parapneumonic empyema · chapter 58 parapneumonic empyema joseph s. mclaughlin and mark j. krasna...

CHAPTER 58

Parapneumonic Empyema

JosephS. McLaughlin and Mark J. Krasna

Pleuraleffusion is a common accompaniment of the inflam-mationof bacterial pneumonia. These "uncomplicated effu-sions"arenonpurulent,have a negative Gram's stain resultforbacteria,have negative results by culture, and are gener-allyfreeflowing.Light (1985, 1991, 2002) and associates(1972,1980),using biochemical parameters, noted a pHgreaterthan 7.30, a normal glucose level, and a lactic aciddehydrogenase(LDH) concentration less than 1,000 lUlL.Mostparapneumonic effusions resolve with appropriate an-tibiotictreatment and resolution of the pulmonary infection.

Thoracicempyema occurs when bacteria invade the nor-mallysterile pleural space. The process was described byIAndrewsand colleagues, reporting for the American Tho-

.cicSociety (ATS) in 1962, as a continuum of three stages'able58-1).Stage1 is characterizedby the presence of an exudative

.£fusionfrom increased permeability of the inflamed and:wollenpleuralsurfaces.This stage corresponds to the un-omplicatedparapneumonic effusion of Light and col-

gues(1980) and is initially sterile. Fibrin is deposited andIlymorphonuclearleukocytes are present in small num-rs.Withbacterial invasion, the process blends into the fi-opurulentstage 2, true empyema or Light's "complicated":!euraleffusion.Initially, the fluid is still relatively clear and

ow,but the white blood cell count is greater than 500,lispermicroliter, the specific gravity is greater than 1.018,dtheprotein level is greater than 2.5 gldL. The pH is lesshn7.2and the LDH levels reach 1,000 lUlL. Although fi-

depositsand early angioblastic and fibroblastic prolifer-onare seen in later phases of the exudative stage, these

essesaccelerate and heavy fibrin deposition takes placebothpleuralsurfaces,particularly the parietal pleura. The)Jsionbecomes purulent, with a white cell count above000cellsper IJL.Biochemically, the pH decreases to lev-below7.0, the glucose decreases to less than 50 mgldL,theLDH increases to greater than 1,000 lUlL. Stage 3. asearly as 1 week after infection with collagen orga-tionanddepositionon both pleural surfaces and entrap-Ioftheunderlying lung. This process is mature in 3 to 4

weeks, and the organized collagen on the pleural surface istermed a "peel." The effusion at this point is grossly puru-lent, and at least 75% of the volume is sediment on standing.Chronicity is characterized by dense fibrosis, contractionand trapping of the lung, atelectasis, and prolonged pulmo-nary infection and reduction of the size of the hemithorax.Fibrothorax with invasion of the chest wall and narrowing ofthe intercostal spaces may be thought of as the end stage ofthis process.

Complications of the empyema process may take placeearly or late. Necrosis of the visceral pleural surface, asnoted by Hankins and associates (1978), may result in bron-chopleural fistula heralded by sudden expectoration of, attimes, copious amounts of purulent sputum. Marks andEickhoff (1970) noted that the necrosis of the parietal pleuraand the chest wall and skin results in empyema necessitatis.These conditions are usually seen in patients treated with an-tibiotics for pneumonia who have unrecognized empyema.Often, these patients have persistent, low-grade fever untilthe empyema drains through the chest wall or into the lung.Rarely, osteomyelitis of the ribs or spine may occur. Inva-sion of the mediastinum with pulmonary esophageal fistulaand pericarditis have also been reported. Metastatic spreadis unusual, but ScheId (1998) has suggested that up to 12%of brain abscesses are from pleuropulmonary disease, in-cluding empyema.

COMMUNITY-ACQUIRED PNEUMONIA

According to the American Thoracic Society (2001),community-acquired pneumonia (CAP) is the sixth mostcommon cause of death in the United States and is the mostcommon infectious cause of death. There are approximately5.6 million cases of CAP in the United States each year, 1.1million requiring hospitalization. The outpatient mortalityrate ranges from 1% to 5%, and the overall mortality rate is12%. Guidelines for the evaluation and treatment of CAP

819

820 · XI. THE PLEURA

Table 58-1. American Thoracic SocietyClassification of Empyema

Stage 1. Exudative with swelling of the pleural membranesStage 2. Fibrinopurulent with heavy fibrin depositsStage 3. Organization with ingrowth of fibroblasts and deposition

of collagen

From Andrews NC, et al: Management of nontuberculousempyema: a statement of the subcommittee on surgery. Am RevRespir Dis 85:935, 1962.With permission.

using an "evidence-based" approach have been publishedby the American Thoracic Society (1995, 2001), Mandelland associates (2000) for the Canadian Infectious DiseasesSociety and the Canadian Thoracic Society, Bartlett and hisassociates (2000) for the Infectious Diseases Society ofAmerica, and the British Thoracic Society (1993). All ofthese official statements modify past practices and recom-mendations significantly.

As noted by Bartlett and associates (2000),up to 50% of thepathogens responsible for CAP are never identified despitecareful and extensive testing. There is no single test that canidentify all potential pathogens, and all diagnostic tests havetheir limitations. Sputum Gram's stains and cultures are oftenunable to identify Streptococcuspneumoniae and frequentlyencountered pathogens such as Mycoplasma pneumoniae,Chlamydiapneumoniae, and Legionella species and respira-tory viruses.These may be detectedat a later time by serologictesting. Furthermore, "atypical" infections can occur and in-volve either more than one bacterial species or a bacterialpathogen and a virus. Thus, an empiric approach to the initialtreatment of community-acquired infection is recommended,based on the patient's status as definedby Bartlett and Mundy(1995): (a) outpatientswith no historyof cardiopulmonarydis-ease and no modifying factors; (b) outpatientswith cardiopul-monarydisease [congestiveheart failureor chronicobstructivepulmonary disease (COPD)] or other modifying factors suchas risk factors for drug-resistant Streptococcus pneumoniae(DRSP) or Gram-negativebacteria; (c) inpatientsnot admittedto the intensivecare unit (ICU) who (i) have cardiopulmonarydisease and/or other modifying factors (includingresiding in anursingcare facility) or who (ii)have no cardiopulmonarydis-ease and no other modifyingfactors; and (d) ICU-admittedpa-tients with either risk or no risk for Pseudomonasaeruginosa.Antibiotic therapy is based on the probable bacterial etiologyof pneumonia in the four groups. Parenthetically,appropriateefforts shouldbe made to identify the agent or agents involvedand specificantibiotictherapydirected to their eradication.

Patients in group I (outpatients) are most commonly in-fected with Streptococcus pneumoniae, Mycoplasma pneu-moniae, Chlamydia pneumoniae, and respiratory viruses.Less common are Legionella species and Haemophilus in-fluenzae (in cigarette smokers). The mortality rate of thisgroup is less than 1% to 5%. Roughly 50% to 90% of thesepatients have no identifiable etiologic agents. Patients in

group II (outpatients with comorbidity) have congestiveheart failure or COPD, but no risk factors for DRSP or

Gram-negative bacteria (including residing in a nursing carefacility). Streptococcus pneumoniae remains the most likelypathogen, but DRSP is a consideration in this group andmodifies the antibiotic therapy. The mortality rate in group11is less than 5%, but according to Fine and associates (1996,

1997), as many as 20% of patients treated initially as outpa-tients will require hospitalization. Group III patients are ad-mitted to the hospital due to severe pneumonia. They aredivided into two groups: those with cardiopulmonary diseaseor modifying factors (including residing in a nursing care fa-cility) and those with no cardiopulmonary disease and no

modifying factors. Modifying factors and cardiopulmonarydisease increase the mortality rate significantly. Group lirApatients suffer mortality rates ranging from 5% to 25%.Group IV patients have severe pneumonia defined by thestudy group as those admitted to an ICU. They are dividedfurther into patients with no risk for Pseudomonas aerugi-nosa infection and those with risk for Pseudomonas aerugi-

nosa infection. Patients with severe CAP have a mortalityrate of up to 50% (Tables 58-2 through 58-5).

HOSPITAL-ACQUIRED PNEUMONIA

The Official Statement of the American ThoracicSociety(1995) noted that hospital-acquired pneumonia (HAP)is amajor cause of death within the hospital setting. It is gener-ally defined as pneumonia occurring 48 hours after admis-sion and excludes pneumonia existing prior to hospitaladmission. It is thought to occur at a rate of 5 to 10casesper 1,000hospital admissions with increases of 6 to 20timesthat rate among patients who are mechanically ventilated.Craven and Driks (1987), Craven and associates (1991),

Table 58-2. Community-Acquired Pneumonia:Group 1. Outpatients, No Cardiopulmonary Disease,

No Modifying Factors

Organisms Therapy

Streptococcus pneumoniaeMycoplasma pneumoniaeChlamydia pneumoniae

(alone or as mixedinfection)

Haemophilus injluenzaeRespiratory virusesMiscellaneousLegionella speciesMycobacterium tuberculosisEndemic fungi

From American Thoracic Society: Guidelines for the managementof adults with community-acquired pneumonia. Am J RespirCritCare Med 163:1732, 2001. With permission.

Advanced-generation macrolide:azithromycin, cIarithromycin,or doxycycline

Table 58-4. Community-Acquired Pneumonia: Group 3.Inpatients Not in Intensive Care Unit

Organisms

Cardiopulmonarydisease and/or modifying factors(including being from a nursing home)

Streptococcuspneumoniae (including DRSP)Haemophilus influenzaeMycoplasmapneumoniaeChlamydiapneumoniaeMixedinfection (bacteria plus atypical pathogen)Enteric Gram-negativesAspiration(anaerobes)VirusesLegionellaspeciesMiscellaneous

Mycobacterium tuberculosis, endemic fungi,Pneumocystis carinii

No cardiopulmonary disease, no modifying factorsS.pneumoniaeH. influenzaeM. pneumoniaeC.pneumoniaeMixedinfection (bacteria plus atypical pathogen)VirusesLegionellaspeciesMiscellaneous

M. tuberculosis, endemic fungi, P. carinii

Table58-3. Community-Acquired Pneumonia:Group2. Outpatient, with Cardiopulmonary Disease

and/or other Modifying Factors

Organisms Therapy

Streptococcuspneumoniae j3-lactam(oralcefpodoxime,(includingDRSP) cefuroxime, high-dose

Mycoplasmapneumoniae amoxicillin;or amoxicillin!Chlamydiapneumoniae clavulanate;or parenteralMixedinfection(bacteriaplus ceftriaxonefollowedby oral

atypicalpathogenor virus) cefpodoxime)Haemophilusinfluenzae plusEntericGram-negatives Macrolideor doxycyclineRespiratoryviruses orMiscellaneous Antipneumococcal fluoro-Moraxellacatarrhalis, quinolone (used alone)

Legionellaspecies,aspiration, (anaerobes),Mycobacterium

tuberculosis,endemicfungi

i:DRSP,drug-resistant Streptococcus pneumoniae.dapted from American Thoracic Society: Guidelines for the

managementof adults with community-acquired pneumonia. AmRespirCrit Care Med 163: 1735, 200 I. With permission.

Drossand associates (1980), Gross and Van Antwerpen1983),and Fagon and associates (1989) all found that pneu-oniais the second most common nosocomial infection in

e UnitedStates and has the highest mortality and morbidity'ates.The crude mortality rate approaches 70%. However,

58. PARAPNEUMONIC EMPYEMA · 821

many of these patients have other life-threatening conditions,and approximately two thirds of all deaths of patients withHAP can be attributed to these other conditions. The attribut-able mortality rate can be higher if bacteremia is present or ifthe etiologic agent is Pseudomonas aeruginosa or Acineto-bacter species.

Craven (1991), Rouby (1992), Bartlett (1986), andProd'hom (1994) and their associates as well as Schleup-ner and Cobb (1992) found that the bacteria most fre-quently associated with HAP are enteric Gram-negativebacteria and Staphylococcus aureus. Accumulated datasuggest that the etiology is polymicrobial in up to half ofmechanically ventilated patients. The role of viruses hasyet to be defined. As with CAP, initial diagnosis and treat-ment of HAP is based on an empirically based review ofnumerous studies to determine the cause of the pneumoniaand its treatment. It is recognized that these studies havethe basic shortcoming of being unable to identify the etio-logic agents initially in a significant number of patients.Accordingly, patients have been categorized by the ATSinto mild-to-moderate pneumonia and severe pneumonia.These groups are further characterized based on the pres-ence or absence of risk factors and the time [early «5days) or late (>5 days)] of onset of the pneumonia (Table58-6). This "cookbook" approach to therapy must be sup-plemented by appropriate bacteriologic studies and di-rected treatment based on these studies.

Therapy

Intravenous j3-lactam(cefotaxime,ceftriaxone, ampicillin!sulbactam, high-dose ampicillin)

plusIntravenous or oral macrolide or

doxycyclineorIntravenous antipneumococcal

fluoroquinolone alone

Intravenous azithromycin alone ifmacrolide allergic or intolerant:doxycycline and a j3-lactam

orMonotherapy with an antipneumococcal

Fluoroquinolone

DRSP, drug-resistant Streptococcus pneumoniae.Adapted from American Thoracic Society: Guidelines for the management of adults with commu-nity-acquired pneumonia. Am J Respir Crit Care Med 163: 1736, 2001. With permission.


Intravenous ~-lactam (cefotaxime,ceftriaxone)

plus eitherIntravenous macrolide (azithromycin)orIntravenous fluoroquinolone

Table 58-5. Community-Acquired Pneumonia: Group 4. ICU-Admitted Patients

Organisms Therapy

No risks for Pseudomonas aeruginosaStreptococcus pneumoniae (including DRSP)Legionella speciesHaemophilus injluenzaeEnteric Gram-negative bacilliStaphylococcus aureusMycoplasma pneumoniaeRespiratory virusesMiscellaneous

Chlamydia pneumoniae, Mycobacteriumtuberculosis, endemic fungi

Risks for Pseudomonas aeruginosaAll of the above pathogens plus

p. aeruginosaSelected intravenous antipseudomonal ~-

lactam (cefepime, imipenem, meropenem,piperacillinltazobactam) plusintravenous antipseudomonal quinolone(ciprofloxacin)

orSelected intravenous antipseudomonal ~-

lactam (cefepime, imipenem, meropenem,piperacillinltazobactam) plusintravenous aminoglycoside

plus eitherintravenous macrolide (azithromycin)or intravenous nonpseudomonalfluoroquinolone

DRSP, drug-resistant Streptococcus pneumoniae; ICU, intensive care unit.Adapted from American Thoracic Society: Guidelines for the management of adults withcommunity-acquired pneumonia. Am J Respir Crit Care Med 163: 1736, 2001. With permission.

BACTERIOLOGY

The bacterial etiology of empyema has changed overthe years. Before the development of antibiotics, 10% ofpatients who survived pneumonia developed empyema. Ac-

Table 58-6. Modifying Factors that Increase the Risk forInfection with Specific Pathogens

Penicillin-resistant and drug-resistant pneumococciAge >65 years~-Iactam therapy within the past 3 monthsAlcoholismImmune-suppressive illness (including therapy with

corticosteroids)Multiple medical comorbiditiesExposure to a child in a day-care center

Enteric Gram-negativesResidence in a nursing homeUnderlying cardiopulmonary diseaseMultiple medical comorbiditiesRecent antibiotic therapy

Pseudomonas aeruginosaStructural lung disease (bronchiectasis)Corticosteroid therapy (>I0 mg of prednisone per day)Broad-spectrum antibiotic therapy for> 7 days in the past

monthMalnutrition

Adapted from American Thoracic Society: Guidelines for themanagement of adults with community-acquired pneumonia. AmJ Respir Crit Care Med 163:1737, 2001. With permission.

cording to Ehler (1941), Streptococcus and Pneumococcu,were the most frequent organisms. After the introduction0..antibiotics, the incidence of empyema from these organisms was markedly reduced, as was the mortality rate.Staphylococcus became much more prevalent, and in the;1950s and 1960s, Ravitch and Fein (1961) found that thisorganism produced 90% of empyema in children under 2years of age. In more recent times, as pointed out by Varkeyand associates (1981) and Bergeron (1990), penicillin-resis-tant Staphylococcus, Gram-negative bacteria, and anaerobicorganisms have been the predominant microbes. Bartlettand colleagues (1974a) reported that 76% of empyema pa-tients had positive culture results for either anaerobes ex-clusively (35%) or anaerobes in combination with aerobes(41 %). Anaerobic bacteria are normal flora of the mouth,intestine, and female genital tract. They are difficult to cul-ture, being extremely oxygen sensitive. Multiple organ-isms, 50 in the series of Sullivan and colleagues (1973), arefrequently cultured, but the most common is Peptostrepto-coccus. Similar findings were recorded by Ali and Unruh(1990).

Brook and Frazier (1993) independently studied 197pa-tients with culture-positive empyema from two militaryhospitals. Aerobic organisms were isolated in 127 patients(64%), mixed aerobic and anaerobic organisms in 45 pa-tients (23%), and anaerobic organisms in 25 patients (13%).The predominant aerobic or facultative bacterial isolateswere Streptococcus pneumoniae (70), Staphylococcusau-

reus (58), Escherichiacoli (17), Klebsiellapneumoniae(16), and Haemophilus influenzae (12). The predominantanaerobeswere anaerobic cocci (36), pigmented Prevotellaand Porphyromonas (24), Bacteroides fragilis (22), andfusobacterium species (20). Most patients from whomS.pneumoniae and H. influenzae were cultured had pneu-monia.Most patients from whom S. aureus were recoveredhad pneumonia, aspiration pneumonia, or lung abscesses.Recoveryof anaerobic bacteria was associated with the di-agnosis of aspiration pneumonia and lung, dental, andoropharyngealabscesses.

Lemenseand associates (1995) at the Medical Universityof SouthCarolina studied, retrospectively, 43 patients withempyema.Twenty-four of the 43 patients (56%) had para-pneumonicempyema, and 19 patients had empyema from avarietyof causes including gunshot wounds, thoracotomy,and esophageal rupture. Most nonparapneumonic empy-emashad positive cultures (84%), whereas only 11 of 24patientswith parapneumonic empyema (46%) had positivecultureson Gram's stain. Alfageme and associates (1993)notedincreasing numbers of patients with Staphylococcusempyema,particularly in alcoholics.

Thepropensity for developing empyema varies consider-ablywiththe type of bacteria producing the primary pneu-monia,the setting in which the infection is acquired, andthealteration in these produced by antibiotic therapy ad-ministeredfor primary pneumonia or for concurrent condi-tions.For example, as reported by Bartlett (1974a) andFang(1990)and their colleagues, Light (1990), and John-sonandFinegold (1994), S. pneumoniae is responsible for60%to 75% of CAPs, but only 2% of patients with pneu-mococcalpneumonia develop empyema (Tables 58-7 and58-8).I3-hemolyticStreptococcus accounts for 1% to 2% ofCAPs,butup to 10% of adults and 50% of children developempyema.In hospital settings, Staphylococcus and aerobic,Gram-negativebacteria are the most common agents pro-ducingpneumonia. Both have significant potential to pro-duce empyema.

CLINICALFEATURES

Empyemamay be heralded by an exacerbation or recur-renceof the septic course of pneumonia or may present as acontinuationof symptoms and manifestations of the pri-marypneumonic process. Antibiotics have blunted andchangedthe clinical picture, and there may be only subtleprogressionfrom the symptoms and signs of pneumonia tothoseof empyema. The most common presenting symp-tomsof empyema according to Varkey and associates(1981)are shortnessof breath (82%), fever (81%),cough(70%),and chest pain (67%), all of which are common topneumonia.The presentation of these symptoms in a pa-tientwithfebrile respiratory illness or the accentuation orprolongationof these symptoms in a patient with pneumo-niashouldalert the clinician to the possibility of empyema.

58. PARAPNEUMONICEMPYEMA · 823

Table 58-7. Etiology of Pneumonia

Agent Incidence (%)0

Community-acquired pneumoniaStreptococcus pneumoniaeHaemophilus influenzaebStaphylococcus aureusMycoplasma pneumoniaeLegionella pneumophilaChlamydia pneumoniaeStreptococcus pyogenescUnknown

Hospital-acquired pneumoniaAerobic Gram negative

Klebsiella pneumoniaeEscherichia coliPseudomonas aeruginosadS. aureus

60-755-10<51-101-51-5<135

45+

<10

°lncidence varies with series.bEmpyema complicates 10% of H. influenzae pneumonias in chil-dren, but is rare in adults.cEmpyema in 30% or greater. Incidence may be much higher inchildren.dMost common cause of pneumonia in intensive care unit, withsignificant potential for empyema.

The clinician should be aware that the incidence of pleuriticchest pain and leukocytosis are similar whether or notpleural effusion is present.

Empyema from aerobic bacterial pneumonia usually pre-sents as an acute febrile illness with chest pain, sputum pro-duction, and leukocytosis. Anaerobic pleural infection ismore indolent and, in Bartlett's (1974a) series, averaged 10days before presentation. The majority of these patientshave a history of alcoholism or unconsciousness, frequentlyhave lost weight and have mild anemia. Hospital-acquiredinfection and immunosuppression may alter this course.

Physical examination reveals a toxic, anxious patientwith tachycardia and tachypnea. Restricted and guardedchest wall excursion may be present. Percussion of thechest wall may elicit pain and dullness over the empyema

Table 58-8. Incidence of Empyema Accordingto Bacteria-Causing Pneumonia

Incidence ofOrganism Effusion (%) Empyema (%)

AerobicGram positive

Streptococcus 50 <5pneumoniae

Staphylococcus 70 80aureus (children)

Staphylococcus 40 20aureus (adults)

Gram negativeEscherichia coli 50 90Pseudomonas 50 90

Anaerobic 35 90


area. With chronicity, the patient may develop clubbing ofthe fingers, contraction of the chest wall, inanition, and othersigns of chronic illness.

DIAGNOSIS

Traditional Radiographic Evaluation

The presence of a significant pleural effusion in associa-tion with a lower respiratory illness is typical, but this clini-cal picture may also be seen with pulmonary embolism,acute pancreatitis, Dressler's syndrome, tuberculosis, andother conditions. Light and associates (1980) have suggestedthat bilateral decubitus chest radiographs be obtained. Withthe involved side down, the distance between the inside ofthe chest wall and the outside of the lung should not exceed10 mm. If this distance exceeds 10 mm, thoracentesisshould be performed.

When seen by the surgeon, many empyemas are advancedand empyemic collections are loculated. Because the poste-rior lateral diaphragmatic angle is the most dependent posi-tion of the thorax, most empyemas are found in this area. Theinverted D or pregnant lady sign, as coined by LeRoux andDodds (1964), on the lateral view is classic (Fig. 58-1). Thedifferentiation of lung abscess with effusion from empyemaand bronchopleural fistula may also be suggested by "plain"radiographic examination (Fig. 58-2). Friedman and Helle-kant (1977) noted that the air fluid levels oflung abscess areusually of the same dimensions in posterior/anterior and lat-eral views, whereas empyema air fluid levels are rarely thesame in these views.

Computed Tomography

Computed tomographic (CT) scanning is now used uni-versally to identify underlying parenchymal disease and todistinguish empyema from lung abscess (Fig. 58-3). Fluid

density and the presence of loculations can be determined,the latter being an important factor in treatment planning.CT-guided thoracocentesis is a highly accurate and safetechnique that is used routinely.

Sonography

Alternately, sonography of the chest may be used. Sonog-raphy can demonstrate pleural fluid collections, loculations,and parenchyma involvement and may be used to guide tho-racocentesis. The choice of this technique is institutionallydependent.

MANAGEMENT

Effective management of empyema requires (a) controlof infection and sepsis by appropriate antibiotic therapy,(b)evacuation of pus from the pleural space, and (c) oblitera-tion of the empyema cavity. Once the diagnosis is estab-lished, treatment must proceed with all possible haste. InBartlett's (1974b) series of 43 patients with anaerobicempyemas, five patients (12%) died. He attributed all fivedeaths to a delay in appropriate drainage. In Ashbaugh's se-ries (1991), delay in instituting drainage increased themor-tality rate from 3.4% to 16%.

Colice and colleagues (2000) published the AmericanCollege of Chest Physicians' Consensus Statement of theTreatment of Parapneumonic Effusions. The study groupexamined 789 citations from a Medline search. Twenty-four articles were identified for full review by the panel.Meta-analysis was performed and recommendations pre-sented, primarily addressing techniques of drainage. How-ever, it should be noted that, as with all meta-analyses, ithas the significant shortcoming of incomplete comparativeanalysis despite extensive data. Drainage tube size couldnot be compared, and specific antibiotic therapy was not

B

Fig. 58-1. A. Posteroanterior radiographof patient with encapsulated pleural effu-sion. B. Inverted D or pregnant lady signon lateral view is classic, if not typical. ofencapsulated empyema.

6,L

Fig. 58.2. A, B. Chest radiographs reveal fluid-filled cavities.Thediameters of the cavities are essentially equal in anteroposte-riorand lateral views. This symmetry is typical of lung abscessesandfluid-filledcysts (arrows) within the lung. C, D. Chest radio-graphsreveal a normal study result, then pneumonia two weekslater. (Continues)

58. PARAPNEUMONICEMPYEMA· 825

B

o


E

G

F

H

Fig. 58-2. (Continued) E, F. This study reveals an asymmetric air fluid space typical of bronchopleural fistulaand empyema. G, H. A decubitus study confirms the large empyema and the computed tomography scan con-firms the large air- and fluid-filled space outside the lung.

addressed. Thus, the following, while drawing from thesedata, is a compilation of these data plus other studies andthe authors' experiences.

Antibiotic Therapy

Appropriate antibiotics should be administered promptly.Many patients, when fIrst seen by the surgeon, will have al-

ready been treated with antibiotics and, according to LeMenseand colleagues (1995), approximately 50% of cultures withparapneumonic empyema will be negative. The yield ishigher with empyema secondary to HAP. Despite these facts,aerobic and anaerobic cultures of the blood and empyemafluid should be obtained. If the cultures become positive, an-tibiotic therapy should be guided by these results. Initial an-tibiotic therapy is usually based upon empiric guidelines forthe treatment of CAP or HAP.

A

Thoracentesis

The appearance of the fluid provides a first and majorsourceof information concerning diagnosis and treatment.Straw-colored,clear, or slightly cloudy fluid is found in ster-ile"parapneumonic"effusions. This condition correspondstoearlystage 1of the ATS classification and the "uncompli-cated"effusion of Light (1980), the "benign" effusion ofPotts(1976),and the "low pH" pleural effusion of VanWay(1988)and their associates. Twenty to 70% of these willclearwithappropriate treatment of the primary pneumonia.Theproblem and dilemma is that benign-appearing fluidmayheraldan impending "frank" empyema. Therefore, thefluidmustbe examined by Gram's stain, cultured for aero-bicandanaerobicbacteria, and tested for antibiotic sensitiv-ity.Biochemicalchanges (Le., low pH, low glucose, andhighLDH) indicate and precede positive culture and, ac-cordingto Light (1991) and others, can serve as a guide totherapy.Cloudyor frankly purulent fluid is diagnostic. Aer-obicpushas little or no offensiveodor.Anaerobicpus isusuallyfoul smelling.

Needleaspiration of the fluid collection should be per-formedwith an 18-gaugeneedle under local anesthesia andradiographicor sonographic guidance. If the fluid appearsbenign,isnot loculated,and can be removed totally or nearlyso, thoracentesismay be all that is necessary to controlthediseaseprocess.Traditionally, leukocyte counts, Gram's


B

Fig. 58-3. Patient with acute lower respiratory illness. A. Patientupright radiograph reveals density in right lower lung field. B.Computed tomography scan reveals large empyema collectionwith atelectatic lobe and consolidation.

stains, cultures, and glucose and LDH determinations wererecommended. Of these, only culture and pH determinationsare of practical value. A pH of less than 7.1 indicates anempyema, and positive cultures may serve as a late guide forantibiotic therapy. Radiographic examination should be re-peated in 24 hours to ascertain the status of the effusion. Ifthe volume has increased or if the patient's status has not im-proved, a closed chest tube thoracostomy is indicated. If tho-racentesis demonstrates cloudy or purulent fluid, or if theeffusion is greater than one half of the hemithorax (even iffree flowing), fibrin loculations are probably present andclosed chest tube drainage is required. The presence of obvi-ous pus also indicates the need for closed chest tubedrainage. It is noted that, in meta-analysis, the mortality rateof patients treated primarily by thoracentesis is 10%and 46%of patients require additional procedures.

Chest Thbe Drainage

Closed chest tube drainage is the usual first step in thetreatment of acute empyema. However, this approach maytake a number of forms. Traditionally, a No. 36F or largertube was placed in the most dependent area of the empyemacavity as determined by previous thoracentesis and radio-graphic examination. Underwater seal drainage with moder-ate suction (-20 cm Hz0) is applied to drain the purulent

Table 58.9. Surgical Treatment of Empyema: Proportion of Deathswith 95% Confidence Interval (CI) in Individual Cohorts and Pooled

by Primary Management Approach

DeathProportion (%)

No drainageTherapeutic

thoracentesisTube thoracostomy 408Fibrinolytics 94VPlTS 42Surgery 159

VPlTS,video-assisted thoracic surgery.Pldapted from CoHce GL, et al: Medical and surgical treatment of parapneumonic effusions. Chest118:1158,2000. With permission.


Procedure No. at Risk

61175

fluid and to obliterate the space. In meta-analysis, Colice andcolleagues (2000) found that such treatment is associatedwith a 9% mortality rate and that secondary intervention, in-cluding the use of additional chest tubes, is necessary in 40%of patients (Tables 58-9 and 58-10). During the past decade,image-guided drainage with much smaller tubes (12-16F) inassociation with the use of fibrinolytic agents has supersededtraditional chest tube drainage. Meta-analysis indicates alower (4.3%) mortality rate with secondary interventionbeing required in approximately 15% of patients. Numerousstudies have compared the efficiency of fibrolytic agents, no-tably streptokinase and urokinase. Urokinase, although muchmore expensive and although temporarily recalled from themarketplace because of viral contamination, appears to bethe most efficacious, primarily due to its lack of systemic ef-fect. Both agents are efficient in breaking down fibrin locula-tions and increasing fluid drainage. Moulton (1989) and Lee(1991) and their associates reported success rates of 90%.Moulton and colleagues (1995) subsequently updated theirseries and reported successful drainage in 111 of 118 cases(94%). Two patients died of sepsis with incomplete drainage.Five patients underwent decortication (three recovered andtwo died postoperatively). Fifty-three patients (45%) re-quired placement of more than one drain. The mean duration

No. Died 95% CI, %

1.8, 166.2, 15.8

418

6.610.3

36423

8.84.34.81.9

6.3, 12.01.2, 10.50.6,16.20.6,16.2

of drainage was 6.3 days. The mean number of urokinasein-stallations was five. The mean total dose of urokinase usedper case was 466,000 units. No complications were encoun-tered.

Bouros and associates (1997) compared streptokinaseand urokinase in a double-blind study of 50 consecutivepa-tients with either complicated pleural effusion or frankempyema who had inadequate drainage through the chesttube «70 mL per 24 hours). Clinical and radiologic im-provement was noted in all but two patients (8%) in eachgroup who required surgical intervention. The mean vol-ume of drainage was significantly increased (urokinase,420 mL per 24 hours; streptokinase, 380 mL per 24 hours).The mean number of installations was six in both groupsand the mean lengths of stay were approximately 11days.Two patients in the streptokinase group suffered high fever.The total cost for drug therapy was approximately twice asmuch in the urokinase group. The researchers concludedthat fibrinolytic therapy is a valuable adjunct to chest tubedrainage and that, despite the cost, urokinase may be desir-able because of local and systemic reactions to streptoki-nase. These reactions included fever, chills, and chest pain.In Laisaar and colleagues' (1996) series, enzymatic treat-ment was effective in 72% of patients, but 28% required

Table 58-10. Surgical Treatment of Empyema: Proportion of PatientsNeeding a Second Intervention with 95% CI in Individual Cohorts and Pooled by

Primary Management Approach

No at No. with Second InterventionProcedure Risk Second Intervention Proportion (%) 95% CI, %

61175

36.1,62.338.7,54.0

No drainageTherapeutic

thoracentesisTube thoracostomy 434Fibrinolytics 94VPlTS 42Surgery 159

VPlTS,video-assisted thoracoscopic surgery.Pldapted from CoHce GL, et al: Medical and surgical treatment of parapneumonic effusions. Chest118:1158,2000. With permission.

3081

49.246.3

17514o

17

40.314.90.0

10.7

35.7,45.18.4,23.70.0, 8.46.3, 16.6

further treatment such as decortication or lung resection.Chin and Um (1997), in a controlled study of 52 patients(29 treated with drainage only and 23 treated with drainageplus streptokinase daily), noted increased drainage in thestreptokinasegroup, but no difference in hospital stay dura-tion,mortality, or need for decortication.

The technique is simple. A 12 to 16 F chest tube is placedin the effusion guided by CT, ultrasonography, or fluo-roscopy.For moderate-sized unilocular collections, a singledrainis usually all that is necessary. If collections are largeor if multiple loculations are present, more than one drainmaybe placed. Following drain placement, all fluid is aspi-rated. Follow-up imaging is then performed to determinethe final position of the drains and the location and amountof any remaining pleural fluid. If significant loculations re-main, additional drains should be placed. The drains arethenattached to -20 cm water suction via a closed under-waterseal system. The drains are irrigated daily with sterilesalineto maintain patency. The decision to proceed with fi-brinolyticagents is based primarily on the amount of fluidremainingafter catheter placement. If all fluid was drainedduring the first 2 days following drain placement and thelung has expanded, no fibrinolytic agents are recom-mended.If significant amounts of fluid remain, fibrinolyt-ics are injected through the drains into the pleural cavity(assumingthat no bronchopleural fistula is present). Gener-ally,100,000 units of urokinase or 250,000 units of strep-tokinaseare dissolved -in 250 mL of normal saline solution.Followinginstallation of 75 to 100 mL of the fibrinolyticsolution,the catheters are clamped and the patient is in-structedto intermittently change position from supine toeachdecubitus position in order to facilitate mixing of thesolutionwith the pleural fluid. After 1 to 4 hours, the cathe-tersareunclamped and as much fluid as possible aspirated.Thenet amount of fluid is recorded. The procedure shouldbe repeated after 1 to 2 hours of suction drainage. Threeseparateirrigations daily are recommended. Three criteriaareusedto determine when to remove the chest tubes. Clin-

icalimprovement with resolution of fever and leukocytosisis desirable.However, if the pleural space is considered ad-equatelydrained and an alternate source of fever is apparent(i.e.,persistent pneumonia), it is not considered absolutelynecessarythat the white blood cell count be normal and thepatientafebrile.The second criterion is that almost all of thepleural fluid was drained as determined by radiographicevaluation.If any fluid remains and the clinical response isincomplete,the drainage process should be continued. Thethirdcriterionis that there should be no more than 20 mL of

netdrainoutput over a 24-hour period.

OpenDrainage

Open drainage was the classic method of drainingempyemacavities prior to the antibiotic era. Historically,Grahamand Bell (1918), when reporting to the Empyema


Commission, demonstrated that open drainage was an effec-tive treatment for empyema. They demonstrated that opendrainage must not be performed during the acute phase of thedisease since an open pneumothorax was produced with re-duced ventilation, hypoxia, and death. Once the lung is fixedto the chest wall and ventilation is assured, open drainage issafe. In times past, this was heralded by increased sedimentin the fluid drained by thoracocentesis from the empyemacavity. When the sediment reached 75%, open drainage wassaid to be safe.

At the present time, open drainage through a thoracos-tomy tube may be performed in situations where the lunghas not expanded completely, a space exists, and circum-stances dictate that no further therapy is indicated. Once thespace has become stable (usually 2-3 weeks following chesttube insertion), there are no gross up and down movementsof the water column, and pneumothorax fails to occur whenthe chest is open to atmospheric pressure, the chest tube iscut off a few centimeters from the skin and anchored inplace with a safety pin and tape. The tube may be withdrawna few centimeters per week as space is obliterated and gran-ulation tissue fills the tract and drainage decreases.

Eloesser (1935) was able to maintain patency of the opendrainage tract achieved by rib resection with a modified skinflap technique. According to Ali and colleagues (1996), thisis a useful procedure in those instances where a large spaceexists and prolonged drainage is anticipated.

Video-Assisted Thoracoscopic Decortication

As evidenced by Angelillo Mackinlay (1996) and Landre-neau (1996) and their colleagues as well as one of us (M.K.)(1998) and both authors (1996) and their associates, video-assisted thoracic surgery (VATS) has become the primarymodality for treating complicated empyema after initialtherapy (with or without chest tube drainage) in many insti-tutions. VATS allows adhesiolysis and debridement withbetter exposure than minithoracotomy. Decortication forlung expansion prior to fibrosis can be accomplished withequal facility. Although chest tube drainage combined withenzymatic debridement is effective, VATStherapy results ina higher (90%) success rate, shorter hospital stay, and lesscost. Wait and co-workers (1997) concluded that it can rou-tinely salvage patients in whom chest tube drainage with en-zymatic debridement was not successful.

The VATSprocedure is described in Chapter 18. A dual-lumen endotracheal tube is inserted for single-lung ventila-tion. Its position is confirmed by bronchoscopy. Bloodpressure, central venous pressure, arterial oxygen saturation(Sa02), and end tidal C02 are monitored continuously. Thepatient is placed in the lateral decubitus position. A two- orthree-port triangular approach is used with the ports laterserving as chest tube drainage sites. The empyema cavity isentered, loculations broken up, adhesions lysed with endo-shears and cautery, fibrin and purulent material removed


mechanically and by suction, and the chest cavity irrigatedwith copious amounts of saline. The lung surface is wipedclean and the lung reexpanded. Morbidity is low and chesttubes can usually be removed sequentially beginning in 3 to4 days depending on the virulence of the infection, the pres-ence of a bronchopleural fistula, and the state of the patient.If complete reexpansion is not achieved, the remaining, de-pendent chest tube can be cut off and converted to anempyema tube.

VATS is highly effective in the treatment of parapneu-monic empyema. The combined mortality rate in AngelilloMackinlay's (1996) and Wait's (1997) and their associates'series is 4.8% and, most important, no additional proce-dures were necessary.

Thoracotomy, Minithoracotomy, and Decortication

Although VATShas the advantages of greater patient tol-erance, open thoracotomy with debridement and decortica-tion is an effective means of dealing with empyema. Moststudies of this technique predate the widespread use ofVATS.Van Way and associates (1988) reported on 22 pa-tients in which debridement and decortication were per-formed and in which drainage was established throughlimited thoracotomy. The process resolved completely in alland there were no deaths in this series. Miller (1990) de-scribed 52 patients treated with open thoracotomy. Good re-sults were obtained in 50 patients, and there were nooperative deaths. MandaI and associates (1998) treated 179consecutive adult patients with primary empyema. Ninetyunderwent closed thoracostomy, with a cure rate of 62%and a mortality rate of 11%. Twenty-four of these required asecond procedure. Seventy-six patients underwent decorti-cation as either a primary or secondary procedure, with acure rate of 88% and a mortality rate of 1.3%.Thus, 42% ofthese patients required decortication. In a meta-analysis of159 patients undergoing surgical procedures, the mortalityrate was 1.9% and 11% of the patients required at least oneadditional intervention.

Empyemectomy

Empyectomy is rarely performed. It requires an ex-trapleural dissection of the parietal pleura and tedious dis-section of the sac from the lung. Just as in decortication of achronically collapsed and trapped lung, lung damage requir-ing undesirable and unnecessary resection is often the result.

CHRONIC EMPYEMA

Chronicity refers to a state of continued infection associ-ated with both fibrosis and a pleural space that is often com-promised by bronchopleural fistulae. Fortunately, this is an

~--

uncommon occurrence following appropriate treatment. Inthe past, such spaces were treated by some form of thora-coplasty. This generally took the form of a modified Shedeprocedure. Ribs were resected over the cavity, the parietalpleura was removed, and the muscle bundles were allowedtofall against the visceral pleura. The procedure was effective,but mutilating. Hankins and associates (1978) concludedthatmuscle translocation into the cavity with occasional rib re-section is the more desirable and effective procedure.

TREATMENT OF EMPYEMA IN CHILDREN

Empyema in infants and children is usually associatedwith pneumonia. Its incidence has diminished greatly in re-sponse to-the successful treatment of pneumonia with an-tibiotics. In addition, its bacterial etiology has changedconcomitantly with the evolution of microbial resistance.Nearly two decades ago, Foglia and Randolph (1987) foundthat the most frequent causes of empyema were Haemo-phi/us injluenzae, ~-hemolytic streptococci, Streptococcuspneumoniae, and anaerobes. At present, Hardie and co-workers (1996) have found that Streptococcus pneumoniae,often penicillin and erythromycin resistant, is the mostcommon organism. Gustafson and associates (1990) foundthat anaerobic bacteria produce effusions that loculatequickly and are difficult to drain, whereas staphylococcaleffusions are most commonly unilocular and relativelyeas-ier to drain.

The clinical scenario has changed as well, with empyemaseen more often in older children as opposed to infants.Hardie and colleagues' (1996) study of CAP complicatedby empyema, revealed that 40% of empyemas were due toStreptococcus pneumoniae, 15% of which were resistant topenicillin, and 44% of cultures were negative. None of theempyemas were associated with Staphylococcus aureus orHaemophi/us injluenzae. Only one empyema was causedby group A Streptococcus. Miller (1990) noted that the cul-ture-negative status of these children is now a commonfinding because of pretreatment with antibiotics in the com-munity setting.

Campbell and Nataro (1999) noted that host defenses areweakened in the nutrient-rich purulent space due to (a) lackof phagocytic surface for optimal white cell function, (b)low levels of opsonins and complement, and (c) conditionsof acidosis, hypoglycemia, and hypoxia. Furthermore,Bryant and Salmon (1996) found that many antibiotics donot work optimally in pleural empyema despite efficientpenetration into the pleural space.

Fever is the most common presenting symptom in child-hood empyema. Fever, cough, and dyspnea are present in75% of children on admission. Decreased breath sounds,tactile and vocal fremitus, and tachypnea and tachycardiaare often present according to Middlekamp (1964) andMcLaughlin (1984) and their associates and Chonmaitreeand Powell (1983). Other symptoms are grunting respira-

tion, intercostal retraction, and lethargy. The polymorphicleukocytecount is virtually always elevated. ChonmaitreeandPowell (1983) noted pulmonary lobar infiltrate in 52%of patients with empyema, segmental infiltrate in 33%, andbilateral patchy infiltrate in 14%. Foglia and Randolph(1987)found that CT scanning is extremely useful in clari-fyingthe extent of parenchymal lung and pleural involve-ment.However,Donnelly and Klosterman (1997), in studiesof pediatric patients with empyema evaluated by CT scan,observedthat uncomplicated (free-flowing) effusions, can-notbe differentiated from effusions in which significant fi-brinhas been developed.

The treatment of empyema in children remains contro-versial,but the goals of therapy were clearly outlined byMayoand associates (1982) (Table 58-II).

The possibility of empyema in children as heralded byclinicalsigns and symptoms and radiographic evidence ofpleuralinvolvement or infusion requires prompt attention.ACTscanshould be performed if the presence of fluid is inquestionand thoracentesis should be performed for visual,biochemical,and cultural evidence of infection.

Antibiotics administered intravenously, according toGram's stain, culture, and sensitivity studies, and pleuraldrainageare the mainstays of treatment. Foglia and Ran-dolph(1987) demonstrated that most children treated withappropriateantibiotics and chest tube drainage recover. Thiscoursedepends on the disease stage, type of bacteria, anddegreeof lung trapping. If rapid recovery does not occur,additionalintervention is indicated. Raffensperger and col-leagues(1982) recommended rninithoracotomy or full tho-racotomy.Gustafson and colleagues (1990) noted thataggressivecare, including early thoracotomy, led to early re-coveryand excellent long-term results. Rizalar and associ-ates (1997) reported on 32 patients treated by "early"decorticationfor postpneumonic empyema with no deathsandno recurrences. In older children, VATS deloculation,decortication,if necessary, and drainage were used by deCamposand associates (1997) with good results. GandhiandStringel(1997) performed VATSdecortication in ninechildren(mean age 4, range 21 months to 13 years) withsuccessfuloutcomesin all. One child required a blood trans-fusionbecauseof bleeding following difficult and extensivedecortication.Open drainage is never indicated in children

Table 58-11. Goals of Therapyin Empyema in Children

SavelifeEliminatethe empyemaReexpandthe trapped lungRestoremobility ofthe chest wall and diaphragmReturnrespiratory function to normalEliminatecomplications or chronicityReducethe duration of hospital stay

FromMayo P, Saba SP, McElvein RB: Acute empyema in childrentreatedby open thoracotomy and decortication. Ann Thorac Surg34:401,1982. With permission.

58. PARAPNEUMONIC EMPYEMA · 831

because of late skeletal deformities. We do not believe thatthe use of enzymatic debridement is indicated in infants andsmall children. However, Rosen and associates (1993) andKomecki and Sivan (1997) reported success with adjuvantuse of enzymatic debridement. A controlled trail is needed todefine the use of enzymatic debridement in children.

REFERENCES

Alfageme I, et al: Empyema of the thorax in adults: etiology, microbio-logic findings and management. Chest 103:839, 1993.

Ali I, Unruh H: Management of empyema thoracis. Ann Thorac Surg50:355, 1990.

Ali SM, Siddiqui AA, McLaughlin JS: Open drainage of massive tubercu-lous empyema with progressive reexpansion of the lung: an old conceptrevisited. Ann Thorac Surg 62:218, 1996.

American Thoracic Society: Hospital-acquired pneumonia in adults: diag-nosis, assessment of severity, initial antimicrobial therapy and preventa-tive strategies. Am J Respir Crit Care Med 153:1711, 1995.

American Thoracic Society: Guidelines for the management of adults withcommunity-acquired pneumonia. Am J Respir Crit Care Med 163: 1730,200\.

Andrews NC, et al: Management of nontuberculous empyema: a statementof the subcommittee on surgery. Am Rev Respir Dis 85:935, 1962.

Angelillo Mackinlay TA, et al: VATS debridement versus thoracotomy inthe treatment of loculated postpneumonia empyema. Ann Thorac Surg61:1626,1996.

Ashbaugh DG: Empyema thoracis. Factors influencing morbidity andmortality. Chest 99: 1162, 199\.

Bartlett JG, Finegold SM: Anaerobic infections of the lung and pleuralspace. Am Rev Respir Dis 110:56, 1974a.

Bartlett JG, et al: Bacteriology of empyema. Lancet 1:338, 1974b.Bartlett JG, et al: Bacteriology of hospital-acquired pneumonia. Arch In-

tern Med 146:868,1986.

Bartlett JG, Mundy LM: Community-acquired pneumonia. N Engl J Med333:1618,1995.

Bartlett JG, et al: Practice guidelines for the management of community-acquired pneumonia in adults. Infectious Diseases Society of America.Clin Infect Dis 31:347, 2000.

Bergeron MG: The changing bacterial spectrum and antibiotic choice. InDeslauriers J, Lacquet LK (eds): International Trends in General Tho-racic Surgery. Vol. 6. St. Louis: CV Mosby, 1990.

Bouros D, et al: Intrapleural streptokinase versus urokinase in the treat-ment of complicated parapneumonic effusions: a prospective, double-blind study. Am J Respir Crit Care Med 155:291, 1997.

British Thoracic Society: Guidelines for the management of community-acquired pneumonia in adults admitted to hospital. Br J Hosp Med49:346, 1993.

Brook I, Frazier EH: Aerobic and anaerobic microbiology of empyema. Aretrospective review in two military hospitals. Chest 103:1502, 1993.

Bryant RE, Salmon CJ: Pleural empyema. Clin Infect Dis 22:747, 1996.Campbell JD, Nataro JP: Pleural empyema. Pediatr Infect Dis J 18:725, 1999.Chin NK, Lim TK: Controlled trial of intrapleural streptokinase in the

treatment of pleural empyema and complicated parapneumonic effu-sions. Chest 111:275, 1997.

Chonmaitree T, Powell KR: Parapneumonic pleural effusion and empyemain children. Review of a 19-year experience, 1962-1980. Clin Pediatr22:414, 1983.

Colice GL, et al: Medical and surgical treatment of parapneumonic effu-sions: an evidence-based guideline. Chest 118:1158, 2000.

Craven DE, Driks MR: Pneumonia in the intubated patient. Semin RespirInfect 2:20, 1987.

Craven DE, Steger KA, Barber TW: Preventing nosocomial pneumonia: stateof the art and perspectives for the 1990's. Am J Med 91(suppl):44, 199\.

de Campos JR, et al: Thoracoscopy in children and adolescents. ChestJJJ:494,1997.

Donnelly LF, Klosterman LA: CT appearance of parapneumonic effusionsin children: findings are not specific for empyema. AJR 169:179,1997.

Ehler AA: Non-tuberculous thoracic empyema: a collective review of theliterature from 1934 to 1939. Int Abstr Surg 72: 17, 1941.


Eloesser L: An operation for tuberculous empyema. Surg Gynecol Obstet60:1096,1935.

Fagon JY, et al: Nosocomial pneumonia in patients receiving continuousmechanical ventilation. Prospective analysis of 52 episodes with use ofa protected specimen brush and quantitative culture techniques. Am RevRespir Dis 139:877,1989.

Fang GD, et al: New and emerging etiologies for community-acquiredpneumonia with implications for therapy. A prospective multicenterstudy of 359 cases. Medicine 69:307, 1990.

Fine MJ, et al: Prognosis and outcomes of patients with community-acquired pneumonia. A meta-analysis. JAMA 275: 134, 1996.

Fine MJ, et al: A prediction rule to identify low-risk patients with commu-nity-acquired pneumonia. N Engl J Med 336:243,1997.

Foglia RP, Randolph J: Current indications for decortication in the treat-ment of empyema in children. J Pediatr Surg 22:28, 1987.

Friedman PJ, Hellekant CAG: Radiologic recognition of bronchopleuralfistula. Radiology 124:289,1977.

Gandhi RR, Stringel G: Video-assisted thoracoscopic surgery in the man-agement of pediatric empyema. JSL Surg 1 :251, 1997.

Graham EA, Bell RD: Open pneumothorax: its relation to the treatment ofacute empyema. Am J Med Sci 156:839,1918.

Gross PA, et al: Deaths from nosocomial infections: experience in a uni-versity hospital and a community hospital. Am J Med 68:219, 1980.

Gross PA, Van Antwerpen C: Nosocomial infections and hospital deaths: acase-control study. Am J Med 75:658, 1983.

Gustafson RA, et al: Role of lung decortication in symptomatic empyemasin children. Ann Thorac Surg 49:940,1990.

Hankins JR, et al: Bronchopleural fistula. Thirteen-year experience with77 cases. J Thorac Cardiovasc Surg 76:755, 1978.

Hardie W, et al: Pneumococcal pleural empyemas in children. Clin InfectDis 22:1057,1996.

Johnson CC, Finegold SM: Pyogenic bacterial pneumonia, lung abscessand empyema. In Murray JF, Nadel JA (eds): Textbook of RespiratoryMedicine. Vol. 1. 2nd Ed. Philadelphia: WB Saunders, 1994.

Kornecki A, Sivan Y: Treatment of loculated pleural effusion with in-trapleural urokinase in children. J Pediatr Surg 32:1473,1997.

Krasna MJ: Thoracoscopic decortication. Surg Laparosc Endosc 8:283, 1998.Krasna MJ, Deshmukh S, McLaughlin JS: Complications of thoracoscopy.

Ann Thorac Surg 61:1066,1996.Laisaar T, Puttsepp E, Laisaar V: Early administration of intrapleural

streptokinase in the treatment of multiloculated pleural effusions andpleural empyemas. Thorac Cardiovasc Surg 44:252, 1996.

Landreneau RI, et al: Thoracoscopy for empyema and hemothorax. Chest109:18, 1996.

Lee KS, et al: Treatment of thoracic multiloculated empyemas with intra-cavitary urokinase: a prospective study. Radiology 179:771, 1991.

LeMense GP, Strange C, Sahn SA: Empyema thoracis. Therapeutic man-agement and outcome. Chest 107:1532,1995.

LeRoux BT, Dodds TC: A Portfolio of Chest Radiographs. Edinburgh: E &S Livingston, 1964.

Light RW: Parapneumonic effusions and empyema. Clin Chest Med 6:55,1985.

Light RW: Parapneumonic effusions and infections of the pleural space. InLight RW (ed): Pleural Diseases. 2nd Ed. Philadelphia: Lea & Febiger,1990.

Light RW: Management of para pneumonic effusions. Chest 100:892,1991.Light RW: Pleural effusion. N Engl J Med 346:1971, 2002.Light RW, et al: Pleural effusions: the diagnostic separation of transudates

and exudates. Ann Intern Med 77:507, 1972.Light RW, et al: Parapneumonic effusions. Am J Med 69:507, 1980.Mandai AK, et al: Outcome of primary empyema thoracis: therapeutic and

microbiologic aspects. Ann Thorac Surg 66: 1782, 1998.Mandell LA, et al: Canadian guidelines for the initial management of com-

munity-acquired pneumonia: an evidence-based update by the CanadianInfectious Diseases Society and the Canadian Thoracic Society. Clin In-fect Dis 31:383, 2000.

Marks MI, EickhoffTC. Empyema necessitatis. Am Rev Respir Dis 101:759,1970.

Mayo P, Saha SP, McElvein RB: Acute empyema in children treated byopen thoracotomy and decortication. Ann Thorac Surg 34:401, 1982.

McLaughlin FJ, et al: Empyema in children: clinical course and long-termfollow-up. Pediatrics 73:587, 1984.

Middlekamp IN, Purkerson ML, Burford TH: The changing pattern ofempyema thoracis in pediatrics. J Thorac Cardiovasc Surg 47: 165, 1964.

Miller II: Empyema thoracis. Ann Thorac Surg 50:343, 1990.Moulton JS, Moore PT, Mencini RA: Treatment of loculated pleural effu-

sions with transcatheter intracavitary urokinase. AIR 153:941,1989.Moulton JS, et al: Treatment of complicated pleural fluid collections with

image-guided drainage and intracavitary urokinase. Chest 108: 1252, 1995.Potts DE, Levin DC, Sahn SA: Pleural fluid pH in parapneumonic effu-

sions. Chest 70:328, 1976.

Prod'hom GP, et al: Nosocomial pneumonia in mechanically ventilated pa-tients receiving antacid, ranitidine or sucralfate as prophylaxis for stressulcer: a randomized controlled trial. Ann Intern Med 120:653, 1994.

Raffensperger JG, et al: Mini-thoracotomy and chest tube insertion forchildren with empyema. J Thorac Cardiovasc Surg 84:497, 1982.

Ravitch M, Fein R: The changing picture of pneumonia and empyema ininfants and children: a review of the experience at the Harriett LaneHome from 1934 through 1958. JAMA 175:1039,1961.

Rizalar R, et al: Postpneumonic empyema in children treated by earlydecortication. Eur J Pediatr Surg 7: 135, 1997.

Rosen H, et al: Intrapleural streptokinase as adjunctive treatment for per-sistent empyema in pediatric patients. Chest 103: 1190, 1993.

Rouby JJ, et al: Nosocomial bronchopneumonia in the critically ill. Histo-logic and bacteriologic aspects. Am Rev Respir Dis 146:1059, 1992.

Scheid WM: Bacterial meningitis, brain abscess and other suppurative in-tracranial infections. In: Harrison's Principles of Internal Medicine.14th Ed. New York: McGraw-Hill, 1998.

Schleupner CJ, Cobb DK: A study of the etiologies and treatment of noso-comial pneumonia in a community-based teaching hospital. Infect Con-trol Hosp EpidemioII3:515, 1992.

Sullivan KM, et al: Anaerobic empyema thoracis. The role of anaerobes in226 cases of culture-proven empyemas. Arch Intern Med 131:521, 1973.

Van Way C III, Narrod J, Hopeman A: The role of early limited thoracotomyin the treatment of empyema. J Thorac Cardiovasc Surg 96:436, 1988.

Varkey B, et al: Empyema thoracis during a ten-year period. Analysis of 72cases and comparison to a previous study (1952 to 1967). Arch InternMed 141:1771,1981.

Wait MA, et al: A randomized trial of empyema therapy. Chest 111:1548,1997.

parapneumonic empyema · chapter 58 parapneumonic empyema joseph s. mclaughlin and mark j. krasna...

Documents