Aspiration pneumonia occurs when a large volume of oropharyngeal or gastric contents are aspirated into the lungs, depositing a large bacterial inoculum. This can overwhelm normal lung defenses and cause pneumonia. Risk factors include dysphagia, altered mental status, vomiting, enteral feeding, and oropharyngeal colonization with more virulent bacteria. Aspiration is common but often does not cause pneumonia due to protective mechanisms; however, large volume macroaspiration can lead to aspiration pneumonia.

1. Aspiration Pneumonia
Nov 7/2016
Review: David M. J of Critical Care

2.  Given this broad use of the term aspiration, classifying the
majority
 of bacterial pneumonias as a consequence of aspiration is
strictly
 correct based on known pathophysiology of community-
acquired
 (CAP) and hospital-acquired pneumonia (HAP) [2–5].
However, when
 a clinician uses the term aspiration pneumonia, he or she is
typically
 implying a subset of bacterial pneumonia that, although
sharing the
 common pathophysiologic mechanism with most other
pneumonias,
 represents a unique entity of a macroaspiration event resulting
in

3.  It is important to understand that aspiration is a
common event
 that may lie within the spectrum of normal
physiology. A large
 proportion of healthy people with normal mental
status aspirate
 during sleep based on the detection of radiolabeled
oral dyes in the
 lungs of healthy volunteers

4.  Therefore, one of
 the most common consequences of aspiration is
actually to have no
 consequence—the inoculum is cleared by the
normal airway and/or
 parenchymal host defenses without overt clinical
syndromes.

5.  Although occurring in otherwise healthy people,
several important
 clinical consequences of aspiration can occur.

7. Chemical Pneumonitis
 Chemical pneumonitis is characterized by
macroaspiration of
 noxious liquids with immediate hypoxemia, fever,
tachycardia, and
 abnormal chest radiograph and lung examination
result. The most
 common noxious fluid is sterile gastric contents,
although others such
 as bile and other agents instilled into the stomach
may also result in
 this syndrome.

8.  Animal experiments helped differentiate the pathophysiology
of
 chemical pneumonitis from subclinical aspiration based on the
pH
 and volume of gastric material needed to stimulate an
immediate and
 severe inflammatory reaction. Based on experiments using
human
 gastric secretions and rabbit lungs, a pH less than 2.4 was
required to
 cause vigorous inflammation. At higher pH, the reaction seen
 microscopically was more similar to the changes caused by
the
 instillation of water into the lungs [21]. In terms of quantity,
 experiments inducing chemical pneumonitis in a dog model
required
 2 mL of hydrochloric acid solution per kilogram to induce the
clinical

9. Bland aspiration
 Not all noninfectious macroaspirations cause an inflammatory
 response in the lung; and therefore, to label these as
pneumonitis
 would be inappropriate. Probably the 2 most common
examples are
 aspiration of blood as a complication of severe epistaxis or
 hematemesis and the aspiration of enteral feedings. Twenty
percent
 of patients undergoing esophagogastroduodenoscopy will
have an
 infiltrate immediately after the procedure in the dependent lung
 [24,25]. Most resolve without antibiotic changes. Most
episodes of
 aspiration with enteral nutrition are also uncomplicated

10.  Although bland aspiration may not initially be infectious, blood
 and enteral feedings represent excellent culture media for
growth of
 either resident bacteria or the small aliquot of bacteria
included in the
 inoculum. Generally, mucociliary clearance and the resident
alveolar
 macrophages can clear the inoculum within hours. The major
issue is
 confusion with an infectious aspiration pneumonia, particularly
when
 the large-volume aspiration is not observed. Prolonged
antibiotic
 treatment is unlikely to prevent this secondary pneumonia but
may
 select for more multidrug-resistant (MDR) pathogens

11. CAP & HAP
 Microaspiration has long been known to be the dominant
 pathophysiologic mechanism behind CAP. Supporting evidence
 includes the finding that most common CAP-causing microorganisms
 colonize the oropharynx or nasopharynx in nonhospitalized patients
 [2,27,28]. Similarly, the pathophysiology underlying HAP, including
 ventilator-associated pneumonia (VAP), has proved to be microaspiration
 of oropharyngeal, upper gastrointestinal, or subglottic
 contents [3,5,29–32]. The distinct microbiology of HAP stems from
 microaspiration occurring after hospitalized patients become colonized
 with the virulent organisms found in intensive care unit and
 hospital environments [4,33–36].
 Given the above evidence of aspiration as a common event,
 development of a parenchymal lung infection depends largely on host
 defense factors [12,37] and the virulence of the aspirated pathogen.
 This interaction helps explain the phenomenon of subclinical
 aspiration without subsequent pneumonia described mostly in
 young healthy volunteers and surgical candidates

12. Aspiration Pneumonia
 Current use of this term most commonly refers to an acute
lung
 infection developing after a large-volume aspiration of
oropharyngeal
 or upper gastrointestinal contents with a high enough pH to
avoid
 chemical pneumonitis (likely pH much greater than 2.5). This
type of
 aspiration deposits a large bacterial load of pathogens from
the oral
 cavity or upper gastrointestinal tract into the lungs. The
possibility of
 infection with these normally nonvirulent, predominantly
anaerobic
 organisms is partly because of the large inoculum
[2,17,21,39–41].
 Confusion surrounding this terminology and the exact
definition

13.  Macroaspiration is the unique pathophysiologic
component of
 what most clinicians call aspiration pneumonia. The
challenge in
 specifically diagnosing aspiration pneumonia is that, for
many
 patients in the community who are at risk for
macroaspiration, the
 events in the days leading up to presentation with fever,
cough, and
 chest radiograph infiltrate are unclear. A common risk for
macroaspiration
 is decreased mental status, but this can be the result of
CAP
 rather than the cause [43]. Because of this reality,
substantial
 diagnostic overlap exists between aspiration, HAP, and

14.  Aspiration pneumonia represents 5% to 15% of
pneumonias in the
 hospitalized population. The ICD-9 code–based
reviews suggest an
 increasing incidence, making it the second most
common diagnosis in
 Medicare patients who are hospitalized [2,44].
However, higher
 reimbursement rates for this ICD-9 code than for
CAP ICD-9 codes may
 falsely increase the frequency in this population.

15. RFs for Aspiration Pneumonia
 Specific predisposing factors for aspiration pneumonia focus
on
 the risk for high frequency and/or large volume of aspiration.
Some
 risks may be more pertinent for the macroaspiration
characteristic of
 aspiration pneumonitis or anaerobic pleuropneumonia than for
 microaspiration. Additionally, factors that influence the resident
 bacterial flora leading to colonization by more virulent
pathogens,
 which are more likely to overwhelm the normal protective
mechanisms,
 also play a role in development of clinical disease

16. RFs for Aspiration Pneumonia
Dysphagia/swallowing dysfunction
 Dysphagia, typically from neurologic disease
(dementia, Parkinson
 disease, multiple sclerosis, poststroke), is
considered the most
 important risk factor for aspiration pneumonia, given
the abovedescribed
 pathogenesis.

17. RFs for Aspiration Pneumonia
Dysphagia/swallowing dysfunction
 It is important to remember that dysphagia itself is not
definitive
 evidence of aspiration. Many high-risk patients will not
complain of
 dysphagia but still aspirate based on advanced testing [51].
The
 poststroke population certainly has a higher prevalence of
pneumonia
 with or without symptomatic dysphagia [52,53]. A lag time of
more
 than 5 seconds between noxious stimuli and cough, as well as
an
 increasing stimuli needed to produce a cough, has been linked
to
 pneumonia in poststroke patients regardless of dysphagia

18. RFs for Aspiration Pneumonia
Dysphagia/swallowing dysfunction
 The swallowing mechanism can also be affected by
chest anatomy.
 Swallowing dysfunction is very common in chronic
obstructive
 pulmonary disease (COPD) patients with
hyperinflation

19. RFs for Aspiration Pneumonia
Dysphagia/swallowing dysfunction
 Certain medications interfere with the swallow reflex
and may
 potentially lead to aspiration [58]. Although sedatives
may suppress
 the patient’s mental status sufficiently to lead to
aspiration,
 antipsychotic medications may actually affect the
swallowing mechanism
 by inhibiting dopamine and therefore lead to
aspiration.
 Accordingly, these drugs have been linked to
pneumonia in a fairly
 large retrospective study

20. RFs for Aspiration Pneumonia
Altered Mental Status
 The association between acute altered mental status
(AMS) and
 aspiration pneumonia has not been studied extensively
despite the
 obvious connection
 Most available case series focus on the association of
 acute AMS with chemical pneumonitis in the setting of
sedation,
 poisoning, and trauma In these populations, vomiting and
 large-volume reflux of gastric contents may also increase
the risk of
 aspiration pneumonia.

21. RFs for Aspiration Pneumonia
Altered Mental Status
 Two specific types of AMS—acute alcohol abuse and
seizures—are
 most likely to lead to the anaerobic pleuropneumonia
syndrome.
 Probably the highest risk of aspiration pneumonia
occurs in the severe
 alcohol abuse population. Acute alcohol ingestion
has multifactorial
 risks for aspiration pneumonia including AMS,
increased risk of
 vomiting, and direct effects of alcohol on normal
neutrophil function.

22. RFs for Aspiration Pneumonia
Esophogeal motility disorders/vomiting
 Esophogeal motility disorders independent of GERD are also
 associated with aspiration and an increased risk of
pneumonia.
 Many are a component of an underlying systemic disease,
such as
 scleroderma or polymyositis, which may also compromise the
host
 immune response itself or secondary to immunosuppressive
treatment.
 Primary esophageal disorders, such as achalasia and
esophageal
 strictures, increase the risk of aspiration of not only liquids but
also
 solids. The latter are a unique form of aspiration risk in which
bronchial impaction and postobstructive pneumonia result
from the

23. RFs for Aspiration Pneumonia
Esophogeal motility disorders/vomiting
 Given the frequency of vomiting, the incidence of
aspiration
 pneumonia/pneumonitis is actually very low.
Protective laryngeal
 reflexes will prevent macroaspiration in the
overwhelming majority
 of circumstances. Macroaspiration with vomiting
almost always
 requires concomitant abnormal mental status, such
as anesthesia
 induction, acute alcohol intoxication, or
narcotics/sedatives.

24. RFs for Aspiration Pneumonia
Esophogeal motility disorders/vomiting
 Another unique syndrome is vomiting associated with
small bowel
 obstruction. In this situation, the stomach is no longer
sterile but
 instead is filled with fluid that has significant overgrowth
of bowel
 flora. Narcotics and antiemetics may compromise mental
status at the
 time of vomiting. The result is a fulminant aspiration
pneumonia due
 to gram-negative bowel pathogens, rather than the
predominant
 gram-positive/anaerobic oral flora.

25. RFs for Aspiration Pneumonia
Enteral feeding
 The risk for aspiration pneumonia with enteral tube
feeding has
 been extensively studied, especially in the more
critically ill. Smalland
 large-bore nasogastric tubes, postpyloric tube feeds,
gastric tube
 feeds, and jejunal tube feeds have all been
associated with aspiration
 pneumonia in patients with and without endotracheal
and tracheostomy
 tubes.

26. RFs for Aspiration Pneumonia
Enteral feeding
 Exact risk is difficult to characterize given the wide
 variety of incidences reported, small sample sizes,
and lack of
 standard definitions regarding aspiration and
aspiration pneumonia
 [65–75]. Regardless of the deficiencies in
epidemiologic data,
 aspiration pneumonia is common enough in this
population that it
 should be a consideration for all patients on tube
feeds.

27. RFs for Aspiration Pneumonia
Enteral feeding
 Certain
 patients appear to be at greater risk. Intuitively, GERD and
decreased
 gastric motility are implicated when tube feeds are aspirated
Decreased gastric motility, typically defined by high gastric
 residual volume, has also been suggested as a risk factor for
aspiration
 in tube-fed patients [65,67]. However, the criteria for high
gastric
 residual volume vary widely between studies from 50 to
greater than
 500 mL at every 4-hour checks. A potentially independent risk
factor
 is that patients with high gastric residuals may also be at
increased
 risk of vomiting

28. RFs for Aspiration Pneumonia
Oropharyngeal colonization
 Microbiologic factors also influence the risk of aspiration
pneumonia.
 Pathophysiologically, risk of pneumonia relates to the
body’s
 ability to combat the bacteria that routinely reach the
lower
 respiratory tract. Unusual or more virulent microbes may
be more
 difficult to eradicate by the normal host defenses. By far,
the most
 important influence on alterations in normal
oropharyngeal flora is
 use of systemic antibiotics.

29. RFs for Aspiration Pneumonia
Oropharyngeal colonization
 An independent association of poor oral hygiene with
aspiration
 pneumonia is also supported by the literature [56,77]. The
microbial
 density is increased in patients with gingival disease even if
the
 spectrum has not shifted, increasing the likelihood of
pneumonia
 developing in association with an episode of aspiration due to
the
 greater inoculum. This suggests that edentulous patients are
at lower
 risk for aspiration pneumonia. In edentulous patients, the
tongue is
 more important as a focal point for colonization. Abe et al [78]
 associated tongue-coating scores in an edentulous elderly

30. RFs for Aspiration Pneumonia
Oropharyngeal colonization
 Many of the studies of oral hygiene have also
demonstrated greater
 colonization by more virulent organisms in patients
with poor oral
 hygiene. This is especially true for the colonization of
gram negatives
 and respiratory pathogens in the intensive care unit [

31. RFs for Aspiration Pneumonia
Other Risks
 Other risks
 General risk factors like male sex and smoking may
increase risk
 for aspiration pneumonia based on case-controlled and
cohort studies
 [48]. Diabetes mellitus has been repeatedly associated
with pneumonia
 in patients who have had an acute stroke [48].
 Much has been discussed regarding the increased risk of
 pneumonia as a whole in patients being treated with
proton pump
 inhibitors and/or histamine receptor–2 antagonists
[80,81].

32. RFs for Aspiration Pneumonia
Other Risks
 Although
 these medications may not increase the risk of aspiration, they
change
 the gastrointestinal environment such that natural host
defenses,
 which include gastric acid secretion, cannot reduce bacterial
burden. If
 a subsequent aspiration event occurs, patients appear more
likely to
 deliver an inoculum of bacteria high enough to cause clinical
infection.
 Conversely, the frequent use of proton pump inhibitors or
histamine
 receptor–2 antagonist, particularly in the hospitalized
population,
 may be associated with a lower incidence of aspiration
pneumonitis

33. Diagnosis
 In clinical practice, aspiration pneumonia is
 most often coded as the diagnosis when a new chest
radiograph
 infiltrate in a dependent pulmonary segment is found
in patients with
 risk factors for aspiration. In a bed-bound patient, the
dependent
 pulmonary segments are the posterior segments of
the upper lobes
 and the superior segments of the lower lobes. In
ambulatory patients,
 lower lobes are classically involved, especially the
right

34. Diagnosis
 Clinical features can help distinguish aspiration pneumonia
from
 chemical pneumonitis and other lung infections. As opposed to
 chemical pneumonitis, the aspiration event in aspiration
pneumonia
 is rarely witnessed [17]. The large volume of stomach contents
 required to cause chemical pneumonitis usually makes it a
more
 obvious event. Furthermore, the clinical course of chemical
pneumonitis
 is hyperacute hypoxemia, occurring almost immediately (within
 hours) and resulting in either devastating lung injury or
resolution
 within 48 hours. These patients are likely to also have
bronchospasm,
 frothy sputum, and chest radiographs with bilateral patchy
infiltrates
including nondependent areas

35. Diagnosis
 Because of this difficulty, efforts have been made to
use biomarkers
 to distinguish aspiration pneumonia from other
aspiration syndromes.
 El-Solh et al [85] attempted to use procalcitonin to
 distinguish aspiration pneumonitis from aspiration
pneumonia in
 the intensive care unit setting, given data to suggest
that procalcitonin
 is a helpful marker for bacterial causes of sepsis

36. Diagnosis
 Unfortunately, no difference between procalcitonin
levels was demonstrated in
 culture-negative and culture-positive patients.

37. Diagnosis
 Biomarkers more specific to aspiration have also been
studied.
 Pepsinogen in tracheal secretions or BAL was very suggestive
of
 aspiration as part of the pathogenesis of posttransplant BO
and VAP.
 Bronchoalveolar lavage amylase levels have been
demonstrated to
 correlate with clinical risk factors for aspiration, as well as with
 positive cultures [87–89]. This relationship may even be true in
 patients with VAP [90]. Bronchoalveolar lavage amylase can
also
 function as an end point for studies of interventions to
decrease risk of
 aspiration in ventilated patients

38. Microbiology
 The unique pathophysiology of aspiration
pneumoniamay lend itself
 to unique pathogens. However, the microbiology,
and therefore the
 treatment, has seen significant changes over the last
40 to 50 years.

39.  The original teaching was that anaerobic bacteria
were by far the
 most common pathogens in aspiration pneumonia
based on well-done
 microbiology studies undertaken in patientswith
aspiration pneumonia
 acquired in and out of the hospital fromthe 1960s to
1980s.*

40.  These anaerobic infections
 commonly included greater than one pathogen, with
Bacteroides
 species, Prevotella, Fusbacterium species, and
peptostreptococci predominating
 (Table 2). Most of these patients had the anaerobic
 pleuropneumonia syndrome described above.

42.  As homogenous as these initial results appeared, evidence
accumulated
 that aspiration pneumonia occurring after hospitalization had a
microbiologic
 spectrum that included more Staphylococcus aureus, aerobes,
and gram-negative bacilli [17,84,92,94]. The pathogens that
dominate aspiration
 pneumonia microbiology after a macroaspiration event after
 hospitalization are similar to those of many nosocomial
infections.
 Although very limited, data fromreliable cultures in
nonintubated patients
 do suggest a higher frequency of anaerobes than in intubated
patients; but
 the frequency is substantially lower than that of the prior

43.  Recent studies reveal much different results even for
patients
 presenting from the community. El-Solh et al [97]
reported a series of
 patients with suspected aspiration pneumonia who
underwent
 bronchial sampling after intubation. Of the 54
patients with a
 bacterial diagnosis, 20% grew only anaerobes, with
an additional
 11% that included anaerobes as part of mixed flora.

44.  In contrast,
 common causative organisms in this study were
Escherichia coli,
 S aureus, and Klebsiella pneumoniae. Tokuyasu et al [98]
described this
 trend further in a series of elderly Japanese patients with
clinically
 diagnosed aspiration pneumonia. Of 111 organisms
isolated in 62
 individuals, only 22 (20%) were anaerobes. Anaerobes
were heavily
 outweighed by gram-negative bacilli (almost all enteric
gramnegatives),
 found in 51.6% of patients

45.  Even the etiology in patients with lung abscess has changed.
 Takayanagi et al [99] reported bacterial etiologies in 122
patients
 diagnosed with community-acquired lung abscess, likely a
result of
 untreated aspiration. In this population, 74% grew aerobes
only, 12%
 grew anaerobes only, and 14% grew mixed flora. Of the 107
aerobic
 cases, 79% were Streptococcus species. In a very similar
study, Wang et
 al [100] reported that only 40 (44%) of 90 community-acquired
lung
 abscesses grew any anaerobes, with only 13% purely
anaerobic. Of the
 remaining cases, 33% were caused by K pneumoniae (almost
all being
 pure K pneumoniae isolates).

46.  The latter finding suggests that
 aspiration may not even play a role in some cases of
lung abscess,
 but rather more virulent CAP pathogens. The
combination of lung
 abscess and empyema has also been reported with
communityacquired
 methicillin-resistant S aureus (MRSA) pneumonia

47.  These 2 distinct bodies of literature, taken
chronologically, reveal a
 fading importance of anaerobic bacteria in aspiration
pneumonia and even community-acquired lung abscess.
A second implication of this
 etiologic shift is that the principles of typical nosocomial
microbiology
 apply to patients with aspiration pneumonia if
macroaspiration
 occurs after hospitalization. This etiologic overlap
between aspiration
 pneumonia and HAP has been progressively evident
since the 1970s

48. Treatment
 As one would expect, empirical treatment of aspiration
pneumonia
 has evolved, given the above changes in the microbiology of
the
 infection [102]. Intravenous penicillin was the drug of choice in
the
 past, as anaerobes constituted the vast majority of infections
with few
 penicillinase-producing bacterial strains [103,104]. A
randomized
 controlled trial (RCT) of 39 patients with lung abscesses
compared
 penicillin with clindamycin in the early 1980s [105]. Although a
small
 group of patients, the treatment failure rate and cure rate were
much
 better for clindamycin, with all 13 followed patients being cured
vs 8

49.  The failure of penicillin to cure anaerobic
 infections was better characterized several years
later in a Spanish
 RCT of confirmed anaerobic lung infections [106]. In
this cohort of 37
 patients, 47 anaerobes were isolated. Ten of these
47, all Bacteroides
 species, were penicillin resistant, whereas none
were clindamycin
 resistant. None of the 5 patients with penicillin-
resistant bacteria
 randomized to penicillin responded to therapy.

50.  Metronidazole has also been studied in anaerobic lungs
infections.
 Sanders et al [107] described a poor cure rate in 13
patients with
 pleuropulmonary (11 of 13 being lung abscesses)
infections with
 confirmed anaerobic bacteria. Similarly, Perlino [108]
reported higher
 cure rates with clindamycin when compared to
metronidazole in
 cases of lung abscess and pneumonia with confirmed
anaerobic flora
 in a small RCT of 13 patients.

51.  It is important to understand that these studies included
mainly
 classic anaerobic pleuropneumonia syndrome with
anaerobes confirmed
 on culture and that most were completed decades ago. In
the
 currently uncommon patient with aspiration resulting in
classic
 anaerobic pleuropneumonia, prior results and the likely
greater
 incidence of penicillin resistance suggest that
clindamycin may be
 the optimal agent

52.  Recent studies have focused on pneumonia in patients with
risk
 factors for aspiration. Kadowaki et al [109] randomly assigned
100
 elderly Japanese patients with suspected aspiration
pneumonia to
 clindamycin, a carbapenem (penipenem/betamiprom), low-
dose
 ampicillin/sulbactam (1.5 g twice daily), or high-dose
ampicillin/
 sulbactam (3 g twice daily). The investigators found little
variance in
 efficacy (N75% cure rate in all groups) and adverse events.
Interestingly,
 no anaerobes were actually cultured. Of note, clindamycin was
the

53.  Another study of elderly Japanese
 with aspiration pneumonia [98] demonstrated a clinical
efficacy rate of
 61.3% with another carbapenem, meropenem. This lower
efficacy than
 that found by Kadowaki et al [109] may be due to greater
severity of
 illness in the study patients. Once again, nosocomial
pathogens rather
 than anaerobes were the most common documented
etiologies; and 33
 of these 62 patients had MRSA growing in their
postantibiotic sputum
 culture.

54.  A recent randomized German study compared high-dose
 ampicillin/sulbactam (3 g thrice times daily) to the standard
CAP
 antibiotic moxifloxacin for the treatment of aspiration
pneumonia and
 lung abscess in 96 elderly patients [110]. Clinical response
rates were
 identical at 66.7%, and adverse reaction rates were very
similar.
 Microbiology was consistent with the more recent data
described
 above, with less than 10% of bacteria cultured being
anaerobes. Of note,
 higher (although not statistically significant) mortality was seen
in the
 ampicillin/sulbactamgroup, with 14 patients dying compared to
6 in the
 moxifloxacin group.

55.  These recent data from Japan and Germany have
demonstrated
 effective treatment strategies for aspiration
pneumonia in the face of new microbiology patterns.
Based on this limited evidence,
 clindamycin, a carbapenem, ampicillin/sulbactam,
and moxifloxacin
 all appear to be reasonable first-line therapies in
modern-day
 community-acquired aspiration pneumonia

56.  For patients with hospital-acquired macroaspiration
pneumonias,
 use of broad-spectrum combination therapy is
recommended if MDR
 risk factors are present. Probably the single most
important risk factor
 for MDR pathogens is prior antibiotic treatment. If none,
the
 antibiotics listed for community-acquired aspiration
pneumonia are
 adequate. The longer the prior course and the broader
the spectrum of
 agent, the greater the likelihood of MDR pathogens.

57.  In nonventilated
 patients, anaerobes may still play a role; and
cefepime should be
 avoided. For ventilated patients, the high oxygen
tension is sufficient
 to kill anaerobes; and any β-lactam should be
appropriate, although
 changing β-lactam class may be prudent.

58. Prevention
Dietary Changes
 Dietary interventions have been studied in patients with dysphagia.
 In a small study involving patients with dysphagia secondary to
 neurodegenerative disease (pseudobulbar dysphagia) [111], more
 aspiration pneumonia occurred in those on a pureed diet compared
to
 a mechanical soft diet with thickened liquids. However, the utility of
 dietary intervention has been questioned. Depippo et al [112]
 randomized 115 poststroke patients to 3 groups according to speech
 therapist intervention: Group A was given advice based on swallow
 testing, but the ultimate dietwas determined by the patient and
family;
 Group B was prescribed a specific diet based on swallow testing;
and
 Group C was prescribed a specific diet and directly observed for
 compliance daily. No statistically significant differences between the
 groups were found in any end point.

59. Prevention
Drugs to Protect the Airway
 A number of small studies have reviewed
pharmacologic intervention
 to protect the airway via the cough reflex. The most
 interesting drugs studied are angiotensin-converting
enzyme inhibitors
 (ACEIs) because of their role in degrading substance
P and
 bradykinin, stimulants of the cough reflex. A
reduction in aspiration
 pneumonia in patients on an ACEI has been
suggested in one casecontrol
 study [113] in elderly Japanese patients

60. Prevention
Drugs to Protect the Airway
 Further, investigators
 have categorized an increased risk for pneumonia in
patients with
 certain ACE gene polymorphisms that are
associated with higher ACE
 levels: homozygous deletion of an alu repeat within
intron 16 (ACE
 DD). The risk of pneumonia was markedly reduced
in a case-control
 study of patients without this genotype who were
taking an ACEI,
 whereas it was unaffected in patients with the
genotype

61. Prevention
Enteric Feeding Tubes
 Because enteric tube feeding presents a risk for
aspiration, there has
 been considerable effort to compare types of tube
feeds tominimize this
 risk. The most important comparisons are between
gastric and
 postpyloric feedings. Given the gastric dysmotility
caused by critical
 illness, gastroparesis, and commonmedications,
postpyloric feeds have
 been commonly postulated to be superior

62. Prevention
Enteric Feeding Tubes
 Two small prospective
 trials have found no difference [115,116] in
pneumonia rates. To the
 contrary, a very small randomized trial, with almost
no cases of
 aspiration pneumonia, and another prospective trial
found advantages
 to jejunal feeds [117,118]. Comparisons have also
been made between
 nasogastric tube and percutaneous endoscopic
gastrostomy tube feeds
 in a variety of clinical settings.

63. Prevention
Enteric Feeding Tubes
 Several randomized controlled trials have
 failed in demonstrating a difference in pneumonia
complication rates
 between the 2 feeding strategies. Percutaneous
endoscopic gastrostomy
 tubes are more likely to achieve goal feeds but come
at a much higher
 cost

64. Prevention
Enteric Feeding Residual Volumes
 Many institutions monitor residual volumes from tube
feeds to
 know when aspiration risk is increased. Residual
volumes of 500 mL are considered high enough to hold
tube feeds [65]. However, the
 inaccuracy of this method has been well documented
[122].
 Furthermore, results from a recent randomized clinical
trial suggest
 that using strict residual volumes (250 mL) to understand
when to
 hold nasogastric tube feeds does not affect the incidence
of VAP

65. Prevention
Oral Care
 Oral care has been shown to assist in preventing
aspiration
 pneumonia as expected given the evidenced
discussed above. Data are
 again limited, but encouraging quality oral care offers
potential
 benefit with almost no morbidity

66. Prevention
Prophylactic Antibiotics
 For patients at risk of aspiration around the time of
endotracheal
 intubation, several studies have shown that a short
course (≤24
 hours) of “prophylactic” β-lactam antibiotics may
decrease the risk of
 subsequent VAP [125,126]. An extremely elevated
BAL amylase may
 better select patients for this intervention.

67. Prevention
Head Elevation
 One very important preventative measure surrounds
preventing
 aspiration in the hospitalized, critically ill patient.
Increased aspiration
 in the supine position was evident after a Spanish study
that detected
 enterically administered dye aspirated into the lungs of
mechanically
 ventilated patients. Aspiration rates not only were higher
in patients
 who were supine but were dependent on how much time
was spent
 in the supine position

68. Prevention
Head Elevation
 The same investigators then confirmed
 the importance of this phenomenon by demonstrating
drastically
 reduced rates of HAP in mechanically ventilated patients
in the
 semirecumbent position compared to supine [128]. A
subsequent
 study did not show a benefit of semirecumbant position
when
 compared to elevation of as little as 10° from supine
[129]. However,
 the risk benefit ratio of elevation of the head of the bed in
ventilated
 patients is so favorable that it has become standard
practice

69. Thank You!