Nosocomial PneumoniaLessons Learned

Girish B. Nair, MDa, Michael S. Niederman, MDb,c,*

a Pulmonary and Critical Care Medicine, Winthrop-University Hospital, Mineola, NY 11501,ion Plaza North,ony Brook, Stony

Brook, NY 11794, USA

KEYWORDS

Nosocomial pneumonia Hospital-acquired pneumonia Ventilator-associated pneumonia Health careassociated pneumonia

KEY POINTS

Nosocomial pneumonia is the leading cause of death from hospital-acquired infection. Mechanical ventilation is the most important risk factor for the development of nosocomialpneumonia.

Health careassociated pneumonia (HCAP) affects a heterogeneous population, and notall patients require empiric broad-spectrum antibiotic therapy directed at multidrug-resistant pathogens.

There is no gold-standard test for the diagnosis of ventilator-associated pneumonia (VAP),but use of the Clinical Pulmonary Infection Score and procalcitonin can assist in antibioticde-escalation.

New Centers for Disease Control and Prevention streamlined surveillance definitions ofVAP may be useful in the diagnosis of sick patients with ventilator-associated complica-tions, but are not specific for the diagnosis of pneumonia.

Patients with risk factors for multidrug-resistant pathogens and diagnosed with hospital-acquiredpneumonia/VAP require combination therapywith broad-spectrumantimicrobials.

Patients should be reassessed 72 hours after initiation of treatment for VAP and HCAP,and antibiotics should be de-escalated, based on available culture and clinical data.

Nonresponders should be evaluated for treatment failure or complications from infection. Use of ventilator bundles in the intensive care unit is associated with a significant reduc-tion in the incidence of VAP.* Corresponding author. Department of Medicine, Winthrop-University Hospital, 222 StationPlaza North, Suite 400, Mineola, NY 11501.E-mail address: mniederman@winthrop.org

Crit Care Clin 29 (2013) 521546USA; b Department of Medicine, Winthrop-University Hospital, 222 StatSuite 400, Mineola, NY 11501, USA; c Department of Medicine, SUNY at St Ventilator-associated complications Prevention Antimicrobial treatmenthttp://dx.doi.org/10.1016/j.ccc.2013.03.007 criticalcare.theclinics.com0749-0704/13/$ see front matter 2013 Elsevier Inc. All rights reserved.

BACKGROUND

Nosocomial pneumonia (NP) is the second most common cause of nosocomial in-fections.1 NP is defined as lower respiratory infection that develops in a hospitalizedpatient after 48 hours of admission, and was not present or incubating at the time ofadmission (Table 1).2 NP includes hospital-acquired pneumonia (HAP), ventilator-associated pneumonia (VAP), and health careassociated pneumonia (HCAP).VAP is a subset of HAP, which develops at least 48 hours after endotracheal intu-bation. HCAP is pneumonia that develops in patients with exposure to the healthcare environment, such as dialysis patients and nursing home patients, and wasincluded in the 2005 American Thoracic Society/Infectious Diseases Society ofAmerica (ATS/IDSA) guidelines as part of NP, because these patients may harbormultidrug-resistant (MDR) microorganisms. However, it has been recently appreci-ated that such patients represent a very heterogeneous population, and thatsome can be treated with antimicrobials targeted at community-acquiredpathogens.3

Nair & Niederman522NP imposes significant economic costs and has the highest mortality among allnosocomial infections (ranging from 20% to 50%). There are significant additionalhospital costs of US$10,019 to $40,000 in patients who develop VAP.4,5 In 2010,The National Healthcare Safety Network (NHSN) reported the incidence of VAPto range from 0.0 to 5.8 per 1000 ventilator days in various hospital settings.6

The reported incidence is lower in comparison with previous years, possiblyreflecting the use of various preventive strategies, particularly ventilator bundles,a set of daily interventions for intubated patients. The lower incidence may also beinfluenced by the recent initiatives of the Center for Medicare and Medicaid ser-vices to set up a national benchmark of Zero VAP and to consider VAP as anon-reimbursable event. However, health care institutions in many instancesare underreporting episodes of VAP, even though antibiotic prescription and clin-ical diagnosis remains prevalent.7 Emerging streamlined surveillance definitionswere introduced by the Centers for Disease Control and Prevention (CDC), andmay help narrow the discrepancy that currently exists between clinical and surveil-lance definitions.8

This review discusses various issues pertaining to diagnosis, as well as newer un-derstandings, in the treatment and prevention of NP.

Table 1Definitions of various terms associated with nosocomial pneumonia

Hospital-acquired pneumonia Alveolar infection that occurs 48 h or more afteradmission and that was not incubating at the timeof admission

Ventilator-associated pneumonia Alveolar infection that occurs more than 48 h afterendotracheal intubation

Health careassociated pneumonia Alveolar infection in patients who were hospitalized inan acute care hospital for more than 2 d within 90 dof the pneumonia; those who resided in a long-termcare facility; those who had home infusion therapy,chemotherapy, or wound care within the past 30 d;those who attended a hospital or hemodialysisclinic; those with family member harboring

multidrug-resistant organism

LCOD13Resaltado

LCOD13Resaltado

LCOD13Resaltado

LCOD13Resaltado

LCOD13Resaltado

LCOD13Resaltado

endotracheal tube bypasses the lung defense mechanisms in the upper respiratory

Nosocomial Pneumonia 523tract, and can serve as direct conduit between the upper and lower respiratory tract.The endotracheal tube impairs mucociliary function, causes injury to the mucosa fa-voring bacterial binding, and allows pooling of secretions in the subglottic area abovethe endotracheal tube cuff, which can then be aspirated. The oropharynx of hospital-ized patients is colonized with gram-negative organisms in 35% to 75% of patientswithin 3 to 5 days of admission, depending on the severity and type of underlyingillness. Recent antibiotic therapy, endotracheal intubation, smoking, malnutrition, gen-eral surgery, dental plaque, and therapies that elevate the gastric pH increase airwaycolonization.10 However, not all colonization results in lower respiratory tract infectionin intubated patients. Ventilator-associated tracheobronchitis (VAT) is an increasinglyrecognized condition whereby patients have clinical signs (fever, leukocytosis, andpurulent sputum), microbiologic findings (Gram stain with bacteria and leukocytes,with either a positive semiquantitative or quantitative sputum culture), but absenceof a new infiltrate on chest radiograph.11 At present there is controversy regardingthe clinical importance of VAT; some investigators suggesting that early and aggres-sive therapy may prevent progression to VAP, whereas others have suggested that it isan illness independent of VAP and does not always need to be treated.12,13 However,most studies have shown that VAT prolongs the duration of mechanical ventilation andstay in the intensive care unit (ICU), and may increase mortality, but that these adverseconsequences can be mitigated by therapy.14

The interaction between host defenses and the ability of microorganisms to colonizeand invade the respiratory tract determines whether the patient will develop pneu-monia. Pneumonia usually develops in patients with underlying comorbid illness orin those who cannot mount an adequate immune response (such as patients on immu-nosuppressive medications or those with critical illness), or is the result of an over-whelming inoculum of virulent pathogens (as in massive aspiration), so that hostimmunity cannot adequately contain the infection. In some instances the patient hasno preexisting immunity and the organism so virulent that infection can developeven in an intact immune system (as with influenza).2,15,16 Different genotypes thatdetermine the host immune response have been identified in vivo. These genotypesmake some patients prone to infection, whereas in others acute lung injury developsbecause of an excessive immune response.17

ETIOLOGIC AGENTS

The timely identification and institution of antibiotics is crucial in the treatment of NP,and requires a good understanding of the etiologic agents prevalent in a specific ICUas well as the local antibiotic resistance patterns. The ATS/IDSA 2005 guidelinesfocused on the assessment of risk factors for specific pathogens and the empiriccoverage for drug-resistant pathogens, dividing NP into two groups based on thetime of onset since admission, defining early-onset NP (within 4 days after admis-PATHOGENESIS

Microorganisms can reach the lower respiratory tract by any of several mechanisms:aspiration (primarily from a previously colonized oropharynx), inhalation (mostlyLegionella, tuberculosis, and viruses), hematogenously (right-sided endocarditis), orby direct extension from adjacent infected foci. Upper respiratory and digestive tractcolonization with pathogenic microorganisms with resultant microaspiration is themost common mechanism associated with the development of pneumonia.9 Thesion) and late-onset NP (at day 5 or later after admission).

A wide variety of microorganisms originating from the health care environment orfrom the patients flora itself can cause NP. Data from the SENTRY surveillance pro-gram, a multinational study, indicated that the top 6 etiologic agents causing NP(Staphylococcus aureus [28%], Pseudomonas aeruginosa [21.8%], Klebsiella spp[9.8%], Escherichia coli [6.9%], Acinetobacter spp [6.8%], and Enterobacter spp[6.3%]) caused 80% of all infections (Fig. 1).18 Early-onset infections are usuallycaused by a group of core organisms: Streptococcus pneumoniae, Haemophilusinfluenzae, methicillin-sensitive S aureus, and nonresistant gram-negatives (E coli,Klebsiella spp, Enterobacter spp, Proteus spp, and Serratia marcescens) (Table 2).The risk of developing infection with MDR pathogens such as methicillin-resistant Saureus (MRSA), P aeruginosa, and Acinetobacter spp depends on the presence ofrisk factors, namely prolonged hospitalization (5 days), recent hospitalization (within

antibiotic resistance patterns, because the wrong empiric coverage can have a worse

Nair & Niederman524outcome than correct therapy, especially in infections involving P aeruginosa, Acine-tobacter spp, and MRSA.21

The ATS/IDSA guidelines included HCAP patients in the NP guidelines because oftheir risk of acquiring MDR pathogens. However, recent data point to the contrary,and show that not all nursing home patients develop pneumonia with MDR organisms,especially those with good functional status (as defined by activities of daily living) andthose who had not received antibiotics in the previous 6 months.22 Data from the CAP-NETZ (German Competence Network for Community-Acquired Pneumonia) trial fromGermany showed that HCAP patients older than 65 years have etiologic agents similarto those causing community-acquired pneumonia, except for an increased incidenceof S aureus pneumonia.23

RISK FACTORS FOR DEVELOPMENT OF NP AND FOR MORTALITYVAP and HAP

The ATS/IDSA 2005 guidelines identified modifiable and nonmodifiable risk factors forthe development of NP. Host factors such as extremes of age, factors that promoteoropharyngeal colonization and increased aspiration, and factors that impair adequate90 days), recent antibiotic therapy, residence in a nursing home, or need for chroniccare outside the hospital (Box 1).2,19 Fungal and rare pathogens such as Aspergillusspp, Pneumocystis jiroveci,Candida spp,Nocardia spp, and viruses such as cytomeg-alovirus should be suspected in immunosuppressed and transplant patients.10 A sig-nificant proportion, up to 40% of infections, is polymicrobial in nature, with more than1 infective pathogen involved.20 Careful considerations should be given to the localFig. 1. Relative incidence of various pathogens causing nosocomial pneumonia.

Nosocomial Pneumonia 525Table 2Bacterial pathogens causing nosocomial pneumonia and preferred antimicrobial choices

Etiologic Agents Condition Antimicrobial Choices

Core PathogensStreptococcus pneumoniaeHaemophilus influenzaeEnterobacter spp

Early-onset nosocomialpneumonia not at riskfor MDR

Monotherapy withA second, or nonpseudomonalthird-generation cephalosporin

orpulmonary toilet may result in pneumonia (Box 2). Mechanical ventilation is the mostimportant risk factor for the development of VAP, and the risk is greatest in the first5 days after intubation.24 Other risk factors for the development of VAP are age olderthan 60, a high mortality risk Acute Physiology and Chronic Health Evaluation(APACHE) II or Simplified Acute Physiology (SAPS) II score, malnutrition, acute lunginjury (ie, acute respiratory distress syndrome [ARDS]), burns, recent abdominal orthoracic surgery, multiple organ failure, a low Glasgow Coma Scale (GCS) score, priorantibiotic therapy, elevation of gastric pH (by antacids or histamine type 2 blockingagents), large volume aspiration, use of a nasogastric tube, use of inadequate

Escherichia coliKlebsiella sppProteus sppSerratia marcescensMethicillin-sensitiveStaphylococcus aureus

b-Lactam/b-lactamase inhibitorcombination

orErtapenemorQuinolone (levofloxacin ormoxifloxacin)

orClindamycin and aztreonam(if penicillin allergic)

MDR PathogensPseudomonas aeruginosaAcinetobacter sppStenotrophomonasmaltophilia

Klebsiella pneumoniae(ESBL1)

Methicillin-resistantS aureus (MRSA)

Late-onset HAP/VAP orhave risk factors fordevelopment of MDRorganisms

Combination TherapyAntipseudomonal cephalosporin(cefepime, ceftazidime)

orAntipseudomonal carbapenem(imipenem or meropenem)

orb-Lactam/b-lactamase inhibitor(piperacillintazobactam)

orAztreonam if penicillin allergicPLUSAntipseudomonal fluoroquinolone(ciprofloxacin or levofloxacin)

orAminoglycoside (amikacin/gentamicin/tobramycin)

MRSA VancomycinLinezolidTelavancin

Acinetobacter spp CarbapenemPolymyxin B and E (colistin)Tigecycline in combination withcarbapenem, sulbactam

Aminoglycoside

Abbreviations: ESBL, extended-spectrum b-lactamase; HAP, hospital-acquired pneumonia; MDR,multidrug-resistant; VAP, ventilator-associated pneumonia.

Box 1

Risk factors for multidrug-resistant organisms

Current hospitalization of at least 5 days

Antibiotic treatment in the prior 90 days

High frequency of antibiotic resistance in the community/specific hospital unit

Immunosuppressive disease/therapy

Presence of multiple risk factors for HCAP

Hospitalization for 2 days or more in the preceding 90 days

Residence in a nursing home or extended-care facility

Home infusion therapy (including antibiotics)

Chronic dialysis within 30 days

Home wound care

Family member with MDR pathogen

Nair & Niederman526endotracheal tube cuff pressure, prolonged sedation and paralysis, maintaining pa-tients in the supine position in bed, use of total parenteral nutrition rather than enteralfeeding, and repeated reintubation.2,5 In a recent study investigating the incidence ofpneumonia in patients who had therapeutic hypothermia following out-of-hospital car-diac arrest, a higher incidence of early-onset pneumonia was noted.25 In multivariateanalysis, therapeutic hypothermia was the single most important risk factor for thedevelopment of early-onset pneumonia, although there was no increased incidenceof subsequent VAP. Endotracheal tube cuffs play an important role in the pathogen-esis of pneumonia, and polyurethane cuffs (PUC) appear to reduce the risk of pneu-monia to a greater extent than polyvinyl chloride cuffs, particularly if the PUC doesnot promote the formation of longitudinal folds that allow the aspiration of secretions

that pool above the cuff. One recent study also demonstrated that with the use of a

Box 2

Risk factors for the development of NP

Host Factors

Age older than 60, malnutrition, immunosuppression, a high mortality risk APACHE II or SAPS IIscore, ARDS, severe acute or chronic illnesses, and burns

Factors Increasing Oropharyngeal and Gastric Colonization

Endotracheal intubation, prior antibiotic therapy, elevation of gastric pH (by antacids orhistamine type 2 blocking agents), structural lung disease, cross-contamination by respiratoryequipment handling

Conditions Favoring Aspiration or Reflux

Low GCS score, supine position, endotracheal intubation, insertion of nasogastric tube, use ofinadequate endotracheal tube cuff pressure, repeated reintubation

Impaired Pulmonary Toilet

Surgical procedures that involve the head and neck, being immobilized as a result of trauma orillness, use of paralytics

Nosocomial Pneumonia 527pneumatic device to maintain endotracheal tube cuff pressure, it is possible to mini-mize aspiration of secretions that pool above the endotracheal tube cuff.26

Mortality in Patients with VAP or HAP

When patients develop VAP and die, it remains uncertain as to whether it was theresult of pneumonia or if they were so ill that they would have died anyway, and studiesof this question have focused on the attributable mortality of VAP. Several olderstudies reported that many of the deaths were directly attributable to pneumonia(15%50%), and that VAP increased the odds ratio of death by 2-fold.4 However,one recent large study has suggested that VAP contributes little to excess mortalityin critically ill patients.27 In this study of 4479 ventilated patients, 685 had VAP, andthe mortality of 24.1% was not substantially different from the 23.1% mortality inthe non-VAP patients. However, these studies were done in the presence of therapy,and it is difficult to assess the impact of pneumonia on mortality because mortalitywould certainly occur if no therapy, or the wrong therapy, were given. In fact, inappro-priate therapy is a well-known risk for attributable mortality.28 In addition, up untildeath, patients with VAP have longer duration of mechanical ventilation, and longerlength of stay in the hospital and ICU.5

The crude mortality from VAP ranges from 20% to 50%.2 Factors associated withmortality include severity of illness, limited activity before the onset of mechanicalventilation, development of shock, acute renal failure, and worsening of hypoxemiaduring the period of mechanical ventilation.29 As mentioned, inappropriate therapyis a mortality risk, along with infection involving an MDR pathogen, delay in the initi-ation of therapy, failure to de-escalate therapy, prolonged mechanical ventilation,coma on admission, chronic liver disease, nonsurgical primary diagnosis, bilateral in-filtrates, septic shock, multiple system organ failure, and use of prior antibiotictherapy.

Health CareAssociated Pneumonia

Health careassociated pneumonia, as mentioned earlier, represents a diverse popu-lation of patients with recent exposure to the health care environment who are at riskfor infection with MDR pathogens (see Box 1). HCAP includes patients with recenthospitalization (for 2 or more days within 90 days of the infection); residents of anursing home or long-term care facility; hemodialysis patients, and individuals whohave received recent intravenous antibiotic therapy, chemotherapy, or wound carewithin the past 30 days. HCAP patients with severe illness or poor functional statusas defined by activities of daily living score, or who are taking immunosuppressivemedications, are at increased risk of developing MDR infections, and need broad-spectrum empiric antimicrobial therapy. However, not all HCAP patients harbor riskfactors, and patients who do not are usually infected with pathogens similar to thosecausing community-acquired pneumonia.

DIAGNOSTIC CONSIDERATIONS

NP is clinically diagnosed in patients who have been hospitalized for at least 48 hoursand subsequently develop a new or progressive infiltrate on chest imaging, accompa-nied by at least 2 of the following 3 systemic signs: fever, leukocytosis, and purulentsputum. This clinical diagnosis by itself is fraught with low specificity, requiringmore objective measures. The ATS/IDSA guidelines recommended a chest radio-graph, an assessment of oxygenation, blood cultures, and culture sampling of lower

respiratory tract secretions as part of the initial evaluation.2 Lower respiratory tract

Nair & Niederman528cultures can be obtained bronchoscopically or by nonbronchoscopic techniques, andshould be collected before starting therapy (using a technique and culture method thatthe clinician is expert in performing and interpreting).In critically ill patients, numerous other conditions such as ARDS, congestive heart

failure, atelectasis, or pulmonary infarctionmaymimic pneumonia, and the radiographicsigns can be nonspecific. However, there is no gold-standard test to accurately diag-nose VAP. Moreover, it is difficult to distinguish between microbial colonization andactive infection, especially in ventilated patients. The situation is further complicatedby the recognition of VAT, an illness characterized by fever, purulent sputum, andhigh counts of bacteria in tracheal aspirate culture, but no parenchymal lung infiltrate.30

Recognition of VATmay be important because it is an illness thatmay need therapy, andif not recognized and treated may lead to VAP. Nevertheless, as already mentioned,early diagnosis and initiation of appropriate antibiotic therapy for VAP is important,because delay in appropriate treatment is associated with adverse outcomes. On theother hand, antimicrobial resistance to commonly used antibiotics is rapidly spreading,and reliable diagnosis with judicious use of specific antibiotics holds the key not only forsuccessful treatment but also for prevention of antimicrobial resistance.

CDC Definition

The CDC definition of VAP requires a constellation of radiographic and clinicalsigns, with definitive diagnosis requiring microbiological confirmation. Radiograph-ically a new, persistent, or progressive infiltrate or cavitation should be seen on 2 ormore serial chest radiographs. Clinically the patient should show either fever(temperature >38C [100.4F] with no other recognized cause), leukopenia (

Nosocomial Pneumonia 529oxygenation, degree of radiographic abnormality, and presence of pathogens in thesputum), with each scored on a scale from 0 to 2, pneumonia being diagnosed by atotal score of at least 6 (out of a maximum of 12).33 This score was modified furtherby Singh and colleagues34 to include radiographic progression in order to improvediagnostic accuracy. A CPIS greater than 6 at baseline and at 72 hours is consideredsuggestive of pneumonia and the need for a full course of antibiotic therapy. Diagnosisof VAP based on the CPIS score has poor sensitivity in patients with ARDS and thosein surgical ICUs. Further investigations showed that the CPIS has low sensitivity andspecificity for diagnosing VAP, and is most useful in treatment decisions for patientswith a low likelihood of VAP, for whom it may guide the duration of therapy andresponse to treatment.35,36 Data from 2 independent, large studies demonstratethat of all the measures included in the CPIS score, oxygenation, as measured bythe of PaO2/FiO2 ratio, consistently proved to be a reliable marker, signifying responseto therapy and outcome in VAP patients, with a worse outcome in patients who did nothave an improvement in oxygenation by day 3 of therapy.35,37

New Streamlined Surveillance Definitions

The inconsistency between the reported incidence of VAP based on the surveillancediagnosis and the routine prescription of antibiotics for pneumonia has recentlybeen rigorously examined.8 With a goal of improving the efficacy and objectivity ofVAP diagnosis, the CDC has proposed a new focus on ventilator-associated compli-cations, centered on worsening oxygenation and systemic signs of infection. This pro-posal includes an algorithm that excludes the use of chest radiography and alsostratifies ventilator-associated complications based on specific parameters.38 Insteadof having only 2 arms, VAP versus no VAP, the newer guidelines identify patients atvarious stages along the spectrum. The new definition categorizes patients into 3groups: ventilator-associated complications (VAC), infection-related ventilator compli-cations (IVAC), and probable versus possible VAP. Each group contains specific pa-rameters that reveal illness progression from one group to the next (Box 3). In arecent study enrolling more than 600 mechanically ventilated patients comparingVAC with the CDC definition for VAP, patients who were identified as VAC had highermortality than VAC-negative patients (odds ratio [OR] 2.0, 95% confidence interval [CI]1.33.2), and had other complications at a rate similar to that in patients with VAP, andVAC was easier to assess (mean 1.8 vs 39 minutes per patient).38 However, VAC is notalways the result of infection, and the frequency of VAC was 23%, compared with9.3% of these patients meeting the CDC definition of VAP. Patients who were identi-fied as VAC were probably much sicker as evidenced by worsening oxygenation, andthus had higher mortality than those without VAC, whereas those with stable VAP (withno desaturation) identified by the CDC definition might have been missed by the VACdefinition. Although the VAC definition can identify sicker patients, the demand for aproper diagnostic tool to distinguish patients with VAP is still lacking, and furtherstudies are needed in this regard. Most importantly, the incidence of VAC may notbe influenced by efforts at VAP prevention, hence the use of VAC as an indicator ofquality of care currently remains unproven. In another study of VAC, 150 of 543 venti-lated patients met this definition, pointing out the high frequency of VAC in the ICU.39

Recently this concept has been extended in defining an objective, streamlined defini-tion of VAP, which takes 3.5 minutes to diagnose, compared with 39 minutes for thetraditional CDC definition.40 This definition does include radiographic findings, butfewer patients satisfied these criteria in comparison with those who were defined ashaving VAP using the traditional CDC criteria (30 vs 57 of 599 patients). In a subse-

quent study, Klompas and colleagues41 have argued that omission of radiographic

Box 3

New CDC surveillance definitions

Ventilator-associated complication

If the patient develops any one of the following criteria of worsening oxygenation:

1. Minimum daily FiO2 values increase 0.20 (20 points) over baseline and remain at or abovethat increased level for 2 calendar days

2. Minimum daily positive end-expiratory values increase 3 cm H2O over baseline and remainat or above that increased level for 2 calendar days

Infection-related ventilator-associated complication

If the patient has VAC and also fits both of the two following criteria:

1. Temperature greater than 38C or white blood cell count greater than 12,000 or less than4000/mm3

2. A new antimicrobial agent is started and is continued for 4 or more calendar days

Possible VAP

If one of the following criteria is met:

On or after calendar day 3 of mechanical ventilation and within 2 calendar days before or afterthe onset of worsening oxygenation, ONE of the following criteria is met:

1. Purulent respiratory secretions

Defined as secretions from the lungs, bronchi, or trachea that contain >25 neutrophils and

Nosocomial Pneumonia 531findings and other clinical signs from the streamlined VAP definition does not diminishthe predictive value of surveillance methods in identifying patients at risk for adverseoutcomes during mechanical ventilation.

Role of Biomarkers

Recently biomarkers have been proposed to aid in the identification of VAP.42 Ideallythe level of the biomarker should correlate with the disease process. Concentration ofseveral markers, such as soluble triggering receptor expressed onmyeloid cells type 1(sTREM-1), procalcitonin (PCT), Pancreatic stone protein (PSP), surfactant protein-D,Interluekin-1b, and C-reactive protein (CRP), in serum, BAL, and exhaled breathcondensate have been examined.sTREM-1 is a glycoprotein belonging to the immunoglobulin G subclass. Increased

secretion by monocytes is seen when triggered by lipopolysaccharide, and it helps todownregulate the deleterious effect of proinflammatory cytokines. In a prospectivestudy of 148 mechanically ventilated patients, a BAL sTREM-1 level higher than5 pg/mL had sensitivity of 98% and specificity 90% in diagnosing VAP (area underthe curve [AUC] 0.93; 95% CI 0.920.95; P

Nair & Niederman532Antibiotic choicesSeveral factors such as severity of infection, patient-specific risk factors, and totalnumber of days in the hospital before pneumonia onset influence the etiologic patho-gens and, thus, the choice of antimicrobial agents. Antibiograms with local patternsof resistance should also be considered when choosing therapy. For example, in someinstitutions ciprofloxacin may be an effective choice for P aeruginosa, whereas inothers it is not an effective agent, and an aminoglycoside may be required. Irrespec-tive of the risk factors, the initial treatment regimen for early-onset HAP and VAPshould include coverage of the core group of organisms that includes H influenzae,S pneumoniae, antibiotic-sensitive gram-negative bacilli (Enterobacter spp, E coli,Klebsiella spp, Proteus spp, and Serratia marcescens), as well as methicillin-sensitive S aureus. The recommended regimen for less severe infection in patientsnot at risk for the aforementioned MDR pathogens would include a monotherapyregimen of: a second-generation or nonpseudomonal third-generation cephalosporin,a b-lactam/b-lactamase inhibitor combination, ertapenem, or a quinolone (levofloxa-cin or moxifloxacin).2 If the patient is penicillin allergic, therapy may comprise a quino-lone or the combination of clindamycin and aztreonam (see Table 2).Patients with either late-onset infection or the presence of any of the other MDR risk

factors should have empiric coverage for MDR gram-negative and gram-positive path-ogens, in addition to the core pathogens already mentioned. Thus a combination of anantipseudomonal b-lactam (imipenem, meropenem, piperacillin/tazobactam, aztreo-nam, or cefepime) plus an aminoglycoside or quinolone (ciprofloxacin or high-dose lev-ofloxacin) is commonly used (see Table 2). In patients at risk for MRSA, therapy shouldalso include a third drug, either linezolid or vancomycin. In a large multicenter trial ofnearly 400 patients with documented MRSA pneumonia, comparing linezolid with opti-mally dosed vancomycin, linezolid led to a significantly higher rate of clinical responsethan vancomycin, but with no difference in mortality and less nephrotoxicity.49 Thelack of difference in mortality could be related to the fact that patients who failedvancomycin could be salvaged with linezolid, and any survival in this setting was attrib-uted to vancomycin. In those infected with Acinetobacter spp or extended-spectrumb-lactamaseproducing Enterobacteriaceae, treatment should include a carbapenem.However, there is a concern about growing antimicrobial resistance against carbape-nems in patients with Acinetobacter, and if this is seen a combination of agents,including tigecycline with sulbactam, colistin, or an aminoglycoside, can be used. Tige-cycline should not be used asmonotherapy, because in a large double-blindmulticentertrial comparing tigecycline with imipenem, the tigecycline-treated VAP patients had asignificantly lower cure rate. Therefore if this agent is used, it should be as part of a com-bination regimen alongwith a carbapenem, sulbactam, colistin, or an aminoglycoside.50

In some instances, adjunctive therapy with an inhaled antibiotic, such as colistin or anaminoglycoside, may be valuable. Occasionally patients are infected with Legionellaspp and may require treatment with a second-generation quinolone.

Combination therapyWhen therapy is directed against MDR pathogens, it is important to use at least 2 an-tibiotics with different mechanisms of action, to prevent emergence of panresistanceand antimicrobial competition at the bacterial binding site of action. The role of com-bination therapy has been established for bacteremic pseudomonal infection, espe-cially in critically ill and neutropenic patients. However, the role of combinationtherapy in nonbacteremic NP is more controversial.In a multicenter study of VAP by Heyland and colleagues,51 for those patients whohad infection with Pseudomonas spp, Acinetobacter spp, and MDR gram-negative

Nosocomial Pneumonia 533bacilli, the use of combination therapy with meropenem and a quinolone resulted inbetter microbiological eradication of the infecting organisms compared with merope-nem alone (64.1% vs 29.4%, P 5 .05) and more frequent initial appropriate therapy inthe combination group than in the monotherapy group (84% vs 11%). Combinationtherapy was not associated with a reduction in mortality for all patents with VAP, prob-ably because the incidence of infection with MDR pathogens was low. Hence, for pa-tients with NP at risk for infection with MDR pathogens, combination therapy shouldbe used early in the clinical course as soon as the diagnosis is suspected becauseof the broader coverage in comparison with a single agent, resulting in a higher fre-quency of appropriate therapy than with the use of a single agent.51 In choosing acombination regimen it is also necessary to use a different antibiotic (preferablybelonging to a different class) from which the patient has been exposed previously(within the past 14 days), as repeated use results in resistance to that class, especiallyif the pathogen is P aeruginosa.52

Selecting the optimal second agent One unresolved question is, which agent shouldbe added to a b-lactam when using dual therapy directed against gram-negatives?Combination therapy with an aminoglycoside is usually the judicious first step,reserving quinolones for subsequent ICU infection.53 Several studies suggest thataminoglycosides may have more intrinsic activity than antipseudomonal quinolones,and for this reason are preferable as the second agent.54 In fact, in recent years theactivity of quinolones against P aeruginosa has declined, and they may not offer asmuch additional coverage when used with a broad-spectrum b-lactam as can beachieved by using an aminoglycoside. Other studies have also confirmed that in thetreatment of VAP the empiric use of aminoglycoside therapy (usually with a b-lactam)leads to better outcome than when a quinolone is used.55 One other reason to preferan aminoglycoside to a quinolone is that quinolone use in the ICU may select for multi-drug resistance, making it difficult to treat a second infection if the first episode of ICUinfection was treated with a quinolone.56 Therefore, if quinolones are used in the ICUfor nosocomial infections, they may best be reserved for a later infection rather than fora first infection.However, with the use of aminoglycosides attention should be given to the narrow

therapeutic to toxic ratio, and potential for nephrotoxicity, particularly in elderly pa-tients. Moreover, aminoglycosides have poor penetration into bronchial secretionsespecially at low pH, a situation common with severe bacterial infection. Despite theserestrictions, aminoglycosides plus a b-lactam provide better antimicrobial coveragethan a quinolone added to a b-lacatam.54 In a study by Beardsley and colleagues,54

when ciprofloxacin was added to a b-lactam the combination covered only 85% ofgram-negatives, compared with 95% when amikacin was added to a b-lactam.Administration of the total 24-hour dose into a single dose improved efficacy whileminimizing toxicity. The concerns about aminoglycoside toxicity can also be mitigatedby minimizing the duration of therapy with these agents. The agents are added primar-ily to provide coverage while waiting for antimicrobial sensitivity data, and once it isknown which b-lactam is active against the etiologic pathogens, therapy can bestreamlined to b-lactam monotherapy, usually after 3 to 5 days of combinationtherapy.

Effectiveness of combination therapy One other potential benefit of empiric combina-tion therapy is that it may lead to more rapid killing of bacteria than occurs with mono-therapy, and may correct for the relative errors of monotherapy. In a retrospective

study of 1223 matched pairs of patients with septic shock receiving either appropriate

Nair & Niederman534monotherapy or appropriate combination therapy, the use of combination therapywas associated with reduced mortality, particularly for patients with respiratory tractinfections.57 The benefit of combination therapy applied to gram-positives and gram-negatives, bacteremic and nonbacteremic infection, and to multiple types of antimi-crobial agents. However, the benefit of the second agent declined as the timebetween the first and second agent increased, and was less clear when monotherapywas with a rapidly bactericidal agent (such as a carbapenem).Long durations of combination treatment can potentially cause drug toxicity and

promote antibiotic resistance, and may not improve mortality. In a recent multicenter,prospective study, patients who were treated concordant with the ATS/IDSA guide-lines (usually with a combination regimen) had a significantly higher mortality in com-parison with discordant therapy (34% vs 20%).58 However, a closer analysis of thestudy showed that patients who were on the concordant arm were sicker and moreoften had episodes of severe sepsis (91% vs 76%), more prior antibiotics (75% vs56%), significantly higher APACHE II scores (21 vs 20, P 5 0$048), and less de-escalation of antibiotics compared with the noncompliant group. The ATS/IDSA2005 guidelines emphasized not only appropriate broad-spectrum coverage butalso de-escalation when culture data are available, and the study did not consider afailure to de-escalate as noncompliance with the guidelines. Other investigatorshave attempted to validate the same NP guidelines, and Ferrer and colleagues59 foundthat these guidelines were useful in a study of 276 patients with ICU-acquired pneu-monia. Initial therapy was appropriate 83% of the time if the guidelines were followed,and only 64% of the time if they were not (P 5 .01) but, interestingly, 238 of the 276patients fell into the category at risk for MDR pathogens. In addition, of the 38 notfound to be at risk for MDR pathogens, 16% had P aeruginosa and 16% hadMRSA, demonstrating the low frequency of patients who are not at risk for theseorganisms.

Appropriate dosage of antibioticsOne other consideration in therapy for VAP is to not only use the correct antibiotic, butto use it at the right dose. Box 4 summarizes the recommended doses of antibioticsfor critically ill ventilated patients with pneumonia, based on doses used in prospec-tive, randomized controlled studies. Use of too low a dose can lead to clinical failureand the selection of antimicrobial resistance. It may be important to consider dosingregimens and methods of administration when targeting MDR pathogens. In criticallyill patients with sepsis, renal clearance of antibiotics may be enhanced, and doses ofantibiotics may need to be increased.60 In addition, when aminoglycosides andquinolones are used, they kill bacteria in a concentration-dependent fashion, andonce-daily dosing can optimize the killing of organisms. On the other hand, with theb-lactams, killing depends on how long the antibiotic concentration exceeds the min-imum inhibitory concentration (MIC) of the target pathogen, and killing can be opti-mized by prolonged or continuous infusion methods. Few data exist to show thatchanging methods of administration can improve clinical outcome, but Nicasio andcolleagues61 have reported that use of an optimized dosing clinical pathway (incorpo-rating prolonged infusions of cefepime and meropenem) led to a reduction ofinfection-related mortality in 94 patients with VAP.

De-escalationOnce the culture or Gram-stain data are available from either quantitative or semi-quantitative lower respiratory tract cultures, and when there has been a chance to

observe the clinical response to empiric therapy, it is possible to adjust therapy using

Nosocomial Pneumonia 535Box 4

Dosages of commonly used antimicrobials in patients with HAP/VAP

Early-Onset Infection/Not at Risk for MDR Pathogen

Monotherapy with any of the following agents:

Ampicillin/sulbactam: 1.53 g intravenously every 6 hours Ceftriaxone: 12 g intravenously every 24 hours Ertapenem: 1 g intravenously every 24 hours Levofloxacin: 750 mg intravenously every 24 hours Moxifloxacin: 400 mg intravenously every 24 hoursTreatment of Late-Onset HAP

Combination therapy with:a de-escalation strategy.62 This approach involves reducing the spectrum of the ther-apy and the number of drugs used, to focus on specific pathogen(s) identified in cul-ture. De-escalation is used at a variable rate, with studies reporting the use of thispractice in anywhere from 22% to 74% of episodes of VAP.63 De-escalation is morelikely to occur if there is a protocol for its implementation (vs standard care), if the initialtherapy is appropriate and involves a broad-spectrum multidrug regimen (rather thanmonotherapy). The rate of de-escalation is higher if cultures are positive rather thannegative, even though the finding of a negative culture, collected before starting ther-apy, should be a reason to narrow and focus therapy. It is unclear whether the methodof diagnosis has any impact on the rate, but de-escalation is less likely to occur if thereis a high frequency of MDR pathogens causing pneumonia.63 In addition to a narrow-ing of the drug spectrum and a reduction in the number of drugs used, de-escalationcan involve reducing the duration of therapy. The foundation of this approach is to usean empiric, broad-spectrum regimen that assures a high rate of initially appropriatetherapy, in contrast to using an initial narrow-spectrum therapy that could be

Any one of the following

Cefepime: 12 g intravenously every 812 hours Ceftazidime sodium: 2 g intravenously every 8 hours Imipenem/cilastatin: 500 mg intravenously every 6 hours; or 1000 mg intravenously every8 hours

Meropenem: 1 g intravenously every 8 hours Piperacillin/tazobactam: 4.5 g intravenously every 6 hoursPlus any one of the following

Ciprofloxacin: 400 mg intravenously every 812 hours Levofloxacin: 750 mg intravenously every 24 hours Amikacin: 20 mg/kg intravenously every 24 hours Gentamicin: 7 mg/kg intravenously every 24 hours Tobramycin: 7 mg/kg intravenously every 24 hoursIf MRSA is suspected, add one of the following

Linezolid: 600 mg intravenously every 12 hours Vancomycin HCl: 15 mg/kg intravenously every 12 hours

Nair & Niederman536inappropriate, necessitating subsequent use of more drugs, with an increased risk ofdeath because of a delay in starting the correct therapy.The use of this strategy could help to prevent the emergence of multiresistant or-

ganisms, avoid superinfection, and prevent antibiotic side effects, while reducingthe cost of treatment. Patients who have had antibiotic de-escalation have higher sur-vival rates than patients who have had no de-escalation, but this may be related to theefficacy of initial therapy rather than the practice of de-escalation.62,64,65 However, allthe available data suggest that if the patient is improving with initial therapy, the use ofde-escalation is not harmful and will not contribute to subsequent treatment failure.Patients who have received an empiric broad-spectrum regimen, but are not foundto have MDR pathogens on culture, can be switched to monotherapy (with any ofthe following: doripenem, imipenem, meropenem, piperacillin/tazobactam, cefepime,ciprofloxacin, high-dose levofloxacin) to finish the antibiotic course. Patients withculture-proven P aeruginosa require combination therapy with a b-lactam and an ami-noglycoside for no more than 5 days, after which the patient can be switched tomono-therapy with an agent to which the organism is sensitive.2

Duration of treatmentThe duration of antibiotic treatment for patients with HAP should be based on the clin-ical response (reflected by a drop in the CPIS, improved oxygenation, or decreasedlevels of biomarkers), and all patients should be treated for at least 72 hours. If thelower respiratory tract cultures are negative, it may be possible to stop therapy (espe-cially if an alternative diagnosis such as atelectasis or heart failure is made), or toshorten the duration of therapy. In instances where cultures are positive but showthat the initial empiric regimen was appropriate, and patient has a good clinicalresponse, it may be possible to reduce the duration of therapy to as little as 7 to8 days.66 In the PRORATA trial, investigators used a PCT-based algorithm to deter-mine the duration of therapy, and those in the PCT group had a shorter duration oftherapy compared with those managed by clinical evaluation alone, with no differencein mortality between the 2 groups.46 However, those receiving initially inappropriatetherapy, patients with complications such as empyema or lung abscess, and thoseinfected with P aeruginosa or Acinetobacter spp may require a longer course oftreatment.

Localized treatmentsIn nonresponders despite adequate systemic antibiotics or in patients infected withhighly resistant organisms, localized treatment by the use of aerosolized antibiotics(gentamicin, tobramycin, colistin, and ceftazidime) may be valuable. This type of ther-apy helps to achieve high local concentrations of medication, which can overcome theproblems of poor lung penetration of antibiotics such as the aminoglycosides, andeven treat organisms that would otherwise be resistant at the levels of antibioticachieved with systemic therapy.67 In a recent double-blind phase II trial, use of aninvestigational drug-device combination of amikacin, formulated for inhalation in pa-tients with a high frequency of MDR gram-negatives causing VAP, led to shorter dura-tion of systemic antibiotics, earlier clinical response, and less need to escalateantibiotics.68 Thus the use of adjunctive aerosol therapy may augment the efficacyof systemic therapy, and could form a strategy to avoid prolonged duration of sys-temic therapy in the setting of MDR pathogen VAP. In one study, patients with suscep-tible or intermediately sensitive strains of P aeruginosa causing VAP were randomizedto receive aerosol therapy alone (with ceftazidime and amikacin) and without systemic

therapy, or intravenous ceftazidime and amikacin.69 After 8 days of antibiotic

Nosocomial Pneumonia 537administration, both groups had comparable efficacy (successful treatment in 70% vs55%), treatment failure (15% vs 30%), and superinfection with other microorganisms(15% vs 15%). In both of these recent studies the aerosol was delivered with avibrating mesh plate, which generates small particle sizes and is associated withretention of greater than 60% of the administered dose in the lung. These resultsare promising, and suggest that aerosolized antibiotics could provide us with an addi-tional alternative in the treatment of infections with MDR pathogens. Despite the datashowing efficacy of aerosolized therapy alone, most clinicians would use thisapproach along with intravenous agents.

Ventilator-Associated Tracheobronchitis

As mentioned earlier, VAT represents a continuum of microbial colonization of the tra-chea and bronchi, which can progress to VAP. The microbiology is essentially thesame as for VAP and, hence, can include MDR pathogens. VAT is associated withlonger duration of mechanical ventilation, and systemic antimicrobial treatment signif-icantly improves the outcome in patients diagnosed with VAT.70 In a prospective studyin a surgical and medical ICU, the frequency of VAT was less than the that of VAP, butpatients diagnosed with VAT frequently progressed to VAP, which was frequentlycaused by MDR organisms.12 In an unblinded, multicenter study, patients diagnosedwith VAT were randomized to receive systemic antimicrobials or not. The study wasterminated early because ICU mortality was significantly lower in patients on antimi-crobials (18% vs 47%, P 5 .047).14 In another study, the use of aerosolized therapywith both vancomycin and gentamicin appeared to be an effective therapy for VAT,and may have prevented some patients from developing VAP.71 Patients diagnosedwith VAT should be assessed for the same MDR risk factors as patients with VAP,and treatment should be initiated early, with the same antimicrobial agents as usedin VAP. Treatment can be monitored clinically and microbiologically with serial endo-tracheal aspirate specimens.11

Treatment Considerations in Health CareAssociated Pneumonia

According to the 2005 ATS/IDSA guidelines, all HCAP patients should receive empiriccoverage for the same drug-resistant pathogens as with HAP/VAP. However, not allpatients with HCAP are at risk for MDR pathogens, but those who are have multiplerisk factors: severe infection, previous use of antimicrobial agents, poor functional sta-tus, recent hospitalization, and immune suppression.3,22 Shorr and colleagues72 havetried to develop a prediction rule for patients who are likely to have MDR pathogenswith HCAP, and the identified risks include recent hospitalization, ICU admission, he-modialysis, and nursing home residence. Supporting the fact that some patients withHCAP are not at risk for MDR pathogen infection are recent data showing efficacy forantibiotic regimens in HCAP that are often monotherapy, not broad spectrum, andmore closely resemble therapy for community-acquired pneumonia than therapy forVAP.7376

Recent data from a prospective multicenter trial fromGermany showed that patientswith HCAP had more severe pneumonia, assessed by CRB-65 score and higher mor-tality among those older than 65 years, in comparison with patients with community-acquired pneumonia, but had similar etiologic agents and a similar frequency ofmechanical ventilation.23 Nonetheless, in this study there was no association betweenexcess mortality and potential MDR pathogens. In another multicenter study thatenrolled more than 15,000 HCAP patients treated with guideline-concordant HCAP(GC-HCAP) therapy or guideline-concordant community-acquired therapy, the stron-

gest predictors of 30-day mortality were recent hospital admission (OR 2.49, 95% CI

Nair & Niederman5382.122.94) and GC-HCAP therapy, the latter being a mortality risk factor (OR 2.18,95% CI 1.862.55).76 Thus, the investigators found that in nonsevere HCAP patients,GC-HCAP therapy was not associated with improved survival when compared withGC-CAP therapy. By contrast, Falcone and colleagues74 compared HCAP patientsreceiving a GC-HCAP regimen with historical controls receiving a discordant regimen,and found that GC-concordant patients had a significantly shorter duration of antibi-otic therapy (median 15 vs 12 days), a shorter duration of hospitalization (median 18 vs14 days), and a lower mortality rate (17.8% vs 7.1%). From these studies, it is clear thatthere are conflicting data about the best therapy for HCAP, but it is safe to assume thatwhile not all HCAP patients need a VAP-like regimen, those with severe pneumoniaand risks for MDR pathogens should have empiric broad-spectrum coverage. Onthe other hand, those with less severe infection and no risk factors for MDR pathogens(poor functional status, recent hospitalization, recent antibiotic therapy, immunosup-pressive therapy) can be treated with monotherapy using a third-generation cephalo-sporin, a second-generation fluoroquinolone, or ertapenem. Even if a patient receivesan initial broad-spectrum regimen, it is possible to de-escalate to monotherapy,particularly if cultures are negative, and in one study this practice was achieved in70% of culture-negative HCAP patients, usually with a quinolone.77

NONRESPONDERS

Patients who fail to respond after appropriate initial antibiotic coverage should be eval-uated for complications of the illness or its therapy (empyema, lung abscess, drugfever, antibiotic-induced colitis, bronchopleural fistula) or a secondary site of infection.Alternative diagnoses such as inflammatory lung disease, atelectasis, heart failure,malignancy, and pulmonary embolus should also be considered, as well as infectionwith a resistant or unusual pathogen, such as a fungal infection. Evaluation generallyrequires further imaging studies including computed tomography scanning of thechest, or invasive diagnostic tests such as bronchoscopy with biopsy or pulmonaryangiography.A comprehensive clinical approach to NP management is shown in Fig. 2.

PREVENTION

A large number of measures have been studied in an attempt to reduce the incidenceof pneumonia in intubated patients (Box 5). Overall it is possible to reduce the inci-dence of VAP, but this requires the use of multiple interventions together, which areoften administered in the form of a ventilator bundle. The use of these bundles hasbeen effective at reducing the incidence of VAP, particularly if the implementation ofthe bundle is closely monitored.78,79 The incidence of VAP can be reduced substan-tially if there is a high rate of compliance with all the interventions in the bundle.80

The most commonly used elements in ventilator bundles are elevation of the headof the bed, daily interruption of sedation, daily weaning trials, oral decontaminationwith chlorhexidine, deep venous thrombosis prophylaxis, and intestinal bleeding pro-phylaxis. Some of the commonly used measures are characterized as follows.

Use of oral antiseptics. In a prospective, randomized, double-blind study of 350patients undergoing cardiac surgery, use of chlorhexidine gluconate 0.12% oralrinse reduced the incidence of pneumonia by 69% in comparison with controls(5/173 vs 17/180; P

Fig

Nosocomial Pneumonia 539Step1: Clinicalsuspicion for Nosocomial

pneumonia

Obtain radiographicconfirmation and lower respiratory

tract culture

Assess risk factors for multidrug

resistant pathogensConsider local

antibiogram

Step 2: Initiate empiric antibiotics

per ATS/IDSA 2005guidelines

Consider time of onset of disease

Fever, purulenttracheobronchial secretions,

worsening oxygenationstatus, Leukocytosis or

leukopeniaIn fact, the entire regimen of selective digestive decontamination may not be anymore effective than the use of oral antiseptics.

Subglottic drainage aspiration (SSD). A specially adapted endotracheal tube hasbeen designed to allow suctioning of secretions that pool above the endotra-cheal tube cuff, thereby interrupting the aspiration of secretions into the tracheo-bronchial tree. A recent randomized controlled trial showed a significantreduction in microbiologically confirmed VAP with the use of SSD whencompared with the control population (14.8% vs 25.6%; P 5 .02), but no differ-ence in duration of mechanical ventilation or mortality.83 In another randomizedstudy of more than 700 patients, comparing continuous aspiration of subglotticsecretions (CASS) versus controls after cardiac surgery, the CASS group had alower VAP rate (26.7% vs 47.5%; P 5 .04) and fewer days in the ICU (7 vs16.5 days; P 5 .01).84 Widespread use of this device has not occurred, in partbecause the tubes that use SSD have a smaller inner diameter than comparablysized tubes without SSD, making it more difficult to perform endotracheal suc-tioning with these newer tubes.

Ventilator circuit changes. Less frequent circuit changes help prevent bacterialcolonization of the circuit. In a study of 72 consecutive patients, comparing

Step 3: Re-evaluatein 72 hours

Clinical improvement, positive culture results, imporvement in CPISscore and biomarkers

Consider De-escalationand treatment for

7-8 days in the absenceof MDR pathogens

Negative cultureresults and clinicallystable or have otherpossible explanationfor initial worsening

Consider stoppingtreatment.

Rule out complications,improper dosage ofantibiotics, resistant

pathogens.

Clinical worsening andNon responders

. 2. Suggested simplified approach in the management of nosocomial pneumonia.

Nair & Niederman540 Hand washing Respiratory therapy equipment Maintain endotracheal cuff pressure, avoid changing ventilator circuit no more than48 hours apart

Compliance with use of ventilator bundles Subglottic secretion drainage Isolate patients with resistant organisms Avoid endotracheal intubation if possible (noninvasive ventilation) Reduce use of nasogastric tubes (place orally, and if possible postpyloric) Daily interruption of sedation Restricted blood transfusion policy Avoid immunosuppressive medications if possible Start enteral feeding only after >2448 hours after intubation Selective oral decontamination Deep vein thrombosis prophylaxis Influenza and pneumococcal vaccine: consider hospital-based programsBox 5

Preventive strategies used in NPventilator circuit changes every 48 hours to no change, the incidence of pneu-monia, bacterial colonization, and days on the ventilator showed no statisticaldifference when no change was the standard of care.85

Maintenance of sustained tracheal cuff pressure. Microaspiration of gastric con-tents is facilitated by the underinflation of the tracheal cuff pressure, and can beeffectively prevented by maintaining the tracheal cuff pressure by use of an auto-mated pneumatic device, or by close monitoring of the cuff pressure.26

Use of silver-coated endotracheal tube. Inhibition of the biofilm formation byusing an endotracheal tube coated with silver might be a useful strategy inVAP prevention, but consistent data on its use are still emerging. In a multicentertrial involving intubated patients, those who had silver-coated endotrachealtubes had lower rates of microbiologically confirmed VAP than the control groupwith uncoated tubes (4.8% vs 7.5%; P 5 .03). However, there were no differ-ences in the duration of intubation or mortality between the 2 groups.86 This de-vice is not widely used because it remains quite costly, and with the use ofventilator bundles it may have little incremental impact.

Many other interventions have been studied for the prevention of VAP, and arereviewed elsewhere. However, a promising approach involves modification of theendotracheal tube, its cuff composition, or its shape, as well as attention to improvedsuctioning of endotracheal secretions, or the removal of biofilm from the inside of theendotracheal tube.87

SUMMARY

NP is the leading cause of death from hospital-acquired infection. Mechanical ventilation is themost important risk factor for the development of NP.

5. Rello J, Ollendorf DA, Oster G, et al. Epidemiology and outcomes of ventilator-

1

1

Nosocomial Pneumonia 541associated pneumonia in a large US database. Chest 2002;122(6):211521.6. Dudeck MA, HT. National Healthcare Safety Network (NHSN) report, data sum-

mary for 2010, device-associated module. CDC; 2010. Available at: http://www.cdc.gov/nhsn/PDFs/dataStat/NHSN-Report_2010-Data-Summary.pdf.

7. Niederman MS. Hospital-acquired pneumonia, health care-associated pneumonia,ventilator-associated pneumonia, and ventilator-associated tracheobronchitis:definitions and challenges in trial design. Clin Infect Dis 2010;51(Suppl 1):S127.

8. Klompas M. Is a ventilator-associated pneumonia rate of zero really possible?Curr Opin Infect Dis 2012;25(2):17682.

9. Bonten MJ, Bergmans DC, Ambergen AW, et al. Risk factors for pneumonia, andcolonization of respiratory tract and stomach in mechanically ventilated ICU pa-tients. Am J Respir Crit Care Med 1996;154(5):133946.

0. Rotstein C, Evans G, Born A, et al. Clinical practice guidelines for hospital-acquired pneumonia and ventilator-associated pneumonia in adults. Can JInfect Dis Med Microbiol 2008;19(1):1953.

1. Craven DE, Chroneou A, Zias N, et al. Ventilator-associated tracheobronchitis:the impact of targeted antibiotic therapy on patient outcomes. Chest 2009; HCAP affects a heterogeneous population, and not all patients require empiricbroad-spectrum antibiotic therapy directed at MDR pathogens.

There is no gold-standard test for the diagnosis of VAP, but use of the CPIS andPCT can assist in antibiotic de-escalation.

NewCDC streamlined surveillance definitions of VAPmay be useful in the diag-nosis of sick patients with ventilator-associated complications, but are not spe-cific for the diagnosis of pneumonia.

Patients with risk factors for MDR pathogens and diagnosed with HAP/VAPrequire combination therapy with broad-spectrum antimicrobials.

Patients should be reassessed 72 hours after initiation of treatment for VAP andHCAP, and antibiotics should be de-escalated, based on available culture andclinical data.

Nonresponders should be evaluated for treatment failure or complications frominfection.

Use of ventilator bundles in the ICU is associated with a significant reduction inthe incidence of VAP.

REFERENCES

1. Vincent JL, Rello J, Marshall J, et al. International study of the prevalenceand outcomes of infection in intensive care units. JAMA 2009;302(21):23239.

2. Niederman MS, Craven D, Bonten MJ, et al. Guidelines for the management ofadults with hospital-acquired, ventilator-associated, and healthcare-associatedpneumonia. Am J Respir Crit Care Med 2005;171(4):388416.

3. Brito V, Niederman MS. Healthcare-associated pneumonia is a heterogeneousdisease, and all patients do not need the same broad-spectrum antibiotic ther-apy as complex nosocomial pneumonia. Curr Opin Infect Dis 2009;22(3):31625.

4. Safdar N, Dezfulian C, Collard HR, et al. Clinical and economic consequencesof ventilator-associated pneumonia: a systematic review. Crit Care Med 2005;33(10):218493.135(2):5218.

Nair & Niederman54212. Dallas J, Skrupky L, Abebe N, et al. Ventilator-associated tracheobronchitis in amixed surgical and medical ICU population. Chest 2011;139(3):5138.

13. Craven DE, Hjalmarson K. Prophylaxis of ventilator-associated pneumonia:changing culture and strategies to trump disease. Chest 2008;134(5):898900.

14. Nseir S, Favory R, Jozefowicz E, et al. Antimicrobial treatment for ventilator-associated tracheobronchitis: a randomized, controlled, multicenter study. CritCare 2008;12(3):R62.

15. Skerrett SJ, Niederman MS, Fein AM. Respiratory infections and acute lunginjury in systemic illness. Clin Chest Med 1989;10(4):469502.

16. Campbell DG, Niederman MS, Broughton WA, et al. Hospital-acquired pneu-monia in adults: diagnosis, assessment of severity, initial antimicrobial therapy,and preventive strategies. A consensus statement, American Thoracic Society,November 1995. Am J Respir Crit Care Med 1996;153(5):171125.

17. Waterer GW, Quasney MW, Cantor RM, et al. Septic shock and respiratory fail-ure in community-acquired pneumonia have different TNF polymorphism asso-ciations. Am J Respir Crit Care Med 2001;163(7):1599604.

18. Jones RN. Microbial etiologies of hospital-acquired bacterial pneumonia andventilator-associated bacterial pneumonia. Clin Infect Dis 2010;51(Suppl 1):S817.

19. Trouillet JL, Chastre J, Vuagnat A, et al. Ventilator-associated pneumoniacaused by potentially drug-resistant bacteria. Am J Respir Crit Care Med1998;157(2):5319.

20. Fagon JY, Chastre J, Domart Y, et al. Nosocomial pneumonia in patientsreceiving continuous mechanical ventilation. Prospective analysis of 52 epi-sodes with use of a protected specimen brush and quantitative culture tech-niques. Am Rev Respir Dis 1989;139(4):87784.

21. Kollef MH. Inadequate antimicrobial treatment: an important determinant ofoutcome for hospitalized patients. Clin Infect Dis 2000;31(Suppl 4):S1318.

22. El Solh AA, Pietrantoni C, Bhat A, et al. Indicators of potentially drug-resistantbacteria in severe nursing home-acquired pneumonia. Clin Infect Dis 2004;39(4):47480.

23. Ewig S, Klapdor B, Pletz MW, et al. Nursing-home-acquired pneumonia inGermany: an 8-year prospective multicentre study. Thorax 2012;67(2):1328.

24. Cook DJ, Walter SD, Cook RJ, et al. Incidence of and risk factors for ventilator-associatedpneumonia in critically ill patients.Ann InternMed1998;129(6):43340.

25. Perbet S, Mongardon N, Dumas F, et al. Early-onset pneumonia after cardiacarrest: characteristics, risk factors and influence on prognosis. Am J RespirCrit Care Med 2011;184(9):104854.

26. Nseir S, Zerimech F, Fournier C, et al. Continuous control of tracheal cuff pres-sure and microaspiration of gastric contents in critically ill patients. Am J RespirCrit Care Med 2011;184(9):10417.

27. Bekaert M, Timsit JF, Vansteelandt S, et al. Attributable mortality of ventilator-associated pneumonia: a reappraisal using causal analysis. Am J Respir CritCare Med 2011;184(10):11339.

28. Heyland DK, Cook DJ, Griffith L, et al. The attributable morbidity and mortality ofventilator-associated pneumonia in the critically ill patient. The Canadian CriticalTrials Group. Am J Respir Crit Care Med 1999;159(4 Pt 1):124956.

29. Tejerina E, Frutos-Vivar F, Restrepo MI, et al. Incidence, risk factors, andoutcome of ventilator-associated pneumonia. J Crit Care 2006;21(1):5665.

30. Craven DE, Hjalmarson KI. Ventilator-associated tracheobronchitis and pneu-

monia: thinking outside the box. Clin Infect Dis 2010;51(Suppl 1):S5966.

Nosocomial Pneumonia 54331. Niederman MS. The argument against using quantitative cultures in clinical trialsand for the management of ventilator-associated pneumonia. Clin Infect Dis2010;51(Suppl 1):S939.

32. Heyland D, Cook D, Dodek P, et al. A randomized trial of diagnostic techniquesfor ventilator-associated pneumonia. N Engl J Med 2006;355(25):261930.

33. Pugin J, Auckenthaler R, Mili N, et al. Diagnosis of ventilator-associated pneu-monia by bacteriologic analysis of bronchoscopic and nonbronchoscopic blindbronchoalveolar lavage fluid. Am Rev Respir Dis 1991;143(5 Pt 1):11219.

34. Singh N, Rogers P, Atwood CW, et al. Short-course empiric antibiotic therapyfor patients with pulmonary infiltrates in the intensive care unit. A proposedsolution for indiscriminate antibiotic prescription. Am J Respir Crit Care Med2000;162(2 Pt 1):50511.

35. Luna CM, Blanzaco D, Niederman MS, et al. Resolution of ventilator-associatedpneumonia: prospective evaluation of the clinical pulmonary infection score asan early clinical predictor of outcome. Crit Care Med 2003;31(3):67682.

36. Zilberberg MD, Shorr AF. Ventilator-associated pneumonia: the clinical pulmo-nary infection score as a surrogate for diagnostics and outcome. Clin InfectDis 2010;51(Suppl 1):S1315.

37. Shorr AF, Cook D, Jiang X, et al. Correlates of clinical failure in ventilator-associated pneumonia: insights from a large, randomized trial. J Crit Care2008;23(1):6473.

38. Klompas M, Khan Y, Kleinman K, et al. Multicenter evaluation of a novel surveil-lance paradigm for complications of mechanical ventilation. PLoS One 2011;6(3):e18062.

39. Hayashi Y, Morisawa K, Klompas M, et al. Toward improved surveillance: theimpact of ventilator-associated complications on length of stay and antibioticuse in patients in intensive care units. Clin Infect Dis 2013;56(4):4717.

40. Klompas M, Kleinman K, Khan Y, et al. Rapid and reproducible surveillance forventilator-associated pneumonia. Clin Infect Dis 2012;54(3):3707.

41. Klompas M, Magill S, Robicsek A, et al. Objective surveillance definitions forventilator-associated pneumonia. Crit Care Med 2012;40(12):315461.

42. Fagon JY. Biological markers and diagnosis of ventilator-associated pneumonia.Crit Care 2011;15(2):130.

43. Gibot S, Cravoisy A, Levy B, et al. Soluble triggering receptor expressed onmyeloid cells and the diagnosis of pneumonia. N Engl J Med 2004;350(5):4518.

44. Anand NJ, Zuick S, Klesney-Tait J, et al. Diagnostic implications of soluble trig-gering receptor expressed on myeloid cells-1 in BAL fluid of patients with pul-monary infiltrates in the ICU. Chest 2009;135(3):6417.

45. Luyt CE, Guerin V, Combes A, et al. Procalcitonin kinetics as a prognosticmarker of ventilator-associated pneumonia. Am J Respir Crit Care Med 2005;171(1):4853.

46. Bouadma L, Luyt CE, Tubach F, et al. Use of procalcitonin to reduce patientsexposure to antibiotics in intensive care units (PRORATA trial): a multicentrerandomised controlled trial. Lancet 2010;375(9713):46374.

47. Tang H, Huang T, Jing J, et al. Effect of procalcitonin-guided treatment in pa-tients with infections: a systematic review and meta-analysis. Infection 2009;37(6):497507.

48. Iregui M, Ward S, Sherman G, et al. Clinical importance of delays in the initiationof appropriate antibiotic treatment for ventilator-associated pneumonia. Chest

2002;122(1):2628.

Nair & Niederman54449. Wunderink RG, Niederman MS, Kollef MH, et al. Linezolid in methicillin-resistantStaphylococcus aureus nosocomial pneumonia: a randomized, controlledstudy. Clin Infect Dis 2012;54(5):6219.

50. Freire AT, Melnyk V, Kim MJ, et al. Comparison of tigecycline with imipenem/cil-astatin for the treatment of hospital-acquired pneumonia. Diagn Microbiol InfectDis 2010;68(2):14051.

51. Heyland DK, Dodek P, Muscedere J, et al. Randomized trial of combinationversus monotherapy for the empiric treatment of suspected ventilator-associated pneumonia. Crit Care Med 2008;36(3):73744.

52. Trouillet JL, Vuagnat A, Combes A, et al. Pseudomonas aeruginosa ventilator-associated pneumonia: comparison of episodes due to piperacillin-resistantversus piperacillin-susceptible organisms. Clin Infect Dis 2002;34(8):104754.

53. Niederman MS. Reexamining quinolone use in the intensive care unit: use themright or lose the fight against resistant bacteria. Crit Care Med 2005;33(2):4434.

54. Beardsley JR, Williamson JC, Johnson JW, et al. Using local microbiologic datato develop institution-specific guidelines for the treatment of hospital-acquiredpneumonia. Chest 2006;130(3):78793.

55. Fowler RA, Flavin KE, Barr J, et al. Variability in antibiotic prescribing patternsand outcomes in patients with clinically suspected ventilator-associated pneu-monia. Chest 2003;123(3):83544.

56. Nseir S, Di Pompeo C, Soubrier S, et al. First-generation fluoroquinolone useand subsequent emergence of multiple drug-resistant bacteria in the intensivecare unit. Crit Care Med 2005;33(2):2839.

57. Kumar A, Zarychanski R, Light B, et al. Early combination antibiotic therapyyields improved survival compared with monotherapy in septic shock: apropensity-matched analysis. Crit Care Med 2010;38(9):177385.

58. Kett DH, Cano E, Quartin AA, et al. Implementation of guidelines for manage-ment of possible multidrug-resistant pneumonia in intensive care: an observa-tional, multicentre cohort study. Lancet Infect Dis 2011;11(3):1819.

59. Ferrer M, Liapikou A, Valencia M, et al. Validation of the American ThoracicSociety-Infectious Diseases Society of America guidelines for hospital-acquired pneumonia in the intensive care unit. Clin Infect Dis 2010;50(7):94552.

60. Udy AA, Roberts JA, De Waele JJ, et al. Whats behind the failure of emergingantibiotics in the critically ill? Understanding the impact of altered pharmacoki-netics and augmented renal clearance. Int J Antimicrob Agents 2012;39(6):4557.

61. Nicasio AM, Eagye KJ, Nicolau DP, et al. Pharmacodynamic-based clinicalpathway for empiric antibiotic choice in patients with ventilator-associated pneu-monia. J Crit Care 2010;25(1):6977.

62. Niederman MS. The importance of de-escalating antimicrobial therapy in pa-tients with ventilator-associated pneumonia. Semin Respir Crit Care Med2006;27(1):4550.

63. NiedermanMS, Soulountsi V. De-escalation therapy: is it valuable for themanage-ment of ventilator-associated pneumonia? Clin Chest Med 2011;32(3):51734.

64. Kollef MH, Morrow LE, Niederman MS, et al. Clinical characteristics and treat-ment patterns among patients with ventilator-associated pneumonia. Chest2006;129(5):12108.

65. Soo Hoo GW, Wen YE, Nguyen TV, et al. Impact of clinical guidelines in the man-

agement of severe hospital-acquired pneumonia. Chest 2005;128(4):277887.

Nosocomial Pneumonia 54566. Chastre J, Wolff M, Fagon JY, et al. Comparison of 8 vs 15 days of antibiotic ther-apy for ventilator-associated pneumonia in adults: a randomized trial. JAMA2003;290(19):258898.

67. Michalopoulos A, Kasiakou SK, Mastora Z, et al. Aerosolized colistin for thetreatment of nosocomial pneumonia due to multidrug-resistant Gram-negativebacteria in patients without cystic fibrosis. Crit Care 2005;9(1):R539.

68. Niederman MS, Chastre J, Corkery K, et al. BAY41-6551 achieves bacteri-cidal tracheal aspirate amikacin concentrations in mechanically ventilatedpatients with Gram-negative pneumonia. Intensive Care Med 2012;38(2):26371.

69. Lu Q, Yang J, Liu Z, et al. Nebulized ceftazidime and amikacin in ventilator-associated pneumonia caused by Pseudomonas aeruginosa. Am J RespirCrit Care Med 2011;184(1):10615.

70. Nseir S, Di Pompeo C, Pronnier P, et al. Nosocomial tracheobronchitis in me-chanically ventilated patients: incidence, aetiology and outcome. Eur Respir J2002;20(6):14839.

71. Palmer LB, Smaldone GC, Chen JJ, et al. Aerosolized antibiotics and ventilator-associated tracheobronchitis in the intensive care unit. Crit Care Med 2008;36(7):200813.

72. Shorr AF, Zilberberg MD, Reichley R, et al. Validation of a clinical score for as-sessing the risk of resistant pathogens in patients with pneumonia presenting tothe emergency department. Clin Infect Dis 2012;54(2):1938.

73. Grenier C, Pepin J, Nault V, et al. Impact of guideline-consistent therapy onoutcome of patients with healthcare-associated and community-acquired pneu-monia. J Antimicrob Chemother 2011;66(7):161724.

74. Falcone M, Corrao S, Licata G, et al. Clinical impact of broad-spectrum empir-ical antibiotic therapy in patients with healthcare-associated pneumonia: a multi-center interventional study. Intern Emerg Med 2012;7(6):52331.

75. Chalmers JD, Taylor JK, Singanayagam A, et al. Epidemiology, antibiotic ther-apy, and clinical outcomes in health care-associated pneumonia: a UK cohortstudy. Clin Infect Dis 2011;53(2):10713.

76. Attridge RT, Frei CR, Restrepo MI, et al. Guideline-concordant therapy and out-comes in healthcare-associated pneumonia. Eur Respir J 2011;38(4):87887.

77. Schlueter M, James C, Dominguez A, et al. Practice patterns for antibiotic de-escalation in culture-negative healthcare-associated pneumonia. Infection2010;38(5):35762.

78. Resar R, Pronovost P, Haraden C, et al. Using a bundle approach to improveventilator care processes and reduce ventilator-associated pneumonia. JtComm J Qual Patient Saf 2005;31(5):2438.

79. Morris AC, Hay AW, Swann DG, et al. Reducing ventilator-associated pneu-monia in intensive care: impact of implementing a care bundle. Crit Care Med2011;39(10):221824.

80. Caserta RA, Marra AR, Durao MS, et al. A program for sustained improvement inpreventing ventilator associated pneumonia in an intensive care setting. BMCInfect Dis 2012;12:234.

81. DeRiso AJ 2nd, Ladowski JS, Dillon TA, et al. Chlorhexidine gluconate 0.12%oral rinse reduces the incidence of total nosocomial respiratory infection andnonprophylactic systemic antibiotic use in patients undergoing heart surgery.Chest 1996;109(6):155661.

82. Pileggi C, Bianco A, Flotta D, et al. Prevention of ventilator-associated pneu-

monia, mortality and all intensive care unit acquired infections by topically

applied antimicrobial or antiseptic agents: a meta-analysis of randomizedcontrolled trials in intensive care units. Crit Care 2011;15(3):R155.

83. Lacherade JC, De Jonghe B, Guezennec P, et al. Intermittent subglottic secre-tion drainage and ventilator-associated pneumonia: a multicenter trial. Am J Re-spir Crit Care Med 2010;182(7):9107.

84. Bouza E, Perez MJ, Munoz P, et al. Continuous aspiration of subglottic secre-tions in the prevention of ventilator-associated pneumonia in the postoperativeperiod of major heart surgery. Chest 2008;134(5):93846.

85. Dreyfuss D, Djedaini K, Weber P, et al. Prospective study of nosocomial pneu-monia and of patient and circuit colonization during mechanical ventilationwith circuit changes every 48 hours versus no change. Am Rev Respir Dis1991;143(4 Pt 1):73843.

86. Kollef MH, Afessa B, Anzueto A, et al. Silver-coated endotracheal tubes andincidence of ventilator-associated pneumonia: the NASCENT randomized trial.JAMA 2008;300(7):80513.

87. Lorente L, Blot S, Rello J. New issues and controversies in the prevention ofventilator-associated pneumonia. Am J Respir Crit Care Med 2010;182(7):8706.

Nair & Niederman546

Nosocomial PneumoniaKey pointsBackgroundPathogenesisEtiologic agentsRisk factors for development of NP and for mortalityVAP and HAPMortality in Patients with VAP or HAPHealth CareAssociated Pneumonia

Diagnostic considerationsCDC DefinitionClinical Pulmonary Infection ScoreNew Streamlined Surveillance DefinitionsRole of Biomarkers

Antimicrobial managementTreatment Considerations in HAP and VAPTiming of antibioticsAntibiotic choicesCombination therapySelecting the optimal second agentEffectiveness of combination therapy

Appropriate dosage of antibioticsDe-escalationDuration of treatmentLocalized treatments

Ventilator-Associated TracheobronchitisTreatment Considerations in Health CareAssociated Pneumonia

NonrespondersPreventionSummaryReferences