococci (MIC ≥0.1 μg/mL), penicillin concentrations in the CSF will be inadequate to kill the organism.
Cefotaxime and Ceftriaxone
In patients with meningitis caused by penicillin-nonsusceptible pneumococci, the treatment drug of choice is cefotaxime or ceftriaxone, providing the organism is susceptible. All other cephalosporins have higher MICs for penicillin-nonsusceptible organisms and should not be administered for meningitis. In a recent study, concentrations of ceftriaxone in the CSF were 0.9 to 30 μg/mL 2.8 hours after administration of recommended IV doses.32 These concentrations are high enough for successful therapy of meningitis caused by susceptible organisms.
Therapeutic failures occur when cefotaxime or ceftriaxone is used for strains resistant to these antibiotics (MIC ≥ 2 μg/mL).9,28,29,51-53 When the MIC is 1.0 μg/mL for cefotaxime or ceftriaxone, indicating an intermediate level of resistance, some patients have been treated successfully with cefotaxime or ceftriaxone alone.54 The response is unpredictable, however, possibly because of individual variations in the CSF concentrations of cefotaxime or ceftriaxone or in the E test interpretations.
Administration of 100 mg/kg of ceftriaxone once a day is effective for the treatment of meningitis caused by penicillin-susceptible S. pneumoniae.32,55,56 For penicillin-nonsusceptible but ceftriaxone-susceptible organisms, administration twice a day may be preferred. If a once-a-day regimen is chosen, the full dose must be given at the same time every day.
The recommended dosage of cefotaxime for meningitis is 200 to 225 mg/kg/d IV in three to four divided doses. Some experts recommend initiating therapy with dosages as high as 300 mg/kg/d given in four divided doses until susceptibility results are available.57,58 The higher serum and CSF concentrations achieved by high-dose therapy may be beneficial for patients with organisms in the intermediate category. Forresistant strains, however, these concentrations may not be high enough for therapy to be successful.
A limited number of alternatives are available for the treatment of meningitis caused by pneumococci that are nonsusceptibleto penicillin and cefotaxime and ceftriaxone. In children, vancomycin enters the CSF in the presence of inflamed meninges to provide concentrations of 2.0 to 5.9 μg/mL within 2.8 hours of an IV dose.32 Although all strains of pneumococci currently tested have been susceptible in vitro to vancomycin, failures of vancomycin to treat meningitis or to prevent its development or progression have been reported when dosages less than 60 mg/kg/d were administered.59,60 These reports are supported by studies in a rabbit model that suggest bactericidal levels of vancomycin in the CSF may be difficult to maintain.61 For these reasons, vancomycin should not be used alone for therapy of meningitis.
Combination Vancomycin–Cefotaxime (or Ceftriaxone) Therapy
For pneumococci that are nonsusceptible to cefotaxime and ceftriaxone, combination therapy of vancomycin plus cefotaxime or ceftriaxone produces a synergistic effect in vitro,62 in the animal model,61 and in the CSF of children with meningitis.32 Thus, for initial empirical therapy, this combination should be highly effective and may prevent the emergence of resistance to either drug alone.63
Rifampin and Combination Therapy Using Rifampin With Vancomycin or Cefotaxime (or Ceftriaxone)
Rifampin is active against most, but not all, penicillin-nonsusceptible pneumococci.26Rifampin should never be used alone, because strainsresistant to rifampin have been isolated, and in some settings, resistance can develop rapidly during therapy.64Whereas β-lactam antibiotics are rapidly bactericidal, rifampin is only slowly bactericidal against S pneumoniae in vitro.62 In one rabbit model of meningitis, the combination of vancomycin and rifampin was slowly bactericidal, and 13 of 15 animals had sterile CSF cultures after 24 hours.61 For the combination of ceftriaxone and rifampin in the same model, a bactericidal effect was achieved in 24 hours, even against a strainresistant to ceftriaxone (MIC of 4 μg/mL).65In this model, the addition of rifampin to ceftriaxone was as effective as the addition of vancomycin to ceftriaxone in treating cefotaxime- and ceftriaxone-nonsusceptible pneumococcal meningitis.
There is a discrepancy between the poor in vitro bactericidal activity of rifampin when combined with cefotaxime, ceftriaxone, or vancomycin and the apparent efficacy of these combinations in the rabbit pneumococcal meningitis model.66 More information is needed to clarify the therapeutic efficacy of rifampin-containing antibiotic combinations for drug-nonsusceptible pneumococcal meningitis. Until such information is available, rifampin should be added only if the organism is demonstrated to be susceptible, and there is a delay in the expected clinical or bacteriologic response.
Chloramphenicol, when administered at 50 to 100 mg/kg/d IV in four divided doses, achieves effective concentrations in the CSF for penicillin-susceptible pneumococci. For children with penicillin-nonsus-ceptible, chloramphenicol-susceptible pneumococcal strains, however, treatment with chloramphenicol has often not been successful.67 These failures may occur, because strains having chloramphenicol MICs indicating susceptibility may have minimal bactericidal concentrations (MBCs) of 8 μg/mL or greater, indicating resistance to killing. Adequate bactericidal activity cannot be achieved in the CSF when the MBC values are this high. Thus, chloramphenicol should not be used to treat penicillin-nonsusceptible pneumococci unless the causative strain is known to have a chloramphenicol MBC value of 4 μg/mL or less.67 Most laboratories need at least 3 to 4 days to obtain this information, because the organism must be sent to a reference laboratory. Inadequate data are available to support the use of chloramphenicol in combination with other antimicrobial agents for management of penicillin-nonsusceptible pneumococcal meningitis.
Imipenem and Meropenem
Imipenem is a carbapenem antibiotic with a broad spectrum of activity against a variety of organisms, including penicillin-nonsusceptible pneumococci. Although most strains of S pneumoniae are susceptible to imipenem, resistant strains have been isolated.19,25 The potential epileptogenic properties of imipenem in patients with disease of the central nervous system, including meningitis, preclude routine use of this antibiotic.68
Meropenem is a carbapenem antimicrobial similar to imipenem and recently approved by the Food and Drug Administration for treatment of bacterial meningitis in children 3 months of age or older. The epileptogenic potential is much less than that of imipenem, and in most cases, meropenem will be active against isolates nonsusceptibleto penicillin. Although clinical experience is limited, for meropenem-susceptible isolates, meropenem alone or in combination may provide a satisfactory alternative for patients who do not tolerate vancomycin.69
Use of Dexamethasone for Patients With Pneumococcal Meningitis
Current recommendations suggest that dexamethasone beconsidered for children with pneumococcal meningitis.70 The effectiveness of dexamethasone for preventing sequelae of pneumococcal meningitis is unproved, and expert opinion is divided on its use.71-73 No large, single prospective controlled study of the use of dexamethasone to prevent hearing loss or other sequelae in children with pneumococcal meningitis has been performed.
Cell wall components from organisms that have been rapidly lysed by antibiotics are believed to induce the release of cytokines into the subarachnoid space. These cytokines seem to play a seminal role in the initial events of meningeal inflammation.71 To prevent or modify cytokine release in pneumococcal meningitis, dexamethasone, if used should be given shortly before or at the time of antibiotic administration.71,72
In the rabbit model of meningitis, dexamethasone interferes with the penetration of ceftriaxone and vancomycin into the CSF when either is administered alone.43,65 When the antibiotics were given together with dexamethasone in one model, the combination was effective against a pneumococcal strain with a ceftriaxone MIC of 1 μg/mL.65 For a resistant strain, however, dexamethasone reduced the bactericidal efficacy of the combination. Dexamethasone did not affect the penetration of rifampin, and the combination of ceftriaxone plus rifampin effectively eradicated this same resistant strain.65
The penetration of ceftriaxone and vancomycin into the CSF in children32 is better than that demonstrated in the rabbit model.65 In children with bacterial meningitisreceiving dexamethasone, the CSF ceftriaxone concentrations have been reported to be 0.9 to 30 μg/mL32 and 0.7 to 9.2 μg/mL.74 These concentrations are similar to those reported previously in the absence of dexamethasone.75 The vancomycin levels in the CSF were 2.0 to 5.9 μg/mL and represent 20% of serum levels,32 much better than the 3% penetration reported for the animal model.65 Rifampin levels in the CSF were 0.3 to 1.9 μg/mL.32 The CSF of children who received dexamethasone and the antibiotic combinations of ceftriaxone plus vancomycin or ceftriaxone plus rifampin, when incubated in vitro with pneumococcal strains resistant to ceftriaxone, had significantly enhanced bactericidal activity when compared with the CSF from a child who received dexamethasone and ceftriaxone alone, indicating a significantly enhanced effect from either combination.32
Dexamethasone therapy can decrease fever, giving a false impression of clinical improvement, even though CSF sterilization has not been achieved.9,25,51 If dexamethasone is administered, careful and frequent observation of the patient is indicated.28,51,59
Assessment of Efficacy of Therapy After 24 to 48 Hours
Penicillin-nonsusceptible organisms do not cause disease of greater severity than disease caused by penicillin-susceptible strains.33 To assess the efficacy of therapy, the clinical course should be followed closely. A positive blood culture 24 to 36 hours after initiating therapy can indicate a hidden focus of infection, a resistant organism, or continued bacterial replication in the CSF. Another lumbar puncture should be considered within 24 to 48 hours to document eradication of the pathogen, particularly if (1) the organism is demonstrated to be potentiallynonsusceptible to penicillin by quantitative (MIC) or oxacillin disk testing, the cefotaxime or ceftriaxone susceptibility results are not yet available, and the clinical condition has not improved or has worsened; or (2) the child has received dexamethasone, which might interfere with the interpretation of the clinical response. Some experts believe that a second lumbar puncture is not necessary if therapy was initiated with vancomycin plus cefotaxime (or ceftriaxone), and the clinical response has been good, particularly if dexamethasone was not used.
Continuation of Antimicrobial Therapy
To curtail continued emergence of antimicrobial resistance, it is imperative that antibiotic therapy be reviewed as soon as the quantitative susceptibility test results are available (Table4). If the organism is susceptible to penicillin, cefotaxime, or ceftriaxone, vancomycin should be discontinued and penicillin, cefotaxime, or ceftriaxone continued for the usual course of therapy.40 If resistance to penicillin and cefotaxime and ceftriaxone is documented, vancomycin plus cefotaxime or ceftriaxone should be continued for the full course of therapy. If the patient's clinical condition has not improved or has worsened while receiving this combination of antibiotics or if a follow-up CSF examination indicates failure to reduce the number of organisms substantially or to eradicate the organism, some experts would add rifampin to the vancomycin and cefotaxime or ceftriaxone combination.11 Other experts would discontinue vancomycin and continue therapy with ceftriaxone plus rifampin, providing the organism is susceptible to rifampin.10
NONMENINGEAL INVASIVE INFECTIONS
Although more than 100 nonmeningeal invasive infections (excluding otitis media) caused by penicillin-nonsusceptiblepneumococci have been reported since 1977,7,15,33,76-78reports of treatment failures using standard therapy have been exceedingly rare. This is most likely a result of the high concentrations of antibiotics achieved in serum. This is particularly true for cefotaxime and ceftriaxone, in which serum concentrations after recommended doses often exceed by 100-fold the MIC for S pneumoniae strains with an MIC of 0.1 to 1 μg/mL for these antibiotics.32
Only one prospective study comparing the outcomes of nonmeningeal invasive infections caused by both susceptible andnonsusceptible pneumococci has been performed in children. In this study, the outcome of nonmeningeal infections caused by penicillin-nonsusceptible organisms correlated with the severity of the illness on admission and the presence of underlying disease and not on the susceptibility of the organism to penicillin.33 This finding may result from the fact that most of the reported infections were caused by organisms with an MIC of 0.1 to 1 μg/mL. Of interest, four patients with pneumonia attributable to organisms with MICs in this range received only oral amoxicillin (peak serum levels of 6 to 14 μg/mL), and the disease in all patients responded rapidly.33 Some infections caused byresistant pneumococci with an MIC of 4 μg/mL or greater have responded to ampicillin (100 mg/kg/d, IV).33 Carefully controlled prospective studies evaluating outcome variables for patients with nonmeningeal infections caused by pneumococcinonsusceptible to the β-lactam antibiotics are needed.
For immunocompetent patients with possible invasive pneumococcal infections who do not have meningitis and are not critically ill on admission, standard antibiotic therapy that may include cefuroxime should continue to be used. Therapy does not need to be altered if the organism is reported to have a penicillin MIC of 0.1 to 1 μg/mL and the patient is responding well. If the organism isresistant to penicillin (MIC ≥ 2.0 μg/mL), changes in antimicrobial therapy should be based on the clinical response and not on the MIC value. Additional initial antibiotic coverage for potential penicillin-nonsusceptible strains could be considered for patients who are critically ill, including those with myopericarditis, severe multilobar pneumonia with hypoxia, or hypotension.
Immunocompromised children may be at increased risk for severe infections caused by drug-nonsusceptible pneumococci, because they often receive frequent courses of antimicrobials for therapy or prophylaxis, allowing for the selection of resistant strains. Immunocompromising conditions that place children at risk for severe pneumococcal disease include sickle cell disease, other hemoglobinopathies, congenital or acquired immunoglobulin deficiencies, agammaglobulinemia, nephrotic syndrome, human immunodeficiency virus infection, immunosuppressive medications, or congenital or acquired asplenia.
For children who are immunocompromised as a result of asplenia or other altered immune function, it is not yet known whether serum levels of penicillin, cefotaxime, or ceftriaxone are adequate to treat nonmeningeal pneumococcal infections caused bynonsusceptible strains. For critically ill children who are immunocompromised, some experts would initiate empiric therapy with vancomycin and cefotaxime or ceftriaxone until susceptibility results are available.79 Subsequent therapy should be based on test results for antibiotic susceptibility and the patient's clinical course.
Invasive pneumococcal infections are rare but well described in neonates.80 Two neonates with penicillin-nonsusceptible pneumococcal infections have been reported.33,81 The initial empirical antibiotic regimen for neonatal sepsis and meningitis should be adequate for all pneumococcal infections with the exception of meningitis caused by a cefotaxime- or ceftriaxone- and penicillin-nonsusceptible strain. For patients in this age group with CSF smears that show gram-positive diplococci and bacterial antigen tests supporting a diagnosis of pneumococcal meningitis, consideration should be given to the initial empirical addition of vancomycin to the therapeutic regimen.
INFECTION CONTROL MEASURES
Nosocomial acquisition of pneumococcal infections by children and their caregivers is rare in the United States. Antimicrobial-nonsusceptible pneumococcal infections are transmitted by the same routes as susceptible organisms. The CDC recommends Standard Precautions for all patients with invasive pneumococcal infections.82 Standard Precautions synthesize the major features of the former categories of Universal Precautions and Body Substance Isolation. They apply to blood, all body fluids except sweat, nonintact skin, and mucous membranes and “are designed to reduce the risk of transmission of microorganisms from both recognized and unrecognized sources of infection in hospitals.”82 Airborne, Droplet and Contact Precautions are not recommended for invasive pneumococcal infections. For the details of the use of Standard Precautions, the CDC guidelines82 or an Infection Control Practitioner should be consulted. Nasopharyngeal cultures of family members or intimate contacts are not indicated.
1. Children with moderate to severe bacterial infections should have cultures obtained from appropriate, potentially infected, normally sterile body fluids to determine the cause of the infection and allow for susceptibility testing of the organism.
2. Quantitative (ie, MIC) susceptibility testing for penicillin and cefotaxime or ceftriaxone should be performed on all pneumococci isolated from normally sterile body fluids whenever possible. If the patient has meningitis and the organism is nonsusceptible to penicillin, cefotaxime, and ceftriaxone, disk or MIC susceptibility testing for vancomycin, meropenem, and rifampin should be performed. If the patient has a nonmeningeal infection and the organism isresistant to penicillin, cefotaxime, and ceftriaxone, disk, or MIC susceptibility testing for vancomycin, rifampin, clindamycin, trimethoprim-sulfamethoxazole, erythromycin, imipenem, meropenem, and chlor-amphenicol should be considered.
If quantitative susceptibility testing is not available in the laboratory, screening for penicillin resistance using the 1-μg oxacillin disk and NCCLS guidelines should be performed for all pneumococcal isolates from normally sterile body sites. Organisms with an oxacillin disk zone size less than 20 mm should be referred for quantitative susceptibility testing.3 Isolation of S pneumoniae with a vancomycin MIC greater than 1 μg/mL should be reported immediately to the state health department.
Management of Children With Bacterial Meningitis Possibly Caused byS pneumoniae
4. Vancomycin plus cefotaxime or ceftriaxone should be administered initially to all children older than 1 month with definite or probable bacterial meningitis. (Some experts believe that vancomycin need not be used if there is compelling evidence that the cause is an organism other than S pneumoniae, such as Gram-negative diplococci on a smear of CSF during an outbreak of meningococcal disease.)
Because pneumococcal meningitis may occur in infants younger than 1 month, consideration should be given to the addition of vancomycin to the usual antibiotic combination for neonatal sepsis especially if: (1) the CSF smear shows characteristic Gram-positive diplococci or (2) bacterial antigen testing supports a diagnosis of pneumococcal meningitis.
For children with immediate hypersensitivity to the β-lactam antibiotics, the combination of vancomycin plus rifampin should be considered.5 A lumbar puncture should be considered after 24 to 48 hours to evaluate therapy if (1) the organism is penicillin-nonsusceptible by quantitative (MIC) or oxacillin disk testing, the results from cefotaxime and ceftriaxone quantitative susceptibility testing are not yet available, and the child's condition has not improved or has worsened or (2) the child has received dexamethasone, which might interfere with the ability to interpret the clinical response.
6 Once the results of susceptibility testing are available, modifications of therapy should be made as indicated in Table 4. If the organism is susceptible to penicillin or cefotaxime or ceftriaxone,vancomycin should be discontinued, and penicillin or cefotaxime or ceftriaxone should be continued. Vancomycin plus cefotaxime or ceftriaxone should be continued only if the organism isnonsusceptible to penicillin and to cefotaxime or ceftriaxone.
7 Addition of rifampin or substitution of rifampin for vancomycin after 24 to 48 hours of therapy could be considered if the organism is susceptible to rifampin and if (1) after 24 to 48 hours, despite therapy with vancomycin plus cefotaxime or ceftriaxone, the clinical condition has worsened; (2) the follow-up Gram-stained smear or culture of CSF indicates failure to eradicate or to substantially reduce the number of organisms; or (3) the organism has an unusually high cefotaxime or ceftriaxone MIC of 4 μg/mL or greater. Consultation with an infectious disease specialist should also be considered.
Management of Children With Probable Aseptic Meningitis
8. During the summer and fall, children will be hospitalized for severe headache, vomiting, fever, a stiff neck, and CSF pleocytosis caused by enteroviruses. Although antibiotics are not always indicated, some practitioners will choose to treat these children with ceftriaxone or cefotaxime until the CSF and blood cultures are negative even when all laboratory results suggest aseptic meningitis. Vancomycin should not be used under these circumstances unless the child appears toxic and/or is hypotensive.
Management of Nonmeningeal Invasive Pneumococcal Infections Requiring Hospitalization
9. For nonmeningeal invasive infections in the previously well child who is not critically ill, antimicrobials currently in use to treat S pneumoniae and other potential pathogens should be initiated at usually recommended doses.
10. For a select group of children with invasive infections potentially attributable to S pneumoniae who are critically ill, additional initial antibiotic coverage for possible penicillin- and cefotaxime- or ceftriaxone-nonsusceptible strains could be considered. Such patients might include those with myopericarditis or severe multilobar pneumonia with hypoxia or hypotension. If vancomycin is administered, it should be discontinued as soon as antibiotic susceptibilities demonstrate effective alternative agents.
11. If the organism has an MIC of 2 μg/mL or greater to penicillin, cefotaxime, and ceftriaxone, therapy should be adjusted based on the clinical response, susceptibilities to other antimicrobials, and results of follow-up cultures of blood and other body fluids. The local, county, and state health departments should be notified of these organisms and the help of an infectious disease specialist considered.
12. For the child with severe β-lactam allergy, initial management for a potential pneumococcal infection could include vancomycin or clindamycin, in addition to antimicrobial coverage for other potential pathogens as indicated. Continuation of therapy should be guided by the results of susceptibility testing. Vancomycin should not be continued if the organism is susceptible to other appropriate, non-β-lactam antibiotics. Consultation with an infectious disease specialist should be considered.
Management of Nonmeningeal Invasive Pneumococcal Infections in the Immunocompromised Host
13. No modifications need to be made to the current management of possible pneumococcal infections in immunocompromised children, providing they are not critically ill. For critically ill patients, consideration should be given to initiating therapy with vancomycin and cefotaxime or ceftriaxone. Vancomycin should be discontinued as soon as susceptibility results indicate effective alternative antimicrobials.
Use of Vancomycin
14. At the time vancomycin therapy is initiated in children, urinalysis should be performed, and serum creatinine and blood urea nitrogen concentrations should be determined. Renal function should then be followed closely and appropriate dosage modifications made if renal function is compromised.
15. Physicians should consider determining trough serum concentrations for children who have renal impairment, are critically ill, are receiving concurrent ototoxic or nephrotoxic drugs, or are younger than 3 months. Trough concentrations of vancomycin, which can be obtained before the third dose, should be 10 to 15 μg/mL or less.
16. A peak serum vancomycin concentration can be obtained 30 to 60 minutes after completion of a 30-minute infusion for the patient whose condition does not improve or for the patient with an organism resistant to penicillin, cefotaxime, and ceftriaxone. Therapeutic peak serum vancomycin concentrations for meningitis fall within the range of 35 to 40 μg/mL.
Committee on Infectious Diseases, 1995 to 1996
Neal A. Halsey, MD, Chairperson
P. Joan Chesney, MD
Michael A. Gerber, MD
Donald S. Gromisch, MD
Steve Kohl, MD
S. Michael Marcy, MD
Melvin I. Marks, MD
Dennis L. Murray, MD
James C. Overall, Jr, MD
Larry K. Pickering, MD
Richard J. Whitley, MD
Ram Yogev, MD
Georges Peter, MD
Caroline B. Hall, MD
Robert Breiman, MD
National Vaccine Program Office
Stephen C. Hadler, MD
Centers for Disease Control and Prevention
M. Carolyn Hardegree, MD
Food and Drug Administration
Richard F. Jacobs, MD
American Thoracic Society
Noni E. MacDonald, MD
Canadian Paediatric Society
Walter A. Orenstein, MD
Centers for Disease Control and Prevention
N. Regina Rabinovich, MD
National Institutes of Health
Benjamin Schwartz, MD Centers for Disease Control and Prevention
George H. McCracken, Jr, MD
Sheldon L. Kaplan, MD
James H. Jorgensen, PhD
The recommendations in this statement do not indicate an exclusive course of treatment or serve as a standard of medical care. Variations, taking into account individual circumstances, may be appropriate.
- Spratt B
- Doern GV,
- Brueggemann A,
- Holley HP Jr.,
- et al.
- Barry AL,
- Pfaller MA,
- Fuchs PC,
- Packer RR
- Thornsberry C,
- Brown SD,
- Yee C,
- et al.
- Jorgensen JH,
- Doern GV,
- Maher LA,
- Howell AW,
- Redding JS
- Spika JS,
- Facklam RR,
- Plikaytis BD,
- Oxtoby MJ,
- The Pneumococcal Surveillance Working Group
- ↵Kaplan SL, U.S. Pediatric Multicenter Pneumococcal Surveillance Group. Surveillance of pneumococcal infections in children. In: Program and Abstracts of the 36th Interscience Conference on Antimicrobial Agents and Chemotherapy (New Orleans, LA). Washington DC: American Society for Microbiology; 1996:252 #K16
- Paris MM,
- Ramilo O,
- McCracken GH Jr.
- ↵Paris MM, Medley F, Pastor P, Murphy TV. Population-based surveillance for penicillin-and cephalosporin-resistant Streptococcus pneumoniae in Dallas County. In: Program and Abstracts of the 35th Interscience Conference on Antimicrobial Agents and Chemotherapy (San Francisco, CA). Washington, DC: American Society for Microbiology; 1995:89
- Mason EO,
- Kaplan SL,
- Lamberth LB,
- Tillman J
- Tan TQ,
- Mason EO,
- Kaplan SL
- Rathore MH,
- Jenkins SG
- ↵Leggiadro RJ, Davis Y, Hill B, Facklam RR, Tenover FC. Increasing rates of invasive drug-resistant Streptococcus pneumoniae (DRSP) in Memphis. In: Abstracts of the Infectious Disease Society of America 33rd Annual Meeting (San Francisco, CA). September 16–18, 1995:72
- Quinonez JM,
- Mancao MY
- Hagland LA,
- Istre GR,
- Pickett DA,
- Welch DF,
- Fine DP,
- The Pneumococcus Study Group
- Zangwill KM,
- Vadheim CM,
- Eriksen E,
- et al.
- Fairchok, MP, Ashton WS, Fischer GW
- Klugman KP
- McDougal LK,
- Rasheed JK,
- Biddle JW,
- Tenover FC
- ↵National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial susceptibility testing. In: Sixth Informational Supplement: NCCLS Document M100-S6. Villanova, PA: National Committee for Clinical Laboratory Standards; 1995;15(14)
- Klugman K,
- Friedland IR,
- Bradley JS
- Macias EA,
- Mason EO Jr.,
- Ocera HY,
- LaRocco MT
- Skulnick M,
- Small GW,
- Lo P,
- et al.
- Nolte FS,
- Metchock B,
- Williams T,
- Diem L,
- Bressler A,
- Tenover FC
- Centers for Disease Control
- Wenger JD,
- Hightower AW,
- Facklam RR,
- et al.
- Cabellos C,
- Martinez-Lacasa J,
- Martos A,
- et al.
- Schaad UB,
- Krucko J,
- Pfenninger J
- Rorabaugh ML,
- Berlin LE,
- Heldrich F,
- et al.
- Schaad UB,
- McCracken GH Jr.,
- Loock CA,
- Thomas ML
- John CC
- Tan TQ,
- Schutze GE,
- Mason EO,
- Kaplan SL
- Yogev R,
- Shulman ST,
- Chadwick EG,
- David AT,
- Glogowski W
- Congeni BL,
- Bardley J,
- Hammerschlag MR
- Viladrich PF,
- Cabellos C,
- Pallares R,
- et al.
- Viladrich PF,
- Gudiol F,
- Linares J,
- et al.
- Friedland IR,
- Paris MM,
- Ehrett S,
- Hickey SM,
- Olsen KD,
- McCracken GH Jr.
- Friedland IR,
- Paris MM,
- Shelton S,
- McCracken GH Jr.
- Paris MM,
- Hickey SM,
- Uscher MI,
- Shelton S,
- Olsen KD,
- McCracken GH Jr.
- Tubau F,
- Cabellos C,
- Linares J
- Friedland IR,
- Klugman KP
- Klugman KP,
- Dagan R,
- The Meropenem Meningitis Study Group
- ↵Committee on Infectious Diseases, American Academy of Pediatrics. Sect 5. Antimicrobials and related therapy. In: Peter G, ed. 1997 Red Book. Report of the Committee on Infectious Diseases. 24th ed. Elk Grove Village, IL: American Academy of Pediatrics; 1997;000:000
- Prober CG
- Long SS
- ↵Arditi M, The US Pediatric Multicenter Pneumococcal Surveillance Group (PMPSG). Clinical and microbiological characteristics of pneumococcal meningitis in children. In: Program and abstracts of the 36th Interscience Conference on Antimicrobial Agents and Chemotherapy (New Orleans, LA). Washington DC: American Society for Microbiology; 1996:251 #K10
- Gaillard JL,
- Abadie V,
- Cheron G,
- et al.
- Latif R,
- Dajani AS
- Copyright © 1997 American Academy of Pediatrics