Clinical presentation and diagnostic evaluation of ventilator-associated pneumonia

INTRODUCTION — Ventilator-associated pneumonia (VAP) is a type of hospital-acquired pneumonia (HAP) that develops after more than 48 hours of mechanical ventilation [1]. VAP is a common and serious problem in the intensive care unit that is associated with an increased risk of death. Accurate diagnosis is important so that appropriate treatment can be instituted early while simultaneously avoiding antibiotic overuse and consequently, antibiotic resistance.

Patients with severe HAP who require mechanical ventilation after the onset of infection do not meet the definition of VAP. This is termed ventilated hospital-acquired pneumonia (VHAP) [1]. However, the microbiology, diagnostic evaluation, and outcomes of VHAP are more similar to VAP than HAP [1-4]. The clinical presentation and diagnosis of VAP are reviewed here. The epidemiology, pathogenesis, and risk factors for VAP as well as its prevention and treatment are discussed separately. (See "Treatment of hospital-acquired and ventilator-associated pneumonia in adults" and "Risk factors and prevention of hospital-acquired and ventilator-associated pneumonia in adults" and "The ventilator circuit".)

CLINICAL PRESENTATION

Clinical features — Most patients with VAP present with a gradual or sudden onset of the following more than 48 hours after intubation [5]:

●Symptoms – dyspnea (few patients have symptoms since most are nonverbal on mechanical ventilation)

●Signs – fever, tachypnea, increased or purulent secretions, hemoptysis, rhonchi, crackles, reduced breath sounds, bronchospasm

●Ventilator mechanics – reduced tidal volume, increased inspiratory pressures

●Laboratory findings – worsening hypoxemia, leukocytosis

●Imaging – new or progressive infiltrate on chest radiograph or computed tomography (CT) (see 'Chest imaging' below)

Features may also be accompanied by systemic abnormalities, such as encephalopathy or sepsis. (See "Sepsis syndromes in adults: Epidemiology, definitions, clinical presentation, diagnosis, and prognosis", section on 'Sepsis'.)

Importantly, as isolated findings, none of these features are sensitive or specific for the diagnosis of VAP [6]. In particular, fever and respiratory distress are common among intubated patients with a wide variety of etiologies that must also be considered. (See "Fever in the intensive care unit" and "Assessment of respiratory distress in the mechanically ventilated patient".)

Chest imaging — We obtain a chest radiograph on all patients with suspected VAP [1,7]. A diagnosis of VAP cannot be made without the identification of an infiltrate on chest imaging, particularly one where the filtrate is new or progressive. Common radiographic abnormalities in VAP include alveolar infiltrates, air bronchograms, and silhouetting of adjacent solid organs. The chest radiograph can also help determine the severity of the disease (multilobar versus unilobar) and identify complications, such as pleural effusions or cavitation.

Chest imaging abnormalities alone are insufficient to diagnose VAP because the findings are nonspecific (ie, imaging is frequently abnormal even in the absence of VAP). This was illustrated by an observational study in which only 43 percent of patients who had clinical and radiographic evidence of VAP at the time of their death were subsequently confirmed to have VAP by postmortem examination [7].

DIAGNOSTIC EVALUATION — VAP should be suspected in patients with a new or progressive pulmonary infiltrate on imaging PLUS supportive clinical findings of infection (eg, fever, secretions, leukocytosis) (see 'Clinical presentation' above). The diagnosis is confirmed when lower respiratory tract sampling identifies a pathogen. (See 'Diagnosis' below.)

Advanced imaging — Chest CT, without contrast, is not routine in patients with suspected VAP but may be useful in patients with a normal chest radiograph who have clinical symptoms of respiratory tract infection (eg, patients with fever plus leukocytosis and purulent tracheobronchial secretions). CT may also help identify a target lobe for sampling. Chest CT may also be appropriate in those with a previous CT diagnosis of pneumonia to look for new or progressive changes or the development of a pleural effusion. However, in mechanically ventilated patients, pulmonary infiltrates are frequently identified and may be associated with multiple causes. Therefore, imaging determination of VAP in the critical care setting remains nonspecific [7,8]. Lung ultrasonography may be more useful for ruling out pneumonia, although subpleural consolidations and dynamic air bronchograms may support VAP [9]. Bedside ultrasound can also be employed to diagnose VAP, although whether it leads to improved outcomes over conventional chest imaging is unknown [10].

Respiratory tract sampling

Timing of sampling and empiric antibiotics — Respiratory samples are ideally obtained prior to the initiation of antibiotics or change of antibiotic therapy (in those already receiving antibiotics), because antibiotic therapy reduces the sensitivity of both the microscopic analysis and culture [11-14]. However, not uncommonly, severe illness or delays in sampling requires that empiric antibiotic therapy be initiated prior to diagnostic sampling. Nonetheless, once respiratory specimens have been obtained, empiric antibiotic therapy is indicated for suspected VAP, targeting the suspected organism using local antibiograms (algorithm 1), the details of which are discussed separately. (See "Treatment of hospital-acquired and ventilator-associated pneumonia in adults", section on 'Empiric therapy'.)

Respiratory tract sample type — Most experts agree that in patients with suspected VAP, the lower respiratory tract should be sampled and peripheral blood cultures should be sent. However, experts disagree on how the respiratory sample should be obtained (invasive versus noninvasive sampling) and whether or not cultures should be quantitative or nonquantitative.

Our preferred approach — The authors favor the approach set out by the 2017 guidelines issued by The European Respiratory Society (ERS)/European Society of Intensive Care Medicine (ESCIM)/European Society of Clinical Microbiology/Infectious Diseases (ESCMID)/Asociación Latinoamericana del Tórax (ALAT), which state a preference for invasive sampling methods (eg, mini-bronchoalveolar lavage [BAL], bronchoscopic BAL, or protected specimen brush [PSB]) with quantitative cultures [2]. This preference is based upon the high value that we place on diagnostic accuracy and the potential to reduce antibiotic exposure, thereby promoting good antibiotic stewardship when compared with noninvasive sampling (ie, endotracheal aspiration) and nonquantitative cultures. In contrast, other experts prefer the approach set forth by The Infectious Diseases Society of America/The American Thoracic Society, which state a preference for noninvasive sampling with semiquantitative cultures for the diagnosis of VAP [1,15]; this preference is based upon evidence that demonstrates no difference in mortality or length of stay with either approach [16]. (See 'Alternative approach' below.)

However, practice varies widely and is often institution-, physician-, and patient-specific. The wide variation in practice is partially explained by a lack of gold standard for the diagnosis of VAP [17] and differences in opinion regarding the relative importance of specific clinical outcomes (mortality, length of stay, antibiotic resistance). Additional factors include the ability of the patient to tolerate invasive sampling, and the presence of other indications for bronchoscopy (eg, endobronchial obstruction, bleeding, cancer, invasive pneumonia in an immunosuppressed patient, suspicion for a second organism [eg, patients with bronchiectasis], noninvasive methods are unhelpful), available expertise, and institutional availability and cost of quantitative cultures. As an example, while bronchoscopy may be preferred in an immunosuppressed patient with hemoptysis, endotracheal aspiration may be preferred in patients in whom the suspicion for VAP is low or in whom the risk of bronchoscopy is high (eg, patients with barotrauma, high peak pressures, severe hypoxemia). (See 'Society guideline links' below.)

Invasive respiratory sampling — Invasive methods of sampling generally involve nonbronchoscopic methods (eg, mini-BAL) or bronchoscopic methods (eg, bronchoscopic BAL or PSB) [18-26].

●Bronchoscopic BAL is our preferred method of lower respiratory tract sampling. The rationale for this approach is that BAL, compared with PSB (and probably mini-BAL), is a larger sample that obtains a dominant alveolar component with minimal airway contamination. When compared with noninvasive sampling techniques (endotracheal aspirates), bronchoscopic sampling has been shown in several studies to reduce antibiotic overuse and result in more rapid de-escalation of antimicrobial therapy without impacting mortality or length of stay [27-30] (see 'Efficacy' below). A larger sample is also especially helpful in immunocompromised patients where additional microbiologic studies are typically warranted. In addition, cytospin and differential cell count can be performed on BAL fluid which may be helpful in non-neutropenic hosts where the identification of less than 50 percent neutrophils would make acute bacterial pneumonia unlikely. Additional advantages include the ability to inspect the airways for purulence and exclude conditions like malignancy and hemorrhage. PSB is an alternative to BAL. Rarely, both BAL and PSB are obtained, in which case many experts do the PSB first.

•BAL involves the infusion and aspiration of sterile saline through a flexible bronchoscope that is wedged in an affected bronchial segmental or subsegmental orifice (eg, fourth order bronchus (figure 1)). The technique of bronchoscopy and BAL are discussed in detail separately. (See "Basic principles and technique of bronchoalveolar lavage" and "Flexible bronchoscopy in adults: Preparation, procedural technique, and complications".)

•PSB is a brush that is contained within a protective sheath, which minimizes the likelihood that the brush will be contaminated during bronchoscopy. The procedure involves placing the bronchoscope tip next to the affected bronchial segmental orifice, pushing the sheath through the bronchoscope under direct visual guidance, and then advancing the brush out of the sheath and into the airway. Specimens are collected by brushing the airway wall, withdrawing the brush into the sheath, and then removing the sheath from the bronchoscope. The distal end of brush can be cut and sent for microscopic analysis and culture.

One disadvantage of bronchoscopic sampling is that it is only performed by physicians with expertise in the procedure and cannot be performed by ancillary staff. In addition, it is more invasive and ventilated patients are at risk of complications including worsening hypoxemia and barotrauma. However, indirect data suggest that the invasive approach incurs minimal added risk to the patient, although these data are likely influenced by selection bias in favor of bronchoscopy [31,32]. (See "Diagnosis, management, and prevention of pulmonary barotrauma during invasive mechanical ventilation in adults" and "Flexible bronchoscopy in adults: Overview".)

●Mini-BAL is performed by advancing a catheter through the endotracheal tube blindly until resistance is met, infusing sterile saline through the catheter (typically three 50 mL aliquots), and then aspirating using the syringe (the catheter is estimated to be located in the distal endobronchial airway [eg, second or third order bronchus (figure 1)]). We generally reserve mini-BAL for those in whom bronchoscopy is too risky or not available. This procedure requires less expertise and training than bronchoscopy and it is often performed by ancillary staff (eg, nurses and respiratory therapists) rather than clinicians. Complications are rare but include hypoxemia, endobronchial perforation, and barotrauma. While mini-BAL samples are likely to be less contaminated by the airway than endotracheal aspirates, they are less likely to contain a deep alveolar sample [20,33]. Supporting its role as an alternative to bronchoscopic-BAL is a meta-analysis of six studies where patients underwent both mini- and bronchoscopic-BAL in succession [34]. The reported sensitivity of mini-BAL for VAP was 0.9 (95% CI 0.78-1) and specificity was 0.83 (95% CI 0.72-0.94). However, interpretation is limited due to variations in volume of saline used and heterogeneity of patients studied.

Microscopic analysis and quantitative culture — All respiratory tract samples should be sent for microscopic analysis and our preference is that quantitative cultures be obtained:

●Microscopic analysis – This involves semi-quantitative analysis of polymorphonuclear leukocytes and other cell types, as well as the Gram stain. While not diagnostic of VAP per se, the microscopy results return before cultures and can be helpful in determining a possible pathogen and alter antibiotic selection [35]. The presence of abundant neutrophils is consistent with VAP and the bacterial morphology may suggest a likely pathogen (eg, Gram-negative rods). In a prospective cohort study of 39 patients who underwent BAL, VAP was correctly excluded in all patients in whom neutrophils were fewer than 50 percent of the total nucleated cells [36].

●Quantitative cultures – Bacteria can be counted on any respiratory specimen. VAP is supported when an established threshold of bacterial growth is exceeded [37]. Only bacteria that are pulmonary pathogens should be counted. As examples, Staphylococcus epidermidis and most Gram-positive bacilli (except actinomycosis and nocardia) should not be counted.

Typical thresholds include the following:

•Endotracheal aspirates – ≥1,000,000 colony forming units (cfu)/mL

•Bronchoscopic- or mini-BAL – 10,000 cfu/mL

•PSB – 1000 cfu/mL

These thresholds are considered sufficiently high that patients with tracheobronchial colonization are unlikely to be mistakenly detected as having VAP (ie, low false-positive rate) [13,38].

Lower thresholds are accepted by some experts if it is felt that the risk of missing VAP (ie, a false-negative result) exceeds the risk of unnecessary treatment (ie, a false-positive result) [39]. For example, in a prospective cohort study of 122 patients with suspected VAP, thresholds between 1000 and 10,000 cfu/mL for BAL specimens and between 100 and 1000 cfu/mL for PSB specimens increased the sensitivity (77 to 87 percent for BAL and from 67 to 81 percent for PSB) without significantly decreasing the specificity (from 77 to 73 percent for BAL and from 88 to 84 percent for PSB) [40].

Generally speaking, compared with quantitative cultures derived from bronchoscopic specimens, quantitative cultures derived from nonbronchoscopic specimens have similar or higher diagnostic sensitivity (ie, low false negative rate) but a lower specificity (ie, higher false-positive rate) [20,22]. For example in a prospective cohort study of 38 patients, the sensitivity was highest for endotracheal aspirates (87 percent), and lower for mini-BAL (67 percent), bronchoscopic BAL (58 percent), and PSB (42 percent) [20].

Quantitative cultures are not routinely performed in most laboratories, unless specifically requested. They are generally considered more labor-intensive and more costly than qualitative or semiquantitative cultures.

Quantifying anaerobes typically follow the same rules, but take longer and need specific laboratory expertise such that few laboratories perform it.

Efficacy — Data that support invasive sampling with quantitative cultures include mostly observational data and small randomized trials that report an increased number of antibiotic-free days and a reduction in the over-identification of VAP, when compared with less invasive techniques:

●In one study of 413 patients with suspected VAP, compared with noninvasive sampling (eg, endotracheal aspiration) with nonquantitative cultures, an invasive quantitative strategy (BAL or PSB) combined with an algorithm for treatment de-escalation resulted in a reduction in mortality at day 14 (16 versus 26 percent) as well as reduced antibiotic use (mean number of antibiotic-free days, 5 versus 2.2 at 14 days and 11.5 versus 7.5 at 28 days) and organ failure scores [28]. Randomized trials and a meta-analysis performed since then have not shown any difference in mortality, length of stay, or antibiotic-free days [41,42]. However, there were no antimicrobial de-escalation protocols in those studies which may have affected the data on antibiotic use. These data are discussed in detail separately. (See 'Noninvasive respiratory sampling' below.)

●In a prospective trial of 155 patients with suspected VAP, bronchoscopic quantitative methods of sampling were associated with a reduction in the number of samples in which bacteria were identified (43 versus 86 percent) and led to a reduction in antibiotic use when compared with noninvasive sampling methods [43].

●In an observational study of 89 patients with clinically suspected VAP and a negative (<10⁴ colony forming units/mL) quantitative BAL, early discontinuation of antibiotics (within one day) resulted in fewer superinfections (23 versus 43 percent), respiratory superinfections (10 versus 29 percent), and multidrug resistant superinfections (8 versus 36 percent), compared with patients in whom antibiotics were discontinued later than one day [44].

Alternative approach — Noninvasive sampling (ie, endotracheal aspirates) with semiquantitative cultures (or less commonly, qualitative cultures) is the alternative approach preferred by The Infectious Diseases Society of America/The American Thoracic Society [1,15]. This preference is based upon the lack of clear evidence demonstrating superior outcomes (mortality or length of stay) with this approach compared with invasive sampling and quantitative cultures (described above) [16]. (See 'Our preferred approach' above.)

Noninvasive respiratory sampling — Tracheobronchial aspiration (ie, endotracheal aspirate) is performed by advancing a catheter through the endotracheal tube until resistance is met and suction is applied (likely located in trachea or main stem bronchus (figure 1)). The sample is directly aspirated into a sterile specimen trap that can be sent for microbiologic analysis (picture 1).

A clinician is not necessary to perform or supervise tracheobronchial aspiration. This method is cheap, safe, efficient, and facilitates serial sampling. However, it may be less accurate for sampling the alveolar component of the lower respiratory tract and lead to the over diagnosis of VAP. It may also be less suitable for patients with invasive pneumonia who are immunocompromised. (See 'Efficacy' above.)

Microscopic analysis and nonquantitative culture

●Microscopic analysis is standard for all respiratory samples and is described above. (See 'Microscopic analysis and quantitative culture' above.)

●Semiquantitative cultures are typically reported as showing heavy, moderate, light, or no bacterial growth. The amount of growth suggests VAP has not been firmly established, but most experts consider moderate or heavy growth to be positive. Compared with quantitative cultures, semiquantitative cultures are less likely to distinguish patients whose airways are colonized from those who have VAP [16,45]. As a result, false-positive results are more likely, which can potentially lead to over treatment of VAP. Semiquantitative cultures are generally less labor intensive than quantitative cultures and less costly.

●Qualitative cultures do not specify the amount of bacterial growth. VAP is considered to be present when a sample is positive. Their diagnostic utility has not been directly compared with semiquantitative or quantitative cultures.

Efficacy — Noninvasive sampling with semiquantitative cultures has not been shown to affect important clinical outcomes when compared with quantitative invasive techniques [16,27-29,41,46,47]. This was best illustrated in a meta-analysis of three randomized trials (1240 patients), which found that semiquantitative cultures from noninvasive samples did not alter mortality, days of mechanical ventilation, length of intensive care unit stay, or antibiotic use when compared with quantitative invasive strategies [16]. However, many of the included studies did not use a de-escalation strategy for antimicrobials which may have biased the data favoring the noninvasive strategy.

Endotracheal aspiration typically has the highest sensitivity for the diagnosis of VAP but lower specificity when compared with invasive strategies, which may potentially lead to the over diagnosis and over treatment of VAP; these data are discussed above. (See 'Efficacy' above.)

Lung biopsy criteria — Lung biopsy is not routinely performed in patients with suspected VAP since a diagnosis of VAP can be made in most patients using lower respiratory tract sampling and cultures. Lung biopsy may be reserved for patients in whom infiltrates are progressive despite antibiotic therapy or patients in whom a non-infectious etiology is suspected.

The purpose of acquiring tissue under these circumstances is to identify a pathogen that may have been missed with previous sampling or a pathogen that is difficult to culture (eg, fungus, herpes viruses) or to identify a noninfectious process masquerading as infection (eg, cancer, cryptogenic organizing pneumonitis, lymphangitis, interstitial pneumonitis, vasculitis).

Choosing between transthoracic needle, transbronchial, or thoracoscopic biopsy is often clinician-specific with most experts electing to perform thoracoscopic biopsy in ventilated patients. Biopsy in a ventilated patient is a high risk invasive procedure and the sensitivity is unknown. (See "Overview of minimally invasive thoracic surgery".)

Role of polymerase chain reaction techniques — Molecular methods have evolved to assist in the rapid diagnosis and antimicrobial management of pathogens in patients with pneumonia including VAP [48]. However, they are not routinely or uniformly performed and test interpretation can be challenging. Polymerase chain reaction (PCR) is a fast and inexpensive technique that amplifies small segments of microbial DNA for identification. Multiplex PCR assays allow multiple tests to be performed simultaneously, which is important in critically ill patients where the list of potential pathogens can be broad. PCR methods can allow for the rapid identification of specific pathogens in respiratory samples facilitating both empiric antibiotic prescription and subsequent modification [48-51]. Commercially available multiplex PCR platforms are available and have demonstrated rapid and reasonably accurate detection of pathogens in patients with suspected VAP to facilitate antibiotic management [52,53]. Further study is needed to help identify for the clinician when and how PCR can be used.

DIAGNOSIS — VAP is a clinical diagnosis made in a patient who has been mechanically ventilated for ≥48 hours who develops a new or progressive lung infiltrate on imaging with clinical evidence that the infiltrate is of infectious origin (eg, fever, purulent sputum, leukocytosis, and decline in oxygenation), together with a positive pathogen identified on microbiologic respiratory sample [1,15]. A positive microbiologic sample in a patient with a normal chest radiograph should raise the suspicion for tracheobronchitis. VAP cannot be confirmed or excluded until the culture results are complete, which generally takes two to three days; thus the diagnosis is retrospective during which time empiric therapy is ongoing. (See 'Clinical definition of ventilator associated pneumonia' below and "Complications of the endotracheal tube following initial placement: Prevention and management in adult intensive care unit patients", section on 'Tracheobronchitis'.)

Importantly, we and others do not include use of the entities of ventilator-associated conditions (VAC) and infection-related ventilator-associated complications (IVACs), which were introduced by the United States Centers for Disease Control and Prevention (CDC) for the purposes of surveillance and quality improvement [54,55]. These definitions do not aid diagnosis and treatment decisions for individual patients. (See 'CDC definitions for ventilator associated events (VAE-VAC-IVAC-VAP)' below.)

Common pathogens isolated in patients with VAP are Staphylococcus aureus, Pseudomonas aeruginosa, and other gram-negative bacilli. The Clinical Trials Transformation Initiative (CTTI) performed a prospective study in United States hospitals in 2016 [56]. Microbiologic testing was collected and recorded in 479 of 539 patients (89 percent) fulfilling study criteria for HAP and VAP in the ICU setting. A bacterial pathogen was identified from at least one source in 308 of the 479 specimens (64 percent). Staphylococcus aureus (22 percent of patients) and Pseudomonas aeruginosa (11 percent of patients) were the most frequently isolated bacterial pathogens individually with Enterobacterials being identified in 25 percent of patients. Common pathogens and their treatment are discussed separately. (See "Epidemiology, pathogenesis, microbiology, and diagnosis of hospital-acquired and ventilator-associated pneumonia in adults", section on 'Microbiology' and "Treatment of hospital-acquired and ventilator-associated pneumonia in adults", section on 'Empiric therapy'.)

Clinical definition of ventilator associated pneumonia — VAP is diagnosed when a patient who has been mechanically ventilated for ≥48 hours develops a new or progressive infiltrate with associated signs and symptoms of infection (eg, new onset of fever, purulent sputum, leukocytosis, decline in oxygenation, altered respiratory mechanics) (see 'Clinical presentation' above) in whom positive respiratory specimens are present (ie, increased neutrophils are seen in the microscopic analysis and growth of a pathogen in culture exceeds a predefined threshold). (See 'Microscopic analysis and quantitative culture' above and 'Microscopic analysis and nonquantitative culture' above.)

CDC definitions for ventilator associated events (VAE-VAC-IVAC-VAP) — The CDC National Healthcare Safety Network implemented ventilator-associated events (VAE) surveillance in January 2013 [54]. This term was proposed to provide a more uniform and consistent manner of reporting cases of ventilator-associated complications. The definitions below are not intended to be used clinically so their impact on VAP prevention management is uncertain.

The VAE system is a three-tier surveillance definition that uses objective, readily available data to identify complications, including VAP, that occur in mechanically ventilated adult patients [57,58]:

●Ventilator-associated condition (VAC) – The first tier definition, VAC, identifies patients with a period of sustained respiratory deterioration (changes in positive end-expiratory pressure [PEEP] ≥3 cm H₂O or fraction of inspired oxygen [FiO₂] ≥0.2 [ie, 20 points] for two days) following a sustained period of stability or improvement on the ventilator (greater than or equal to two days) (algorithm 2).

●Infection-related ventilator-associated complication (IVAC) – The second tier definition, IVAC, requires that patients with VAC also have an abnormal temperature (below 36°C or above 38°C) or white blood cell count (≤4000 or ≥12,000 cells/mm³) and be started on one or more new antibiotics that continue for four or more days (algorithm 3).

●Possible and probable VAP – The third tier definitions, possible and probable VAP, require that patients with IVAC also have laboratory and/or microbiological evidence of respiratory infection. Possible VAP is defined as Gram stain evidence of purulent pulmonary secretions or a pathogenic pulmonary culture in a patient with IVAC (algorithm 4). Probable VAP is defined as Gram stain evidence of purulence plus quantitative or semiquantitative growth of a pathogenic organism beyond specified thresholds (algorithm 5). Probable VAP can also be triggered by positive tests for respiratory viruses, Legionella species, pleural fluid cultures, and suggestive histopathology with or without an abnormal Gram stain result. (See 'Clinical definition of ventilator associated pneumonia' above.)

The effect of these criteria on clinical outcomes is unknown. One prospective study has reported that less than one-third of VAPs are identified by the VAE criteria and that most VAPs are non-preventable events [58]. In addition, patients with VAEs appear to have greater morbidity and mortality when compared with patients without a VAE.

Ventilated hospital-acquired pneumonia — Ventilated hospital-acquired pneumonia (VHAP) represents pneumonia occurring in patients hospitalized for more than 48 hours with the subsequent need for mechanical ventilation [1]. The importance of VHAP is demonstrated by its greater overall prevalence, and in some studies greater mortality, compared with VAP [1-4]. In general, the epidemiology, pathogens, and outcomes of VHAP are more similar to VAP than hospital-acquired pneumonia (HAP) [1,2,4].

THERAPEUTIC ADJUSTMENT AFTER CULTURE RESULTS — Once cultures return, the patient should be re-evaluated to determine if additional diagnostic evaluation or changes in management are warranted. The following suggestions are ideal but a component of individualization and clinical judgement is necessary:

●Patients with positive cultures demonstrating pathogenic organisms who have improved on empiric therapy probably have VAP; the antimicrobial therapy should be pathogen-targeted. (See "Treatment of hospital-acquired and ventilator-associated pneumonia in adults", section on 'Tailoring therapy'.)

●Patients with positive cultures demonstrating pathogenic organisms who have not improved are still likely to have VAP but may have an explanation for lack of clinical improvement including inappropriate antimicrobial therapy or a complication of the VAP (eg, abscess, empyema) or a second source of infection. Alternatively, patients may have a noninfectious etiology for their symptoms. The antimicrobial regimen may need to be adjusted (eg, intensified regimen such as a second agent with improved pulmonary penetration), complication addressed (eg, drainage of pleural fluid), or other potential causes for failing to improve be sought (eg, additional sampling). (See 'Differential diagnosis' below and 'Lung biopsy criteria' above.)

●Patients with negative cultures who have improved are unlikely to have VAP assuming cultures were obtained prior to antimicrobial therapy and antimicrobial therapy should be discontinued. Some experts continue antibiotics empirically when cultures were taken after initiation of antimicrobial therapy and quickly de-escalate according to the clinical response and suspicion for VAP. (See "Treatment of hospital-acquired and ventilator-associated pneumonia in adults", section on 'Tailoring therapy'.)

●Patients with negative cultures who have not improved may or may not have VAP and further management should be individualized according to the suspicion for VAP; if cultures were obtained on antimicrobial therapy and the suspicion for VAP remains strong, antibiotics are frequently continued or altered. If the suspicion is low, then other diagnoses, sites of infection, or other organisms (eg, viruses or fungi) should be sought and the decision to continue antibiotic therapy individualized. In cases where antibiotics are continued, they should be stopped after no more than seven days.

DIFFERENTIAL DIAGNOSIS — There are many causes of pulmonary infiltrates, fever, respiratory abnormalities, and leukocytosis other than ventilator-associated pneumonia (VAP). The following conditions can present with this constellation of findings:

●Aspiration pneumonitis – Aspiration pneumonitis refers to chemical aspiration without infection; it is distinguished from VAP by history (ie, witnessed aspiration), microscopic analysis and culture of respiratory secretions (ie, negative), and clinical course (ie, self-limited). However, some cases of aspiration can be complicated by secondary infection. (See "Aspiration pneumonia in adults".)

●Pulmonary embolism with infarction – Pulmonary embolism can mimic VAP if it causes pulmonary infarction; it is distinguished from VAP when imaging (eg, CT pulmonary angiography [CTPA]) reveals pulmonary embolism. (See "Clinical presentation, evaluation, and diagnosis of the nonpregnant adult with suspected acute pulmonary embolism".)

●Acute respiratory distress syndrome – Acute respiratory distress syndrome (ARDS) is distinguished from VAP by history (ie, risk factors for ARDS may be present) and the microscopic analysis and culture of respiratory secretions (ie, negative). However, the two conditions may coexist. (See "Acute respiratory distress syndrome: Clinical features, diagnosis, and complications in adults".)

●Pulmonary hemorrhage – Both pulmonary hemorrhage and VAP may cause hemoptysis. Pulmonary hemorrhage tends to present with frank bleeding while VAP often appears as blood mixed with purulent secretions, but this distinction is imperfect. Definitively distinguishing pulmonary hemorrhage from VAP requires that the cause of the hemoptysis be identified, typically on bronchoscopy. (See "Evaluation of nonlife-threatening hemoptysis in adults".)

●Lung contusion – Pulmonary contusion is due to trauma, but it may be difficult to distinguish from VAP because the clinical and radiographic manifestations are similar and often delayed following the trauma. Pulmonary contusion is distinguished from VAP by history (ie, recent trauma) and the absence of organisms on microscopic analysis and culture of respiratory secretions. (See "Overview of inpatient management of the adult trauma patient", section on 'Pulmonary contusion'.)

●Infiltrative tumor – The manifestations of a diffuse infiltrative cancer are similar to VAP. Diffuse infiltrative cancer is distinguished from VAP by history (ie, history of malignancy); culture and microscopic analysis are negative but may be positive for malignant cells. Lung biopsy may be necessary. (See "Pulmonary tumor embolism and lymphangitic carcinomatosis in adults: Diagnostic evaluation and management".)

●Radiation pneumonitis – Radiation-induced lung injury may cause acute pneumonitis or chronic fibrosis. The former develops approximately 4 to 12 weeks after irradiation, with symptoms and signs that mimic VAP; it is distinguished from VAP by history (ie, prior irradiation) and negative microscopic analysis and culture of respiratory secretions. (See "Radiation-induced lung injury".)

●Drug reaction – Pulmonary drug toxicity can result from many different drugs, most notably antineoplastic agents (eg, cyclophosphamide, methotrexate). The clinical manifestations of pulmonary drug toxicity can be identical to VAP and the timing of the onset of symptoms and signs is highly variable (ie, days to months after receiving the medication). Pulmonary drug toxicity is distinguished from VAP by history (ie, received a potentially toxic agent within the past months) and negative microscopic analysis and culture of respiratory secretions. (See "Pulmonary toxicity associated with systemic antineoplastic therapy: Clinical presentation, diagnosis, and treatment".)

●Cryptogenic organizing pneumonia – Clinical features of cryptogenic organizing pneumonia (COP) may be identical to VAP; it is distinguished from VAP by history (ie, risk factors for COP may be present, such as a recent viral infection) and negative microscopic analysis and culture of respiratory secretions. Definitive diagnosis of COP requires lung biopsy. (See "Cryptogenic organizing pneumonia".)

●Vasculitis – Vasculitis (eg, systemic lupus erythematosus) is a rare cause of progressive pulmonary infiltrates that may present in a patients with a known diagnosis of an autoimmune disorder and can only be distinguished by biopsy.

TESTS OF LIMITED VALUE — Biomarkers including procalcitonin, C-reactive protein (CRP), and soluble triggering receptor (sTREM-1), and the clinical pulmonary infection score (CPIS), are additional diagnostic tests that have been investigated. However, they have little role in the evaluation of suspected VAP [1].

Biomarkers

Procalcitonin — Unlike patients with suspected community acquired pneumonia, where the role of procalcitonin can facilitate the decision to start antimicrobials, evidence is conflicting in patients with in suspected VAP [59,60]. Until higher quality studies resolve the uncertainty, we believe that serum procalcitonin levels should not be used for this purpose.

Procalcitonin may be useful in patients with confirmed VAP when making the decision to discontinue antibiotic therapy [61] and it may be a useful prognostic marker [62-64]. (See "Procalcitonin use in lower respiratory tract infections", section on 'Ventilator-associated pneumonia'.)

Others — Other biomarkers, such as C-reactive protein (CRP) and soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), were initially considered promising markers for improving diagnostic strategies for VAP. However, subsequent studies reported that the measurement of such biomarkers in BAL fluid has minimal diagnostic value [65-67].

Clinical Pulmonary Infection Score (CPIS) — The CPIS combines clinical, radiographic, physiologic, and microbiologic data into a numerical result (table 1). Initial validation of the CPIS found that a score greater than six correlated with VAP [68]. However, subsequent studies failed to confirm this. In one prospective cohort study, the CPIS identified VAP with a sensitivity and specificity of only 60 and 59 percent, respectively [69].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Hospital-acquired pneumonia and ventilator-associated pneumonia in adults".)

SUMMARY AND RECOMMENDATIONS

●Ventilator-associated pneumonia (VAP) is a type of hospital-acquired pneumonia that develops after 48 hours or more of mechanical ventilation. Patients with severe hospital-acquired pneumonia who require mechanical ventilation after the onset of infection do not meet the definition of VAP. (See 'Introduction' above.)

●The presenting clinical findings of VAP are nonspecific with the gradual or sudden onset of a new or progressive pulmonary infiltrate on imaging and supportive clinical findings of infection including fever, purulent tracheobronchial secretions, leukocytosis, increased respiratory rate, decreased tidal volume, increased minute ventilation, and decreased oxygenation. Common radiographic abnormalities in VAP include alveolar infiltrates, air bronchograms, and silhouetting of adjacent solid organs. Clinical features alone are nonspecific for the diagnosis of VAP. (See 'Clinical presentation' above.)

●In patients with suspected VAP, most experts agree that the lower respiratory tract should be sampled and peripheral blood cultures should be sent obtained prior to the initiation of antibiotics or change of antibiotic therapy (in those already receiving antibiotics). Once the respiratory specimens have been obtained, empiric antibiotic therapy should be administered. (See 'Respiratory tract sampling' above and 'Timing of sampling and empiric antibiotics' above and "Treatment of hospital-acquired and ventilator-associated pneumonia in adults".)

●Experts disagree on how the respiratory sample should be obtained (invasively or noninvasively) and whether or not cultures should be quantitative or nonquantitative. No approach has demonstrated superiority over the other with regards to clinical outcomes including mortality or length of stay. (See 'Respiratory tract sample type' above.)

•We favor invasive sampling methods (eg, mini-bronchoalveolar lavage [BAL], bronchoscopic BAL, or protected specimen brush [PSB]) with quantitative cultures rather than noninvasive sampling (ie, endotracheal aspiration) with nonquantitative cultures. This preference is based upon the high value that we place on diagnostic accuracy and the potential to reduce antibiotic exposure, thereby promoting good antibiotic stewardship. (See 'Our preferred approach' above.)

•Noninvasive sampling using endotracheal aspiration with semiquantitative cultures is an acceptable alternative approach. While this approach is less invasive, it may over diagnose VAP and potentially lead to inappropriate use of antimicrobials. (See 'Alternative approach' above.)

•Practice varies widely and is often institution-, physician-, and patient-specific. Factors that influence the decision include ability of the patient to tolerate invasive sampling, the presence of other indications for bronchoscopy, available expertise, and cost.

•Procalcitonin and other biomarkers such as C-reactive protein (CRP) and soluble triggering receptor (sTREM-1) are not recommended. Lung biopsy is not routinely performed but may be reserved for patients in whom infiltrates are progressive despite antibiotic therapy or patients in whom a noninfectious etiology is suspected. (See 'Tests of limited value' above and 'Lung biopsy criteria' above.)

●VAP is a clinical diagnosis based upon the identification a new or progressive lung infiltrate on imaging with clinical evidence that the infiltrate is of infectious origin (eg, fever, purulent sputum, leukocytosis, and decline in oxygenation), together with a positive pathogen identified on microbiologic respiratory sample. VAP cannot be confirmed or excluded until the culture results are complete, which generally takes two to three days. Once cultures return, the patient should be reevaluated to determine if additional diagnostic evaluation or changes in management are warranted. (See 'Diagnosis' above and 'Therapeutic adjustment after culture results' above.)

●There are many causes of pulmonary infiltrates, fever, respiratory abnormalities, and leukocytosis other than VAP. These include aspiration pneumonitis, pulmonary embolism with infarction, acute respiratory distress syndrome, pulmonary hemorrhage, pulmonary contusion, infiltrative tumor, radiation pneumonitis, pulmonary drug toxicity, cryptogenic organizing pneumonia (COP), and vasculitis. (See 'Differential diagnosis' above.)