key: cord-0040165-oy8tqk1n
authors: Niederman, Michael S.
title: Antibiotic Use in the Mechanically Ventilated Patient
date: 2009-05-15
journal: Mechanical Ventilation
DOI: 10.1016/b978-0-7216-0186-1.50043-0
sha: eb031247df61e2e8ab8d574055f8903b10cab5f1
doc_id: 40165
cord_uid: oy8tqk1n

nan

with short-term ventilation. The risk of VAP is 3% per day for the first 5 days of ventilation, 2% per day for days 6 to 10, and 1% per day for the next 5 days. 3 Although most chronically tracheostomized and ventilated patients develop respiratory infection, the risk is greater for developing tracheobronchitis than for pneumonia, and patients who survive to the point of long-term ventilation have a relatively low daily risk of developing pneumonia, although the cumulative risk is high. 4, 5 When antibiotics are used for most patients, initial therapy is aimed at a broad spectrum of likely pathogens and is thus empiric because the infecting pathogen is often not known. Therapy can be more specifically focused later, on the basis of the results of diagnostic tests. In some cases, initial empiric therapy must be continued because no etiologic pathogen is identified. When a pathogen is recovered, the term appropriate refers to the use of at least one antimicrobial agent that is active in vitro against the etiologic pathogen. 1 The term adequate includes not only appropriate therapy, but also the use of that agent in the correct dose, by the right route, given in a timely fashion, and with penetration to the site of infection. Timely and appropriate antibiotic therapy can improve survival in patients with severe CAP and nosocomial pneumonia, and the benefits are most evident in patients who are not otherwise terminally ill. 1,6-10 However, even with the use of the correct agents, not all patients recover. That some ventilated patients die in spite of microbiologically appropriate therapy is a reflection of the degree of antibiotic efficacy, as well as a reflection of host response capability (which may, in part, have a genetic determination) and the fact that not all deaths are the direct result of infection. In some patients, death is the result of underlying serious illness; the percentage of deaths that result from infection is termed the attributable mortality. In patients with VAP, this rate has been estimated to be as high as 50% to 60%. 10 However, the use of timely and appropriate antimicrobial therapy can reduce attributable mortality to as low as 20%.

The need for initial therapy to be accurate has led to the frequent and prolonged use of multiple antibiotics in critically ill patients; this approach, in turn, has promoted the development of multidrug-resistant (MDR) pathogens, which are common in many infections. 11 Modern management of these patients requires initial broad-spectrum therapy, followed by a narrowing and focusing of therapy as clinical and culture data become available. This "de-escalation" approach is often accompanied by efforts to use short-duration therapy. 12 Numerous guidelines for empiric therapy for both CAP and nosocomial pneumonia have been developed, but several caveats should be remembered. 1, 6 First, although current guidelines for empiric therapy are evidence based, very few outcome studies have been conducted to demonstrate the utility of these guidelines in improving mortality and other outcomes. Second, guidelines must be reevaluated relative to local patterns of antibiotic susceptibility. In the case of CAP, the emergence of penicillin-resistant pneumococcus, community-acquired methicillin-resistant Staphylococcus aureus (MRSA), and epidemic viral illness (influenza, severe acute respiratory syndrome [SARS] ) may affect the selection of initial therapy, particularly if resistance is prevalent in a specific community. In the setting of VAP, each hospital has its own unique flora and antibiotic susceptibility patterns; knowledge of such patterns is essential. 1, 13 

When an antibiotic interferes with the growth of bacteria, it does so by undermining the integrity of the cell wall or by interfering with bacterial protein synthesis or common metabolic pathways. The effect is termed either bactericidal or bacteriostatic. These broad categorizations may not apply for a given agent against all organisms, however. 14 Bactericidal antibiotics kill bacteria, generally by inhibiting cell wall synthesis or by interrupting a key metabolic function of the organism. Agents of this type include the penicillins, cephalosporins, aminoglycosides, fluoroquinolones, vancomycin, daptomycin, rifampin, and metronidazole. Bacteriostatic agents inhibit bacterial growth, do not interfere with cell wall synthesis, and rely on host defenses to eliminate bacteria. Agents of this type include the macrolides, tetracyclines, sulfa drugs, chloramphenicol, linezolid, and clindamycin. The use of specific agents is dictated by the susceptibility of the causative organism, at a given location, to individual antibiotics. However, when neutropenia is present, or if there is accompanying endocarditis or meningitis, the use of a bactericidal agent is preferred. Thus, for most patients with pneumonia, it is not essential to choose a bactericidal agent. One additional consideration is that certain organisms can produce toxins, and antibiotics that inhibit protein synthesis (linezolid, clindamycin) may have an advantage in toxin-mediated illnesses, such as those caused by certain strains of S. aureus, when compared with cell wall-active bactericidal antibiotics. 15 The antimicrobial activity of a specific agent is often described by the terms minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC). MIC refers to the minimum concentration of an antibiotic that inhibits the growth of 90% of a standard sized inoculum, leading to no visible growth in a broth culture. At this concentration, not all the bacteria have necessarily been killed. MBC refers to the minimum concentration needed to cause a 3-logarithmic decrease (99.9% killing) in the size of the standard inoculum; generally, all pathogenic bacteria are killed at this concentration. The MIC is used to define the sensitivity of a pathogen to a specific antibiotic, under the assumption that the concentration required for killing (the MIC) can be reached in the serum in vivo. However, these terms must be interpreted cautiously in patients with pneumonia because the MIC data do not consider the penetration of an agent into lung tissues and sites of infection. Thus, if an agent achieves infection site concentrations that exceed serum concentrations, the efficacy may be better than predicted by the MIC data. Concerns about antimicrobial resistance have led to the emergence of a newer term, the mutant prevention concentration (MPC). 16 The MPC is defined as the lowest concentration of an antimicrobial that prevents bacterial colony formation from a culture containing greater than 10 10 bacteria. At lower than MPC concentrations, spontaneous mutants can persist and be enriched among the organisms that remain during therapy. The clinical relevance of this concept is still uncertain.

The concentration of an antibiotic in the lung depends on the permeability of the capillary bed at the site of infection (the bronchial circulation), the degree of protein binding of the drug, and the presence or absence of an active transport site for the antibiotic in the lung. 17 In the lung, the relevant site to consider for antibiotic penetration is controversial, but concentrations in lung parenchyma, epithelial lining fluid, and cells such as macrophages and neutrophils are probably important in pneumonia, whereas bronchial concentrations may be more important in tracheobronchitis. In addition, certain organisms are intracellular, such as Legionella pneumophila and Chlamydophila pneumoniae; for these organisms, macrophage concentrations may be important. Conditions at the site of infection may also be important, and some agents can be inactivated by certain local conditions. Aminoglycosides have reduced activity at acidic pH, which may be present in infected lung tissues. In addition, certain bacteria develop resistance by producing destructive enzymes (such as β-lactamases), by altering the permeability of the outer cell wall, by changing the target site of antimicrobial action, or by pumping (efflux) of the antimicrobial from the interior of the cell. In all these conditions, a high local concentration of antimicrobial agent may help offset the bacterial resistance mechanisms.

The concentration of an antibiotic in lung parenchyma depends on its penetration through the bronchial circulation capillaries. The bronchial circulation has a fenestrated endothelium, so antibiotics penetrate in proportion to their molecular size and protein binding. Small molecules that are not highly protein bound pass readily into the lung parenchyma. Penetration through the bronchial circulation is inflammation dependent, whereas the pulmonary vascular bed has a nonfenestrated epithelium, which is inflammation independent; penetration is best for lipophilic agents. 17 These lipophilic drugs include chloramphenicol, the macrolides (including the azalides and ketolides), linezolid, clindamycin, the tetracyclines, the quinolones, and trimethoprim-sulfamethoxazole. Agents that are poorly lipid soluble are inflammation dependent for their entry into the epithelial lining fluid and include the penicillins, cephalosporins, aminoglycosides, vancomycin, carbapenems, and monobactams. 18 If the volume of distribution exceeds 3 L, this finding implies distribution outside the plasma. 16 Some lipophilic agents, such as the macrolides and quinolones, are distributed extensively to body tissues, and the serum levels underestimate their effect at sites of infection, especially for drugs that achieve a high intracellular concentrations in phagocytes. The volume of distribution can also be increased by obesity. Therefore, dosing based on ideal body weight may lead to underdosage, but basing doses of hydrophilic antibiotics on total body weight may result in overdosage. 16 

For some drugs, such as the aminoglycosides, penetration into the epithelial lining fluid and lung secretions is not optimal, and systemic therapy of pneumonia is often less effective than treatment of bacteremia. For this reason, these agents have been studied for delivery through the aerosol route. Local administration of antimicrobials has been used in the therapy of bronchiectasis, especially in the setting of cystic fibrosis, and in the treatment of VAP. Direct delivery of antibiotics is usually achieved by nebulization. This approach achieves high intrapulmonary concentrations, and it may do so without substantial systemic absorption. Thus, the risk of systemic toxicity is reduced. The use of this approach in mechanically ventilated patients is somewhat anecdotal, and not carefully studied, but it has been proposed for patients with either infectious tracheobronchitis or VAP. 19, 20 Both infections can involve highly resistant gram-negative bacteria, and the local delivery of antibiotics may effectively treat some pathogens that cannot be eradicated by systemic therapy.

In mechanically ventilated patients, local antibiotic administration, by instillation or nebulization, has been used to prevent pneumonia. In general, this approach is not recommended, because even when it has been successful, there has been concern about the emergence of MDR gram-negative bacteria in patients who subsequently do develop infection, and these organisms may be difficult to treat. Only one prospective randomized trial has examined the impact of the adjunctive use of locally instilled tobramycin with intravenous agents in the management of VAP. 21 Although the addition of endotracheal tobramycin did not improve clinical outcome compared with placebo, microbiologic eradication was significantly greater in the patients receiving aerosolized antibiotics.

In spite of these data, uncontrolled case series have shown that when patients have VAP caused by MDR Pseudomonas aeruginosa or Acinetobacter species, aerosolized aminoglycosides, polymyxin, or colistin may be helpful as adjuncts to systemic antibiotics. 1, 19, 20 One side effect of aerosolized antibiotics has been bronchospasm, which can be induced by the antibiotic or the associated diluents present in certain preparations. A specially formulated preparation of tobramycin for aerosol administration is available and may avoid this complication.

Although the optimal method of administration of aerosol therapy is unknown, most studies have shown that nebulization can be effective and can achieve more uniform distribution than direct instillation. When aerosol therapy is used in mechanically ventilated patients, it must be carefully synchronized with the ventilator cycle, and the optimal delivery device is not yet defined. In an animal model, investigators found that by using an ultrasonic nebulizer placed in the inspiratory limb of the ventilator circuit, proximal to the Y-connector, up to 40% of the administered dose could be retained in the lung; the tissue concentrations achieved were given systemically, and systemic absorption was minimal. 22, 23 To optimize delivery, inspiratory time may need to be as high as 50% of the ventilatory cycle, and routine humidification should be stopped during antibiotic administration. In ventilated patients, the ventilator may need to be set with a tidal volume of 8 to 10 mL/kg, with no humidification system in use during the use of the ultrasonic nebulizer, which should be set to deliver 8 L/minute.

Pharmacokinetics is the study of the absorption, distribution, and elimination of a drug in the body. This information can be used to describe the concentration in serum. Pharmacokinetics also includes the study of the concentration at other sites of the body, including the site of infection and the relationship between drug concentrations and their pharmacologic or toxic effect. 16 For antibiotics, this means the relationship of antibiotic concentrations at the site of infection, compared with the MIC of the target organism. Pharmacodynamics refers to the action of a drug on the body, including its therapeutic effect.

Some antibiotics kill bacteria in relation to how long their concentration stays at levels higher than the MIC of the target organism (time-dependent killing), whereas other agents are effective in relation to the peak concentration achieved (concentration-dependent killing). 16 If antibiotic killing is time dependent, dosing schedules should be chosen to achieve the maximal time greater than the MIC of the target organism. However, for many organisms, the concentration of the antibiotic needs to be greater than the MIC for only 40% to 50% of the dosing interval, and possibly for as little as 20% to 30% of the interval in the case of carbapenems. Antibiotics of this type include the β-lactams (penicillins and cephalosporins), carbapenems, aztreonam, macrolides, and clindamycin. The rate of killing is saturated once the antibiotic concentration exceeds four times the MIC of the target organism. Continuous infusion of β-lactams is under study to optimize time-dependent killing with these agents.

When bacterial killing is concentration dependent, activity is determined by the degree of concentration achieved at the site of infection and the size of the area under the curve (AUC; of drug concentration plotted versus time) in relation to the MIC of the target organism. Alternatively, the action of these agents can be described by how high the peak serum concentration (Cmax) is in relation to the organism MIC. Classic agents of this type include the aminoglycosides and the fluoroquinolones, but the ketolides are also concentration-dependent antibiotics. 16 For these types of agents, the optimal killing of bacteria is defined by the ratio of AUC to MIC, often referred to as the area under the inhibition curve (AUIC). The target AUIC for gram-negative bacteria is 125 or greater. 24 For both the aminoglycosides and quinolones, some studies have shown that efficacy can also be defined by the Cmax/MIC ratio, aiming for a target of 12 for quinolones against pneumococcus. 25 Optimal use of these agents would entail infrequent administration but with high doses, the underlying principle behind the once-daily administration of aminoglycosides. With once-daily aminoglycoside dosing, the patient achieves a high peak concentration (maximal killing) and a low trough concentration (minimal nephrotoxicity); this regimen relies on the postantibiotic effect (PAE) to maintain the efficacy of the antibiotic after the serum (or lung) concentrations fall to less than the MIC of the target organism. If an antibiotic has a PAE, it is capable of suppressing bacterial growth even after its concentration falls to less than the MIC of the target organism. In clinical practice, the use of once-daily aminoglycoside dosing has had variable benefits in both efficacy and toxicity. 26 Most of the agents that kill in a concentration-dependent fashion have a prolonged PAE, whereas agents with little or no PAE against gram-negative bacteria are generally also agents that kill in a time-dependent fashion, hence they are given several times daily.

Although it is necessary to initiate prompt and accurate antibiotic therapy for the patient with severe pneumonia, initial therapy is usually empiric because it is often impossible to identify a specific etiologic agent when the pneumonia is first diagnosed. The choice of empiric therapy is based on knowledge of the likely etiologic pathogen for each clinical setting, modified by knowledge of local patterns of bacteriology and the prevalence of specific types of antimicrobial resistance in a given region or a specific hospital setting. The American Thoracic Society and the Infectious Diseases Society of America have developed algorithms for initial empiric therapy of severe pneumonia arising in both the community and the hospital. 1, 6, 27 Severe Community-Acquired Pneumonia

The primary etiologic pathogen for severe CAP is pneumococcus, and patients requiring mechanical ventilation to treat CAP should be treated for possible drug-resistant Streptococcus pneumoniae (DRSP), along with other likely pathogens including L. pneumophila. Other organisms to be treated include Haemophilus influenzae, Mycoplasma pneumoniae, C. pneumoniae, aspiration organisms (which usually are enteric gram-negative organisms more than anaerobes), and aerobic gram-negative bacilli (including P. aeruginosa). 6, 27 There has been some controversy about whether enteric gram-negative bacteria are common in CAP, and the identified risk factors have included features that would reclassify some affected patients as having HCAP and not CAP. Thus, patients admitted from a nursing home or dialysis center, or those who have been hospitalized in the previous 90 days, should be treated by the HCAP algorithm. Risk factors for gram-negative organisms include aspiration, pulmonary comorbidity, and recent antibiotic therapy (>7 days in the past month). Concern for P. aeruginosa infection is increased in patients with bronchiectasis, malnutrition, human immunodeficiency virus infection, or corticosteroid therapy (>10 mg/day). 6, 27, 28 In the patient with severe CAP following influenza, another consideration is S. aureus, and there has been some concern with severe CAP in this setting caused by community-acquired MRSA. 15, 29 Patients with CAP who are admitted to intensive care units (ICUs) are divided into those at risk for P. aeruginosa and those who are not. All patients require combination therapy. Monotherapy is not recommended for any ICU-admitted CAP patient because of the absence of efficacy data for any single agent in this setting. For patients not at risk for P. aeruginosa, therapy should be with a selected β-lactam combined with either a macrolide or a quinolone. 6, 27 Recommended β-lactams include ceftriaxone, cefotaxime, and ampicillin-sulbactam. Ceftriaxone can be given from 1 g daily to 2 mg twice daily; the latter is recommended for severe pneumococcal infection, especially with associated meningitis. All patients require the addition of either a macrolide or a quinolone for possible atypical pathogen (Legionella spp., M. pneumoniae, or C. pneumoniae) infection, either as the sole cause or as part of a mixed infection. Some studies, including those involving severe CAP, have shown that the addition of this type of coverage is associated with a lower mortality rate than when other regimens are used. [30] [31] [32] The recommended macrolide is intravenous azithromycin (500 mg/day for 7 to 10 days) because it is better tolerated than erythromycin. The recommended quinolone is moxifloxacin (400 mg/day), regardless of renal function or levofloxacin (750 mg/day), for patients with normal renal function. In patients with abnormal renal function, levofloxacin doses should be adjusted, after using the same initial starting dose for all patients. In addition, several retrospective studies of patients with bacteremic pneumococcal pneumonia demonstrated that dual therapy including a β-lactam combined with a macrolide or a quinolone is associated with improved outcome, compared with single-agent β-lactam therapy. 33, 34 For the penicillin-allergic patient, therapy should be with an antipneumococcal quinolone (levofloxacin or moxifloxacin) in addition to aztreonam.

When risk factors for P. aeruginosa are present, the patient should be treated with two antipseudomonal agents in addition to coverage for DRSP and Legionella. Effective antimicrobials that can be used for this type of patient with severe CAP are selected β-lactams (cefepime, piperacillin-tazobactam, imipenem, meropenem), in combination with an antipseudomonal quinolone (ciprofloxacin, high-dose levofloxacin: at 750 mg/day). Alternatively, the foregoing β-lactams can be combined with an aminoglycoside and either azithromycin or a nonpseudomonal quinolone (moxifloxacin). In the penicillinallergic patient, aztreonam can be combined with an aminoglycoside and an antipneumococcal fluoroquinolone (Box 39.1).

Monotherapy with an antipneumococcal fluoroquinolone for severe CAP has not yet been proven safe and effective. Moxifloxacin is efficacious for CAP, even in elderly patients, but few patients with severe CAP have been studied. Although in the Community-Acquired Pneumonia Recovery in the Elderly (CAPRIE) study, which compared moxifloxacin with levofloxacin for CAP in elderly patients, the cure rate for moxifloxacin (94.7%) was greater than that for levofloxacin (84.6%), few patients in this study were mechanically ventilated. 35 In a recent study of severe CAP that compared levofloxacin with combination ceftriaxone-ofloxacin therapy, equivalent clinical responses were observed in both treatment groups (79.1% with levofloxacin compared with 79.5% with combination therapy). 36 However, patients with shock were excluded from the study, and in patients with mechanical ventilation, treatment with levofloxacin resulted in a lower clinical cure rate (63% compared with 72% with combination therapy). 36 Current guidelines recommend that if quinolones are used for severe CAP, they should be used as a replacement for a macrolide and should be part of a combination regimen, usually with a β-lactam. 6, 27 However, in a recent report, the use of initial empiric therapy with a β-lactam with a fluoroquinolone for severe CAP was associated with increased short-term mortality (odds ratio, 2.71; 95% confidence interval, 1.2 to 6.1), in comparison with other guidelinerecommended antimicrobial regimens. 37 If a specific organism is later recovered from culture, then therapy can be modified and focused, with the previous caveat about bacteremic pneumococcal pneumonia kept in mind. In addition, if the patient's history contains certain epidemiologic clues (travel, comorbid illness, animal exposure), therapy should be modified to cover the suspected organism. For example, those patients with chronic obstructive pulmonary disease should be treated for H. influenzae and Moraxella catarrhalis. In addition, if a sputum Gram stain is obtained and shows gram-positive cocci in clusters, particularly in a patient with severe pneumonia following a recent viral or influenza infection, therapy for S. aureus should be added.

Although community-acquired MRSA is not common in CAP, it has been reported following influenza or viral infection in otherwise healthy patients with severe, bilateral, necrotizing pneumonia. 15, 29 This organism is different from nosocomial MRSA because it occurs in previously healthy people, produces the Panton-Valentine leukocidin (a virulence factor that causes tissue necrosis), and is usually of the USA 300 clonal type. 29 Treatment for this type of pneumonia is uncertain, but vancomycin alone may not be effective. One case series suggested that therapy include an antibiotic that inhibits toxin, such as clindamycin (added to vancomycin) or linezolid (used alone). 15 All the foregoing recommended therapies are effective for DRSP, which is now common in the United States. Pneumococcus resistant to penicillin (>0.1 mg/L) may account for up to 40% of clinical isolates in the United States, and it is more common in patients who are immunocompromised and in those who have received β-lactam antibiotics in the past 3 months. 6, 38 Most of the penicillin resistance seen in patients with pneumonia is of the intermediate type and is not high level (minimal inhibitory concentration ≥2.0 mg/L). This observation may explain the findings that outcome in CAP is generally not worsened by the presence of penicillin-resistant organisms, compared with penicillinsensitive organisms. Effective therapy has been achieved with ceftriaxone and with cefotaxime, which are probably more likely to be effective than cefuroxime. 39, 40 The antipneumococcal quinolones moxifloxacin and levofloxacin (at the 750-mg dose) are also effective. If Legionella infection is documented, a quinolone may be the most reliable therapy. 41 In mechanically ventilated patients with severe CAP, consideration should be given to adjunctive therapy with corticosteroids. Three clinical situations may necessitate this approach. First, if a patient has pneumococcal pneumonia with meningitis, therapy with corticosteroids, started before antibiotics, has improved the likelihood of a good neurologic outcome. 42 Second, when a patient with severe CAP is hypotensive, relative adrenal insufficiency is common, and physiologic replacement doses of corticosteroids may be helpful. 43 Finally, one prospective, randomized trial of severe CAP showed improved outcomes, including mortality, when patients received a continuous infusion of low-to moderate-dose corticosteroids. 44 In addition to corticosteroids, the use of activated protein C may be helpful for severe CAP, although 448 Section 5

Adjuncts to Ventilator Therapy

• Administer the first dose of antibiotic therapy within 4-6 hours of arrival to the hospital.

• Treat all patients for pneumococcus (including DRSP) and Legionella, and consider coverage of Haemophilus influenzae, enteric gram-negative bacteria (including Pseudomonas aeruginosa), Staphylococcus aureus (including MRSA), and atypical pathogens (Mycoplasma pneumoniae, Chlamydophila pneumoniae). • Limit macrolide monotherapy to outpatients or inpatients with no risk factors for DRSP, enteric gram-negative bacteria, or aspiration, who do not have severe illness. • Use EITHER a selected β-lactam with a macrolide (azithromycin) OR quinolone (levofloxacin or moxifloxacin) for patients not at risk for P. aeruginosa infection.

• If penicillin allergic, use an antipneumococcal quinolone PLUS aztreonam. • For those at risk for P. aeruginosa infection, use an antipseudomonal β-lactam PLUS either ciprofloxacin or levofloxacin OR combine with an aminoglycoside AND either a macrolide or an antipneumococcal quinolone (levofloxacin or moxifloxacin).

• If penicillin allergic, use aztreonam PLUS an aminoglycoside PLUS an antipneumococcal quinolone. • To cover DRSP, the selected acceptable intravenous β-lactams include ceftriaxone, cefotaxime, ertapenem, and ampicillin/sulbactam. • Antipseudomonal β-lactams include cefepime, imipenem, meropenem, and piperacillin/tazobactam. • Do not administer quinolone monotherapy to any patient with intensive care unit-admitted CAP. • The newer antipneumococcal quinolones, in order of decreasing antipneumococcal activity, are gemifloxacin (oral only), moxifloxacin (oral and intravenous), and levofloxacin (oral and intravenous). • Vancomycin and linezolid should be used rarely and only in patients with severe CAP and either meningitis (vancomycin) or severe necrotizing pneumonia after influenza.

• If community-acquired, toxin-producing, MRSA is suspected, use EITHER linezolid alone OR consider adding clindamycin to vancomycin. the documented benefit was minimal in patients who received accurate and appropriate antibiotic therapy, compared with those who did not. 45 

The hospital-acquired pneumonia (HAP) group includes patients with VAP and those with severe nosocomial pneumonia who require mechanical ventilation to support them at a time of severe infection. All these patients are at risk for infection with a group of core pathogens, but some are also at risk for infection by additional organisms, particularly MDR pathogens, based on the presence of risk factors. 1 Although these general patterns apply to most patients, bacteriologic features vary from one hospital to another, from one ICU to another, and from one time period to another. Therefore, it is necessary to have a knowledge of local microbiologic data when adapting treatment recommendations to a specific clinical setting. 13 Some patients, particularly those with acute respiratory distress syndrome, can have polymicrobial infection, in which multiple bacterial pathogens act synergistically. Pure anaerobic pneumonia is uncommon in patients with HAP, including those with aspiration risk factors, and in this latter population, enteric gramnegative organisms are still the dominant concern. 28, 46 Although fungal and viral pathogens can cause HAP, this situation is relatively uncommon, based on current data. Candida is generally an uncommon pathogen, but a common colonizer, and it rarely causes pneumonia. Conversely, Aspergillus can be an important cause of HAP, particularly in patients who have had a prolonged hospital stay with antibiotic and corticosteroid therapy.

Once there is a clinical suspicion of HAP in a mechanically ventilated patient, the patient should be treated with antibiotics, unless the suspicion of infection is low and a Gram stain of an endotracheal suction aspirate is negative. The antibiotic choice is either a narrowspectrum regimen or a broad-spectrum regimen, the latter directed at MDR pathogens. The narrow-spectrum approach is used if the patient has pneumonia that started in the first 4 days of hospitalization, if no other risk factors for MDR pathogens are present, and if the patient does not have HCAP. All other patients, including those with HCAP, receive broad-spectrum therapy. 1 Narrow-spectrum therapy is directed at the core pathogens and is generally achieved with a single agent. Options are ceftriaxone, ampicillin-sulbactam, ertapenem, levofloxacin, moxifloxacin, and ciprofloxacin. If the patient is allergic to penicillin, a quinolone can be used, or the patient can be given the combination of clindamycin and aztreonam. 1 When the patient has risk factors for MDR pathogens, therapy is directed not only at core pathogens, but also at P. aeruginosa, Acinetobacter species, and, in many instances, MRSA. To provide this spectrum of coverage, therapy includes at least two and often three antibiotics. The recommended therapy is to use either an aminoglycoside or an antipseudomonal quinolone (ciprofloxacin or levofloxacin) in combination with an antipseudomonal β-lactam. The choices of β-lactams are cefepime, ceftazidime, imipenem, meropenem, and piperacillintazobactam. In choosing among these options, it is important to use a different agent from any the patient received in the previous 14 days, because repeated use of the same agent may lead to clinical failure related to selection of resistance as a result of recent exposure 47 (Box 39.2).

If there are concerns about MRSA because of risk factors, a high local prevalence, or the presence of clusters of gram-positive organisms on a Gram stain of a tracheal aspirate sample, then a third agent, either linezolid or vancomycin, should be added. If culture data from tracheal aspirates become available, then therapy can sometimes be more specific. For example, if MRSA is documented in the ventilated patient, some data suggest an advantage of linezolid over vancomycin. 48 For Acinetobacter species, the drug of choice is a carbapenem, but if resistance to this class of antibiotics is present, the recommended therapy is colistin, although tigecycline may become an option in the future, once more clinical data in this setting are available.

The value of combination therapy is controversial. In general, no strong data show that the use of an aminoglycoside with a β-lactam is more effective than β-lactam monotherapy, unless the patient is neutropenic or has pseudomonal bacteremia. 1, 49 Thus, when the mechanically ventilated patient is given combination therapy, the major justification is to provide a broad enough spectrum of therapy to treat MDR pathogens effectively in a patient with risk factors. It is important to use the correct dose of antibiotics in critically ill ventilated patients with pneumonia. Based on clinical trial data, the correct doses for patients with normal renal function include the following: cefepime, 1 to 2 g every 8 to 12 hours; imipenem, 500 mg every 6 hours or 1 g every 8 hours; meropenem, 1 g every 8 hours; piperacillin-tazobactam, 4.5 g every 6 hours; levofloxacin, 750 mg/day, or ciprofloxacin, 400 mg every 8 hours; vancomycin, 15 mg/kg every 12 hours leading to a trough level of 15 to 20 mg/L; linezolid, 600 mg every 12 hours; and aminoglycosides of 7 mg/kg/day of gentamicin or tobramycin and 20 mg/kg of amikacin. 1 To use antibiotics effectively and responsibly for these patients, it is necessary to implement a de-escalation strategy, which implies the prompt use of broadspectrum empiric therapy whenever there is a clinical suspicion of infection with MDR pathogens. The reasons for this approach are to avoid delaying therapy and to use a broad-enough spectrum regimen to cover the likely pathogens. 1, 12 However, to avoid overuse of antibiotics, a key decision point comes on day 2 to 3, when culture and clinical response data become available. It then becomes possible to narrow and focus therapy or, in some cases, to stop therapy altogether or aim for a short duration of treatment. Based on culture data, de-escalation is often possible. Therefore, if an aminoglycoside was used with a β-lactam, the maximal benefit may have been achieved after 5 days of dual therapy, and the aminoglycoside can usually be stopped at that point. 1, 50 If a nonresistant gram-negative organism is identified, therapy can immediately be de-escalated to a single agent. Drugs shown effective for critically ill mechanically ventilated patients are ciprofloxacin, levofloxacin, imipenem, meropenem, piperacillin-tazobactam, and cefepime. De-escalation can be done as soon as culture data are available. It can also be done in patients who have negative cultures, and even in those with MDR pathogens, provided an effective single agent is identified from sensitivity testing results. 12 This approach has led to less total antibiotic use, and in some instances, reduced mortality. 51 Another way to use less antibiotic is to shorten the duration of therapy. Several studies have shown that it is possible to treat VAP effectively with 6 to 8 days of therapy, provided the initial therapy was appropriate. 51, 52 The optimal duration of therapy for infections caused by P. aeruginosa and MRSA is still uncertain, but prolonged therapy may be no better than short-duration therapy, in the absence of bacteremia.

Section 5

Adjuncts to Ventilator Therapy

• Initiate therapy as soon as there is a clinical suspicion of infection.

• Obtain a lower respiratory tract culture (tracheal aspirate, protected brush, bronchoalveolar lavage) before initiation of antibiotic therapy. Samples can be obtained bronchoscopically or otherwise and cultured quantitatively or semiquantitatively. Do not delay therapy for the purpose of collecting a culture. • Choose a narrow-spectrum agent for patients with no risk factors for MDR pathogens and no HCAP.

• Options include ceftriaxone, ampicillin/sulbactam, ertapenem, levofloxacin, moxifloxacin and ciprofloxacin.

• For penicillin allergy, use a quinolone OR the combination of clindamycin and aztreonam. • Choose a broad-spectrum regimen with at least two drugs for patients with risk factors for MDR pathogens.

Use a knowledge of local microbiology to guide choices. • Use an aminoglycoside or an antipneumococcal quinolone (ciprofloxacin or high-dose levofloxacin) PLUS an antipseudomonal β-lactam such as cefepime, ceftazidime, imipenem, meropenem, or piperacillin-tazobactam. If there is concern about MRSA, add EITHER linezolid OR vancomycin. • Use the correct therapy in recommended doses.

• Recommended doses for patients with normal renal function are: cefepime, 1-2 g every 8-12 hours; imipenem, 500 mg every 6 hours or 1 g every 8 hours; meropenem, 1 g every 8 hours; piperacillin-tazobactam, 4.5 g every 6 hours; levofloxacin, 750 mg/day, or ciprofloxacin, 400 mg every 8 hours; vancomycin, 15 mg/kg every 12 hours leading to a trough level of 15-20 mg/L; linezolid, 600 mg every 12 hours; and aminoglycosides of 7 mg/kg per day of gentamicin or tobramycin and 20 mg/kg of amikacin. • Choose an empiric therapy that uses agents from a class of antibiotics different from those the patient has received in the past 2 weeks. • The drug of choice for Acinetobacter infection is a carbapenem, but colistin should be considered if there is carbapenem resistance. In the future, tigecycline may be an appropriate choice. • Consider linezolid as an alternative to vancomycin, especially in patients with renal insufficiency, those receiving other nephrotoxic medications, and those with proven MRSA ventilator-associated pneumonia. • Adjunctive aerosolized aminoglycosides can be used for patients with highly resistant gram-negative pathogens, but systemic therapy should be continued.

HCAP, health care-associated pneumonia; MDR, multidrug resistant; MRSA, methicillin-resistant Staphylococcus aureus.

There are no well-controlled studies of the need for antibiotic therapy of nosocomial tracheobronchitis in mechanically ventilated patients. Nosocomial tracheobronchitis is usually defined as a condition in which the patient has fever, increased (usually purulent) sputum, and a positive endotracheal aspirate culture during mechanical ventilation, but in the absence of radiographic pneumonia. 4 These patients usually have infection with P. aeruginosa, and their mortality and length of stay may be higher than in ventilated patients without this complication. In spite of adverse outcomes, it remains uncertain whether specific antimicrobial therapy is needed. Anecdotal experience has shown that in some patients who do not have signs of systemic sepsis, therapy with only aerosolized antibiotics (usually aminoglycosides or colistin) may be effective. 19, 20 Duration of Therapy and Expected Response

In severely ill patients with CAP or nosocomial pneumonia, patients should rapidly improve with effective therapy. Patients who do improve can be treated with antibiotics for 7 to 10 days, whereas those who do not need careful reevaluation. In patients with severe pneumonia, some improvement usually occurs by day 3, and thus this becomes the time point for deciding whether the patient has made an appropriate response to therapy. 6, 27 Treatment failure in severe CAP can occur in up to 15% of all patients and is present either early or late. 53 Most studies of clinical response in CAP examined patients who were not mechanically ventilated. These studies focused on improvements in symptoms of cough, sputum production, and dyspnea, along with the ability to take medications by mouth, and an afebrile status for at least two occasions 8 hours apart. In a ventilated patient, serial measurements of oxygenation are a good indicator of response, and radiographic improvement usually lags behind clinical improvement.

When a patient with severe CAP fails to respond to therapy in the expected time interval, it is necessary to consider infection with a drug-resistant or unusual pathogen (Mycobacterium tuberculosis, Bacillus anthracis [anthrax], Coxiella burnetii, Burkholderia pseudomallei, Pasteurella multocida, endemic fungi, or hantavirus), a pneumonic complication (lung abscess, endocarditis, empyema), or a noninfectious process that mimics pneumonia (bronchiolitis obliterans with organizing pneumonia, hypersensitivity pneumonitis, pulmonary vasculitis, bronchoalveolar cell carcinoma, lymphoma, pulmonary embolus). The evaluation of the nonresponding patient should be individualized, but it may include computed tomography of the chest, pulmonary angiography, bronchoscopy, and occasionally open lung biopsy.

If the patient has VAP or nosocomial pneumonia, clinical improvement should also occur by day 2 to 3, and serial measurement of the Clinical Pulmonary Infection Score may be the best way to evaluate clinical response. 54 Of all clinical parameters, serial improvement in oxygenation is the best measure of a good response to therapy, and usually this occurs by day 3 in patients who are likely to survive. 54 For the patient who is not responding, the first step is to check respiratory tract cultures, just to be sure that the therapy is active against the pathogen isolated. In addition, more cultures and diagnostic testing are needed to rule out infection with an unusual pathogen (fungus), another diagnosis (inflammatory lung disease), or another site of infection or pneumonia complication (central line infection, empyema, antibiotic-induced colitis). When a patient is not responding to initial therapy, a change in antibiotics, combined with an aggressive diagnostic reevaluation, should be carried out no later than day 3.

Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcare-associated pneumonia

Epidemiology and outcomes of health-care-associated pneumonia: Results from a large US database of culture positive patients

Incidence of and risk factors for ventilator-associated pneumonia in critically ill patients

Effect of ventilator-associated tracheobronchitis on outcome in patients without chronic respiratory failure: A case-control study

Respiratory tract colonization and infection in patients with chronic tracheostomy: A one-year study in patients living at home

Guidelines for the management of adults with community-acquired lower respiratory tract infections: Diagnosis, assessment of severity, antimicrobial therapy and prevention

Timing of antibiotic administration and outcomes for Medicare patients hospitalized with community-acquired pneumonia

Impact of BAL data on the therapy and outcome of ventilator-associated pneumonia

A five-year study of severe community-acquired pneumonia with emphasis on prognosis in patients admitted to an intensive care unit

The attributable morbidity and mortality of ventilator-associated pneumonia in the critically ill patient: The Canadian Critical Trials Group

National Nosocomial Infections Surveillance System: Overview of infections caused by gram-negative bacilli

Nosocomial pneumonia: The importance of a de-escalating strategy for antibiotic treatment of pneumonia in the ICU

Variations in etiology of ventilator-associated pneumonia across four treatment sites: Implications for antimicrobial prescribing practices

The importance of bactericidal drugs: Future directions in infectious disease

Pleuropulmonary complications of Panton-Valentine leukocidin-positive community-acquired methicillin-resistant Staphylococcus aureus: Importance of treatment with antimicrobials inhibiting exotoxin production

Application of pharmacokinetics and pharmacodynamics to antimicrobial therapy of respiratory tract infections

Antibiotic penetration into lung tissues

Penetration of gentamicin into alveolar lining fluid of critically ill patients with ventilator-associated pneumonia

Deposition of instilled versus nebulized tobramycin and imipenem in ventilated intensive care unit patients

Aerosolized colistin for the treatment of nosocomial pneumonia due to multidrug-resistant gram-negative bacteria in patients without cystic fibrosis

Double-blind study of endotracheal tobramycin in the treatment of gram-negative bacterial pneumonia: The Endotracheal Tobramycin Study Group

Lung deposition and efficiency of nebulized amikacin during Escherichia coli pneumonia in ventilated piglets

Lung tissue concentrations of nebulized amikacin during mechanical ventilation in piglets with healthy lungs

Pharmacodynamics of intravenous ciprofloxacin in seriously ill patients

Pharmacodynamics of levofloxacin: A new paradigm for early clinical trials

Once-daily aminoglycoside dosing in immunocompetent adults: A meta analysis

Antimicrobial therapy of community-acquired pneumonia

Community-acquired pneumonia due to gram-negative bacteria and Pseudomonas aeruginosa: Incidence, risk and prognosis

Severe community-onset pneumonia in healthy adults caused by methicillin-resistant Staphylococcus aureus carrying the Panton-Valentine leukocidin genes

Associations between initial antimicrobial therapy and medical outcomes for hospitalized elderly patients with pneumonia

Empiric antibiotic therapy and mortality among Medicare pneumonia inpatients in 10 Western states

Impact of initial antibiotic choice on clinical outcomes in community-acquired pneumonia: Analysis of a hospital claims-made database

Linezolid vs vancomycin: Analysis of two double-blind studies of patients with methicillin-resistant Staphylococcus aureus nosocomial pneumonia

Addition of a macrolide to a B-lactam-based empirical antibiotic regimen is associated with lower in-patient mortality for patients with bacteremic pneumococcal pneumonia

Community-acquired pneumonia recovery in the elderly (CAPRIE): Efficacy and safety of moxifloxacin therapy versus that of levofloxacin therapy

Comparison of levofloxacin and cefotaxime combined with ofloxacin for ICU patients with community-acquired pneumonia who do not require vasopressors

The impact of empiric antimicrobial therapy with a beta-lactam and fluoroquinolone on mortality for patients hospitalized with severe pneumonia

Predicting antimicrobial resistance in invasive pneumococcal infections

Prospective observational study of bacteremic pneumococcal pneumonia: Effect of discordant therapy on mortality

An international prospective study of pneumococcal bacteremia: Correlation with in vitro resistance, antibiotics administered, and clinical outcome

Clinical outcomes for hospitalized patients with Legionella pneumonia in the antigenuria era: The influence of levofloxacin therapy

European Dexamethasone in Adulthood Bacterial Meningitis Study Investigators: Dexamethasone in adults with bacterial meningitis

Cortisol levels in patients with severe community-acquired pneumonia

Hydrocortisone infusion for severe community-acquired pneumonia: A preliminary randomized study

Severe community-acquired pneumonia as a cause of severe sepsis: Data from the PROWESS study

Microbiology of severe aspiration pneumonia in institutionalized elderly

Pseudomonas aeruginosa ventilator-associated pneumonia: Comparison of episodes due to piperacillin-resistant versus piperacillinsusceptible organisms

Linezolid vs vancomycin: Analysis of two double-blind studies of patients with methicillin-resistant Staphylococcus aureus nosocomial pneumonia

Beta-lactam monotherapy versus beta-lactam-aminoglycoside combination therapy for sepsis in immunocompetent patients: Systematic review and meta-analysis of randomised trials

Strategy of antibiotic rotation: Long-term effect on incidence and susceptibilities of gramnegative bacilli responsible for ventilator-associated pneumonia

Clinical characteristics and treatment patterns among patients with ventilator-associated pneumonia

Impact of clinical guidelines in the management of severe hospital-acquired pneumonia

Risk factors for treatment failure in community acquired pneumonia: Implications for disease outcome

Resolution of ventilator-associated pneumonia: Prospective evaluation of the clinical pulmonary infection score as an early clinical predictor of outcome