Empyema Thoracis in Children: A 26-Year Review of the Montreal

Children’s Hospital Experience

By W. Chan, E. Keyser-Gauvin, G.M. Davis, L.T. Nguyen, and J.-M. Laberge
Montreal, Quebec

0 The appropriate management of pediatric empyema or in combination with thoracentesis,12 tube thoracos-
thoracis remains controversial. The authors reviewed 47 tomy, fibrinolytics, 7.8thoracoscopy,3-5minithoracotomy,’ l
cases of empyema thoracis over a 26-year period. The
open window drainage,12 or formal thoracotomy9J0 and
management of empyema included initial diagnostic thora-
centesis and classification as acute, fibropurulent, or chronic. decortication. The greatest discrepancies in treatment
If the empyema was “acute,” therapeutic tap, tube thoracos- occur between nonsurgical and surgical specialists. In
tomy, or no surgical intervention was performed. “Fibropuru- view of the controversy and evolving management of
lent”empyemas were uniformly treated with tube thoracos- thoracic empyema we decided to compare our manage-
tomy. The lung was decorticated when the empyema was
ment policy and results with those published in the
encased by a thick peel, had recurred and was multilocu-
lated, was refractory and the patient remained clinically literature.
unwell, or had occurred as a complication of previous
thoracotomy. All patients with acute empyemas responded MATERIALS AND METHODS
to antibiotics irrespective of drainage (average duration of
During a 26-year period (1968 to 1994) 47 patients with empyema
fever, 17 days; average stay in hospital, 27 days). Of the
thoracis were Identified through medical records as having been treated
fibropurulent empyemas in our review, complete drainage
at the Montreal Children’s Hospital, and their charts were reviewed
was attained in seven of 39 (IS%), and decortication was not
retrospectively. The management of empyema included an inittal
required in any empyema that was completely drained.
diagnostic thoracentesis and classification as acute, fibropurulent, or
Loculations persisted in 25 of 39 (64%) after tube thoracos-
chronic (Table 1). A similar staging system categorized as exudative,
tomy but nonetheless resolved. The remaining seven of 39
fibropurulent, and organizing has been defined by the American
(18%) with persistent loculations required formal decortica-
Thoracic Society and is considered the standard.13 If the empyema
tion. Of the patients with fibropurulent empyemas that
thoracis was “acute,” a diagnostic tap, a therapeutic tap, or a tube
responded to tube thoracostomy, the average duration of
thoracostomy was performed. “Fibropurulent” empyemas were uni-
fever was 13 days and hospitalization, 23 days. Of those
formly treated with tube thoracostomy. The lung was decorticated when
requiring decortication the average duration of fever was 24
either the empyema was encased by a thick peel, had recurred and was
days and hospitalization, 40 days. These results will allow a
multiloculated, was refractory and the patient remained unwell, or had
baseline for comparison of new strategies (fibrinolytics and
occurred as a complication of prevtous thoracotomy.
early thoracoscopy) that may reduce days of fever, hospital-
Data collected included age, sex, ethnic origin, presenting symptoms
ization, and risk of formal decortication.
and signs, initial imaging studies, classification of empyema stage,
Copyright o 1997 by W.B. Saunders Company
culture results, drainage procedures, operative procedures, number of
days with fever, number of days in hospital. number of days treated with
INDEX WORD: Empyema.
antibiotics, and complications.

T HE MANAGEMENT of empyema thoracis in chil-

dren has evoked considerable controversy.‘,* The
literature provides many options but assists little in
RESULTS
The average age of the patients was 5 years, 7 months
(range, neonatal to early adulthood). The male to female
establishing the ideal treatment.‘-” Generally, recommen-
ratio was 1:l (24 boys:23 girls). Thirty-seven patients
dations have been based on institutional traditions, per-
were white, seven were Inuit, and three fell into other
sonal experience, and limited case reviews. Decisions
categories.
about individual cases are further influenced by varying
The majority of patients, 94% (44 of 47) presented
criteria such as the patient age, clinical status, antibiotic
primarily with respiratory symptoms of cough, chest
response, the stage and duration of the empyema, and the
pain, dyspnea, and respiratory distress. Eleven percent (5
organism cultured. I2 Treatments include antibiotics alone
of 47) also had symptoms of abdominal pain. Six percent
(3 of 47) presented only with nonspecific symptoms of
irritability or poor feeding. Fever was present in the
From the Departments of General Pediatric Surgery and Pulmonary
Medicine, Montreal Children’s Hospital, McGill University, Montreal,
majority, 81% (38 of 47) and in 6% (3 of 47) it was the
Quebec. only abnormal sign noted. Physical examination local-
Presented at the 28th Annual Meeting of the Canadian Association of ized signs to the chest in 89% (42 of 47) of patients. Of
Paediatric Surgery, Ha&x, Nova Scotia, August 18-20, 1996. those with physical findings present, dullness to percus-
Address reprint requests to J.-M. Laberge, MD, Montreal Children’s sion was the only noted abnormality in 40% (17 of 42).
Hospital, 2300 Tupper St, C-1137, Montreal, Quebec, Canada H3H
IP3.
Initial chest x-ray findings included pneumonia and
Copyright o 1997 by WB. Saunders Company effusion (57%); obliteration of the hemithorax (21%);
OOZZ-3468/97/3206-0017$03.00/O pneumonia, effusion, and pneumatocele ( 13 %); pneumo-

870 Journal of Pediatrrc Surgery, Vol32, No 6 (June), 1997: pp 870-872

EMPYEMA THORACIS

Table 1. Classification of Empyema Thoracis ema encased by a thick peel and were reclassified as
Classrfrcation Charactenstxs chronic. and two had recurrent empyemas that were not
Acute Pleural flurd is clear to slightly cloudy and serous amenable to chest tube drainage. Of those patients with
Sterile fluid incomplete drainage of fibropurulent empyema. 22% (7
Has at least one of the following: of 32) required decortication.
pH c7.20
One patient who required decortication but did not fit
Glucose ~40 mg/dL
LDH >I,000 IU/dL
into the mentioned categories was a patient in whom an
Protern >2.5 g/dL empyema developed after open thoracotomy for a trau-
Specific gravity >I.018 matic hemothorax. The surgeon elected for open drainage
WBC >500 cells/mm3 rather than a trial of tube thoracostomy.
Fibropurulent Fluid is thicker and opaque (ie, pus) or positive
Patients with empyema were treated with both oral and
culture
Chronic An organizrng peel forms around the lung
intravenous antibiotics, and there was no specific policy
determining the length of each type of treatment. Wide
Abbreviation: LDH, lactate dehydrogenase.
ranges in antibiotic usage over the years and treatment
nia and pyopneumothorax (6%); and tension pyopneumo- course was highly dependent on the attending physician.
thorax (2%). Only 10 of 47 patients had additional Those with acute empyemas received antibiotics an
radiological tests including ultrasound scan (7 of 47), CT average of 30 days. Those with fibropurulent empyemas
scan (5 of 47), or both (2 of 47). with complete tube drainage received antibiotics an
At presentation, 15% (7 of 47) were acute empyemas average of 39 days, and those without complete drainage,
(Table 1). Fibropurulent empyemas made up the remain- an average of 30 days. Patients requiring thoracotomy
der at 83% (39 of 47). One patient could not be classified received antibiotic coverage for an average of 50 days.
based on the available information. Complications associated with empyema occurred in
The predisposing condition for the empyema was 42% of our patients (20 of 47, Table 2). There were no
pneumonia in 89% (42 of 47), trauma in 4% (2 of 47), deaths and no long-term complications identified. Scolio-
perforated appendicitis in 4% (2 of 47), and infected sis resolved during follow-up in all four patients in whom
congenital chylothorax in 2% (1 of 47). this was noted.
Initial cultures grew Staphylococcus aureus in 8 of 47, Average follow-up was 103 days (range, 7 to 720 days)
Haemophilus injluenzae in 6 of 47, Streptococcus in 6 of for patients with fibropurulent empyema. One patient
47, Staphylococcus pyogenes in 6 of 47, Pneumococcus with fibropurulent empyema was lost to follow-up.
in 3 of 47, Pseudomonas in 2 of 47, Bacteroides in 2 of DISCUSSION
47. Escherichia coli in 2 of 47, Eikenella in 1 of 47. and
Fusobacterium necrophorum in 1 of 47. Culture was Empyema remains a relatively common problem in the
negative in 11 of 47 and multiple organisms grew in 2 pediatric population with significant morbidity and con-
of 47. sumption of hospital resources. Therapeutic options have
All patients with acute empyemas (7 of 7) responded to increased in the last several years to include home
antibiotics alone or in combination with drainage. Diag-
Table 2. Complications of Empyema
nostic thoracentesis was solely performed in two of seven
patients, therapeutic thoracentesis in one of seven, and Thoracic
Pneumothorax 4
tube thoracostomy in four of seven. Patients with acute
Pneumatocele 4
empyema had fever for an average of 17 days, and the Tension pyopneumothorax 1
average hospital stay was 27 days. Lung abscess 1
Of the patients with fibropurulent empyemas, 82% (32 Hemoptysrs 1
of 39) responded to antibiotics and thoracostomy drain- Cardiac arrhythmia 1
Pencardial effusion 1
age alone. Of the responders 22% (7 of 32) had complete
Respiratory arrest 1
drainage with no remaining loculations (average days Scoliosis 4
with fever, 16; in hospital, 25). The remaining 78% of Gastrointesbnal
responders (25 of 32) did so despite incomplete drainage Upper GI bleeding 1
of loculations (average days with fever, 12; in hospital, lleus 1
Septic
23).
Fulminant sepsis 2
Eighteen percent (7 of 39) of the patients with fibropu- Candidiasis 1
rulent empyemas did not respond to antibiotics and chest DIC 1
tube drainage and required decortication (average days Metabolic
with fever, 24; in hospital, 40). Three of these patients did SIADH 2
not improve clinically with drainage, two had an empy- NOTE. n = 47.
872 CHAN ET AL

intravenous antibiotics, intrathoracic instillations of fibri- SUSPECTED EMPYEMA

nolytics,7,s and thoracoscopic drainage.‘e5 The indications
Diagnostic Thoracentesis
and timing of these newer therapies remain unclear. Some
investigators strongly advocate nonoperative treatment,‘j
others suggest newer technologies,3.4,5.7,*and still others Acute Fibropurulent
suggest that early open operative drainage is the best (pus or gram stain +vs
choice.9-11 or culture +ve)
I
Our review was undertaken to assess whether our Antibiotics
1
results were different than those published and whether a k Drainage Antibiotics + Chest Tube
J \r
change in our practice was warranted. Also, we wanted to No residual Loculations
know if we could determine who in our population could on CXR Remain
benefit from the newer therapies. / Thoracosoopy
1
or Fibrinolytics
Our data suggest that patients with acute empyemas Continue Conservative or both
Management
are likely to experience resolution with antibiotics alone 1
Unsuccessful
or in combination with thoracic drainage. The majority of
patients with fibropurulent empyema (82%) will be c
Consider Open Decortication
successfully treated with antibiotics and tube drainage.
Patients with fibropurulent empyema who have no remain- Fig 1. Suggested management for empyema in children.

ing loculations seen on chest x-ray have minimal risk of

requiring thoracotomy and decortication. Patients with
fibropurulent empyema who have remaining loculations
after tube drainage are still likely to be successfully
treated with antibiotics; however, there is an increased
risk that thoracotomy and decortication may be required.
In our series of 32 patients who did not achieve complete
drainage as seen by chest x-ray, 7 patients (22%) subse-
quently required operative drainage. We cannot assessthe
prognostic usefulness of computed tomography and ultra-
sonography because they were performed in only 10 of
47 patients.
It is the fibropurulent group that may benefit the most
from the newer therapies. Even though resolution occurs
in the majority of children with fibropurulent empyema,
the hospitalization is prolonged when compared with
published results using the newer modalities. Average
published days in hospital postthoracoscopy range from 7
to 13 days.3-5Average published days in hospital postfibri-
nolytics, if successful, range from 17 to 19 days. Incom-
plete drainage of fibropurulent empyema in our series
resulted in an average hospitalization of 23 days and 40
days if decortication was required. Twenty-two percent of
patients who were incompletely drained required decorti-
cation versus 0% to 23% of those treated with fibrinolyt-
its* and 0% of those treated with thoracoscopy.3-5
Thoracoscopic drainage and debridement is appealing
to surgeons but remains an operation that requires a
general anesthetic. Our study provides a basis for a
rational approach to empyemas (Fig 1). Our findings
clearly indicate that acute empyemas and fibropurulent
empyemas that respond to chest tube drainage do not
require more aggressive surgical treatment. Patients with
persistent loculations seen on chest x-ray may benefit
from fibrinolytics and early thoracoscopy. The advocated
decrease in hospitalization, costs, and need for formal
decortication with these newer strategies should be
compared with newer strategies of conservative therapy,
ie, outpatient intravenous antibiotics. A randomized mul-
ticenter study may be the only way to establish the
optimal treatment of empyema in children.

REFERENCES
1. Campbell PW III: New developments m pediatric pneumonia and 8. Rosen H, Nadkarm V, Theroux M: Intrapleural streptokinase as
empyema. Curr Opin Pediatr 7:278-282, 1995 adjunctive treatment for persistent empyema in pediatric patients. Chest
2. Schropp IQ: Empyema and intrathoracic infection in children. 103:1190-1193,1993
Pedtat Thorac Surg 3:443-460, 1993 9. Miller JI: Empyema thoracis. Ann Thorac Surg 50:343-344, 1990
3. Kern JA. Rodgers BM: Thoracoscopy in the management of 10. Gustafson RA, Murray GF, Warden HE: Role of lung decortica-
empyema in children. J Pediatr Surg 28:1128-1132, 1993 tion in symptomatic empyemas m children. Ann Thorac Surg 49:940-
4. Silen ML, Weber TR: Thoracoscopic debndement of loculated 947, 1990
empyema thoracis in children. Ann Thorac Surg 59: 1166.1168, 1995 11. Van Way C III, Narrod J. Hopeman A: The role of early limited
5. Stovroff M, Teague G. Heiss KF, et al: Thoracoscopy in the tboracotomy in the treatment of empyema. J Thorac Cardiovasc Surg
management of pediatric empyema. J Pediatr Surg 30:1211-1215, 1995 96:436-439. 1988
6. Mangete EDO, Kombo BB. Legg-Jack TE: Thoracic empyema: A 12. Miller JI: Infections of the pleura, in Shields TW (ed): General
study of 56 patients. Arch Dis Child 69:587-588, 1993 Thoracic Surgery, Philadelphia, PA. Lea & Febiger, 1989. pp 633-649,
7. Robinson LA. Moulton AL, Fleming WH, et al: Intrapleural Chap 59
fibrinolytic treatment of multiloculated thoracic empyemas. Ann Thorac 13. American Thoracic Society: Management of nontuberculous
Surg 57:802-8 14, 1994 empyema. Ann Rev Respir Dis 85:935-936,1962