Robert. S. Dittus, MD, MPH

Definition of VAP

Although early articles used variable criteria for diagnosing VAP, recent studies have traditionally defined VAP as an infection occurring more than 48 hours after hospital admission in a mechanically ventilated patient with a tracheostomy or endotracheal tube.2 In early 2007, the CDC revised their definition for diagnosing VAP.3 These latest criteria state there is no minimum period that the ventilator must be in place in order to diagnose VAP. This important change must be kept in mind when examining future studies.

The term VAP is more specific than the term health careassociated pneumonia. The latter encompasses patients residing in a nursing home or long‐term care facility; hospitalized in an acute care hospital for more than 48 hours in the past 90 days; receiving antibiotics, chemotherapy, or wound care within the past 30 days; or attending a hospital or hemodialysis clinic.

The CDC published detailed criteria for diagnosing VAP in its member hospitals (Tables 1 and 2).3 Because diagnosing VAP in infants, children, elderly, and immunocompromised patients is often confusing because of other conditions with similar signs and symptoms, the CDC published alternate criteria for these populations. A key objective during development of our algorithms was to consolidate and simplify these diagnostic criteria for ICU clinicians.

Etiology

The most commonly isolated VAP pathogens in all patients are bacteria.21 Most of these organisms normally colonize the respiratory and gastrointestinal tracts, but some are unique to health care settings. Tracheal intubation disrupts the body's natural anatomic and physiologic defenses and facilitates easier entry of these pathogens. Typical organisms include Staphylococcus aureus, Pseudomonas aeruginosa, Enterobacter species, Klebsiella pneumoniae, Acinetobacter species, Escherichia coli, and Haemophilus influenzae.22, 23 Unfortunately, the prevalence of antimicrobial resistance among VAP pathogens is increasing.24 Risk factors for antibiotic resistance are common to ICU patients and include recent antibiotics, hemodialysis, nursing home residence, immunosuppression, and chronic wound care.5 Polymicrobial infections are frequently seen in VAP, with up to 50% of all VAP episodes caused by more than 1 organism.25

Viral VAP is rare in immunocompetent hosts, and seasonal outbreaks of influenza and other similar viruses are usually limited to nonventilated patients.26 However, influenza is underrecognized as a potential nosocomial pathogen, and numerous nosocomial outbreaks because of influenza have been reported.2731 Although herpes simplex virus is often detected in the respiratory tract of critically ill patients, its clinical importance remains unclear.32

Fungal VAP is also rare in immunocompetent hosts. On the other hand, pulmonary fungal infections are common in immunocompromised patients, especially following chemotherapy and transplantation. Candida species are often isolated from the airways of normal hosts, but most cases traditionally have been considered clinically unimportant because these organisms are normal oropharyngeal flora and rarely invade lung tissue.33, 34 It is unclear whether recent studies suggesting Candida colonization is associated with a higher risk for Pseudomonas VAP will change this conventional wisdom.3537

Immunocompromised patients with suspected VAP are unique because they are at risk not only for typical bacteria (which are the most common causes of VAP) but also for rarer opportunistic infections and noninfectious processes that mimic pneumonia.3840 While assessing these patients, clinicians must consider the status of the underlying disease, duration and type of immunosuppression, prophylactic regimens, and risk factors for noninfectious causes of pulmonary infiltrates.41 Common opportunistic infections include viruses, mycobacteria, fungi, and Pneumocystis. Noninfectious processes include pulmonary edema, drug toxicity, radiation pneumonitis, engraftment syndrome, bronchiolitis obliterans organizing pneumonia, alveolar proteinosis, transfusion‐related lung injury, alveolar hemorrhage, and progression of underlying disease. In general, diagnosing VAP in the immunocompromised patient requires a prompt, comprehensive, and multidisciplinary approach.38

In preterm and term infants, the most common VAP pathogens are gram‐negative organisms such as E. coli and P. aeruginosa. Other less common pathogens are Enterobacter, Klebsiella, Acinetobacter, Proteus, Citrobacter, and Stenotrophomonas maltophilia.42, 43 Infants with a preceding bloodstream infection or prolonged intubation are more likely to develop VAP.43, 44 Unfortunately, gram‐negative bacteria often colonize the airways of mechanically ventilated infants, and tracheal aspirate culture data are difficult to interpret in this population.42

Children are more likely to develop VAP if they are intubated for more than 48 hours. The most common pathogens isolated from tracheal aspirates in mechanically ventilated children are enteric gram‐negative bacteria, P. aeruginosa, and S. aureus.45, 46 Few studies have precisely delineated the pathogenesis of VAP in the pediatric ICU population.

Overall, the causes of VAP vary by hospital, patient population, and ICU type. Therefore, it is essential that ICU clinicians remain knowledgeable about their local surveillance data.21 Awareness of VAP microbiology is essential for optimizing initial antibiotic therapy and improving outcomes.

Early Versus Late VAP

Distinguishing between early and late VAP is important for initial antibiotic selection because the etiologic pathogens vary between these 2 periods.4749 Early VAP (days 14 of hospitalization) usually involves antibiotic‐sensitive community‐acquired bacteria and carries a better prognosis. In contrast, late VAP (5 days after hospital admission) is more likely to be caused by antibiotic‐resistant nosocomial bacteria that lead to increased morbidity and mortality. All patients who have been hospitalized or have received antibiotics during the prior 90 days should be treated as having late VAP because they are at much higher risk for colonization and infection with antibiotic‐resistant bacteria.47 Of note, 2 recent studies suggest that pathogens in the early and late periods are becoming similar at some institutions.50, 51 Overall, the distinction between early and late VAP is important because it affects the likelihood that a patient has antibiotic‐resistant bacteria. If antibiotic‐resistant pathogens are suspected, initial therapy should include empiric triple antibiotics until culture data are available.

Culturing Approaches

Because clinical criteria alone are rarely able to accurately diagnose VAP,52, 53 clinicians should also obtain a respiratory specimen for microbiologic culture. Despite the convenience of blood cultures, their sensitivity for diagnosing VAP is poor, and they rarely make the diagnosis alone.54 Two methods are available for culturing the lungsan invasive approach (eg, bronchoscopy with bronchoalveolar lavage) and a noninvasive approach (eg, tracheal aspirate).

Some investigators believe that adult patients with suspected VAP should always undergo an invasive sampling of lower‐respiratory‐tract secretions.55 Proponents of the invasive approach cite the frequency with which potential pathogens colonize the trachea of ICU patients and create spurious results on tracheal aspirates.22 In addition, several studies have shown that clinicians are more likely to narrow the spectrum of antibiotics after obtaining an invasive diagnostic sample.56 In other words, the invasive approach has been associated with better antimicrobial stewardship.

Other investigators believe that a noninvasive approach is equally safe and effective for diagnosing VAP.57 This clinical approach involves culturing a tracheal aspirate and using a pneumonia prediction score such as the clinical pulmonary infection score (CPIS; Table 3). The CPIS assigns 012 points based on 6 clinical criteria: fever, leukocyte count, oxygenation, quantity and purulence of secretions, type of radiographic abnormality, and results of sputum gram stain and culture.58 As developed, a CPIS > 6 has a sensitivity of 93% and a specificity of 100% for diagnosing VAP.58 However, the CPIS requires that nurses record sputum volume and that the laboratory stains the specimen. When the CPIS has been modified based on the unavailability of such resources, the results have been less impressive.5961 Despite studies showing that a noninvasive clinical approach can achieve adequate initial antibiotic coverage and reduce overuse of broad‐spectrum agents,62, 63 clinicians who use the CPIS must understand its inherent limitations.

A meta‐analysis56 comparing the utility of an invasive versus a noninvasive culturing approach identified 4 randomized trials examining this issue.6669 Overall, an invasive approach did not alter mortality, but patients undergoing bronchoscopy were much more likely to have their antibiotic regimens modified by clinicians. This suggests that the invasive approach may allow more directed use of antibiotics. Recently, the Canadian Critical Care Trials Group conducted a multicenter randomized trial looking at this issue.11 There was no difference between the 2 approaches in mortality, number of ventilator days, and antibiotic usage. However, all patients in this study were immediately treated with empiric broad‐spectrum antibiotics until culture results were available, and the investigators did not have a protocol for stopping antibiotics after culture data were available.

In summary, both invasive and noninvasive culturing approaches are considered acceptable options for diagnosing VAP. Readers interested in learning more about this topic should read the worthwhile Expert Discussion70 by Chastre and colleagues55 at the end of this article. In general, we recommend that ICU clinicians use a combination of clinical suspicion (based on the CPIS or other objective data) and cultures ideally obtained prior to antibiotics. Regardless of the chosen culturing approach, clinicians must recognize that 1 of the most important determinants of patient outcome is prompt administration of adequate initial antibiotics.7175

Initial Antibiotic Administration

Delaying initial antibiotics in VAP increases the risk of death.7175 If a patient receives ineffective initial therapy, a later switch to appropriate therapy does not eliminate the increased mortality risk. Therefore, a comprehensive approach to VAP diagnosis requires consideration of initial empiric antibiotic administration.

Whenever possible, clinicians should obtain a lower respiratory tract sample for microscopy and culture before administering antibiotics because performing cultures after antibiotics have been recently started will lead to a higher rate of false‐negative results.76 Unless the patient has no signs of sepsis and microscopy is completely negative, clinicians should then immediately start empiric broad‐spectrum antibiotics.57 Once the culture sensitivities are known, therapy can be deescalated to a narrower spectrum.77 Recent studies suggest that shorter durations of therapy (8 days) are as effective as longer courses and are associated with lower colonization rates by antibiotic‐resistant bacteria.62, 78

Initial broad‐spectrum antibiotics should be chosen based on local bacteriology and resistance patterns. Clinicians must remain aware of the most common bacterial pathogens in their local community, hospital, and ICU. This is essential for both ensuring adequate initial antibiotic coverage and reducing overall antibiotic days.65 Unrestrained use of broad‐spectrum antibiotics increases the risk of resistant pathogens. Clinicians must continually deescalate therapy and use narrow‐spectrum drugs as pathogens are identified.79

Prevention of VAP

In 2005, the American Thoracic Society published guidelines for the management of adults with VAP.5 These guidelines included a discussion of modifiable risk factors for preventing VAP and used an evidence‐based grading system to rank the various recommendations. The highest evidence (level 1) comes from randomized clinical trials, moderate evidence (level 2) comes from nonrandomized studies, and the lowest evidence (level 3) comes from case studies or expert opinion. Others have also published their own guidelines and recommendations for preventing VAP.8082 Table 4 shows the key VAP preventive strategies.

Some strategies are not recommended for VAP prevention in general ICU patients. Selective decontamination of the digestive tract (ie, prophylactic oral antibiotics) has been shown to reduce respiratory infections in ICU patients,113 but its overall role remains controversial because of concerns it may increase the incidence of multi‐drug‐resistant pathogens.114 Similarly, prophylactic intravenous antibiotics administered at the time of intubation can reduce VAP in certain patient populations,115 but this strategy is also associated with an increased risk of antibiotic‐resistant nosocomial infections.116 Using kinetic beds and scheduled chest physiotherapy to reduce VAP is based on the premise that critically ill patients often develop atelectasis and cannot effectively clear their secretions. Unfortunately, neither of these modalities has been shown to consistently reduce VAP in medical ICU patients.117119

Definition of VAP

Although early articles used variable criteria for diagnosing VAP, recent studies have traditionally defined VAP as an infection occurring more than 48 hours after hospital admission in a mechanically ventilated patient with a tracheostomy or endotracheal tube.2 In early 2007, the CDC revised their definition for diagnosing VAP.3 These latest criteria state there is no minimum period that the ventilator must be in place in order to diagnose VAP. This important change must be kept in mind when examining future studies.

The term VAP is more specific than the term health careassociated pneumonia. The latter encompasses patients residing in a nursing home or long‐term care facility; hospitalized in an acute care hospital for more than 48 hours in the past 90 days; receiving antibiotics, chemotherapy, or wound care within the past 30 days; or attending a hospital or hemodialysis clinic.

The CDC published detailed criteria for diagnosing VAP in its member hospitals (Tables 1 and 2).3 Because diagnosing VAP in infants, children, elderly, and immunocompromised patients is often confusing because of other conditions with similar signs and symptoms, the CDC published alternate criteria for these populations. A key objective during development of our algorithms was to consolidate and simplify these diagnostic criteria for ICU clinicians.

Etiology

The most commonly isolated VAP pathogens in all patients are bacteria.21 Most of these organisms normally colonize the respiratory and gastrointestinal tracts, but some are unique to health care settings. Tracheal intubation disrupts the body's natural anatomic and physiologic defenses and facilitates easier entry of these pathogens. Typical organisms include Staphylococcus aureus, Pseudomonas aeruginosa, Enterobacter species, Klebsiella pneumoniae, Acinetobacter species, Escherichia coli, and Haemophilus influenzae.22, 23 Unfortunately, the prevalence of antimicrobial resistance among VAP pathogens is increasing.24 Risk factors for antibiotic resistance are common to ICU patients and include recent antibiotics, hemodialysis, nursing home residence, immunosuppression, and chronic wound care.5 Polymicrobial infections are frequently seen in VAP, with up to 50% of all VAP episodes caused by more than 1 organism.25

Viral VAP is rare in immunocompetent hosts, and seasonal outbreaks of influenza and other similar viruses are usually limited to nonventilated patients.26 However, influenza is underrecognized as a potential nosocomial pathogen, and numerous nosocomial outbreaks because of influenza have been reported.2731 Although herpes simplex virus is often detected in the respiratory tract of critically ill patients, its clinical importance remains unclear.32

Fungal VAP is also rare in immunocompetent hosts. On the other hand, pulmonary fungal infections are common in immunocompromised patients, especially following chemotherapy and transplantation. Candida species are often isolated from the airways of normal hosts, but most cases traditionally have been considered clinically unimportant because these organisms are normal oropharyngeal flora and rarely invade lung tissue.33, 34 It is unclear whether recent studies suggesting Candida colonization is associated with a higher risk for Pseudomonas VAP will change this conventional wisdom.3537

Immunocompromised patients with suspected VAP are unique because they are at risk not only for typical bacteria (which are the most common causes of VAP) but also for rarer opportunistic infections and noninfectious processes that mimic pneumonia.3840 While assessing these patients, clinicians must consider the status of the underlying disease, duration and type of immunosuppression, prophylactic regimens, and risk factors for noninfectious causes of pulmonary infiltrates.41 Common opportunistic infections include viruses, mycobacteria, fungi, and Pneumocystis. Noninfectious processes include pulmonary edema, drug toxicity, radiation pneumonitis, engraftment syndrome, bronchiolitis obliterans organizing pneumonia, alveolar proteinosis, transfusion‐related lung injury, alveolar hemorrhage, and progression of underlying disease. In general, diagnosing VAP in the immunocompromised patient requires a prompt, comprehensive, and multidisciplinary approach.38

In preterm and term infants, the most common VAP pathogens are gram‐negative organisms such as E. coli and P. aeruginosa. Other less common pathogens are Enterobacter, Klebsiella, Acinetobacter, Proteus, Citrobacter, and Stenotrophomonas maltophilia.42, 43 Infants with a preceding bloodstream infection or prolonged intubation are more likely to develop VAP.43, 44 Unfortunately, gram‐negative bacteria often colonize the airways of mechanically ventilated infants, and tracheal aspirate culture data are difficult to interpret in this population.42

Children are more likely to develop VAP if they are intubated for more than 48 hours. The most common pathogens isolated from tracheal aspirates in mechanically ventilated children are enteric gram‐negative bacteria, P. aeruginosa, and S. aureus.45, 46 Few studies have precisely delineated the pathogenesis of VAP in the pediatric ICU population.

Overall, the causes of VAP vary by hospital, patient population, and ICU type. Therefore, it is essential that ICU clinicians remain knowledgeable about their local surveillance data.21 Awareness of VAP microbiology is essential for optimizing initial antibiotic therapy and improving outcomes.

Early Versus Late VAP

Distinguishing between early and late VAP is important for initial antibiotic selection because the etiologic pathogens vary between these 2 periods.4749 Early VAP (days 14 of hospitalization) usually involves antibiotic‐sensitive community‐acquired bacteria and carries a better prognosis. In contrast, late VAP (5 days after hospital admission) is more likely to be caused by antibiotic‐resistant nosocomial bacteria that lead to increased morbidity and mortality. All patients who have been hospitalized or have received antibiotics during the prior 90 days should be treated as having late VAP because they are at much higher risk for colonization and infection with antibiotic‐resistant bacteria.47 Of note, 2 recent studies suggest that pathogens in the early and late periods are becoming similar at some institutions.50, 51 Overall, the distinction between early and late VAP is important because it affects the likelihood that a patient has antibiotic‐resistant bacteria. If antibiotic‐resistant pathogens are suspected, initial therapy should include empiric triple antibiotics until culture data are available.

Culturing Approaches

Because clinical criteria alone are rarely able to accurately diagnose VAP,52, 53 clinicians should also obtain a respiratory specimen for microbiologic culture. Despite the convenience of blood cultures, their sensitivity for diagnosing VAP is poor, and they rarely make the diagnosis alone.54 Two methods are available for culturing the lungsan invasive approach (eg, bronchoscopy with bronchoalveolar lavage) and a noninvasive approach (eg, tracheal aspirate).

Some investigators believe that adult patients with suspected VAP should always undergo an invasive sampling of lower‐respiratory‐tract secretions.55 Proponents of the invasive approach cite the frequency with which potential pathogens colonize the trachea of ICU patients and create spurious results on tracheal aspirates.22 In addition, several studies have shown that clinicians are more likely to narrow the spectrum of antibiotics after obtaining an invasive diagnostic sample.56 In other words, the invasive approach has been associated with better antimicrobial stewardship.

Other investigators believe that a noninvasive approach is equally safe and effective for diagnosing VAP.57 This clinical approach involves culturing a tracheal aspirate and using a pneumonia prediction score such as the clinical pulmonary infection score (CPIS; Table 3). The CPIS assigns 012 points based on 6 clinical criteria: fever, leukocyte count, oxygenation, quantity and purulence of secretions, type of radiographic abnormality, and results of sputum gram stain and culture.58 As developed, a CPIS > 6 has a sensitivity of 93% and a specificity of 100% for diagnosing VAP.58 However, the CPIS requires that nurses record sputum volume and that the laboratory stains the specimen. When the CPIS has been modified based on the unavailability of such resources, the results have been less impressive.5961 Despite studies showing that a noninvasive clinical approach can achieve adequate initial antibiotic coverage and reduce overuse of broad‐spectrum agents,62, 63 clinicians who use the CPIS must understand its inherent limitations.

A meta‐analysis56 comparing the utility of an invasive versus a noninvasive culturing approach identified 4 randomized trials examining this issue.6669 Overall, an invasive approach did not alter mortality, but patients undergoing bronchoscopy were much more likely to have their antibiotic regimens modified by clinicians. This suggests that the invasive approach may allow more directed use of antibiotics. Recently, the Canadian Critical Care Trials Group conducted a multicenter randomized trial looking at this issue.11 There was no difference between the 2 approaches in mortality, number of ventilator days, and antibiotic usage. However, all patients in this study were immediately treated with empiric broad‐spectrum antibiotics until culture results were available, and the investigators did not have a protocol for stopping antibiotics after culture data were available.

In summary, both invasive and noninvasive culturing approaches are considered acceptable options for diagnosing VAP. Readers interested in learning more about this topic should read the worthwhile Expert Discussion70 by Chastre and colleagues55 at the end of this article. In general, we recommend that ICU clinicians use a combination of clinical suspicion (based on the CPIS or other objective data) and cultures ideally obtained prior to antibiotics. Regardless of the chosen culturing approach, clinicians must recognize that 1 of the most important determinants of patient outcome is prompt administration of adequate initial antibiotics.7175

Initial Antibiotic Administration

Delaying initial antibiotics in VAP increases the risk of death.7175 If a patient receives ineffective initial therapy, a later switch to appropriate therapy does not eliminate the increased mortality risk. Therefore, a comprehensive approach to VAP diagnosis requires consideration of initial empiric antibiotic administration.

Whenever possible, clinicians should obtain a lower respiratory tract sample for microscopy and culture before administering antibiotics because performing cultures after antibiotics have been recently started will lead to a higher rate of false‐negative results.76 Unless the patient has no signs of sepsis and microscopy is completely negative, clinicians should then immediately start empiric broad‐spectrum antibiotics.57 Once the culture sensitivities are known, therapy can be deescalated to a narrower spectrum.77 Recent studies suggest that shorter durations of therapy (8 days) are as effective as longer courses and are associated with lower colonization rates by antibiotic‐resistant bacteria.62, 78

Initial broad‐spectrum antibiotics should be chosen based on local bacteriology and resistance patterns. Clinicians must remain aware of the most common bacterial pathogens in their local community, hospital, and ICU. This is essential for both ensuring adequate initial antibiotic coverage and reducing overall antibiotic days.65 Unrestrained use of broad‐spectrum antibiotics increases the risk of resistant pathogens. Clinicians must continually deescalate therapy and use narrow‐spectrum drugs as pathogens are identified.79

Prevention of VAP

In 2005, the American Thoracic Society published guidelines for the management of adults with VAP.5 These guidelines included a discussion of modifiable risk factors for preventing VAP and used an evidence‐based grading system to rank the various recommendations. The highest evidence (level 1) comes from randomized clinical trials, moderate evidence (level 2) comes from nonrandomized studies, and the lowest evidence (level 3) comes from case studies or expert opinion. Others have also published their own guidelines and recommendations for preventing VAP.8082 Table 4 shows the key VAP preventive strategies.

Some strategies are not recommended for VAP prevention in general ICU patients. Selective decontamination of the digestive tract (ie, prophylactic oral antibiotics) has been shown to reduce respiratory infections in ICU patients,113 but its overall role remains controversial because of concerns it may increase the incidence of multi‐drug‐resistant pathogens.114 Similarly, prophylactic intravenous antibiotics administered at the time of intubation can reduce VAP in certain patient populations,115 but this strategy is also associated with an increased risk of antibiotic‐resistant nosocomial infections.116 Using kinetic beds and scheduled chest physiotherapy to reduce VAP is based on the premise that critically ill patients often develop atelectasis and cannot effectively clear their secretions. Unfortunately, neither of these modalities has been shown to consistently reduce VAP in medical ICU patients.117119

Definition of VAP

Although early articles used variable criteria for diagnosing VAP, recent studies have traditionally defined VAP as an infection occurring more than 48 hours after hospital admission in a mechanically ventilated patient with a tracheostomy or endotracheal tube.2 In early 2007, the CDC revised their definition for diagnosing VAP.3 These latest criteria state there is no minimum period that the ventilator must be in place in order to diagnose VAP. This important change must be kept in mind when examining future studies.

The term VAP is more specific than the term health careassociated pneumonia. The latter encompasses patients residing in a nursing home or long‐term care facility; hospitalized in an acute care hospital for more than 48 hours in the past 90 days; receiving antibiotics, chemotherapy, or wound care within the past 30 days; or attending a hospital or hemodialysis clinic.

The CDC published detailed criteria for diagnosing VAP in its member hospitals (Tables 1 and 2).3 Because diagnosing VAP in infants, children, elderly, and immunocompromised patients is often confusing because of other conditions with similar signs and symptoms, the CDC published alternate criteria for these populations. A key objective during development of our algorithms was to consolidate and simplify these diagnostic criteria for ICU clinicians.

Etiology

The most commonly isolated VAP pathogens in all patients are bacteria.21 Most of these organisms normally colonize the respiratory and gastrointestinal tracts, but some are unique to health care settings. Tracheal intubation disrupts the body's natural anatomic and physiologic defenses and facilitates easier entry of these pathogens. Typical organisms include Staphylococcus aureus, Pseudomonas aeruginosa, Enterobacter species, Klebsiella pneumoniae, Acinetobacter species, Escherichia coli, and Haemophilus influenzae.22, 23 Unfortunately, the prevalence of antimicrobial resistance among VAP pathogens is increasing.24 Risk factors for antibiotic resistance are common to ICU patients and include recent antibiotics, hemodialysis, nursing home residence, immunosuppression, and chronic wound care.5 Polymicrobial infections are frequently seen in VAP, with up to 50% of all VAP episodes caused by more than 1 organism.25

Viral VAP is rare in immunocompetent hosts, and seasonal outbreaks of influenza and other similar viruses are usually limited to nonventilated patients.26 However, influenza is underrecognized as a potential nosocomial pathogen, and numerous nosocomial outbreaks because of influenza have been reported.2731 Although herpes simplex virus is often detected in the respiratory tract of critically ill patients, its clinical importance remains unclear.32

Fungal VAP is also rare in immunocompetent hosts. On the other hand, pulmonary fungal infections are common in immunocompromised patients, especially following chemotherapy and transplantation. Candida species are often isolated from the airways of normal hosts, but most cases traditionally have been considered clinically unimportant because these organisms are normal oropharyngeal flora and rarely invade lung tissue.33, 34 It is unclear whether recent studies suggesting Candida colonization is associated with a higher risk for Pseudomonas VAP will change this conventional wisdom.3537

Immunocompromised patients with suspected VAP are unique because they are at risk not only for typical bacteria (which are the most common causes of VAP) but also for rarer opportunistic infections and noninfectious processes that mimic pneumonia.3840 While assessing these patients, clinicians must consider the status of the underlying disease, duration and type of immunosuppression, prophylactic regimens, and risk factors for noninfectious causes of pulmonary infiltrates.41 Common opportunistic infections include viruses, mycobacteria, fungi, and Pneumocystis. Noninfectious processes include pulmonary edema, drug toxicity, radiation pneumonitis, engraftment syndrome, bronchiolitis obliterans organizing pneumonia, alveolar proteinosis, transfusion‐related lung injury, alveolar hemorrhage, and progression of underlying disease. In general, diagnosing VAP in the immunocompromised patient requires a prompt, comprehensive, and multidisciplinary approach.38

In preterm and term infants, the most common VAP pathogens are gram‐negative organisms such as E. coli and P. aeruginosa. Other less common pathogens are Enterobacter, Klebsiella, Acinetobacter, Proteus, Citrobacter, and Stenotrophomonas maltophilia.42, 43 Infants with a preceding bloodstream infection or prolonged intubation are more likely to develop VAP.43, 44 Unfortunately, gram‐negative bacteria often colonize the airways of mechanically ventilated infants, and tracheal aspirate culture data are difficult to interpret in this population.42

Children are more likely to develop VAP if they are intubated for more than 48 hours. The most common pathogens isolated from tracheal aspirates in mechanically ventilated children are enteric gram‐negative bacteria, P. aeruginosa, and S. aureus.45, 46 Few studies have precisely delineated the pathogenesis of VAP in the pediatric ICU population.

Overall, the causes of VAP vary by hospital, patient population, and ICU type. Therefore, it is essential that ICU clinicians remain knowledgeable about their local surveillance data.21 Awareness of VAP microbiology is essential for optimizing initial antibiotic therapy and improving outcomes.

Early Versus Late VAP

Distinguishing between early and late VAP is important for initial antibiotic selection because the etiologic pathogens vary between these 2 periods.4749 Early VAP (days 14 of hospitalization) usually involves antibiotic‐sensitive community‐acquired bacteria and carries a better prognosis. In contrast, late VAP (5 days after hospital admission) is more likely to be caused by antibiotic‐resistant nosocomial bacteria that lead to increased morbidity and mortality. All patients who have been hospitalized or have received antibiotics during the prior 90 days should be treated as having late VAP because they are at much higher risk for colonization and infection with antibiotic‐resistant bacteria.47 Of note, 2 recent studies suggest that pathogens in the early and late periods are becoming similar at some institutions.50, 51 Overall, the distinction between early and late VAP is important because it affects the likelihood that a patient has antibiotic‐resistant bacteria. If antibiotic‐resistant pathogens are suspected, initial therapy should include empiric triple antibiotics until culture data are available.

Culturing Approaches

Because clinical criteria alone are rarely able to accurately diagnose VAP,52, 53 clinicians should also obtain a respiratory specimen for microbiologic culture. Despite the convenience of blood cultures, their sensitivity for diagnosing VAP is poor, and they rarely make the diagnosis alone.54 Two methods are available for culturing the lungsan invasive approach (eg, bronchoscopy with bronchoalveolar lavage) and a noninvasive approach (eg, tracheal aspirate).

Some investigators believe that adult patients with suspected VAP should always undergo an invasive sampling of lower‐respiratory‐tract secretions.55 Proponents of the invasive approach cite the frequency with which potential pathogens colonize the trachea of ICU patients and create spurious results on tracheal aspirates.22 In addition, several studies have shown that clinicians are more likely to narrow the spectrum of antibiotics after obtaining an invasive diagnostic sample.56 In other words, the invasive approach has been associated with better antimicrobial stewardship.

Other investigators believe that a noninvasive approach is equally safe and effective for diagnosing VAP.57 This clinical approach involves culturing a tracheal aspirate and using a pneumonia prediction score such as the clinical pulmonary infection score (CPIS; Table 3). The CPIS assigns 012 points based on 6 clinical criteria: fever, leukocyte count, oxygenation, quantity and purulence of secretions, type of radiographic abnormality, and results of sputum gram stain and culture.58 As developed, a CPIS > 6 has a sensitivity of 93% and a specificity of 100% for diagnosing VAP.58 However, the CPIS requires that nurses record sputum volume and that the laboratory stains the specimen. When the CPIS has been modified based on the unavailability of such resources, the results have been less impressive.5961 Despite studies showing that a noninvasive clinical approach can achieve adequate initial antibiotic coverage and reduce overuse of broad‐spectrum agents,62, 63 clinicians who use the CPIS must understand its inherent limitations.

A meta‐analysis56 comparing the utility of an invasive versus a noninvasive culturing approach identified 4 randomized trials examining this issue.6669 Overall, an invasive approach did not alter mortality, but patients undergoing bronchoscopy were much more likely to have their antibiotic regimens modified by clinicians. This suggests that the invasive approach may allow more directed use of antibiotics. Recently, the Canadian Critical Care Trials Group conducted a multicenter randomized trial looking at this issue.11 There was no difference between the 2 approaches in mortality, number of ventilator days, and antibiotic usage. However, all patients in this study were immediately treated with empiric broad‐spectrum antibiotics until culture results were available, and the investigators did not have a protocol for stopping antibiotics after culture data were available.

In summary, both invasive and noninvasive culturing approaches are considered acceptable options for diagnosing VAP. Readers interested in learning more about this topic should read the worthwhile Expert Discussion70 by Chastre and colleagues55 at the end of this article. In general, we recommend that ICU clinicians use a combination of clinical suspicion (based on the CPIS or other objective data) and cultures ideally obtained prior to antibiotics. Regardless of the chosen culturing approach, clinicians must recognize that 1 of the most important determinants of patient outcome is prompt administration of adequate initial antibiotics.7175

Initial Antibiotic Administration

Delaying initial antibiotics in VAP increases the risk of death.7175 If a patient receives ineffective initial therapy, a later switch to appropriate therapy does not eliminate the increased mortality risk. Therefore, a comprehensive approach to VAP diagnosis requires consideration of initial empiric antibiotic administration.

Whenever possible, clinicians should obtain a lower respiratory tract sample for microscopy and culture before administering antibiotics because performing cultures after antibiotics have been recently started will lead to a higher rate of false‐negative results.76 Unless the patient has no signs of sepsis and microscopy is completely negative, clinicians should then immediately start empiric broad‐spectrum antibiotics.57 Once the culture sensitivities are known, therapy can be deescalated to a narrower spectrum.77 Recent studies suggest that shorter durations of therapy (8 days) are as effective as longer courses and are associated with lower colonization rates by antibiotic‐resistant bacteria.62, 78

Initial broad‐spectrum antibiotics should be chosen based on local bacteriology and resistance patterns. Clinicians must remain aware of the most common bacterial pathogens in their local community, hospital, and ICU. This is essential for both ensuring adequate initial antibiotic coverage and reducing overall antibiotic days.65 Unrestrained use of broad‐spectrum antibiotics increases the risk of resistant pathogens. Clinicians must continually deescalate therapy and use narrow‐spectrum drugs as pathogens are identified.79

Prevention of VAP

In 2005, the American Thoracic Society published guidelines for the management of adults with VAP.5 These guidelines included a discussion of modifiable risk factors for preventing VAP and used an evidence‐based grading system to rank the various recommendations. The highest evidence (level 1) comes from randomized clinical trials, moderate evidence (level 2) comes from nonrandomized studies, and the lowest evidence (level 3) comes from case studies or expert opinion. Others have also published their own guidelines and recommendations for preventing VAP.8082 Table 4 shows the key VAP preventive strategies.

Some strategies are not recommended for VAP prevention in general ICU patients. Selective decontamination of the digestive tract (ie, prophylactic oral antibiotics) has been shown to reduce respiratory infections in ICU patients,113 but its overall role remains controversial because of concerns it may increase the incidence of multi‐drug‐resistant pathogens.114 Similarly, prophylactic intravenous antibiotics administered at the time of intubation can reduce VAP in certain patient populations,115 but this strategy is also associated with an increased risk of antibiotic‐resistant nosocomial infections.116 Using kinetic beds and scheduled chest physiotherapy to reduce VAP is based on the premise that critically ill patients often develop atelectasis and cannot effectively clear their secretions. Unfortunately, neither of these modalities has been shown to consistently reduce VAP in medical ICU patients.117119