Strategies to prevent ventilator-associated pneumonia, ventilator-associated events, and nonventilator hospital-acquired pneumonia in acute-care hospitals: 2022 Update

The purpose of this document is to highlight practical recommendations to assist acute care hospitals to prioritize and implement strategies to prevent ventilator-associated pneumonia (VAP), ventilator-associated events (VAE), and non-ventilator hospital-acquired pneumonia (NV-HAP) in adults, children, and neonates. This document updates the Strategies to Prevent Ventilator-Associated Pneumonia in Acute Care Hospitals published in 2014. This expert guidance document is sponsored by the Society for Healthcare Epidemiology (SHEA), and is the product of a collaborative effort led by SHEA, the Infectious Diseases Society of America, the American Hospital Association, the Association for Professionals in Infection Control and Epidemiology, and The Joint Commission, with major contributions from representatives of a number of organizations and societies with content expertise.

The purpose of this document is to highlight practical recommendations to assist acute-care hospitals to prioritize and implement strategies to prevent ventilator-associated pneumonia (VAP), ventilator-associated events (VAEs), and nonventilator hospitalacquired pneumonia (NV-HAP) in adults, children, and neonates. This document updates the Strategies to Prevent Ventilator-Associated Pneumonia in Acute-Care Hospitals published in 2014. 1 This expert guidance document is sponsored by the Society for Healthcare Epidemiology (SHEA); it is the product of a collaborative effort led by SHEA, the Infectious Diseases Society of America, the Association for Professionals in Infection Control and Epidemiology, the American Hospital Association, and The Joint Commission, with major contributions from representatives of a number of organizations and societies with content expertise.

Summary of major changes
This section lists major changes from the Strategies to Prevent Ventilator-Associated Pneumonia in Acute-Care Hospitals: 2014 Update 1 including recommendations that have been added, removed, or altered. Recommendations are categorized as "essential practices" that should be adopted by all acute-care hospitals (in 2014 these were "basic practices," renamed to highlight their importance as foundational for hospitals' healthcare-associated infection (HAI) prevention programs) or as "additional approaches" that can be considered for use in locations and/or populations within hospitals when these HAIs are not controlled after implementation of essential practices (in 2014 these were "special approaches"). See Tables 2, 3, and 4 for a complete summary of the recommendations contained in this document.

Essential practices
• Added a recommendation for high flow nasal oxygen or noninvasive positive pressure ventilation as options to avoid intubation, minimize duration of intubation, and prevent reintubations • Added a recommendation for spontaneous awakening trials or sedation protocols as effective strategies to minimize sedation in adults • Reclassified endotracheal tubes with subglottic secretion drainage from an Essential Practice to an Additional Approach • Added a recommendation for daily toothbrushing • Added a recommendation to use caffeine therapy to facilitate extubation in preterm neonates

Additional approaches
• Reclassified endotracheal tubes with subglottic secretion drainage as an additional approach rather than an essential practice for adults and older children • Added a recommendation to consider early tracheostomy • Added a recommendation to consider postpyloric rather than gastric feeding in patients at high risk for aspiration

Not recommended
• Oral care with chlorhexidine • Probiotics • Ultrathin polyurethane endotracheal tube cuffs • Tapered endotracheal tube cuffs • Automated control of endotracheal cuff pressures • Frequent endotracheal cuff pressure monitoring

New
• Section on prevention of nonventilator hospital-acquired pneumonia (NV-HAP)

Intended use
This document was developed following the process outlined in the Handbook for SHEA-Sponsored Guidelines and Expert Guidance Documents. 2 No guideline or expert guidance document can anticipate all clinical situations, and this document is not meant to be a substitute for individual clinical judgment by qualified professionals. This document is based on a synthesis of evidence, theoretical rationales, current practices, practical considerations, writing-group consensus, and consideration of potential harm, where applicable.

Methods
SHEA recruited 2 subject-matter experts in the prevention of VAP, VAE, and NV-HAP to lead a panel of members representing the Compendium partnering organizations: SHEA, the Infectious Diseases Society of America (IDSA), the Association for Professionals in Infection Control and Epidemiology (APIC), the American Hospital Association (AHA), and The Joint Commission, as well as representation by the Centers for Disease Control and Prevention (CDC). SHEA utilized a consultant medical librarian, who worked with each panel to develop a comprehensive search strategy for PubMed and Embase (January 2012-July 2019; updated to August 2021). Articles' abstracts were reviewed by panel members in a doubleblind fashion using the abstract management software Covidence (Melbourne, Australia), and potentially relevant articles were reviewed as full text. The Compendium Lead Authors subsequently voted to update the literature findings, and the librarian reran the search to include articles published through August 2021. Panel members reviewed the abstracts of these articles via Covidence and incorporated relevant references.
Recommendations resulting from this literature review process were classified based on the quality of evidence and the balance between desirable and potentially undesirable effects of various interventions (see Table 1). Panel members met via video conference to discuss literature findings; recommendations; quality of evidence for these recommendations; and classification as essential practices, additional approaches, or unresolved issues. Panel members reviewed and approved the document and its recommendations.
The Compendium Expert Panel, made up of members with broad healthcare epidemiology and infection prevention expertise, reviewed the draft manuscript after consensus had been reached by writing panel members.
Following review and approval by the Expert Panel, the 5 partnering organizations, stakeholder organizations, and the CDC reviewed the document. Prior to dissemination, the guidance document was reviewed and approved by the SHEA Guidelines Committee, the IDSA Standards and Practice Guidelines Committee, and the Boards of SHEA, IDSA, APIC, AHA, and The Joint Commission.
All panel members complied with SHEA and IDSA policies on conflict-of-interest disclosure.
Section 1: Rationale and statements of concern 1. Hospitalized patients are at high risk for pneumonia and other pulmonary complications, particularly patients on mechanical ventilation. a. Hospital-acquired pneumonia is the most common nosocomial infection. 3 It affects ∼1 in 100 patients overall and up to 1 in 10 patients on invasive mechanical ventilation. 3,4 The true incidence of nosocomial pneumonia is difficult to discern, however, because diagnostic criteria vary widely, correlate poorly with histology, are often subjective, and are applied differently by different surveyors. 5-8 b. Many hospitals have reported dramatic decreases in VAP rates over the past 20 years, but the extent to which these declines reflect better care versus stricter application of subjective surveillance criteria remains unclear. 9,10 Clinical surveys suggest that 5%-10% of ventilated patients continue to be treated for VAP, [11][12][13][14] and an independent audit conducted by the Centers for Medicare and Medicaid Services reported that VAP rates were stable between 2005 and 2013. 15 c. Patients on mechanical ventilation are at risk for a variety of serious complications in addition to pneumonia. Examples include acute respiratory distress syndrome, fluid overload, atelectasis, pneumothorax, barotrauma, and pulmonary embolism. The CDC created VAE definitions to better capture the breadth of serious complications associated with mechanical ventilation. 16 Approximately 5%-10% of mechanically ventilated patients develop VAEs. [17][18][19][20][21][22][23][24] d. The coronavirus disease 2019 (COVID- 19) pandemic has been associated with an increase in the incidence of nosocomial pneumonia due to hospital-acquired severe acute respiratory coronavirus virus 2 (SARS-CoV-2) infections and hospital-acquired bacterial superinfections in patients admitted with COVID-19 pneumonia. [25][26][27] Differentiating superinfection from underlying COVID-19 pneumonia and COVID-19-related ARDS is challenging due to considerable overlap in clinical signs. 2. VAP, VAE, and NV-HAP are detrimental to patients and increase costs. a. The attributable mortality of VAP is estimated to be ∼10% but varies considerably by type and severity of underlying illness. 28-32 b. VAP and VAE extend duration of invasive mechanical ventilation, prolong intensive care unit (ICU) and hospital length of stay, and increase mortality risk. 18,[21][22][23]31,33,34 They are also associated with greater use of antimicrobials and higher costs. 19,33 Patients with VAEs are ∼50% more likely to die compared to similar patients with VAP. 35 NV-HAP is associated with a doubling of the length of time until discharge and substantially higher mortality rates compared to similar patients without NV-HAP. 36,37 Section 2: Background on detection of VAP, VAE, and NV-HAP 1. Despite pneumonia's clinical importance, our ability to conduct accurate pneumonia surveillance is very limited. a. Pneumonia is usually defined by clinical, radiographic, and microbiological criteria. These signs are neither sensitive nor specific relative to histopathology. 6,[38][39][40] In addition, the signs and symptoms used to diagnose pneumonia are subjective, which leads to substantial interobserver variability. 8,12,13,[41][42][43] Administrative data are similarly inaccurate. 11,[44][45][46][47] Improvements in VAP rates do not reliably correlate with improvements in outcomes. 48,49 i. The weaknesses of traditional pneumonia surveillance definitions limit their utility for measuring the impact of care improvement programs and for benchmarking quality of care between different healthcare facilities. 50 Section 3: Background on prevention of VAP, VAE, and NV-HAP Framework for evaluating and prioritizing interventions 1. The subjectivity and lack of specificity of diagnosing pneumonia complicate the interpretation of VAP and NV-HAP prevention studies. 90 Subjectivity makes it possible that decreases in observed pneumonia rates are due to stricter interpretation of subjective diagnostic criteria rather than true decreases in disease. Lack of specificity makes it possible that lower pneumonia rates are due to less colonization or decreases in conditions that mimic the presentation of pneumonia without corresponding decreases in true pneumonia cases. 2. VAE criteria are more objective and hence less susceptible to these sources of bias, but the literature on VAE prevention is still relatively sparse. 3. Given the limitations of the pneumonia prevention literature and the relative paucity of VAE prevention literature, we prioritize pneumonia prevention strategies associated with improvements in objective outcomes such as duration of mechanical ventilation, ICU or hospital length of stay, mortality, VAEs, antibiotic utilization, and/or costs in randomized controlled trials. In addition, the potential benefits of different interventions are balanced against their feasibility, costs, and potential harms.
Section 4: Recommended strategies to prevent VAP, VAE, and NV-HAP Recommendations are classified as either: 1) essential practices that improve objective outcomes such as duration of mechanical ventilation, length of stay, mortality, VAEs, antibiotic utilization, and/or costs with little risk of harm that should be adopted by all hospitals. We also recommend interventions that are outcome-neutral but cost saving. Or 2) additional approaches that improve objective outcomes (including VAE) but carry some risk of harm, and interventions that lower VAP or NV-HAP rates, but where insufficient data exist to determine their impact on objective outcomes. Hospitals can consider adopting additional approaches if their VAE, VAP, or NV-HAP rates do not improve despite high performance rates of essential practices. Interventions that do not improve VAE, VAP, or NV-HAP rates nor objective outcomes are not recommended. The quality of evidence rating scheme is summarized in Table 1. Recommended strategies are summarized in Table 2 for adults, Table 3 neonates, and Table 4 for pediatric patients.

Essential practices for preventing VAP and/or VAEs in adult patients
Interventions with little risk of harm that are associated with decreases in duration of mechanical ventilation, length of stay, mtality, antibiotic utilization, and/or costs.
Avoid intubation and reintubation if possible 1. Use high-flow nasal oxygen or non-invasive positive pressure ventilation (NIPPV) as appropriate whenever safe and feasible (Quality of Evidence: HIGH). a. High-flow nasal oxygen may help avert intubation in patients with hypoxemic respiratory failure and prevent reintubation after extubation of critically ill patients and postoperative patients compared to conventional oxygen therapy. [91][92][93][94] High-flow nasal oxygen has also been associated with a trend toward less nosocomial pneumonia in patients with hypoxemic respiratory failure. 95 b. NIPPV is associated with lower rates of intubation, reintubation, VAP, and mortality compared to conventional oxygen therapy in patients with acute hypercapnic or hypoxemic respiratory failure. [96][97][98][99][100] Use caution when considering NIPPV to manage patients with impaired consciousness, acute lung injury, acute respiratory distress syndrome, severe hypoxemia, severe acidemia, or when continuing NIPPV for patients whose dyspnea or gas exchange fails to rapidly respond to NIPPV. Helmet ventilation may be associated with better outcomes than facemask ventilation. 99

HIGH
Highly confident that the true effect lies close to that of the estimated size and direction of the effect. Evidence is rated as "HIGH" quality when there are a wide range of studies with no major limitations, there is little variation between studies, and the summary estimate has a narrow confidence interval.
MODERATE The true effect is likely to be close to the estimated size and direction of the effect, but there is a possibility that it is substantially different. Evidence is rated as "MODERATE" quality when there are only a few studies and some have limitations but not major flaws, there is some variation between studies, or the confidence interval of the summary estimate is wide.

LOW
The true effect may be substantially different from the estimated size and direction of the effect. Evidence is rated as "LOW" quality when supporting studies have major flaws, there is important variation between studies, the confidence interval of the summary estimate is very wide, or there are no rigorous studies.
c. High-flow nasal oxygen and NIPPV appear to be similar in their capacity to prevent intubation, reintubation, and postextubation respiratory failure. Some meta-analyses suggest that high-flow nasal cannula may reduce ICU and hospital length of stay compared to NIPPV, while others do not. 92,102 d. Combining high-flow nasal oxygen with NIPPV immediately after extubation may further decrease the risk of reintubation in patients at high risk for extubation failure compared to using high-flow nasal oxygen alone. 103 2. Placing nonintubated patients with COVID-19 acute hypoxemic respiratory failure in the prone position may lower the risk of intubation compared to standard care (Quality of Evidence: MODERATE). 104 Minimize sedation 1. Minimize sedation of ventilated patients whenever possible (Quality of Evidence: HIGH). 105

Preferentially use multimodal strategies and medications
other than benzodiazepines to manage agitation (Quality of Evidence: HIGH). 106 a. Examples include analgesics for pain, reassurance for anxiety, and antipsychotics, dexmedetomidine, and/or propofol for agitation. 106 Dexmedetomidine and propofol are associated with shorter duration of mechanical ventilation and ICU length of stay compared to benzodiazepines. 107 A randomized trial of light sedation with dexmedetomidine versus propofol found no difference in ventilator-free days or mortality. 108 Dexmedetomidine may decrease need for intubation in patients on noninvasive ventilation. 109 3. Utilize a protocol to minimize sedation (Quality of Evidence: HIGH). 110 a. Potential strategies to minimize sedation include nursedriven protocols for targeted light sedation and daily sedative interruptions (ie, spontaneous awakening trials) for patients without contraindications. 106,110 b. A meta-analysis of 6 randomized trials reported that protocols to minimize sedation were associated with significantly shorter ICU length of stay compared to managing patients without protocols. 110 There was no significant association between the use of protocols to minimize sedation and duration of mechanical ventilation or short-term mortality.
There was insufficient evidence to recommend one protocol over another. c. A small, single-center, randomized trial 111 suggested that patients managed with no sedation (but morphine as needed) versus propofol or midazolam may be extubated sooner and have shorter ICU length of stay, but a subsequent multicenter randomized trial of no sedation versus light sedation with daily sedative interruptions reported no difference in ventilator-free days, ICU-free days, or 90-day mortality. 112 4. Implement a ventilator liberation protocol (Quality of Evidence: HIGH) 113 a. Assess readiness to extubate daily in patients without contraindications (i.e., conduct spontaneous breathing trials). 114-117 b. Ventilator liberation protocols are associated with extubating patients an average of 1 day earlier compared to managing patients without a protocol. 113 c. Protocols to minimize sedation, mobilize patients, and liberate them from mechanical ventilation may be synergistic. 118,119 Maintain and improve physical conditioning 1. Provide early exercise and mobilization (Quality of Evidence: MODERATE). a. Early exercise and mobilization programs may shorten duration of mechanical ventilation, reduce ICU length of stay, lower VAP rates, and increase the rate of return to independent function. 113,[120][121][122][123] There is no consistent association between early mobilization and hospital length of stay or mortality. b. Financial modeling suggests that early mobility programs may be cost saving. 119,124 Elevate the head of the bed 1. Elevate the head of the bed to 30-45°(Quality of Evidence: LOW). a. A meta-analysis of 8 randomized trials reported that elevating the head of the bed was associated with a significant reduction in VAP rates but no difference in duration of mechanical ventilation or mortality. 125 The data on outcomes other than VAP, however, were sparse (combined sample size <500 patients); thus, the impact on these outcomes is uncertain. Given the simplicity, ubiquity, minimal risk, lack of cost, and potential benefit of this intervention, we nonetheless classify it as an essential practice while we await further data.
Provide oral care with toothbrushing but without chlorhexidine 1. Provide daily oral care with toothbrushing but without chlorhexidine (Quality of Evidence: MODERATE). a. Daily toothbrushing is associated with significantly lower VAP rates, shorter duration of mechanical ventilation, and shorter ICU length of stay. 126,127 b. Meta-analyses of randomized trials and observational studies allow for the possibility that oral care with chlorhexidine may increase mortality rates. 75,[128][129][130] This is further discussed below.
Provide early enteral rather than parenteral nutrition 1. Provide early enteral rather than parenteral nutrition (Quality of Evidence: HIGH). a. Early enteral nutrition is associated with a lower risk of nosocomial pneumonia, shorter ICU length of stay, and shorter hospital length of stay compared to early parenteral nutrition. 131 Maintain ventilator circuits 1. Change the ventilator circuit only if visibly soiled or malfunctioning (Quality of Evidence: HIGH). a. Changing the ventilator circuit as needed rather than on a fixed schedule has no impact on VAP rates or patient outcomes but decreases costs. 132 b. Follow manufacturers' instructions for use if they differ from this recommendation. c. Follow CDC/Healthcare Infection Control Practices Advisory Committee guidelines and manufacturers' instructions for use of sterilization and disinfection of respiratory care equipment. 133 Additional approaches for preventing VAP and/or VAEs in adult patients Additional approaches are interventions associated with lower VAP rates that that may also decrease VAE rates, duration of mechanical ventilation, length of stay, and/or mortality but carry some risk of harm. Additional approaches also include interventions that are associated with lower VAP rates, but insufficient data exist to determine their impact on objective outcomes. Hospitals can consider adopting additional approaches if their VAP or VAE rates do not improve despite high performance rates with essential practices.
The following interventions may decrease duration of mechanical ventilation, length of stay, and/or mortality in some populations but not in others, and they may confer some risk of harm in some populations.
1. Consider using selective decontamination of the oropharynx and digestive tract to decrease microbial burden in ICUs with low prevalence of antibiotic-resistant organisms. 128,134,135 Antimicrobial decontamination is not recommended in countries, regions, or ICUs with high prevalence of antibiotic-resistant organisms 136 (Quality of Evidence: HIGH). a. A meta-analysis of 6 cluster randomized trials performed in countries with low levels of antibiotic resistance reported that selective decontamination of the oropharynx with topical antibiotics was associated with a 16% reduction in hospital mortality, and decontamination of the oropharynx and digestive tract with a combination of topical, oral, and parenteral antibiotics was associated with an 18% reduction in hospital mortality. 135 Selective digestive decontamination was more effective than selective oral decontamination alone (OR, 0.90; 95% CI, 0.82-0.97 for hospital death). 135 A broader meta-analysis that included a larger but more heterogenous set of studies had similar findings. 137 i. Oral agents that have been used for digestive decontamination include colistin, tobramycin, and amphotericin B. Parenteral agents include cefotaxime. ii. ICUs that implement this practice should actively monitor its impact on antibiotic utilization, antimicrobial resistance, and Clostridioides difficile infections. iii. There is no consensus on what constitutes "low levels of antibiotic resistance," but an arbitrary threshold that has been used by other guidelines and randomized trials is <5% of bloodstream infections caused by extendedspectrum β-lactamase-producing Enterobacterales. 136,138 b. A cluster randomized trial of selective digestive decontamination (without parenteral antibiotics) versus selective oral decontamination versus oral care with 2% chlorhexidine versus routine care conducted in ICUs with high levels of antibiotic resistance (≥5% of bloodstream infections caused by extended-spectrum β-lactamase-producing Enterobacterales) found no difference between study arms in ICU-acquired bloodstream infections or 28-day mortality rates. 136 Selective oral and digestive decontamination with antibiotics is therefore not recommended in settings with high baseline levels of antibiotic resistance.
The following interventions may lower VAP rates, but current data are insufficient to determine their impact on duration of mechanical ventilation, length of stay, and mortality.
1. Consider using endotracheal tubes with subglottic secretion drainage ports to minimize pooling of secretions above the endotracheal cuff in patients likely to require >48-72 hours of intubation (Quality of Evidence: MODERATE). a. Intermittent and continuous drainage of subglottic secretions has been studied in at least 20 randomized controlled trials. On meta-analysis, the use of endotracheal tubes with subglottic drainage reduced VAP rates by 44%. 139 There was no association, however, between subglottic secretion drainage and duration of mechanical ventilation, ICU length of stay, or hospital length of stay. One meta-analysis reported a significant impact on mortality but abstracted one large study twice. 139,140 The effect on mortality was no longer significant after removing the duplicate study (OR, 0.92; 95% CI, 0.83-1.02). One large trial included VAE as an outcome and found no association between subglottic secretion drainage and VAE rates. 140 Some studies have reported that subglottic secretion drainage is associated with less antibiotic utilization, but others have not. 140-142 b. Reductions in duration of mechanical ventilation with subglottic secretion drainage appear to be limited to patients expected to require >48-72 hours of mechanical ventilation. 143 Endotracheal tubes with subglottic secretion drainage ports should therefore be reserved for patients likely to require >48-72 hours of intubation. Patients requiring emergency intubation in the hospital and preoperative patients at risk for prolonged mechanical ventilation are reasonable candidates. c. Extubation followed by immediate reintubation to exchange a conventional endotracheal tube for a subglottic secretion drainage endotracheal tube is not recommended. 2. Consider early tracheostomy (Quality of Evidence: MODERATE). a. Meta-analysis of 17 randomized trials suggests that early tracheostomy (within 7 days of intubation) may be associated with a 40% decrease in VAP rates, less time on mechanical ventilation, and fewer ICU days but no difference in mortality. 144 b. Decision makers should integrate these potential benefits with each patient's values and preferences when determining whether and when to proceed with tracheostomy. 145 3. Consider postpyloric feeding tube placement in patients with gastric feeding intolerance at high risk for aspiration (Quality of Evidence: MODERATE). a. Postpyloric feeding is associated with less aspiration and less pneumonia compared to gastric-tube feeding. Meta-analyses vary in their assessment of whether postpyloric feeding is associated with decreases in ventilator, ICU, and/or hospital length of stay. 146,147 b. Postpyloric tube placement requires special expertise that is not available in all centers and may incur delay in placement. Postpyloric feeding is considered less physiologic than gastric feeding. 131 c. Postpyloric feeding should therefore be reserved for patients with gastric feeding intolerance and for patients at high risk for aspiration as detailed in nutrition society guidelines. 131,148,149 Approaches that should not be considered a routine part of VAP and/or VAE prevention in adult patients The following interventions are inconsistently associated with lower VAP rates and have no impact or negative impact on duration of mechanical ventilation, length of stay, or mortality.
1. Oral care with chlorhexidine (Quality of Evidence: MODERATE) a. Oral care with chlorhexidine has been studied in multiple randomized controlled trials. 150 The impact of oral care with chlorhexidine on pneumonia rates is unclear. Metaanalyses report significantly lower VAP rates, but this signal is driven by unblinded studies. There was no association between oral care with chlorhexidine and lower VAP rates when the analysis was restricted to double-blinded studies. 129 Meta-analyses of both blinded and unblinded studies also show no impact on duration of mechanical ventilation or ICU length of stay. 129,150 Chlorhexidine's lack of impact on VAP, duration of mechanical ventilation, or ICU length of stay was echoed in a large randomized trial of chlorhexidine de-adoption versus usual care. 151 b. Some meta-analyses of randomized trials and some observational studies report an association between oral care with chlorhexidine and higher mortality rates. 75,[128][129][130] The mortality signal is uncertain, however, because other meta-analyses did not find higher mortality rates, there was no change in mortality observed in a large randomized trial of chlorhexidine de-adoption, and the observational studies may be at risk of residual confounding. 126,151,152 Nonetheless, given chlorhexidine's lack of clear impact on VAP rates and the possibility of harm, routine oral care with chlorhexidine is not recommended. Oral care including toothbrushing without chlorhexidine, however, is considered an essential practice. 2. Probiotics (Quality of Evidence: MODERATE) a. Multiple meta-analyses of randomized controlled trials have reported a possible association between probiotics and lower VAP rates, but these analyses have included many studies at high risk of bias due to lack of blinding. [153][154][155][156] There is no association between probiotics and VAP when restricting the analysis to double-blinded studies. 155 This finding was mirrored in a large, rigorous, multicenter, randomized trial conducted after the most recent meta-analysis. 157 Neither this trial nor the meta-analyses found a significant impact on ICU length of stay, hospital length of stay, or mortality. b. Probiotics should not be used in patients with compromised immune systems or gastrointestinal diseases that increase the risk of gut translocation. Multiple cases of fungemia or bacteremia have been reported in patients administered probiotics as have cases of aerosol transmission of probiotics within ICUs. 158-164 3. Ultrathin polyurethane endotracheal tube cuffs (Quality of Evidence: MODERATE) a. Ultrathin polyurethane cuffs seal more uniformly against the tracheal wall and may therefore allow fewer secretions to seep around the cuff and into the lungs. Two small randomized trials 165,166 reported lower VAP rates but a larger, more rigorous study found no difference in VAP rates, duration of endotracheal intubation, or ICU length of stay. 167 Similarly, there were no significant associations between polyurethane cuffs and VAP rates, duration of mechanical ventilation, ICU length of stay, or mortality on meta-analysis. 168 4. Tapered endotracheal tube cuffs (Quality of Evidence: MODERATE) a. A meta-analysis of 5 randomized trials of tapered versus conical endotracheal tube cuffs found no differences in VAP rates, duration of mechanical ventilation, ICU length of stay, hospital length of stay or mortality. 169

5.
Automated control of endotracheal-tube cuff pressure (Quality of Evidence: MODERATE) a. Automated control of endotracheal-tube cuff pressure was associated with lower VAP rates in 2 small trials but this signal has not been borne out in other trials. [170][171][172][173][174] Indeed, 2 large, multicenter, randomized trials of automated cuff-pressure regulation versus thrice-daily manual cuff-pressure assessments found no difference between arms in VAP, VAE, antibiotic utilization, duration of mechanical ventilation, ICU length of stay, or mortality. 174,175 6. Frequent endotracheal-tube cuff-pressure monitoring (Quality of Evidence: MODERATE) a. A single-center prospective trial found no advantage to more frequent versus less frequent cuff-pressure monitoring. 176 At least 1 laboratory investigation suggests that manual measurement of cuff pressure is associated with loss of cuff pressure and potential leakage of fluid around the cuff. 177 7. Silver-coated endotracheal tubes (Quality of Evidence: MODERATE) a. A large, multicenter, randomized controlled trial found that silver-coated endotracheal tubes reduced VAP rates by 36%. However, the organisms associated with VAP included nonpathogenic colonizers, and there was no impact on mean duration of mechanical ventilation, hospital length of stay, or mortality. 178,179 8. Kinetic beds (continuous lateral rotational therapy and oscillation therapy) (Quality of Evidence: MODERATE) a. A meta-analysis of 15 randomized controlled trials found a significant decrease in VAP rates but no impact on duration of mechanical ventilation or mortality. 180 The metaanalysis researchers warned that the observed reduction in VAP rates might be artifactual given weaknesses in contributing studies' design and execution. 9. Prone positioning (Quality of Evidence: MODERATE) a. Prone positioning is associated with lower mortality rates among patients with moderate-to-severe ARDS, but this signal appears to be independent of VAP. Prone positioning may be indicated for reasons other than VAP prevention. 181-183 10. Chlorhexidine bathing (Quality of Evidence: MODERATE) a. Observational studies have suggested that chlorhexidine bathing may reduce the risk of VAP, but this finding has not been borne out in randomized trials. [184][185][186] Chlorhexidine bathing is, however, beneficial in preventing other healthcare-associated infections. 187 Approaches that definitively are not recommended for VAP or VAE prevention Good-quality evidence suggests that the following interventions neither lower VAP/VAE rates nor decrease duration of mechanical ventilation, length of stay, or mortality.
1. Stress-ulcer prophylaxis (Quality of Evidence: MODERATE) a. Stress-ulcer prophylaxis lowers the risk of gastrointestinal bleeding, but a meta-analyses of randomized trials suggested no impact on nosocomial pneumonia, length of stay, or mortality. [188][189][190][191][192]   For example, the smallest available endotracheal tube equipped with subglottic secretion drainage ports is size 6.0 and therefore is not an option for children under 10 years of age. Similarly, the smallest available silver-coated endotracheal tube is size 6.0.

Essential practices for pediatric patients
The following interventions confer minimal risk of harm, and some data suggest that they may lower VAP rates, PedVAE rates, and/or duration of mechanical ventilation.
Avoid intubation if possible 1. Use noninvasive positive pressure ventilation (NIPPV) or high flow oxygen by nasal cannula whenever safe and feasible (Quality of Evidence: MODERATE).
a. Risks of NIPPV in pediatric patients mirror those for adults with the added issue that pediatric patients often need sedation to tolerate NIPPV. 240 250 The bundle included standardized anatomic reference points and securement methods, protocols for high-risk situations, and multidisciplinary apparent-cause analyses. The bundle was associated with significant reductions in unplanned extubations and episodes of cardiovascular collapse. 3. Avoid fluid overload (Quality of Evidence: MODERATE).
a. Meta-analysis of the association between fluid balance and outcomes in critically ill children suggests that fluid overload is associated with increased risk for prolonged mechanical ventilation (>48 hours). 251 b. Interventional studies on fluid management in critically ill children are sparse. One of the few available studies assessed infants at risk for acute kidney injury and fluid overload following cardiac surgery. These infants were randomized to peritoneal dialysis versus furosemide; those randomized to peritoneal dialysis were less likely to develop fluid overload and less likely to have prolonged ventilator use. 252 The generalizability of these findings to other populations is unknown. c. The Pediatric Surviving Sepsis Campaign and the Pediatric Acute Lung Injury Consensus Conference recommend limiting fluid intake, starting diuretics, and/or early renal replacement therapy for children with ARDS and for children in the postresuscitation phase of sepsis. 253,254 Provide regular oral care 1. Provide regular oral care (Quality of Evidence: LOW). a. Four before-and-after studies of VAP bundles that emphasized oral care reported significant decreases in VAP rates following bundle implementation. 234,255-257 b. The American Dental Association recommends beginning oral hygiene a few days after birth in term infants. Wipe the gums with a gauze pad after each feeding to remove plaque and residual formula that could harm erupting teeth. c. For children aged <3 years, the ADA recommends brushing children's teeth as soon as they begin to come into the mouth using fluoride toothpaste in an amount no more than a smear the size of a grain of rice. 258 A pea-sized amount of fluoride toothpaste is recommended for children aged 3-6 years. 258 d. After oral hygiene, rinse and suction the mouth. Keep the oral mucosa and lips clean, moist, and intact using sponge-tipped applicators dipped in non-alcohol, nonperoxide mouth rinse. 255 Elevate the head of the bed 1. Elevate the head of the bed unless medically contraindicated (Quality of Evidence: LOW). a. Three before-and-after studies of VAP bundles that included head of bed elevation reported lower VAP rates. 234 Endotracheal tube selection and management 1. Use cuffed endotracheal tubes (Quality of Evidence: LOW). a. Pediatric intensivists have historically favored uncuffed tubes due to concern that cuffs may induce subglottic stenosis in pediatric airways. Cuffing has proven safe, however, and may decrease the risk of microaspiration. 262,263 Cuffed tubes are now recommended for term newborns and children. 264 2. Maintain cuff pressure and volume at the minimal occlusive settings to prevent clinically significant air leaks around the endotracheal tube, typically 20-25 cm H 2 O. 171,262,265 This "minimal leak" approach is associated with lower rates of post-extubation stridor 265 (Quality of Evidence: LOW). a. The potential merits of automated manometers for VAP prevention have not been studied in pediatric patients.
3. Suction oral secretions before each position change (Quality of Evidence: LOW). 266 Additional approaches to preventing VAP and VAE in pediatric patients The following interventions are associated with minimal risks of harm and some evidence of benefit in adult patients, but data in pediatric populations are limited.
1. Minimize sedation (Quality of Evidence: MODERATE). a. Daily sedative interruptions decreased duration of mechanical ventilation and ICU length of stay without increases in adverse event rates in 1 small, randomized trial. 267 b. There is nonetheless concern that sedative interruptions will increase the frequency of unplanned extubations and reintubations in younger patients, so this practice may be safest in older pediatric patients.

2.
Use endotracheal tubes with subglottic secretion drainage ports (Quality of Evidence: LOW). a. This intervention has not been studied in children and is only feasible for children aged ≥10 years because the smallest available endotracheal tube with subglottic secretion drainage ports is size 6.0. 3. Consider early tracheostomy (Quality of Evidence: LOW).
a. A small, single-center, retrospective cohort study reported that early tracheostomy (<10 days) was associated with lower VAP rates and shorter ICU length of stay compared with late tracheostomy. 268 b. A propensity-matched analysis of the timing of tracheostomy among children with severe traumatic brain injury reported an association between early tracheostomy and lower pneumonia rates, shorter ICU length of stay, and shorter hospital length of stay. 269 c. A meta-analysis of retrospective cohort studies reported that early tracheostomy was associated with lower mortality rates, fewer ventilator days, and shorter ICU length of stay. 270 d. Tracheostomy complications are more frequent in children versus adults. 271 Approaches that are generally not recommended for VAE and VAE prevention in pediatric patients The following interventions have unknown impact on VAP and PedVAE rates and/or have inadequate data on risks.
1. Prolonged systemic antimicrobial therapy for ventilatorassociated tracheitis (Quality of Evidence: LOW) a. One retrospective study found that prolonged antibiotics for ventilator-associated tracheitis did not protect against VAP but did increase the prevalence of multidrug-resistant organisms. 272 Whether, when, and how long to treat ventilatorassociated tracheitis to prevent VAP in children remains unresolved. 273

Selective oropharyngeal or digestive decontamination
(Quality of Evidence: LOW) a. A meta-analysis of 4 randomized trials in critically ill children published between 1991 and 2001 reported that selective digestive decontamination using a combination of oral and parenteral antibiotics may be associated with a decrease in pneumonia rates but no change in mortality. 274 Trials were small, and the longterm impact on antibiotic resistance was not assessed. 3. Prophylactic probiotics (Quality of Evidence: LOW) a. Probiotics should be considered with caution due to sparse data on impact in children, lack of clear benefit in adults, and case reports of Lactobacillus bacteremia associated with probiotic therapy in pediatric patients, including those not known to be immunocompromised. 157,163,[230][231][232][275][276][277][278] No impact on VAP rates for pediatric patients These interventions may be indicated for reasons other than VAP prophylaxis.
1. Oral care with chlorhexidine (Quality of Evidence: MODERATE) a. Chlorhexidine appears to be safe for developing teeth, 279 but randomized controlled trials have found no difference in VAP rates, length of stay, or mortality in infants and children. [280][281][282][283][284][285] 2. Stress-ulcer prophylaxis (Quality of Evidence: LOW) a. Two small studies found no impact on VAP rates. 286,287 A larger, multicenter, cohort study and a meta-analysis reported that acid-suppressive medications were associated with higher VAP rates. 288,289 Lowers VAP rates but no impact on duration of mechanical ventilation, length of stay, or mortality 1. Silver-coated endotracheal tubes (Quality of Evidence: LOW) a. These tubes have not been studied in children and are only feasible for children aged ≥10 years since the smallest available size is 6.0.

No recommendation
These interventions have limited data from pediatric studies, no impact on VAP rates or outcomes in adults, and unclear impact on costs.
1. Closed/in-line suctioning (Quality of Evidence: LOW) a. Closed suctioning may be associated with fewer transient decreases in oxygenation and increases in heart rate and blood pressure compared to open suctioning, but the clinical significance of these findings is unclear. [290][291][292] b. An observational study of open versus closed suctioning in children did not find any difference in VAP rates, length of stay, or mortality, but the significance of these findings are unclear given the lack of blinding and randomization. 293

Recommendations to prevent NV-HAP
Little robust data exist on interventions to prevent NV-HAP. Most studies are nonrandomized, and many do not report the impact on objective outcomes such as length of stay, mortality, or antibiotic utilization. We classify potential prevention strategies into (1) practices supported by interventional studies suggesting lower NV-HAP rates, (2) practices with insufficient data of benefit or harm, and (3) practices that are not recommended, with evidence of futility or possible harm.
Interventions that may lower NV-HAP rates with little risk of harm Provide regular oral care 1. Oral care is the most commonly studied strategy to prevent NV-HAP. Before-and-after series suggest a possible benefit. [294][295][296][297] Two large, cluster randomized trials conducted in nursing homes did not show a benefit, but their generalizability to acute-care hospitals is unknown. 298,299 Most randomized trials in acute-care hospitals have focused on ICU patients, most of whom were on mechanical ventilation, making it difficult to discern their effect on NV-HAP. 300,301 a. Uncertainty remains regarding the most effective protocols, including types of staff involved (eg, dentistry professionals versus nondentistry professionals), frequency of oral care, whether to include an antiseptic, and if so, what type of antiseptic to use (eg, chlorhexidine gluconate, sodium bicarbonate, hydrogen peroxide, cetylpyridinium chloride, povidone-iodine). b. Notwithstanding the gaps in current evidence, we recommend toothbrushing daily given its benefits for oral health and the possible positive impact on objective outcomes observed in before-and-after studies in nonventilated patients and meta-analyses of randomized trials in ventilated patients. 126,127,[294][295][296][297] Diagnose and manage dysphagia 1. Early diagnosis and treatment of dysphagia may prevent NV-HAP, especially among neurologically impaired post-stroke patients. 302-304 2. Potential approaches to diagnose dysphagia include nurseadministered risk assessment tools, bedside functional evaluations of swallowing, video fluoroscopic study, and fiberoptic endoscopic examination. 3. Potential options to manage dysphagia include changes in method of pill administration, adjustments in consistencies of liquids and solids, supervision or assistance with meals, use of straws, and elevation of the head of bed while eating.
Provide early mobilization 1. Data for early mobilization to prevent NV-HAP among hospitalized patients are sparse. 304,305 One quasi-experiment found that bundling mobilization with other interventions reduced NV-HAP, attributable mortality, and antibiotic utilization, but the relative contribution of mobilization to these benefits is unclear. 305 A randomized trial of engaging families to provide turning plus passive mobilization to post-stroke patients versus turning by nursing staff alone reported a significant decrease in pneumonia rates but did not report impact on length of stay or mortality. 306 A nonrandomized controlled trial reported that mobilizing patients in 2 geriatric and respiratory wards was associated with a significant decrease in pneumonia rates compared to usual care in matched wards, although falls were significantly more frequent in the intervention group than in the control group and data were not provided on length of stay or mortality. 307 In a quasi-experimental study, intensified postoperative physical therapy for elderly patients undergoing hip fracture surgery was associated with less pneumonia and shorter length of stay compared to historical controls. 308 2. Additional trials are needed to better quantify the possible benefits versus fall-related harms of mobility programs. Implementation strategies are needed to increase the feasibility, frequency, and safety of mobilizing acute-care patients. In the meantime, early mobilization of patients should take into the account the risk of falls.  305,314,315 Effective bundles have included heterogeneous combinations of oral hygiene, bed positioning, dysphagia diagnosis and management, mobilizing patients, nasal hygiene, sedation restrictions, incentive spirometry, education for physicians and nurses, and/or electronic order-set bundles. One small randomized trial of usual care versus a bundle comprising dysphagia screening, oral care with chlorhexidine, placing the bed in the reverse Trendelenburg position, and vaccination against influenza and pneumococcus reported no difference in NV-HAP rates, length of stay, or mortality but did report a lower 1-year risk of readmission for respiratory infection. 316 2. The respective contribution of each bundle component, the extent to which bundle components are synergistic versus additive, and the most effective combination of interventions to include in bundles remains unknown.
Interventions with insufficient data to determine impact on NV-HAP Bed positioning 1. Elevating the head of the bed is recommended to prevent VAP and VAE despite sparse evidence because some studies suggest benefit, it is simple, economical, and associated with minimal risk of harm in ventilated patients. Even fewer data, however, are available to inform whether and to what extent this applies to NV-HAP. 2. One randomized trial among critically ill patients with tetanus in Vietnam found that semirecumbent position was associated with no difference in pneumonia rates but more frequent complications including the need for tracheostomy. 317 The generalizability of these findings to nontetanus patients in other settings is unknown.
Stress-ulcer prophylaxis 1. Observational studies suggest an association between stressulcer prophylaxis and risk for NV-HAP but we are not aware of any randomized trials assessing the impact of acid-suppressing medications on NV-HAP outside the ICU setting. 318 Approaches not generally recommended for routine NV-HAP prevention Systemic antibiotic prophylaxis 1. Randomized trials of prophylactic antibiotics in acute stroke patients show no impact on pneumonia rates, functional outcome, or mortality. 319

Section 5: Performance measures
Monitoring and reporting 1. Regular monitoring and internal reporting of patient outcomes and adherence rates to recommended prevention strategies ("process measures") are important quality improvement strategies. 2. Both outcome and process-measure reporting are likely beneficial: improving outcomes is the primary goal of care improvement programs but process of care surveillance can help identify specific processes to target for improvement. 3. Report outcome measures to key organizational stakeholders including frontline care providers, service leaders (medical, nursing, respiratory therapy), and senior hospital administrators. Reporting these data back to providers and leaders has been associated with improvements in both performance rates and outcomes. 320-325 4. Report process measures internally only. External reporting of process measure data is not appropriate given substantial variability in the ways different organizations define, collect, analyze, and present process measure data.
Process measures 1. Process-measure definitions and measurement strategies vary widely. 2. For organizations that collect and report process measures: a. Clearly define measures including data sources, inclusion and exclusion criteria, frequency of monitoring, and numerator and denominator criteria. b. Develop a formal system to document compliance.
i. Compliance can be measured via direct observations or via audits of patient charts, bedside checklists, and/ or electronic medical records. Periodically validate the accuracy of paper and/or electronic documentation. c. Perform assessments regularly.
i. The optimal frequency of assessments (eg, once daily, twice daily, or weekly) is not known but can likely be adjusted based on compliance rates and unit size.
ii. An analysis of a large collaborative quality improvement effort suggests that the following approach can be used to determine the frequency of process measure assessments. 326 118,325,[327][328][329] These approaches can be used as starting points to come up with local strategies to define adherence.
Prevention bundles 1. Prevention bundles are widespread in critical care and have been associated with decreases in VAP, VAE, NV-HAP, and in some cases, length of stay and mortality. 2. A meta-analysis of 13 observational studies in adults found that implementation of ventilator bundles was associated with a 10% decrease in mortality, less time to extubation, and shorter length of stay. 330 All the studies included in this analysis, however, were before-and-after or time-series analyses. It is therefore difficult to discern the extent to which lower VAP rates and better outcomes were due to prevention bundles versus changes in patient mix or unrelated changes in care. 331 Observational analyses of pediatric bundles have reported significant decreases in VAP rates, but few data are available on other outcomes. 332 3. Prevention bundles have only been tested in 1 randomized trial.
Researchers in Brazil randomized 188 ICUs to a multifaceted intervention including checklists for the prevention of VAP, daily goal assessments, and clinician prompts. There were no significant differences between intervention versus control units in VAP rates or mortality rates. 329 These results should be interpreted with caution, however, given that baseline adherence with some process measures was already high (eg, head of bed elevation), adherence with other measures did not improve (eg, light sedation, low tidal volume ventilation), and the intervention period may have been too brief to achieve significant changes. 333 4. There is no consensus on which processes to include in VAP/VAE prevention bundles. Bundles vary widely among different centers. 330,334 5. Ventilator bundle components potentially associated with lower mortality rates include staff education, performance feedback, and in adults, elevating the head of the bed, minimizing sedation, and assessing readiness for extubation. 110,113,330,335 Additional promising strategies include conservative fluid management, low tidal volume ventilation, and early mobility. 77 External reporting 1. VAE is a potentially appropriate metric for public reporting, interfacility comparison, and pay-for-performance programs. Better data on the responsiveness of VAE to quality-improvement programs are necessary, however, before recommending VAEs for interfacility comparisons or pay-for-performance programs. Suitable risk adjustment strategies are also needed. 2. PVAP is not suitable for external reporting because substantial variability in clinical and laboratory practices in the acquisition, processing, and interpretation of culture data preclude meaningful comparisons of PVAP rates between institutions. 3. VAP rates generated using former NHSN surveillance definitions are not appropriate for external reporting in light of their considerable subjectivity. 4. Hospitals in states that have mandatory reporting laws must collect and report data as required by their state. Pennsylvania and South Carolina are currently the only 2 states that require hospitals to report VAEs. Local and state health departments can provide specific information on public reporting requirements.
Section 6: Implementation of VAP, VAE, and NV-HAP prevention strategies Prevention of VAP, VAE, and NV-HAP requires implementing best practices to reduce the risk of infection and nurturing a culture that supports implementation. Accountability is an essential principle for preventing healthcare-associated infections. It provides the necessary translational link between science and implementation. Without clear accountability, scientifically based implementation strategies will be used in an inconsistent and fragmented way, decreasing their effectiveness in preventing HAIs. Accountability begins with the chief executive officer and other senior leaders who must make preventing healthcare-associated infections an organizational priority. Senior leadership is accountable for providing adequate resources for effective implementation of a healthcare-associated infection prevention program. These resources include necessary personnel (clinical and nonclinical), education, and equipment. Engagement, education, execution, and evaluation are common attributes of successful care improvement programs. 321,338 These attributes are elaborated below.

Engage
Develop a multidisciplinary team 1. Multidisciplinary teams set goals, define each step in the implementation process, and monitor progress in reaching goals. 79

Execute
Standardize care processes 1. Standardize care processes through the implementation of guidelines, bundles, protocols or pathways. Standardization helps establish new care processes as "normal behaviors" for staff. 77,79,81,83,320,347,349,353,358 2. Interventions to improve adherence with best practices in the ICU that include protocols with or without education are associated with the greatest improvements in processes of critical care. 367 3. Daily multidisciplinary rounds are widely recommended. 368,369 Rounds should follow a structured format and include discussion about the patients' goals for the day, consideration of what resources and actions are necessary to achieve these goals, and identification of potential barriers and/or safety issues. 322,325,344,363,370,371 Create redundancy 1. Build redundancy or independent checks into care delivery processes to increase the likelihood that best practices are followed. 320 4. Analyze all or a representative sample of VAEs for etiology and preventability. Pneumonia, pulmonary edema, acute respiratory distress syndrome, and atelectasis are the precipitants for most VAEs. 17,19,22,58,60,61 Use these analyses to select and refine prevention strategies that address the most frequent and preventable causes of VAEs in the specific unit of interest. 83,351 Provide feedback to staff 1. Provide regular feedback on process and/or outcome data to staff. 77

Barriers and facilitators for adoption
Qualitative studies have identified several common barriers and facilitators to the implementation of VAE/VAP prevention programs.
1. Barriers related to staff workload and time (ie, competing priorities, data collection burden, not enough time), staff turnover, and lack of leadership support can impede implementation progress. 382,383 2. Barriers may also include a unit culture that does not prioritize preventive care and lack of a structured interdisciplinary approach to minimizing sedation and facilitating liberation from mechanical ventilation. 378

Facilitators for bundle adherence include 'reflective motivation'
(ie, VAE and VAP are perceived as serious and common problems among ICU patients; providers consider prevention measures useful to lower VAE or VAP rates and improve patients' outcomes). Reflective motivation can be achieved through increasing knowledge and understanding or through eliciting positive (or negative) feelings about adopting new practices. 383 4. Goddard et al, 384 using a Theoretical Domains Framework of behavior change, found that the social influences domain (local champions, ICU leadership, engagement between providers and family members) and behavioral regulation domain (feedback and having a unit protocol) may act as barriers or facilitators to early rehabilitation.

Positive examples and resources
Many successful improvement programs have been published and provide additional insights into practical strategies to engage multidisciplinary teams, educate key stakeholders, implement multifaceted interventions, and achieve significant reductions in VAE, VAP, and/or NV-HAP. 77,83,305,325,341,345,358,367 In addition, The Society of Critical Care Medicine's (SCCM) ICU Liberation initiative offers numerous tools and resources to assist hospitals with implementing bundles designed to improve patient outcomes and reduce the risk of long-term consequences of critical illness. 369,385 Conflicts of interest. The following disclosures are a reflection of what has been reported to SHEA. To provide thorough transparency, SHEA requires full disclosure of all relationships, regardless of relevancy to the guideline topic. Such relationships as potential conflicts of interest are evaluated in a review process that includes assessment by the SHEA Conflict of Interest Committee and may include the Board of Trustees and Editor of Infection Control and Hospital Epidemiology. The assessment of disclosed relationships for possible conflicts of interest has been based on the relative weight of the financial relationship (i.e., monetary amount) and the relevance of the relationship (ie, the degree to which an association might reasonably be interpreted by an independent observer as related to the topic or recommendation of consideration).
M.K. has received grant funding from CDC and AHRQ plus royalties from UpToDate, Inc. R.B. has served as an advisor/consultant for Pfizer, Mallinckrodt, Ventec Life Systems, Vyaire and Zoll Medical. K.A.C. has received honoraria from Becton Dickinson, CloroxPro, and Elsevier. L.R.G. has received honoraria from Premier Inc, Care Fusion, Infection Control Today, and Advanced Sterilization Products. All other authors report no conflicts of interest.