747

STATE-OF-THE-ART CLINICAL ARTICLE

Pleural Empyema
RichardE. Bryant and Christopher J. Salmon
Pleural empyema is a serious complication of infection adja-
cent to or within the chest that rarely resolves without appro-
priate medical therapy and drainage procedures [1-3]. Host
defenses are seriously compromised by the anatomy and physi-
ology ofan infected pleural space, and subtleties ofpresentation
may delay recognition and appropriate management. Empyema
is usually a complication of pneumonia but may arise from
infections at other sites. Presentation and microbial etiology are
modified by local trauma or surgery or by underlying conditions
such as malignancy, collagen vascular disease, immunodefi-
ciency disorders, and adjacent infection involving the orophar-
ynx, esophagus, mediastinum, or subdiaphragmatic tissues.
Clinical features depend upon the primary organ or space in-
fected, the microbial pathogen(s), and host defense defects.
Recent advances in imaging and instrumentation have facili-
tated the recognition and management of bacterial empyema
[4-9], and scholarly work in the field has improved our under-
standing of its pathophysiology and clinical presentation [1,
10-14]. Use of the thrombolytic agent urokinase, in conjunc-
tion with precise and timely placement of drainage catheters
under imaging guidance, has made it possible to reduce the
risk of pleural fibrosis and lung entrapment while avoiding
thoracotomy [15- 21]. Likewise, video-assisted thoracoscopic
techniques also provide an effective, less invasive means of
assessing and managing the infected pleural space without full
thoracotomy [22, 23].
Historical Perspective
It is interesting that Aristotle recognized the clinical entity
of empyema and described drainage of pus with incision, cau-
tery, and a metal tube [24]. He also described the risk oflethal
pneumothorax when such interventions were undertaken before
loculation of pleural pus had occurred. Twenty-five centuries
later, appreciation of that risk formed the basis for the recom-
mendation of Dr. Evarts A. Graham and the World War I
Received 19 December 1995; revised 19 January 1996.
Reprints or correspondence: Dr. Richard E. Bryant, Director, Infectious
Diseases Division, Oregon Health Sciences University, 3181 SW Sam Jackson
Park Road, L457, Portland, Oregon 97201.
Clinical Infectious Diseases 1996;22:747-64
1996 by The University of Chicago. All rights reserved.
1058-4838/96/2205-0016$02.00
From the Divisions of Infectious Diseases and Thoracic Imaging.
Oregon Health Sciences University. Portland. Oregon
Empyema Commission that an empyema should not be treated
by open drainage in the "acute pneumonic phase" in order to
lessen the risk of fatal pneumothorax [25]. Thereafter, simple
but ingenious closed drainage systems favorably modified the
risks associated with the pleural evacuation, facilitating earlier
and more efficacious drainage. Although closed chest tube
drainage of empyema had been described by Hewitt in 1875,
it came into widespread use only after Graham's report of 1918
[26].
More recently, sophisticated imaging technologies have
greatly enhanced our ability to identify, sample, and drain col-
lections of infected pleural fluid [4-10, 27]. Despite such rapid
advances in diagnosis and therapy, it is still possible for an
empyema to remain undetected unless the risks of this compli-
cation are appreciated and appropriate diagnostic measures are
used. Although currently available antimicrobial agents can
control some of the systemic manifestations of empyema, the
morbidity and mortality caused by undrained pleural pus are
still high [28- 31]. Optimal treatment requires drainage. This
was recognized by Osler, who underwent a rib resection for
treatment of postpneumonic Haemophilus influenzaeempyema,
which ultimately caused his death [20, 30]. Current technology
has increased the speed and finesse with which pleural empy-
emas can be drained and has improved our understanding of
why it is necessary to drain them [1, 2, 4-13].
Anatomy
The pleura is derived embryologically from the primitive
coelomic cavity [14]. It consists of two mesothelial layers with
their associated vascular, lymphatic, and connective tissue por-
tions. The visceral and parietal pleurae are continuous with one
another at the root of the lung, where the hilar airways and
vessels enter the lung parenchyma, and are closely apposed to
the individual pulmonary lobes, the inner aspect of the thoracic
cage, and the lateral margin of the mediastinum. The resultant
pleural space contains scant fluid and is normally a potential
space that becomes a true space only in disease states that
cause accumulation of pleural fluid (liquid or air).
The visceral pleura is attached to the lung surface and is
contiguous with the subpleural pulmonary interstitium [32]. It
is ~ ,um thick and apparently derives its blood supply
from both pulmonary and systemic arteries, draining to the
pulmonary veins. The visceral pleura individually invests pul-

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

748 Bryant and Salmon em 1996;22 (May)
monary lobes. The interlobar fissures seen radiographically or
by CT are due to the additive thickness of the visceral pleural
layers of the participating lobes. The normal pleural fluid vol-
ume is negligible and invisible by imaging. The major, or
oblique, fissure separates the lower lobe from the upper lobe
of the left lung (or the lower lobe from the upper and middle
lobes on the right side). The minor, or horizontal, fissure sepa-
rates the right middle lobe from the upper lobe.
The parietal pleura is composed of four layers but is slightly
thinner than the visceral pleura [32]. It is surrounded by a thin
layer of extrapleural or subcostal fat, which is surrounded by
the fibroelastic endothoracic fascia that constitutes the bound-
ary of the thoracic cavity. The endothoracic fascia is attached
to the perichondrium of the costal cartilage, the ribs and inter-
costal muscles, and the prevertebral fascia surrounding the ver-
tebral bodies and intervertebral disks. The extrapleural fat layer
is normally j1,m thick but may become radiologically
detectable in normal patients. It increases diffusely in the pres-
ence of empyema, but not in obese patients. The parietal pleura
is supplied and drained by systemic vessels. The lymph of the
pleural space is drained by stoma in the parietal pleura, which
represents the predominant-if not exclusive- mechanism by
which liquid is cleared from the pleural space [33, 34]. The
parietal pleura has abundant sensory innervation and should be
well anesthetized before it is manipulated or punctured [35].
Although the quantity of pleural fluid is small, it efficiently
couples the lung to the diaphragm and chest wall during breath-
ing and lubricates the movement of those structures. Neverthe-
less, little or no functional impairment results when the pleural
space is obliterated either experimentally or because of clinical
necessity [1].
Anatomic anomalies of the pleura are rarely of clinical con-
sequence but can cause confusing radiological patterns [36].
Accessory fissures are very frequently encountered at surgery
or post mortem, but only two types are commonly encountered
in practice. The inferior accessory fissure separates the medial
basal segment of the right lower lobe (or the medial subsegment
of the anteromedial basal segment of the left lower lobe) from
the other basal segments of the lower lobe. Such fissures occur
in of people, usually in incomplete forms invagi-
nating the lower lobe at its diaphragmatic aspect. Superior
accessory fissures are present in 30% of patients. These vari-
ant fissures are roughly horizontal and separate the superior
segment of a lower lobe from the basal segments of that lobe.
They may mimic a horizontal fissure on a chest radiograph.
Pathophysiology
The diagnosis and treatment of bacterial empyema are best
understood in relation to the altered anatomy and pathological
physiology of the pleura and the associated host defense dys-
functions. Pleural effusions develop because of increased hy-
drostatic pressure or decreased oncotic pressure associated with
cardiac, renal, hepatic, or metabolic disease [I, 2]. Other factors
contributing to their development include alterations in pleural
permeability due to noninfectious inflammatory diseases, infec-
tion, toxic injury, malignancy, or trauma [37-41]. The pleural
space is normally sterile yet readily colonized once pleural
fluid has accumulated. Host factors predisposing patients to
empyema include pneumonia and parapneumonic effusions as
well as contiguous infections of the esophagus, mediastinum,
or subdiaphragmatic areas that may extend to the pleura. Both
traumatic and iatrogenic injury to adjacent structures may lead
to secondary infection and involvement of the pleura [37-
41]. Similarly, retropharyngeal, retroperitoneal, vertebral, or
paravertebral infection can extend to the pleura.
Pleural effusions are nutritionally rich culture media in which
WBC defenses are severely impaired. The classic studies of
Wood and co-workers showed that effective phagocytosis of
bacteria by neutrophils requires a structure upon which WBCs
can move and can ingest bacteria prior to development of spe-
cific antibodies [42]. Later in the course of infection, phagocy-
tosis is enhanced by antibodies and opsonic factors. However,
in a fluid-filled environment, bacteria can float away from
phagocytic cells and multiply relatively unimpeded [42]. In
current parlance this defect reflects the fact that "white cells
can't jump" (or swim) and thus cannot efficiently fulfill their
host defense function in a liquid medium, whether in the in-
fected pleura, pericardium, joint, or meninges.
The formation of an empyema has been arbitrarily divided
into an exudative phase, during which pus accumulates; a fi-
bropurulent phase, during which fibrin deposition and locula-
tion of pleural exudate occurs; and an organization phase, dur-
ing which fibroblast proliferation and scar formation cause lung
entrapment [43]. Prompt diagnosis and intervention should cir-
cumvent the second and third phases of empyema formation.
To achieve this goal, physicians need to appreciate the subtle-
ties of clinical expression of pleural empyema and the adverse
effects of the suppurative environment on antimicrobial effi-
cacy and tissue injury in the pleural space.
Bacteria in pleural fluid elicit a complex series of host defense
responses that are incompletely understood despite significant re-
cent advances in our knowledge of the role ofTNF, the cytokine
cascade, and perturbations of endothelial cell and leukocyte inter-
actions during infection[44, 45]. When the inflammatoryresponse
is too little or too late, bacteria may multiply until they reach a
stagnant growthphase, associatedwith concentrationsof
bacteria per mL [46]. Empyema fluid is relatively deficient in
opsoninsand complement andbecomes progressivelymore acidic,
hypoxic, and depleted of glucose as infection proceeds [46,47].
Gram-negative aerobic bacilli may release endotoxins, and strep-
tococci or staphylococci may release enzymes that lyse granulo-
cytes in pleural fluid.
During the inflammatory process, leukocytes release intra-
cellular constituents such as bactericidal permeability-
increasing protein, defensins, lysozyme, cationic proteins, lac-
toferrin, and zinc-binding proteins [48]. The latter two compo-
nents may contribute to suppression of bacterial growth by

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

em 1996;22 (May) Pleural Empyema 749
Table 1. Conditions associated withnontuberculous bacterial empy-
ema [31, 38-41, 60].
Microbial Pathogens
In approximately one-half of patients, empyema develops as
a complication of pneumonia (table I). Therefore, the fre-
Experimental Empyema
Animal models of pleural empyema lack many of the fea-
tures of human disease [55, 56]. Empyema in man is usually
monomicrobial, whereas it is difficult to produce disease in
animals without injection of multiple pathogens and concomi-
tant use of foreign bodies like umbilical tape. Empyema did
not occur after tape placement and injection of guinea pigs
with 4 log10 cfu ofBacteroidesfragilis; however, similar prepa-
rations and injection with 4 IOglO cfu of Staphylococcus aureus
produced empyema in 20% of animals, and concomitant injec-
tion with both B. fragilis and S. aureus produced empyema in
>50% [55]. More than 6 IOglO cfu of E. coli and B. fragilis
are required to produce empyema in 50% ofanimals. Umbilical
tape did not affect lethality of disease induced by E. coli and
B. fragilis, but addition of blood did increase lethality in that
model [56].
Empyema has been produced in rabbits by injection of Strep-
tococcus pneumoniae or Klebsiella pneumoniae into a pleural
exudate induced by turpentine [57]. Those lesions will heal
spontaneously and therefore do not appear analogous to human
disease. That model has been used to assess the effect of strep-
tokinase injection on experimental empyema. Although strepto-
kinase effectively reduced the incidence of adhesion, it in-
creased the volume of effusion, possibly because pleural fluid
was not drained [58]. Shohet and co-workers used the turpen-
tine-induced empyema model to study gentamicin efficacy
against K. pneumoniae infection in the pleural space [59]. Cure
rates were reduced when animals were treated with gentamicin
alone, but 100% of animals were cured when placed in an
oxygen chamber, despite the fact that the pharmacokinetics of
gentamicin were unchanged. These studies add further proof
of the suppressive effect of the abscess environment on the
activity of aminoglycosides used as single-drug therapy.
lowering concentrations of iron and zinc. Pneumococci and
perhaps other organisms may undergo autolysis in overtly puru-
lent empyema fluid, thus accounting for a portion of the 12%-
18% rate of sterility of empyema fluid. Late in the course of
infection, the inflammatory response leads to loculation of pus
and occasionally to its spontaneous drainage by erosion through
the chest wall (i.e., empyema necessitatis, which currently rep-
resents failure of diagnostic, medical, and surgical care).
Bacteria within empyemas are relatively unresponsive to an-
tibiotics. In that milieu bacteria may release ,B-lactamase en-
zymes capable of degrading ,B-lactamase-susceptible ,B-lactam
antibiotics [49]. Similarly, microbial enzymes in pus may de-
grade chloramphenicol. Overtly purulent empyema fluid may
be quite acidic, even in the absence of esophageal rupture.
Since aminoglycoside incorporation by bacteria is ordinarily
oxygen-dependent and acid-inhibitable, aminoglycoside effi-
cacy is suppressed in the hypoxic and acidic milieu of pleural
empyema [50]. Furthermore, the calcium and magnesium con-
centrations in pus, the avid binding of aminoglycosides to the
DNA in pus, and the reduced bacterial metabolism in pus may
inhibit aminoglycoside activity in empyema fluid [50, 51].
Bacteria within abscesses or involved in chronic inflamma-
tory states multiply slowly, with generation times that may
reach 8-24 hours [52]. Tuomanen and co-workers found that
there was a direct relationship between the multiplication rate
of Escherichia coli and their death rate after exposure to cepha-
losporins in vitro-i.e., rapidly multiplying organisms were
killed quickly, whereas slowly growing organisms were killed
less rapidly in proportion to their growth rate [53]. When killing
curves were expressed in relationship to the doubling time of
the bacteria exposed to antibiotics, there was a linear relation-
ship between cell division and the rate at which bacteria were
killed by ,B-lactam agents [53].
The mechanisms by which growth rates of bacteria modify
their susceptibility to ,B-lactamantibiotics are incompletely un-
derstood. Stevens and colleagues demonstrated a progressive
reduction of penicillin-binding proteins in streptococci as they
entered a stagnant phase of growth [54]. It appears likely that
the rate of bacterial division affects the quantity and type of
penicillin-binding proteins that are available to interact with
,B-lactamantibiotics. This may in part explain why bacteria in
pus are refractory to antibiotics and why it is necessary to give
prolonged antibiotic therapy to patients with poorly drained,
suppurative infections [54].
Prolonged therapy may be needed because slowly growing
organisms in pus require prolonged contact with ,B-lactamanti-
biotics in order to induce sufficient cell wall injury to kill
bacteria. Fortunately, this impediment can be circumvented by
abscess drainage, which removes large numbers of metaboli-
cally inert bacteria and their toxins and removes inflammatory
components of the empyemic milieu that are capable of sup-
pressing bacterial responsiveness to antibiotics and injuring
host tissues. In addition, there are both new and better ways
to achieve adequate drainage of pleural pus [6- 9, 15- 21].
Cause
Pulmonary infection
Surgery
Trauma
Esophageal perforation
Complication of thoracentesis/chest tube placement
Subdiaphragmatic infection
Spontaneous pneumothorax
Septicemia
Other or unknown
Total
No. ('Yo)
of patients
301 (56)
119 (22)
20 (4)
21 (4)
21 (4)
15 (3)
7 (1)
8 (I)
30 (5)
542 (100)

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

750 Bryant and Salmon eID 1996; 22 (May)
Childhood Empyema
Table 2. Bacteria isolated from nontuberculous pleural empyema
fluid in various studies.
Nelson reported that 54% of the empyemas in children ,;;;6
months of age were caused by S. aureus, and only 6% were
sterile [75]. Empyemas in children in the age groups of
0.5-2 and 2-5 years were caused by S. aureus in 20%, by
S. pneumoniae in ~ 25%, and by H. influenzae in 20% and
Approximately one-quarter ofempyemas are associated with
trauma or surgery [61, 70]. As shown in table 2, there is a
disproportionate increase in staphylococcal infection and a de-
crease in anaerobic infection in such patients [70]. Ill-advised
or incomplete resection of lung nodules or cavities containing
cryptococci or spontaneous rupture of coccidioidomycosis-
associated lung cavities into the pleura may lead to fungal
empyemas. Similarly, instrumentation or surgery causing injury
or perforation of the esophagus or stomach may lead to
mediastinitis or subdiaphragmatic infection that can extend to
the pleura [2]. Sinus drainage from the skin and pleural involve-
ment are suggestive of infection caused by Actinomyces spe-
cies, Mycobacterium tuberculosis, or Nocardia species. Empy-
ema may also occur with Entamoeba histolytica infection but
is rare in the United States [71-74].
30
5
2
4
3
4
13
10
13
23 8
~
26 8
8
18 37
8
9 5
5 5
5
6 5
12 16
16
27
Percentage of patients with empyema
[reference]
In combined series Following trauma
[2,29,31] (n = 217) [6] (n = 31)
Aerobic
Streptococcus species
Streptococcus pneumoniae
Staphylococcus aureus
Staphylococcus epidermidis
Escherichia coli
Enterobacter species
Proteus species
Klebsiella species
Pseudomonas aeruginosa
Other gram-negative bacillus
Aerobic organisms only
Anaerobic
Bacteroides species
Clostridium species
Actinomyces species
Eubacterium species
Proprionibacterium species
Veillonella species
Fusobacterium species
Microaerophilic streptococci
Peptostreptococcus species
Anaerobic organisms only
No organisms
Bacteria isolated
quency with which certain microbes cause parapneumonic em-
pyema in different patient groups reflects the frequency with
which the vulnerable patients in those groups are exposed to,
become colonized with, and fail to clear aspirated secretions
containing those bacteria. Immunocompromised patients are
prone to pleural involvement with fungal or aerobic gram-
negative bacillary infection [30, 31,41,61]. In patients with a
malignancy, fungal or tuberculous foci may be reactivated and
empyema may develop. Similarly, fungal or mycobacterial em-
pyema may develop in transplant recipients and patients with
AIDS, but usually because of disseminated disease.
The microbe-specific factors favoring development of empy-
ema as a complication of pneumonia have special clinical rele-
vance. In overtly healthy adults, the bacteria most commonly
causing pleural empyema are S. aureus, S. pneumoniae, and
Streptococcus pyogenes [1, 62]. Although pneumococcal pneu-
monia may present with parapneumonic pleural effusions in
40% of patients, empyema occurs in ,;;;5% of patients with
pneumococcal pneumonia [1]. Group A streptococcal pneumo-
nia occurs much less frequently than pneumococcal pneumonia
but is associated with a higher frequency of large pleural effu-
sions that progress rapidly to produce empyema and sepsis
[62,63].
It is well appreciated that klebsiella pneumonia and empy-
ema may occur in alcoholic males with multiple host defense
defects that impair containment of or perception of disease
until it is well advanced [64, 65]. It is not clear which host
defense defects are the most important causes of gram-negative
bacillary pneumonia in such patients, but the proteolytic en-
zyme-mediated removal of fibronectin from the nasopharynx
and the subsequent ability of gram-negative bacilli to colonize
the exposed nasopharyngeal membranes are probably two of
the key determinants of ultimate infection [66]. Likewise, the
fetid mouth and a predisposition to aspiration are clearly the
forerunners of the fetid lung, lung abscesses, and/or anaerobic
empyema [2, 67, 68]. Such infections are usually polymicrobic
and linked to pyorrhea or gingivitis and altered consciousness.
Extensive local tissue injury and bacterial synergistic infection
are hallmarks of anaerobic pneumonia and empyema.
The frequency of aerobic and anaerobic isolates seen in three
combined series is shown in table 2 [2,29,31]. It is likely that
the role and frequency of anaerobic organisms are substantially
underestimated by such reports. Bartlett and Finegold found
exclusively anaerobic organisms in 35% of 83 medical service
patients with empyema, anaerobic plus aerobic pathogens in
41%, and aerobic pathogens alone in 23% [67].
S. aureus is a relatively common cause ofempyema in other-
wise healthy adults, in children, and in patients who have had
chest trauma or surgery. S. aureus pneumonia and empyema
have been linked to prior influenza A virus infection [69].
Empyema complicating traumatic hemothorax predisposes pa-
tients to infection with S. aureus, whereas pneumothoraces or
serous effusions are often secondarily infected by aerobic gram-
negative bacilli.

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

em1996;22 (May) Pleural Empyema 751
10%, respectively [75]. Among older children there was a 4-
7-fold higher incidence of empyema fluid culture sterility that
probably reflected autolysis of pneumococci or possibly death
of Haemophilus strains in purulent secretions.
Prior antibiotic therapy reduces the frequency of positive
cultures [76]. Hoff and co-workers reported that 71% of pa-
tients with sterile empyemas had received antibiotics before
cultures were performed, as compared with 41% of patients
whose empyema fluid contained viable bacteria (P < .05) [76].
These differences would probably be even more striking if the
susceptibility of specific bacteria, the potency and duration of
antibiotic therapy, and the problem of antibiotic "carryover"
into culture media could be subjected to multivariant analysis.
In the past, H. injluenzae infection has occurred principally in
children aged <6 to 8 years, and its incidence undoubtedly
will be reduced by the efficacy of the current Haemophilus
conjugate vaccine. Anaerobic lung and pleural infections are
rare in children [75].
Most of the diagnostic and therapeutic considerations with
regard to empyema are the same for children and adults. How-
ever, children more frequently have pneumatoceles and pneu-
mothoraces associated with staphylococcal infection, as well
as scoliosis as a complication of empyema. Because of the
lower incidence of severe underlying disease in children, they
are better candidates for early thorascopic intervention if antibi-
otic treatment, drainage procedures, and thrombolytic therapy
fail [75-77].
Clinical Presentation and Medical Evaluation
Patients with empyema require careful assessment of their
underlying diseases, the severity and duration of their infection,
and the microbes causing it. Physicians should seek host-
specific and epidemiologic information that helps to identify
risks of infection with specific pathogens. The need for urgent
treatment is linked to the severity of infection and to the sever-
ity of the patient's host defense defects, inasmuch as the sickest
and the most susceptible patients need the most rapid interven-
tion [39, 60, 78, 79].
Initial findings may be nonspecific, although otherwise-nor-
mal patients usually have chest pain, chills or fever, and night
sweats at a higher frequency than do patients with host defense
defects. Weight loss and general disability occur with more
indolent presentation. The occurrence of persistent fever, dia-
phoresis, and/or leukocytosis despite the administration of ef-
fective antibiotics should suggest the presence of an empyema
in patients with pulmonary or adjacent infection. Physical ex-
amination is remarkably nonspecific and may be limited to
findings of effusion. Ahigh index of suspicion and an apprecia-
tion of factors that predispose patients to development of empy-
ema facilitate its recognition (table 3).
It is difficult to demonstrate loculation of a pleural effusion
on physical examination, and the presence of a friction rub is
not distinctive. Radiographic demonstration of a pleural fluid
accumulation may depend on a volume of ~ 2 mL to broaden
the costophrenic angle [80]. Likewise, lateral views may show
a suggestive fluid meniscus, which can be obscured by the
presence of infiltrates or overlapping of the diaphragmatic
shadows. Lateral decubitus views facilitate recognition of
smaller volumes of fluid and can be utilized in the intensive
care unit, where patients are sicker and less tolerant of being
moved. Positional changes permit recognition of the extent of
parenchymal infection and may reveal loculated fluid (refer to
the section on radiology of empyema).
Light has proposed useful criteria for the assessment and
management of parapneumonic effusions and empyema [I, 80]
(table 4). He classifies effusions into seven categories of in-
creasing purulence and configurational complexity. Class 1
consists of parapneumonic collections that are < 1 em in width,
as measured on a lateral decubitus chest radiograph. These can
usually be managed medically and do not require fluid aspira-
tion as long as the patient is doing well and there is radiographic
evidence of improvement. Higher classes require increasing
degrees of intervention. It is important to determine fluid char-
acter and detect loculations promptly and accurately. If thora-
centesis reveals fluid that is culture-negative and devoid of
microorganisms on microscopic examination (class 2 or 3),
then antibiotics with or without serial thoracenteses may be
used. Tube drainage is reserved for smear-positive collections
or those with overt purulence (classes 4 to 7). The presence of
an empyema (classes 6 and 7) is demonstrated by pleural pus
and indicates the need for tube drainage and thrombolytic ther-
apy.
Pleural fluid that is not clearly purulent on inspection should
be cultured and analyzed by chemical, physical, and micro-
scopic procedures. Infected parapneumonic effusions are ini-
tially thin and serous but become more purulent as disease
progresses. The pH, glucose, and lactic dehydrogenase levels
and WBC count in the fluid should be determined, and appro-
priate microscopic examinations of stained smears should be
performed [10, 1\, 80, 81]. Empyemas usually have a pH value
of <7.00, a glucose level of <40 mg/dL, and a lactic dehydro-
genase level of >1,000 IU/L [12-14, 81]. If analysis of a
parapneumonic effusion reveals these biochemical features
and/or gram-stain positivity, then tube drainage should be per-
formed.
If determined properly, the acid pH value correlates best
with the extent and stage of infection and is more helpful
than a neutrophil count or microscopic estimate of purulence.
Improper handling of pleural fluid specimens can cause spuri-
ous elevation of the pH value or a reduction in the glucose
value. Therefore, specimens need to be tightly capped and kept
on ice until tested. Overtly purulent empyemic effusions are
acidic and viscous. Values of5.5 to ~ 6 8 can be found in cases
ofchronic loculated empyemas [82]; comparable pH values can
be found in empyemas caused by esophageal rupture [82].
However, low pH is neither a specific nor a sensitive indicator
of esophageal rupture into the pleural space. Abbott et al. found

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

752 Bryant and Salmon em 1996;22 (May)
Table 3. Clinical, radiological, and laboratory clues to the possibility of pleural empyema.
Clinical clues
History. Chills, fever, dyspnea, chest pain, or referred pain; recent pulmonary or contiguous infection in the oropharynx, mediastinum, or subdiaphragmatic
area; symptoms suggesting adjacent tissue infection extending to the pleura, i.e., dysphagia, dyspepsia, hiccups, or pharyngeal, abdominal, back, or shoulder
pain; recent instrumentation, surgery, or trauma of the chest, oropharynx, esophagus, or abdomen; delayed or incomplete response to appropriate medical
therapy for an infection that could extend to the pleura; comorbid diseases such as alcoholism, malnutrition, immunodeficiency, immunosuppression, or
diabetes.
Physical examination. Diminished breath sounds or basilar dullness to percussion; pleural friction rub; bronchophony or egophony above effusion or
adjacent to pneumonia; tracheal or mediastinal shift; scoliosis following a respiratory infection (in children); focal chest wall heat, erythema, swelling, and/or
pain (rare); draining dermal sinuses (rare); hyperpyrexia, shock, tachypnea (>30 respirations/min), and lor altered consciousness (all of which may be
indicative of disproportionately severe infection).
Clinical course. Rapid onset of clinical deterioration and sepsis with respiratory failure; persistent fever, sepsis, and/or organ failure despite appropriate
antibiotic therapy (in a susceptible patient); worsening clinical and laboratory indicators of infection despite appropriate antibiotic therapy.
Radiological clues (imaging method)
Pleural fluid (conventional or lateral decubitus radiography); pleural effusion loculation (lateral decubitus radiography, ultrasonography, or CT); evidence of
pleural effusion and contiguous infection (ultrasonography or CT); pleural mass (conventional radiography); hemothorax, pleural air, amoebic abscess, or
contiguous infection extending to pleura; bronchopleural fistula and empyema (CT).
Laboratory clues
Pleural fluid. Cloudy, bloody, or purulent; WBC count, ;;>50,000 X 10
9IL
(usually); pH level, ",7.1 (or ;;>0.3 lower than serum pH); lactic dehydrogenase
level, ;;>1,000 lUlL; glucose level, <40 mg/dL; positive smear stains or cultures; fetid (-% of anaerobic empyemas).
Findings indicating severe infection. Neutropenia or neutrophilia with immature forms; hypoxia (partial pressure [arterial] of O
2
, ",60 mm Hg); azotemia,
anemia, or acidosis; thrombocytopenia or disseminated intravascular coagulopathy; multiorgan failure; polymicrobic infection or infection with S. aureus,
aerobic gram-negative bacilli, or anaerobic organisms; amoebic pleural empyema, with parasites seen on smear; overtly purulent tuberculous empyema, with
high-density acid-fast bacilli seen on smear.
the pH value of empyema fluid to be <6.0 in 6 of 10 patients
with a perforated esophagus, but in the other 4 patients it was
neutral or alkaline [83].
The notion that empyema pH values of <6.0 are suggestive
of a ruptured esophagus could be misleading unless considered
in the context of how long the patient has been ill [84]. Al-
though chronic, well-localized empyemas can occasionally
have pH values of <6.0, such levels would not be expected
to occur in newly formed lesions [46, 82]. However, pleural
fluid accumulation due to a ruptured esophagus might be ex-
pected to have a low pH value after a relatively brief illness.
This distinction is critically important because the mortality
associated with a >24-hour delay in treatment of esophageal
rupture is ;;;.50% [80]. The pleural fluid amylase level is ele-
vated in cases of esophageal rupture and helps to confirm that
diagnosis [84, 85]. Imaging and endoscopic procedures are
especially helpful in this setting.
Serum pH measurement can help to assess the significance
of low pleural fluid pH values that reflect systemic acidosis.
Pleural fluid pH values that are at least .3 pH units less than
serum pH values support the need for a drainage procedure
[1]. Spurious elevation of empyema fluid pH values may occur
in patients with urea-splitting proteus infections [86]. Malodor-
ous empyema fluid suggests the presence of anaerobic infection
but is present in only about two-thirds of anaerobic empyemas
[2, 87]. Demonstration of high levels of pleural fluid protein
or specific gravity is rarely helpful.
Microbes may be seen on gram-stained empyema fluid
that is sterile. In other instances organisms are neither seen
in nor grown from frank pus. An acridine orange stain is
occasionally helpful for identifying bacteria whose gram-
stain morphology is distorted by prior antibiotic therapy.
Legionella pneumophila is not well visualized by gram stain
but can be detected by direct fluorescent microscopy or by
culture. Testing urine for Legionella antigen is probably the
most sensitive test for pneumonia caused by L. pneumophila
serogroup 1 [88]. Patients at risk offungal empyema require
appropriate smears and cultures of empyema fluid for detec-
tion of fungi.
Serological tests may assist in the diagnosis of histoplas-
mosis or coccidioidomycosis. Disseminated histoplasmosis
involving the pleura of patients with AIDS may be diagnosed
by serum or urine antigen detection [89]. Similarly, Aspergil-
lus antigen quantitation may be useful in the diagnosis of
aspergillus infection involving the pleura in compromised
patients [90]. Patients suspected of having amebiasis should
undergo CT studies for identification of subdiaphragmatic
disease as well as serological testing for disseminated extra-
intestinal amebiasis [91, 92].
Pleuropulmonary amebiasis may develop after erosion of an
amebic liver abscess through the diaphragm, in association with
sudden respiratory distress, cough, and pleurisy [71]. The lung
may be involved, in which case a hepatobronchial fistula and
amebae visible in copious bronchial secretions may be noted.

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

ern 1996;22 (May) Pleural Empyema 753
Table 4. Classification and treatment of parapneumonic effusions and empyema.
Pleural fluid indices of suppuration Mode of therapy required
Width of Bacteria Loculation(s) Thora-
effusion Lactic in gram of pleural
Clinical status coscopy Full
on lateral dehydro- stain fluid noted :+: decorti-
Class of effusion or decubitus Glucose genase and/or Frank in radiologic Stable or Anti- Needle Tube Throm- decorti- cation
empyema radiograph pH (mgldL) (lUlL) culture pus studies improving Sepsis biotics aspiration drainage bolysis cation procedure
I: Nonconsequential
pleural
effusion <I em NO NO NO NO + +
2: Parapneumonic
effusion >1 em ;;,7.2 >40 <1,000 + + +*
Inflammatory
effusions
3: Minimally
complex 7-7.2 >40 >1,000 + + +*
4: Moderately
complex .;7.0 <40 >1,000 + + + + + +*
5: Extremely
complex <7.0 <40 >1,000 :+:/:+: Multiple + + + +
+*t :+:t
Empyema
6: Not complex NN NN NN ~ / : : t + :+: + + + + +*'
+t
7: Complex NN NN NN / + Multiple + + + + +*' +' :+:
NOTE. Table is modified from [1]. NO ~ not determined; NN ~ determinationnot necessary; - ~ no; + ~ yes; ~ possibly.
* Repeat as needed.
'More invasive procedures may be required if patient's condition fails to respond.
This presentation is associated with a mortality of - 30%,
whereas the mortality associated with serous effusion without
abscess rupture is <5% [71,92]. Since 98%-99% of patients
with amebiasis have detectable antibodies to E. histolytica,
serological diagnosis is often definitive [91]. For patients at
risk of nocardial infection of the lung and pleura, modified
acid-fast stains of purulent secretions should be done and the
laboratory should be notified to hold plates for at least 2 weeks.
For patients with smear- and culture-negative pleural effusions
that appear purulent, the possibility of chylous effusions should
be excluded by the testing of fluid for neutral fat, pH, and
sedimentation values after centrifugation at 5,000g [1].
Pus will have an acid pH and cell fragments will sediment,
whereas chylous effusions will have a neutral pH and remain
opaque after centrifugation. Primary pleural eosinophilia is a
rare condition suggesting paragonimiasis and can be diagnosed
by demonstration of parasites in stool, sputum, or bronchoscopy
secretions and by the finding of elevated serum antibody titers
[73]. Levels of IgE and IgG antibodies to Paragonimus west-
ermani may be significantly higher in pleural fluid than in
serum [93].
Pleural tuberculosis can be confirmed by acid-fast smears of
pleural fluid in fewer than one-quarter of cases but can be
diagnosed by pleural biopsy and culture in >90% of patients
[94]. Chronic pleural tuberculosis may cause platelike pleural
calcification. Liquid culture media and the rapid radiomimetic
culture techniques often provide proof of tuberculosis within
2 weeks. Demonstration of pleural fluid adenosine deaminase
levels of >70 U/L supports the diagnosis of pleural tuberculo-
sis, but the test for that is not available in the United States
[1]. The diagnostic utility of PCR detection of mycobacterial
antigen in pleural fluid is still under investigation. Skin test
conversion and symptoms of weight loss, night sweats, and
fever as well as epidemiologic and sociologic risks of tubercu-
losis are important diagnostic clues.
Pleural fluid from patients with collagen-vascular disease,
subdiaphragmatic infections, malignancy, or pancreatitis will
occasionally mimic bacterial empyemic fluid. Pleural effusions
of rheumatologic or pancreatic origin are rarely frankly exuda-
tive and can be distinguished by serological tests for rheuma-
toid factor or pleural fluid amylase level in most instances.
Rheumatoid effusions usually have an antinuclear antibody titer
of ~ 1:160 or a rheumatoid factor titer of ~ 1:320 and occasion-
ally contain lupus erythematosus cells [95, 96]. Malignant effu-
sions with a pH value of <7 are readily diagnosed by cytologi-
cal examination and are associated with a poor prognosis [82].
Radiology of Empyema
Chest radiography remains the important first study for pa-
tients with pleural disease [97]. As little as 25 mL of pleural
fluid can elevate the hemidiaphragm radiographically, but
blunting of the posterior costophrenic sulcus usually requires
-200 mL [1, 80]. If effusions are "free-flowing" volumes, as
little as 5 mL can be detected on the lateral decubitus view
[98]. The newer modalities of ultrasonography and CT have
greatly facilitated diagnosis and treatment ofpatients with para-
pneumonic effusions and empyema.

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

754 Bryant and Salmon em 1996;22 (May)
Ultrasonography
Ultrasonographic devices are widely available, provide real-
time guidance for thoracentesis or pleural catheter placement
[4], and can be transported to the bedside of unstable or criti-
cally ill patients. This imaging adjunct is particularly useful
for sampling fluid that does not layer freely on decubitus films,
and it reduces the incidence of pneumothorax during thoracen-
tesis [99, 100]. The sonographic appearance of pleural fluid
collections is quite variable, ranging from anechoic (completely
echo-free or sonolucent) to very echogenic. When highly echo-
genic, the collections may be mistaken for consolidated lung
or pulmonary abscess [101]. In such instances it is important
to coordinate sonographic and radiographic interpretations.
Sonography can distinguish solid from liquid pleural abnor-
malities with 92% accuracy (vs. the 68% accuracy of chest
roentgenography). With combined use of radiography and so-
nography, the accuracy rises to 98% [8]. The ability of ultraso-
nography to detect variation in the shape ofpleural fluid collec-
tions during respiration is helpful in excluding a solid lesion.
Similarly, evidence of "fluid bronchograms" in cases of con-
solidated lung is another distinguishing feature detectable by
ultrasonography [102].
Discrete intrapleural septations can be demonstrated sono-
graphically in up to 74% of exudative effusions [103], and
some may appear mobile on real-time examination [104]. Ultra-
sonography may show limiting membranes suggesting the pres-
ence of loculated collections, even when they are invisible
by CT. The presence of septations has prognostic importance
because loculated collections (Light's class 5 or higher; see
table 4) require drainage and are usually larger than simple
collections [105]. Anechoic collections may be exudative or
transudative [7].
Computed Tomography
The development of rapid, newer-generation CT scanners
has revolutionized the evaluation and treatment of thoracic
empyema. Empyemas usually appear well-defined, smooth, and
round or elliptical on CT scans. Their margins are composed of
inflamed visceral and parietal pleura that often have a markedly
thickened appearance and enhance after administration of intra-
venous contrast material. The visceral and parietal layers are
separated by the interposed empyema fluid, giving rise to the
"split pleura sign" of empyema [106]. When air is introduced
into the empyema cavity, either iatrogenically following thora-
centesis or in association with a bronchopleural fistula, the
inner aspect of the visceral and parietal margins is usually
smooth. The extrapleural or subcostal fat external to the thick-
ened parietal pleura and deep to the ribs is also noted to thicken
in both acute and chronic empyema. This clearly discernible
fatty hyperplasia has imaging characteristics similar to those
of subcutaneous fat and is much lower in CT attenuation than
the thickened pleura itself. Conventional chest radiographs can-
not distinguish "pleural thickening" that reflects pleural fluid
accumulation from that due to accentuation of this fatty layer.
Empyema is frequently associated with nearby pulmonary
consolidation and sometimes lung abscess. Alternatively, a
lung abscess can resemble effusion or empyema [106]. Differ-
entiation between these diagnostic possibilities is often diffi-
cult, if not impossible, with use of clinical and conventional
radiographic approaches. Fortunately, CT usually allows defin-
itive diagnosis. Lung abscesses are often poorly defined,
roughly spherical, and surrounded by irregularly consolidated
lung. They often contain one or more cavities with shaggy
intramural contours. When abutting a pleural surface, abscesses
form acute angles with the adjacent chest wall. Because they
arise within and occupy consolidated lung, they rarely appear
to displace adjacent pulmonary structures such as peripheral
airways and vessels. Empyemas may form acute or obtuse
angles yet have the other CT characteristics mentioned in the
preceding paragraph.
Treatment
Effective therapy for an empyema requires control of infec-
tion, drainage of pus, and expansion of the lung. Occasionally,
procedural or surgical correction of adjacent infection is re-
quired. Empirical antimicrobial therapy is initiated on the basis
of its anticipated bactericidal activity against the suspected
microbial pathogens and is changed when the susceptibilities
of the infecting microorganism(s) are known.
Drug delivery to the pleura is not a problem. In general, (3-
lactam agents are given in high doses for 2-4 weeks, but
therapy may need to be prolonged if drainage is not optimal
or if an adjacent abscess or osteomyelitis is present. Nafcillin
is the drug of choice for S. aureus infection, and penicillin
is the drug of choice for penicillin-susceptible streptococcal
infections. Infection due to S. pneumoniae with high-level resis-
tance to penicillin (MIC, >2 flg/mL) and to ceftriaxone or
cefotaxime (MIC, flg/mL) should be treated with vancomy-
cin. Cephalosporin-susceptible pneumococci with intermediate
susceptibility to penicillin should be treated with ceftriaxone
or cefotaxime.
Monotherapy with an aminoglycoside is contraindicated be-
cause of its poor activity in pus and the risks of toxicity [49].
The synergistic activity of aminoglycosides with (3-lactam
drugs has justified their use in combination therapy for Pseu-
domonas aeruginosa, Enterobacter cloacae, Serratia marces-
cens, and Acinetobacter calcoaceticus infections, because (3-
lactam antibiotics appear to overcome the suppressive effect
ofthe hypoxic and acidic abscess environment on incorporation
of the aminoglycoside by bacteria [50]. Ciprofloxacin is a logi-
cal alternative to aminoglycosides for use in combined therapy
against those pathogens in early infection and can be given
orally during later stages of infection. lmipenem is the drug of
choice for E. cloacae infection and should be used in conjunc-

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

em 1996;22 (May) Pleural Empyema 755
tion with at least one other effective antibiotic for treatment of
empyema caused by that pathogen.
Patients with chronic pleural empyemas that are poorly
drained may require prolonged antibiotic therapy. Similarly,
empyemas caused by Actinomyces or Nocardia species, myco-
bacteria, or fungi require protracted therapy. Anaerobic empy-
ema can be treated like anaerobic suppurative infection at other
sites; however, metronidazole is minimally active against strep-
tococci, Actinomyces, and propionibacteria and less apt to be
effective for partially drained infections because it is not metab-
olized to its active derivative in partially oxygenated environ-
ments [50].
Since many anaerobic infections are polymicrobic, therapy
is usually selected from among clindamycin, ,8-lactamase-
inhibitor combination drugs, and imipenem on the basis of the
specific concomitant aerobic and/or microaerophilic organisms
present in pus. Neither chloramphenicol nor tetracycline should
be used to treat anaerobic empyema. Anaerobic and polymi-
crobic anaerobic empyemas that have extended to the pleura
from adjacent sites of infection often require surgical treatment
of the primary site of infection. Patients with pleuropulmonary
amebiasis should receive metronidazole (usually 750 mg po
thrice daily for 10 days), and drainage should be performed as
appropriate [71]. Severely ill patients may also require brief
therapy with dehydroemetine or chloroquine [71, 92]. Those
with polymicrobic pleuropulmonary amebiasis should receive
antibiotics appropriate for the concomitant bacteria and may
require large-chest-tube drainage if the empyema fluid is espe-
cially thick. The treatment of choice for pleuropulmonary para-
gonimiasis is praziquantel [73].
If the causative organisms are susceptible, tuberculosis-
related pleural effusions respond well to the usual antitubercu-
lous regimens and rarely require drainage. In practice, the or-
ganisms' susceptibilities may not be known until patients are
well into their second month of treatment with a four-drug
regimen of isoniazid, rifampin, ethambutol, and pyrazinamide.
Thereafter, patients infected with susceptible strains receive
isoniazid and rifampin for 4 months.
Patients with multidrug-resistant tuberculosis require indi-
vidualized regimens based on antimicrobial susceptibility find-
ings [107]. If possible, all patients should receive directly ob-
served therapy. Patients with pleuropulmonary rifampin-
resistant tuberculosis require ~ 18 months of therapy with two
or more effective drugs. Frankly purulent tuberculous empyema
is (fortunately) rare and usually follows a long history ofunsuc-
cessful medical and/or surgical therapies. The pleura is usually
quite thick, is occasionally calcified, and often has high concen-
trations of mycobacteria. Therapy should be initiated with re-
peated thoracentesis and multidrug regimens [107, 108]. Tube
drainage should be avoided in order to prevent secondary bacte-
rial infection of a tuberculous empyema [109]. Control of infec-
tion may require decortication or tailoring procedures such as
thoracoplasty or surgical correction of associated bronchopleu-
ral fistulae [l08].
Empyema Drainage
Drainage of pus is still a major component of adequate treat-
ment of pleural empyema. Serous pleural fluid that is not locu-
lated, is devoid of microorganisms on microscopic examina-
tion, or has ambiguous indicators of suppuration (such as a
high pH, a glucose concentration of ~ mg/dL, and a lactic
dehydrogenase level of < 1,000 lUlL) can be treated with anti-
biotics and reevaluated, with repetition of thoracentesis in
12-18 hours (table 4). Patients with loculated fluid, frank pus,
or smear-positive purulent fluid with a pH of <7.0, glucose
concentration of <40 mgldL, and lactic dehydrogenase level
of > 1,000 lUlL require drainage procedures. Repeated thora-
centesis is rarely successful in such cases. Small-bore percuta-
neous catheters can be used if fluid is serous and thin.
A chest tube with an underwater seal can be placed by
the pulmonologist, the radiologist, or the surgeon and can be
expected to provide successful drainage in two-thirds of pa-
tients with easily accessible, nonloculated fluid collections.
When no response occurs within 24 hours, loculation should
be excluded by ultrasonography and urokinase should be in-
fused to enhance drainage. Urokinase should be used in all
patients who are considered at risk for multiple loculations or
especially thick or viscous pus, and its infusion can be repeated
daily as needed for several weeks.
Recently published reports support an expanded role for the
interventional radiologist in the management of empyema [35,
llO-112]. Imaging guidance allows precise placement of
small-bore catheters into small pockets of unusual configura-
tion that were previously difficult to drain with tube thoracos-
tomy alone. Thus, 16-French or smaller catheters can be placed
safely with much less discomfort and morbidity for the patient
and are at least as efficacious as 28-French or larger surgical
chest tubes in empyema therapy [27, 35, 113-115]. For the
occasional collection of very thick pus, a 24-French catheter
can be placed over a guide wire by a Seldinger technique [9].
Blind, bedside drainage of empyema via surgical tube thoracos-
tomy that is performed with reference only to the patient's
chest radiograph is often unsuccessful. A large fraction of such
patients may require further invasive treatment [31, 116]. Rela-
tively high morbidity and even mortality as high as 5% have
been reported [60].
Identification of loculation and the extent of an empyema
by CT allows optimal planning for sampling and drainage of
the fluid collections. An appropriate site of percutaneous entry
can be selected and the skin marked for thoracentesis. The
decision to proceed with catheter drainage depends on fluid
characteristics and the size and configuration of the empyema,
and it is facilitated by reference to the scheme proposed by
Light (table 4) [1].
Performance of the drainage procedure under direct CT guid-
ance is most convenient if the empyema collection is small
[112] or in a position that would otherwise be difficult to
access (such as anterior, medial, or intrafissural). Postoperative

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

756 Bryant and Salmon em 1996;22 (May)
empyemas are often small and loculated [110] and therefore
ideal for treatment with small catheters under CT guidance.
Alternatively, after placement of skin markers, drainage can
be completed under fluoroscopy. Some authors favor the use
of ultrasonography in management of pleural fluid collections,
especially those that are large [4, 9, 27, 100]. Ultrasonography
is usually quicker and more convenient to perform but is more
operator-dependent than CT. Empyema collections occasion-
ally may be difficult to identify or to distinguish from nearby
consolidated lung by ultrasonography if they contain air or
thick pus with extensive echogenic debris [27].
Several recent reviews provide detailed descriptions of the
techniques for percutaneous pleural catheter placement [27, 35,
110, 113]. Two general approaches can be taken. The first is
a direct trocar technique, in which the catheter is advanced
under imaging guidance over a stiff coaxial cannula, entering
the collection at the site of a preliminary diagnostic thoracente-
sis [27]. The second method uses a modified Seldinger tech-
nique, in which an 18-gauge needle is placed into the collection
under imaging guidance and a guide wire is advanced through
the needle; serial subcutaneous fascial dilations are then per-
formed over the guide wire with dilators ofprogressively larger
diameter, and finally the catheter is placed and anchored to a
skin dressing.
The choice of technique may be individualized to the pa-
tient's requirements. Collections with tenacious pus and thick
margins often require a direct trocar placement to provide the
mechanical advantage needed for entry and to prevent buckling
of the catheter in the subcutaneous tissues. Trocar placement
is also generally quick to perform and can be done without
assistance. Exchanges over a guide wire require two operators.
Fluid is aspirated through the catheter, followed by local irriga-
tion with saline until clear. The catheter is then attached to a
standard underwater-seal drainage system for continuous suc-
tion at the bedside.
Transcatheter Intrapleural Thrombolytic Therapy
Despite catheter placement and drainage of empyema fluid,
patients may still have residual pockets of undrained fluid and
display signs and symptoms of persistent infection [60, 80].
This is not surprising, since empyema fluid is often loculated.
The act of placing a catheter over a guide wire is helpful in
breaking down at least some loculations, but additional mea-
sures may be required to achieve complete drainage. One ap-
proach has been to place additional catheters [4, 27, 35, 110,
113]. In addition, there has been renewed interest in thrombo-
lytic therapy.
Loculation(s) may form early during development of either
complicated parapneumonic effusion or empyema [1]. Their
appearance indicates that the effusion has progressed to the
fibropurulent stage, with extensive deposition of fibrin on the
pleural surfaces. Pleural fibrosis will occur unless the loculated
collections are drained and appropriate antibiotics are given.
Strange and colleagues used an animal model of pleural empy-
ema to show that the initial dense fibrin layer began to be
replaced by a network of connective tissue elements by the
fifth day [58]. When fibrin was not removed promptly, fibrinous
strands became firmly anchored to the pleural surfaces. The
authors postulated that fibrin deposition enhances the develop-
ment of subsequent fibrosis by creating a diffusional barrier to
oxygen, since pleural fluid hypoxia and lactic acidosis have
been shown to enhance fibroblast collagen production in an
empyema [58]. These findings support the clinical urgency of
draining empyema fluid to prevent the formation of intrapleural
fibrosis, which often requires surgical extirpation. It also pro-
vides a rationale for therapies specifically directed at prevention
of deposition and removal of intrapleural fibrin early in the
course of an empyema.
Streptokinase and streptodornase, derived from streptococcal
sources, were first used to help drain loculated pleural pus by
Tillett and Sherry in 1949 [117]. Initial enthusiasm was later
dampened by concern about allergic reactions to the agents.
Currently, there is renewed enthusiasm for intrapleural fibri-
nolysis in cases of complicated parapneumonic effusions and
empyema, in part because of the availability of urokinase,
which is nonantigenic and nonpyrogenic [15, 16, 19, 20]. Al-
though "purified" forms of streptokinase are now available
[17,21], these have been documented to produce an antibody
response [118]. The antibody response is responsible for the
febrile reaction to the agent, which may falsely mimic persis-
tence of empyemas [20]. Urokinase is produced by the human
kidney and does not cause an allergic or febrile reaction.
We usually perform intrapleural urokinase instillation ac-
cording to methods of Moulton et al. [19] and Robinson et al.
[20]. Each dose consists of 100,000 units given in 100 mL of
0.9% sterile saline solution. The 100-mL aliquot is left in the
pleural space for at least 2 hours. The catheter or chest tube is
then "unclamped" and suction restored. This process can be
repeated as needed.
Robinson and colleagues suggest that the instillation begin
in the evening and continue overnight while the patient sleeps
[20]. Unused portions of the urokinase solution should be re-
frigerated between instillations. While it is possible to give
urokinase by transmurally injecting the solution into an in-
dwelling surgical chest tube, it is more efficient to introduce
it through a radiologically placed pleural catheter via a three-
way stopcock, which is then turned off in the direction of the
catheter. Use of the catheter ensures that the agent reaches and
stays within the pleural collection. The larger "dead space"
of surgical chest tubes decreases the effective dose instilled
into the pleura and may decrease the efficacy of a given instilla-
tion. The pleural catheter may be repositioned as needed (see
figure 1).
Use of urokinase in the pleural space is safe. Systemic side
effectshave not beenreported. The total dose givenintothe pleural
cavity is approximately one-tenththe dose of the agent commonly
given intravascularly to lyseclots.Urokinasehasan averageserum

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

em 1996;22 (May) Pleural Empyema 757
half-life of 20 minutes. The presence of a bronchoplcural fistula
is considered by some authors to contraindicate the use of in-
trapleural urokinase [16, 19]. However, no adverse effects associ-
ated with its use in that setting have been documented, Indeed,
empyema and bronchopleural fistula have been effectively man-
aged without complications with use of streptokinase [18]. Uroki-
nase therapy is much less expensive than surgical debridement and
may successfully circumvent the added morbidity and mortality
associated with thoracoscopy or thoracotomy [20, 119].
The need for further closed tube drainage is assessed by
quantitation of the volume expelled daily and the size of the
pleural cavity. Drainage of <50 mUd and cavities <50 mL
in size are indications for tube withdrawal [10).
When closed chest tube drainage fails, thoracostomy, decor-
tication, or open chest tube drainage may be necessary; such
cases are usually managed surgically. Failure of closed tube
drainage may be due to extensive disease, prior trauma, or
surgery. Hematomas usually require extraction [120] but on
occasion have been successfully lysed with urokinase [19].
There has recently been enthusiasm for video-assisted thora-
costomy decortication as an alternative to prolonged closed
chest tube drainage or full decortication. This approach is par-
ticularly well suited for otherwise healthy children or young
adults whose condition is identified early in its course, and
the procedure should be done shortly after thrombolysis fails.
Thoracostomy should probably be considered in the second
week of illness rather than later, because adhesions may limit
natural dissection planes thereafter. Traditional teachings sug-
gest that formal decortication should be performed prior to the
third week or after the sixth week of empyema formation to
minimize lung injury attributed to a poorly demarcated pleural
peel. That concept has been questioned and would undoubtedly
be modified by the degree oflung entrapment, extent of pleural
adhesions, and response to thrombolysis [121, 122].
Video-Assisted Thoracoscopy for Mechanical
Debridement of Intrapleural Loculations
When a multiloculated empyema fails to respond to in-
trapleural urokinase, the use of video-assisted thoracoscopic
surgery (VATS) may provide a significant new alternative to
thoracotomy with decortication. The development ofminiatur-
ized video technology has made it possible to perform this less
invasive procedure under general anesthesia [22]. The thoraco-
scope uses a light-sensitive silicon chip to generate a real-time
image on a television monitor. It is introduced into the pleural
space through a cylindrical conduit. One or two additional
small (1.0-1.5 em long) skin incisions are made to allow pas-
sage of surgical instruments used under thoracoscopic guidance
and for placement of a chest tube.
VATS can disrupt intrapleural adhesions and achieve com-
plete drainage of loculated effusions refractory to intracavitary
thrombolytic therapy [119, 121-123]. The procedure can usu-
ally be done with endotracheal intubation and general anesthe-
sia, without unilateral (contralateral) ventilation [121). In-
trapleural septations are mechanically lysed under direct
thoracoscopic control. The pleural space is then copiously irri-
gated with saline. When the space is evacuated, the pleura can
be inspected by videothoracoscopy to determine if organized
septa and/or an organized pleural peel is present, indicating the
need for a more extensive procedure or full thoracotomy for
decortication [122, 124). VATS has been proven safe and use-
ful for evacuating empyema and avoiding full thoracotomy in
adult and pediatric patients [122, 124].
Bronchopleural fistulae rarely heal spontaneously and usu-
ally require surgical closure. CT has greatly facilitated their
recognition. Those not healing with closed tube drainage may
be treated with muscle flap transposition, in which the muscle
is used to obliterate the empyema cavity and is sutured directly
to the bronchus or adjacent area.
There are several surgical approaches for chronic empyema
or empyema following pneumonectomy. Rib resection, decorti-
cation, empyemectomy, or permanent external drainage proce-
dures are selected on the basis of the ability of the patient to
tolerate the procedure and on the anticipated likelihood of its
success. Approaches vary by region and personal experience
of the surgeon [30, 39, 125). Postpncumonectomy empyema
may be treated with intracavitary antibiotic solutions, open-
window thoracostomy (as devised by Eloesser for those unable
to tolerate more extensive procedures), or obliteration of empy-
ema cavities by muscle flaps. Thoracoplasty is usually the pro-
cedure of last resort.
A complicated parapneumonic infection adjacent to a neo-
plastic bronchial obstruction may require a special approach.
Traditional techniques can be applied if radiation or chemother-
apy can relieve bronchial obstruction. However, if that is not
possible, chest tube drainage is likely to continue for the dura-
tion of the patient's life and represents a substantial disability
and nuisance for the patient. In some instances this can be
managed with prolonged antibiotic therapy that suppresses
symptoms without eradicating the disease [1].
Prognosis
The morbidity and mortality associated with pleural empy-
ema are affected by the microbial etiology, host defense de-
fects, severity and duration of infection, and adequacy of antibi-
otic therapy and drainage. Patients with polymicrobic or
resistant gram-negative bacillary empyema often are elderly,
have nosocomial infection complicating severe underlying dis-
ease, and may die at rates in the range of 40%-70% [29, 39).
Among otherwise healthy young patients, mortality rates are
2%-15%, depending on the duration and severity of their infec-
tion [30, 76].
Patients with inadequately drained empyemas often die
[126]. Therefore, early cardiothoracic surgical consultation
and a team approach are required. Ultrasonography and CT
should be used to facilitate early recognition of pleural space

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

758 Bryant and Salmon em 1996;22 (May)
Figure 1. This case illustrates the use of ultrasonography and CT in identifying and characterizing thoracic empyema, as well as the use of
imaging-guided catheter drainage and transcatheter intracavitary urokinase for managing empyema. A. contrast-enhanced CT image shows
pleural collection. No septations are visible within the pleural collection since they are the same density as the pleural liquid. B. Ultrasonography
reveals dependent, echogenic debris in the effusion and a network of clearly defined limiting membranes forming loculations. The large
loculation on the right was targeted for thoracentesis and catheter placement. C and D, Posteroanterior and lateral chest radiography was done
following catheter placement and the immediate removal of 150 mL of fluid. Only minimal additional fluid was removed over the next 24
hours, during which time the catheter remained folded within the loculation. Urokinase was then introduced through the catheter to improve
drainage. E and F, chest radiographs obtained after the first urokinase treatment show that the walls of the loculation have been lysed, permitting
the catheter to uncoil. G and H, serial urokinase treatments and drainages were done over 48 hours and the catheter was repositioned. These
maneuvers yielded an additional 1,200 mL of empyema fluid. (Reprinted with permission from Mandell GL, ed. Atlas of infectious diseases.
Philadelphia: Current Medicine, 1996 [in press].)

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

em 1996; 22 (May) Pleural Empyema 759
infection. Patients with compromised host defenses are espe-
cially vulnerable to the adverse effects of undrained pus
(e.g., malnutrition, sepsis, and mult iorgan failure) and there-
fore are in urgent need of adequate drainage early in the
course of their infection [13, 81, 127, 128]. Delay in diagno-
sis generally correlates with an adverse outcome [30, 76, 81,
126]. Hoff and co-workers found that the mean duration of
illness prior to the hospitalization of children for empyema
was 4.8 days for those cured with antibiotics alone, 5.8 days
for those requiring chest tube drainage, and 8.0 days for
those requiring decorti cation [76]. They found scol iosis of
"",5 in 44% of children presenting with empyema. None of

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

760 Bryant and Salmon CID 1996;22 (May)
those children were cured by antibiotics alone, and 17 of 27
required decortication. Children with scoliosis secondary to
empyema had been ill for an average of 7.3 days prior to
admission. Pleural thickening or opacification of a hemitho-
rax correlated with poor prognosis in their series.
There is controversy over the best criteria and techniques for
performance of pleural fluid drainage [126, 127]. The approach
suggested by Light is supported in principle by many authors
(table 4). In general, early intervention by the least noxious
means is preferred [1,13,76,81,126]. Storm and co-workers
found that daily infusion of intrapleural antibiotics and irriga-
tion with saline over a 2-week period reduced the need for rib
resection or decortication to 6% among 51 patients on their
medical service [128]. During that period, 77% of 43 patients
treated on a surgical service in that hospital required resection
or decortication [128]. There are insufficient data to evaluate
this approach, and further studies are needed.
Posttraumatic empyema has a poor prognosis, and patients
appear to benefit from early decortication if sepsis is poorly
contained despite antibiotic therapy, iflung entrapment impairs
ventilatory function, or if pleural drainage is inadequate after
2 weeks of therapy [39, 70]. Tube drainage is rarely successful
for management of an infected hemothorax because clots ob-
struct the tube.
There is a vicious cycle of delayed diagnosis and therapy
for patients with multiple underlying host defense defects,
nosocomial infection, and multiply resistant organisms that ad-
versely affects prognosis. Such patients may have fewer signs
and symptoms of their disease, have increased complications
of malnutrition and multiple organ failure, and respond poorly
to medical and surgical therapy.
References
1. Light RW. Parapneumonic effusions and empyema. In: Light RW, ed.
Pleural diseases. 3rd ed. Baltimore: Williams & Wilkins, 1995:
129-53.
2. Bartlett JG, Finegold SM. Anaerobic infections of the lung and pleural
space. Am Rev Respir Dis 1974; 110:56-77.
3. Finland M, Barnes MW. Duration of hospitalization for acute bacterial
empyema at Boston City Hospital during 12 selected years from 1935
to 1972. J Infect Dis 1978; 138:520-30.
4. O'Moore PV, Mueller PR, Simeone JF, et a1. Sonographic guidance in
diagnostic and therapeutic interventions in the pleural space. AJR Am
J Roentgenol1987; 149:1-5.
5. Mirvis SE, Tobin KD, Kostrubiak I, Belzberg H. Thoracic CT in detecting
occult disease in critically ill patients. AJR Am J Roentgenol
1987; 148:685-9.
6. van Sonnenberg E, Nakamoto SK, Mueller PR, et al. CT- and ultrasound-
guided catheter drainage of empyemas after chest-tube failure. Radiol-
ogy 1984; 151:349-53.
7. Yang P-C, Luh K-T, Chang D-B, Wu H-D, Yu C-l, Kuo SoH. Value of
sonography in determining the nature of pleural effusion: analysis of
320 cases. AJR Am J RoentgenoI1992; 159:29-33.
8. Lipscomb Dl, Flower CDR, Hadfield JW. Ultrasound of the pleura: an
assessment of its clinicaJ value. Clinical Radiology 1981; 32:289-90.
9. Boland GW, Lee MJ, Silverman S, Mueller PRo Interventional radiology
of the pleural space. Clinical Radiology 1995; 50:205-14.
10. Sahn SA. Management of complicated parapneumonic effusions. Am
Rev Respir Dis. 1993; 148:813-7.
II. Potts DE, Levin DC, Sahn SA. Pleural fluid pH in parapneumonic effu-
sions. Chest 1976; 70:328-31.
12. Poe RH, Marin MG, Israel RH, Kallay Me. Utility of pleural fluid
analysis in predicting tube thoracostomy/decortication in parapneu-
monic effusions. Chest 1991; 100:963-7.
13. Houston MC. Pleural fluid pH: diagnostic, therapeutic, and prognostic
value. Am J Surg 1987; 154:333-7.
14. Larsen W1. Human embryology. New York: Churchill Livingstone, 1993.
15. Lahorra 1M, Haaga JR, Stellato T, Flanigan T, Graham R. Safety of
intracavitary urokinase with percutaneous abscess drainage. AJR Am
J Roentgenol 1993; 160:171-4.
16. Lee KS, 1m J-G, Kim YH, Hwang SH, Bae WK, Lee BH. Treatment
of thoracic multiloculated empyemas with intracavitary urokinase: a
prospective study. Radiology 1991; 179:771-5.
17. Lysy Y, Gavish A, Lieberson A, Werczberger A, Reiten R, Dudai M.
Intrapleural instillation of streptokinase in the treatment of organizing
empyema. Isr 1 Med Sci 1989;25:284-7.
18. Mitchell ME, Alberts WM, Chandler KW, Goldman AL. Intrapleural
streptokinase in management of parapneumonic effusions: report of
series and review of literature. 1 FI Med Assoc 1989;76:1019-22.
19. Moulton JS, Moore PT, Mencini RA. Treatment of loculated pleural
effusions with transcatheter intracavitary urokinase. AJR Am J Roent-
geno11989; 153:941-5.
20. Robinson LA, Moulton AL, Fleming WH, Alonso A, Galbraith TA.
Intrapleural fibrinolytic treatment of multiloculated thoracic empyemas
Ann Thorac Surg 1994;57:803-14.
21. Taylor RFH, Rubens MB, Pearson MC, Barnes NC. Intrapleural streptoki-
nase in the management of empyema. Thorax 1994;49:856-9.
22. Kaiser LR, Shrager 18. Video-assisted thoracic surgery: the current state
of the art. AJR Am J Roentgenol 1995; 165:1111-7.
23. O'Brien J, Cohen M, Solit R, et a1. Thoracoscopic drainage and decortica-
tion as definitive treatment for empyema thoracis following penetrating
chest injury. J Trauma 1994; 36:536-40.
24. Symbas PN. Chest drainage tubes. Surg Clin N Am 1989;69:41-6.
25. Peters RM. Empyema thoracis: historical perspective. Ann Thorac Surg
1989;48:306-8.
26. Miller KS, Sahn SA. Chest tubes: indications, technique, management
and complications. Chest 1987;91:258-64.
27. Silverman SG, Mueller PR, Saini S, et a!. Thoracic empyema: manage-
ment with image-guided catheter drainage. Radiology 1988; 169:5-9.
28. Orringer MB. Thoracic empyema-back to basics [editorial]. Chest
1988;93:901-2.
29. Mandai AK, Thadepalli H. Treatment of spontaneous bacterial empyema
thoracis. J Thorac Cardiovasc Surg 1987;94:414-8.
30. Mayo P. Early thoracotomy and decortication for nontuberculous empy-
ema in adults with and without underlying diseases: a 25-year review.
Am Surgeon 1985;51:230-6.
31. Varkey B, Rose HD, Kutty CPK, Politis 1. Empyema thoracis during a
ten-year period: analysis of 72 cases and comparison to a previous
study (1952 to 1967). Arch IntemMed 1981;141:1771-6.
32. 1m J-G, Webb WR, Rosen A, Gamsu G. Costal pleura: appearances at
high-resolution CT. Radiology 1989; 171:125- 31.
33. Broaddus VC, Wiener-Kronish JP, Berthiaume Y, Staub Ne. Removal
of pleural liquid and protein by lymphatics in awake sheep. J Appl
PhysioI1988;64:384-90.
34. Shinto RA, Light RW. Effects of diuresis on the characteristics of pleural
fluid in patients with congestive heart failure. Am J Med 1990; 88:
230-4.
35. Westcott JL. Percutaneous catheter drainage of pleural effusion and em-
pyema. AJR Am J Roentgenol1985; 144:1189-93.
36. Godwin JD,Tarver RD. Accessory fissures of the lung. AJR Am J Roent-
genol 1985; 144:39-47.

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

eID 1996; 22 (May) Pleural Empyema 761
37. Light RW. Physiology of the pleural space. In: Light RW, ed. Pleural
diseases. 3rd ed. Baltimore: Wiliams & Wilkins, 1995:7-17.
38. Weese WC, Shindler ER, Smith 1M, Rabinovich S. Empyema of the
thorax then and now: a study of 122 cases over four decades. Arch
Intern Med 1973; 131:516-20.
39. Yeh TJ, Hall DP, Ellison RG. Empyema thoracis: a review of 110 cases.
Am Rev Respir Dis 1963;88:785-90.
40. Snider GL, Saleh SS. Empyema of the thorax in adults: review of 105
cases. Dis Chest 1968;54:410-5.
41. Smith JA, Mullerworth MH, Westlake GW, Tatoulis J. Empyema thor-
acis: a 14-year experience in a teaching center. Ann Thorac Surg
1991; 51:39-42.
42. Wood WB Jr, Smith MR, Watson B. Studies on the mechanism of'recov-
ery in pneumococcal pneumonia. IV. The mechanism of phagocytosis
in the absence of antibody. J Exp Med 1946; 84:387 -402.
43. Andrews NC, Parker EF, Shaw RP, et al. Management of nontuberculous
empyema. Am Rev Respir Dis 1962;85:935-6.
44. Dinarello CA, Gelfand JA, Wolff SM. Anticytokine strategies in the
treatment of the systemic inflammatory response syndrome. JAMA
1993;269: 1829-35.
45. Light RW. Clinical manifestations and useful tests. In: Light RW, ed.
Pleural diseases. 3rd ed. Baltimore: Williams & Wilkins, 1995:36-74.
46. Bryant RE. Pus: friend or foe? In: Root RK, Trunkey DD, Sande MA,
eds. Contemporary issues in infectious diseases. Vol 6. New surgical
and medical approaches. New York: Churchill Livingstone, 1987:
31-48.
47. Lew PD, Zubler R, Vaudaux P, Farquet JJ, Waldvogel FA, Lambert
P-H. Decreased heat-labile opsonic activity and complement levels
associated with evidence of C3 breakdown products in infected pleural
effusions. J Clin Invest 1979;63:326-34.
48. Densen P, Clark RA, NauseefWM. Granulocytic phagocytes. In: Mandell
GL, Bennett JE, Dolin R, eds. Mandell, Douglas and Bennett's princi-
ples and practice of infectious diseases. 4th ed. Vol l. New York:
Churchill Livingstone, 1995:78-101.
49. Bryant RE. Effect of the suppurative environment on antibiotic activity.
In: Root RD, Sande MD, eds. Contemporary issues in infectious dis-
eases. Vol 1. New dimensions in antimicrobial therapy. New York:
Churchill Livingstone, 1984:313 - 37.
50. Bryant RE, Fox K, Oh G, Morthland VH. ,B-Lactam enhancement of
aminoglycoside activity under conditions of reduced pH and oxygen
tension that may exist in infected tissues. J Infect Dis 1992; 165:
676-82.
51. Vaudaux P, Waldvogel FA. Gentamicin inactivation in purulent exudates:
role of cell lysis. J Infect Dis 1980; 142:586-93.
52. Zak 0, Sande MA. Correlation of the in vitro antimicrobial activity of
antibiotics with results of treatment in experimental animal models
and human infection. In: Sabath LD, ed. Action of antibiotics in pa-
tients. Bern, Switzerland: Hans Huber, 1992:55-67.
53. Tuomanen E, Cozens R, Tosch W, Zak 0, Tomasz A. The rate of killing
of Escherichia coli by ,B-Iactam antibiotics is strictly proportional to
the rate of bacterial growth. J Gen Microbiol 1986; 132: 1297-304.
54. Stevens DL, Yan S, Bryant AE. Penicillin-binding protein expression at
different growth stages determines penicillin efficacy in vitro and in
vivo: an explanation of the inoculum effect. J Infect Dis
1993; 167:1401-5.
55. Mavroudis C, Ganzel BL, Cox SK, Polk HC Jr. Experimental aerobic-
anaerobic thoracic empyema in the guinea pig. Ann Thorac Surg
1987;43:298-302.
56. Mavroudis C, Ganzel BL, Katzmark S, Polk HC Jr. Effect of hemothorax
on experimental empyema thoracis in the guinea pig. J Thorac Cardio-
vase Surg 1985;89:42-9.
57. Sahn SA, Taryle DA, Good JT Jr. Experimental empyema: time course
and pathogenesis ofpleural fluid acidosis and low pleural fluid glucose.
Am Rev Respir Dis 1979; 120:355-61.
58. Strange C, Tomlinson JR, Wilson C, Harley R, Miller KS, Sahn SA. The
histology of experimental pleural injury with tetracycline, empyema,
and carrageenan. Experimental and Molecular Pathology 1989;
51:205-19.
59. Shohet I, Yellin A, Meyerovitch J, Rubinstein E. Pharmacokinetics and
therapeutic efficacy of gentamicin in an experimental pleural empyema
rabbit model. Antimicrob Agents Chemother 1987;31:982-5.
60. Sherman MM, Subramanian V, Berger RL. Management of thoracic
empyema. Am J Surg 1977; 133:474-9.
61. Vianna NJ. Nontuberculous bacterial empyema in patients with and with-
out underlying diseases. JAMA 1971;215:69-75.
62. Welch CC, Tombridge TL, Baker WJ, Kinney RJ. Beta-hemolytic strepto-
coccal pneumonia: report of an outbreak in a military population. Am
J Med Sci 1961;242:157-65.
63. Braman SS, Donat WE. Explosive pleuritis: manifestation of group A
beta-hemolytic streptococcal infection. Am J Med 1986; 81:723-6.
64. Homes RB. Friedlander's pneumonia. Am J RadioI1956;75:728-47.
65. Torres A, Serra-Batlles J, Ferrer A, et al. Severe community-acquired
pneumonia: epidemiology and prognostic factors. Am Rev Respir Dis
1991; 144:312-8.
66. Mackowiak PA, Martin RM, Jones SR, Smith JW. Pharyngeal coloniza-
tion by gram-negative bacilli in aspiration-prone persons. Arch Intern
Med 1978; 138:1224-7.
67. Bartlett JG, Gorbach SL, Thadepalli H, Finegold SM. Bacteriology of
empyema. Lancet 1974; 1:338-40.
68. Bartlett JG. Anaerobic bacterial pleuro-pulmonary infections. Seminars
in Respiratory Medicine 1992; 13:158-66.
69. Kaye MG, Fox MJ, Bartlett JG, Braman SS, Glassroth J. The clinical
spectrum of Staphylococcus aureus pulmonary infection. Chest
1990; 97:788-92.
70. Caplan ES, Hoyt NJ, Rodriguez A, Cowley RA. Empyema occurring in
the multiply traumatized patient. J Trauma 1984;24:785-9.
71. Kubitschek KR, Peters J, Nickeson D, Musher DM. Amebiasis presenting
as pleuropulmonary disease. West J Med 1985; 142:203-7.
72. Ibarra-Perez C. Thoracic complications of amebic abscess of the liver:
report of 501 cases. Chest 1981;79:672-7.
73. Skerrett SJ, Plorde JJ. Parasitic infections of the pleural space. Seminars
in Respiratory Medicine 1992; 13:242-58.
74. Thompson JE Jr, Forlenza S, Verma R. Amebic liver abscess: a therapeu-
tic approach. Rev Infect Dis 1985; 7:171-9.
75. Nelson JD. Pleural empyema. Pediatr Infect Dis 1985;4(3)(suppl):
S31-3.
76. Hoff SJ, Neblett WW, Edwards KM, et al. Parapneumonic empyema in
children: decortication hastens recovery in patients with severe pleural
infections. Pediatr Infect Dis J 1991; 10:194-9.
77. Kern JA, Rodgers BM. Thoracoscopy in the management of empyema
in children. J Pediatr Surg 1993;28:1128-32.
78. Pothula V, Krellenstein OJ. Early aggressive surgical management of
parapneumonic empyemas. Chest 1994; 105:832-6.
79. Cham CW, Haq SM, Rahamim J. Empyema thoracis: a problem with
late referral? Thorax 1993;48:925-7.
80. Light RW. Management of empyema. Seminars in Respiratory Medicine
1992; 13:167-76.
81. Himelman RB, Callen PW. The prognostic value of loculations in para-
pneumonic pleural effusions. Chest 1986;90:852-6.
82. Good JT Jr. Taryle DA, Maulitz RM, Kaplan RL, Sahn SA. The diagnos-
tic value of pleural fluid pH. Chest 1980; 78:55-9.
83. Abbott OA, Mansour KA, Logan WD Jr, Hatcher CR Jr, Symbas PN. A
traumatic so-called "spontaneous" rupture of the esophagus: a review
of 47 personal cases with comments on a new method of surgical
therapy. J Thorac Cardiovasc Surg 1970;59:67-83.
84. Dye RA, Laforet EG. Esophageal rupture: diagnosis by pleural fluid pH.
Chest 1974;66:454-6.
85. Sherr HP, Light RW, Merson MH, Wolf RO, Taylor LL, Hendrix TR.
Origin of pleural fluid amylase in esophageal rupture. Ann Intern Med
1972; 76:985-6.

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

762 Bryant and Salmon ern 1996;22 (May)
86. Pine JR, Hollman JL. Elevated pleural fluid pH in Proteus mirabilis
empyema. Chest 1983;84:109-11.
87. Sullivan KM, O'Toole RD, Fisher RH, Sullivan KN. Anaerobic empyema
thoracis: the role of anaerobes in 226 cases of culture-proven empy-
emas. Arch Intern Med 1973; 131:521-7.
88. Ruf B, Schurmann D, Horbach I, Fehrenbach FJ, Pohle 00. Prevalence
and diagnosis of Legionella pneumonia: a 3-year prospective study
with emphasis on application of urinary antigen detection. J Infect Dis
1990; 162:1341-8.
89. Wheat LJ, Connolly-Stringfield PA, Baker RL, et al. Disseminated histo-
plasmosis in the acquired immune deficiency syndrome: clinical find-
ings, diagnosis, and treatment, and review of the literature. Medicine
(Baltimore) 1990;69:361-74.
90. Patterson TF, Miniter P, Patterson JE, Rappeport JM, Andriole VT. As-
pergillus antigen detection in the diagnosis of invasive aspergillosis.
J Infect Dis 1995; 171:1553-8.
91. Adams EB, MacLeod IN. Invasive amebiasis. I. Amebic dysentery and
its complications. Medicine (Baltimore) 56; 1977:315-23.
92. Eggleston FC, Handa AK, Verghese M. Amehic peritonitis secondary to
amebic liver abscess. Surgery 1982;91:46-8.
93. Ikeda T, Oikawa Y, Owhashi M, Nawa Y. Parasite-specific IgE and IgG
levels in the serum and pleural effusion of paragonimiasis westermani
patients. Am J Trop Med Hyg 1992;47:104-7.
94. Levine H, Metzger W, Lacera D, Kay 1. Diagnosis of tuberculous pleu-
risy by culture of pleural biopsy specimen. Arch Intern Med
1970; 126:269- 71.
95. Halla JT, Schrohenloher RE, Volanakis JE. Immune complexes and other
features of pleural effusions: a comparison of rheumatoid arthritis,
systemic lupus erythematosus, and other diseases. Am Intern Med
1980; 92:748-52.
96. Good JT Jr, King TE, Antony VB, Sahn SA. Lupus pleuritis: clinical
features and pleural fluid characteristics with special reference to pleu-
ral fluid antinuclear antibodies. Chest 1983;84:714-8.
97. Bryant RE. Pleural effusion and empyema. In: Mandell GL, Bennett JE,
DolinR, eds. Mandell, Douglas and Bennett's principles and practice
of infectious diseases. 4th ed. Vol I. New York: Churchill Livingstone,
1995:637-41.
98. Moskowitz H, Platt RT, Schachar R, Mellins H. Roentgen visualization
of minute pleural effusion: an experimental study to determine the
minimum amount of pleural fluid visible on a radiograph. Radiology
1973; 109:33-5.
99. Seneff MG, Corwin RW, Gold LH, Irwin RS. Complications associated
with thoracentesis. Chest 1986;90:97-110.
100. Harnsberger HR, Lee TG, Mukono DH. Rapid, inexpensive real-time
directed thoracentesis. Radiology 1983; 146:545-6.
101. Landay MJ, Conrad MR. Lung abscess mimicking empyema on ultraso-
nography. AJR Am J Roentgenol 1979; 133:731-4.
102. Dome HL. Differentiation ofpulmonary parenchymal consolidation from
pleural disease using the sonographic fluid bronchogram. Radiology
1986; 158:41-2.
103. Hirsch JH, Rogers N, Mack LA. Real-time sonography of pleural opaci-
ties. AJR Am J Roentgenol1981; 136:297-301.
104. Marks WM, Filly RA, Callen PW. Real-time evaluation of pleural lesions:
new observations regarding the probability of obtaining free fluid.
Radiology 1982; 142:163-4.
105. Lee MJ, Saini S, Brink JA, et al. Interventional radiology of the pleural
space: management of thoracic empyema with image-guided catheter
drainage. Seminars in Interventional Radiology 1991;8:29-35.
106. Stark DD, Federle MP, Goodman PC, Podrasky AE, Webb WR. Differ-
entiating lung abscess and empyema: radiography and computed to-
mography. AJR Am J Roentgenol1983; 141:163-7.
107. Iseman MD, Madsen LA. Chronic tuberculous empyema with bron-
chopleural fistula resulting in treatment failures and progressive drug
resistance. Chest 1991; 100:124-7.
108. Neihart RE, Hof DG. Successful nonsurgical treatment of tuberculous
empyema in an irreducible pleural space. Chest 1985; 88:792-4.
109. Magovern CJ, Rusch VW. Parapneumonic and post-traumatic pleural
space infections. Chest Surg Clin North Am 1994;4:561 -82.
110. Merriam MA, Cronan Jl, Dorfman GS, Lambiase RE, Haas RA. Radio-
graphically guided percutaneous catheter drainage of pleural fluid col-
lections. AJR Am J Roentgenol1988; 151:1113-6.
III. Goldberg MA, Mueller PR, Saini S, et al. Importance of daily rounds by
the radiologist after interventional procedures of the abdomen and
chest. Radiology 1991; 180:767-70.
112. Stavas J, van Sonneberg E, Casola G, Wittich GR. Percutaneous drainage
of infected and noninfected thoracic fluid collections. J Thorac Imaging
1987;2:80-7.
113. Hunnam GR, Flower CDR. Radiologically-guided percutaneous catheter
drainage of empyemas. Clin Radiol 1988; 39: 121-6.
114. Kerr A, Vasudevan VP, Powell S, Ligenza C. Percutaneous catheter
drainage for acute empyema: improved cure rate using CAT scan,
fluoroscopy, and pigtail drainage catheters. NY State J Med
1991; 91:4-7.
115. Ali I, Unruh H. Management of empyema thoracis, Ann Thorac Surg
1990; 50:355-9.
116. Light RW. Parapneumonic effusions and empyema. Clin Chest Med
1985;6:55-62.
117. Tillett WS, Sherry S. The effect in patients of streptococcal fibrinolysin
(streptokinase) and streptococcal desoxyribonuclease on fibrinous, puru-
lent, and sanguinous pleural exudations. J Clin Invest 1949;28:173-90.
118. Berglin E, Ekroth R, Teger-Nilsson AC, William-Olsson G. Intrapleural
instillation of streptokinase: effects on systemic fibrinolysis. J Thorac
Cardiovasc Surg 1981;29:124-6.
119. Huller JA, Harari D, Baimbridge MY. The management of empyema
thoracis by thoracoscopy and irrigation. Ann Thorac Surg
1985; 39:517 - 20.
120. Hoover LE, Hsu H-K, Ross MJ, et al. Reappraisal of empyema thoracis:
surgical intervention when the duration of illness is unknown. Chest
1986;90:511-5.
121. Silen ML, Weber TR. Thoracoscopic debridement of loculated empyema
thoracis in children. Ann Thorac Surg 1995; 59: 1166-8.
122. Ridley PD, Baimbridge MY. Thoracoscopic debridement and pleural
irrigation in the management of empyema thoracis. Ann Thorac Surg
1991; 51:461-4.
123. Ferguson MK. Thoracoscopy for empyema, bronchopleural fistula, and
chylothorax. Ann Thorac Surg 1993;56:644-5.
124. Moores DWO. Management of acute empyema [editorial]. Chest
1992; 102:1316-7.
125. Grant DR, Finley RJ. Empyema: analysis of treatment techniques. Can
J Surg 1985;28:449-51.
126. Sahn SA, Light RW. The sun should never set on a parapneumonic
effusion [editorial]. Chest 1989; 95:945- 7.
127. Berger HA, Morganroth ML. Immediate drainage is not required for all
patients with complicated parapneumonic effusions. Chest 1990;
97:731-5.
128. Storm HKR, Krasnik M, Bang K, Fridmodt-Moller N. Treatment of
pleural empyema secondary to pneumonia: thoracentesis regimen ver-
sus tube drainage. Thorax 1992;47:821-4.

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

763
OFFICE OF CONTINUING MEDICAL EDUCATION
UCLA SCHOOL OF MEDICINE
This test affords you the opportunity to assess your knowl-
edge and understanding of the material presented in the preced-
ing clinical article, "Pleural Empyema," by Bryant and
Salmon, and to earn continuing medical education (CME)
credit.
The Office of Continuing Medical Education, UCLA School
of Medicine, is accredited by the Accreditation Council for
Continuing Medical Education to sponsor continuing medical
education for physicians. The Office of Continuing Medical
Education, UCLA School ofMedicine, certifies that this contin-
uing medical education activity meets the criteria for 1 credit
hour in Category I of the Physician's Recognition Award of
the American Medical Association and the California Medical
Association Certificate in Continuing Medical Education.
To earn credit, read the State-of-the-Art Clinical Article care-
fully and answer the following questions. Mark your answers
by circling the correct responses on the answer card (usually
found toward the front of the issue), and mail the card after
affixing first-class postage. To earn credit, a minimum score
of 80% must be obtained.
Certificates of CME credit will be awarded on a per-volume
(biannual) basis. Each answer card must be submitted within
3 months of the date of the issue.
This program is made possible by an educational grant from
Roche Laboratories.
I. Which of the following is not a feature of the CT appear-
ance of pleural empyema?
A. Thickened visceral and parietal pleural layers
B. Pleural enhancement with intravenous contrast
C. Extrapleural fat hypertrophy
D. Collection appears to occupy rather than displace lung
E. Smooth inner pleural margins
2. Which of the following is the chief physiological mecha-
nism for fluid removal from the pleural space?
A. Lymphatic drainage through stoma located in the vis-
ceral pleura
B. Lymphatic drainage through stoma located in the pari-
etal pleura
C. Venous drainage through systemic veins
D. Venous drainage through pulmonary veins
E. Venous drainage through pulmonary and systemic
vems
3. What is the smallest quantity of pleural fluid shown to be
detectable on a lateral decubitus roentgenograph?
A. 5 mL
B. 25 mL
C. 100 mL
D. 250 mL
E. 1,000 mL
4. What quantity is usually required to cause a blunting of
the posterior costophrenic sulcus on an upright lateral chest
roentgenograph?
A. 5 mL
B. 25 mL
C. 250 mL
D. 500 mL
E. 1,000 mL
5. Which of the following parapneumonic pleural fluid char-
acteristics does not suggest the need for pleural fluid drain-
age?
A. pH, <7.1
B. Glucose, <40 mg/dL
C. Lactic dehydrogenase, >1,000 lUlL
D. Bacteria seen on gram stain microscopically
E. Protein, >3.5 g/dL
6. Which infectious cause of empyema is unlikely to necessi-
tate chest tube drainage?
A. Mycobacterium tuberculosis
B. Anaerobic organisms
C. Several species of microorganisms
D. Streptococcus pyogenes
E. Escherichia coli
7. Which infection is associated with marked pleural fluid
eosinophilia?
A. Tuberculosis
B. Anaerobic/polymicrobic infection
C. Entamoeba histolytica infection
D. Paragonimiasis
E. None of the above
8. Which of the following features of empyema fluid is not
helpful in excluding the disease causing pleural fluid accu-
mulation?
A. Pleural fluid amylase level-ruptured esophagus
B. Positive cytology-malignancy
C. Rheumatoid factor or antinuclear antibody titer-rheu-
matoid effusion
D. CT liver abscess-pleural amebiasis
E. pH of <6.5-chylothorax
9. Which of the following is most helpful for diagnosing
pleural tuberculosis in the United States?

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

764 CME Test em 1996;22 (May)
A. Acid-fast smear
B. Culture
C. Smear and culture of pleural biopsy specimen
D. Determination of adenosine deaminase level
10. Which of the following suggests that pleural fluid drainage
is not required?
A. Overt pus
B. Fetid odor
C. Loculation revealed by CT or ultrasonography
D. Persistent fever despite appropriate antimicrobial ther-
apy
E. None of the above

b
y

g
u
e
s
t

o
n

J
u
l
y

1
0
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m