Analysis of Tube Thoracostomy

Performed by Pulmonologists at a
Teaching Hospital*
Nancy A Collop , MD, FCCP; Sola Kim, MD; and Steven A Sahn, MD, FCCP

Study objective: To evaluate all tube thoracostomies (TTs) done by pulmonary/critical care
fellows and attending physicians in the Medical University of South Carolina's health-care system
documenting patient demographics, indication for placement, size and characteristics of the
tube, and associated problems.
Design: Prospective.
Setting: University health-care system, including a university hospital, a Veterans Affairs hospital,
and a county hospital.
Patients: All adult patients requiring consultation by a member of the pulmonary/critical care
staff for a tube thoracostomy.
Results: One hundred twenty-six tube thoracostomies were performed over a 24-month period in
91 patients. The most common initial indication for a TT was pneumothorax (69/103, 67%).
Overall mortality in the patient population was 35% (32/91). Early problems ( <24 hours following
placement) occurred in 3% (4/126); late problems (>24 h after placement) occurred in 8%
(10/126). Problems occurred in 36% (4/ll) of small-bore tube placements vs 9% (10/ll5) of
standard TT placements (p=0.02).
Conclusions: Tube thoracostomy can be safely performed by pulmonologists with relatively few
associated problems. (CHEST 1997; 112:709-13)

Key words: chest tube; pleural effusion; pneumothorax; pulmonologist; tube thoracostomy

Abbreviations: TT = tube thoracostomy

T inubethethoracostomy (TT) is an important modality

treatment of thoracic disease that has
traditionally been performed b y thoracic surgeons .
However, in the last decade, TT has been performed
more frequently by other specialties, including pul-
monologists.
TT, nevertheless, remains an invasive procedure
that has the potential for complications. Complica-
tions related to insertion and infection have been
reported for TT when performed by surgeons for
acute trauma. 1 To our knowledge, there have been
no comprehensive reviews of TT applied in the
pulmonary and critical care arena. As the American
Counsel on Graduate Medical Education suggests in
its directory, pulmonary training programs should
provide training in "thoracostomy tube insertion and
drainage."2 Therefore, w e sought to characterize TTs
performed by trainees under the supervision of
*F rom the Division of Pulmonary and Critical Care Medicine,
Medical University of South Carolin a, Charleston .
Manuscript received October 1.6, 1996; revision accepted March
6, 1997.
Reprint requests: Nancy A Collop, MD, FCC P, Medical Unit;er-
sity of South Carolina, 171 Ashley Ave, Charleston, SC 29425
academic pulmonologists in the pulmonary and crit-
ical care setting to assess indications, success, and
problems.
MATERIALS AND METHODS
We prospectively followed up all patients who had Tis placed
by physicians in th e Pulmonary/Critical Care Division at the
Medical University o f South Carolina from August 1992 through
July 1994. The procedures were performed b y a pulmonary/
critical care fellow under the supervision of a pulmona1y/critical
care attending physician at one of three hospitals in our university
syste m: the Medical University Hospital; the Ralph H. Johnson
Veterans Affairs Medical Center; and the Charleston Me morial
Hospital. Instructi on in TT insertion was given t o all fellows
during their orientation and insertions were observed by attend-
ing physicians until the fellow was thought to be competent in
performance of the procedure. A d etailed intake sheet was
completed by the phys ician who performed the TT. The physi-
cians involved in perform ance o f the TT evaluated each f or
problems associated with the insertion and during the course.
The intake s heets were analyzed b y o neof us and any further
information to complete the review was obtained b y medical
record review. Problems associated with the TT were e valuated
by us and categorized into early and late problems. Data are
expressed as mean::+::SE. Fishe r's Exact Test was used to evaluate
the data.

CHEST / 112 / 3 / SEPTEMBER, 1997 709

 RES ULTS
Demographics and Indications
Data were collected for 24 months. A total of 126
TTs were placed in 91 patients. Five patients were
hospitalized on two separate occasions and required
chest tubes during both hospitalizations. Descriptive
data are displayed in Table l. Of the 96 hospitaliza-
tions, 57 (59%) were at the university hospital, 18
(19%) were at the Veterans Affairs hospital, and 21
(22%) were at the county hospital. The Pulmonary/
C1itical Care Division manages the medical ICU at
all hospitals. The patient was on the pulmonary/
critical care service in 59 (61%) cases; the referring
service was an internal medicine service in 36 (37%)
cases and obstetrics in 1 (l%) case. The mean age of
the patients was 51 years with a median age of 52
years (range, 16 to 87 years ). The duration of
placement was a mean of 5.5 days with a median of
4 days (range, 1 to 12 days ). Twenty-three patients
required more than one chest tube placement during
their hospitalization. Six had TTs placed on both the
right and left sides, 13 patients required two Tis on the
same side, and five required three TTs on the same
side.
The original indication for the TT is shown in
Figure l. The majority of initial TTs were placed for
pneumothoraces (68/103, 66%). The size of the chest
tube placed in relation to its original indication is
shown in Table 2. Catheters (Cook; Cook Critical
Care; Bloomington, Ill), either 7, 8, or 14F, were
utilized in 11 of 126 total cases (9%); most Tis
placed were 24F .
Pleurodesis was performed in 21 patients . Twelve
of21 (57%) were performed for a malignant effusion
(5 for lung cancer, 4 for breast cancer, 1 for lym-
phoma, 1 for pancreatic cancer, and 1 for metastatic
head and neck cancer), 2 pleurodesis procedures
were done for a nonmalignant symptomatic effusion,
and 7 were done for treatment of a pneumothorax.
Table !-Demographics of Patients Receiving TT
(n=126) *
Variable
No. of patie nts
Average age , yr
Ave rage duration of TT, d 5.5:!::0.4
Male patients (%)
ICU patients (%)
Receiving mechan ical ventilation (%)
Placed on left side (%)
Patients requiring > 1 TT (%)
aced in 91 patients.
*One hundred twenty-six TTs w e re pl
710
Deaths and Problems Associated with TT
Mortality in the population studied was 35% (32/
91). Most of the patients died with a chest tube in
place (22132, 69%), but the TT was not believed to
be directly responsible for any deaths. The diagnoses
of the patients who died and their original indica-
tions for TT are listed in Table 3.
Analysis of TT was divided into early problems,
which occurred during the first 24 h (4/126, 3%) and
were related to initial placement and success of
desired effect; and late problems, those occurring
after the first 24 hours (10/126, 8%) (Table 4). Early
problems included a TT not placed in the pleural
space; the tube was a 7F Cook catheter placed for an
iatrogenic pneumothorax. Another TI (24F) placed
for a malignant effusion resulted in the development
of a pneumothorax; it was thought that the tube may
have lacerated the lung during placement. Two small
tubes (BF and 16F) became nonfunctional during
the first 24 h. One tube clotted and caused a
recurrent pneumothorax in a patient receiving me-
chanical ventilation; the other was accidentally dis-
lodged from the pleural space and also resulted in a
recurrent pneumothorax.
Late problems (10/126, 8%) included a superficial
site infection (16F) and a 7F catheter that developed
a le ak around the insertion site, which delayed
removal for 1 day. Eight Tis had problems associ-
ated with either clotting (two), kinking (five), or
dislodgment (one). Only one of those was a small
caliber tube (SF ). Seven of those Tis required
placement of a second tube to resolve the problem.
Small-caliber tubes (Cook catheters size 7, 8, or
14F ) were associated with more problems (4111,
36%) than standard TTs (10/115, 9%) (p=0.02).
We also analyzed all patients who required a
second or third TT on the same side. Of those that
are not detailed above, there were nine TTs placed
in patients who either did not initially completely
reexpand their lung following a pneumothorax or
r
subsequently developed a ecurrent pneumothorax
despite the desired initial effect. One of these pa-
tients, who required three ITs, subsequently had
surgery to remove a bleb. The initial tube sizes
placed for these patients were as follows: 1 to 14F, 3
to 24F, 2to 28F, 2 to 32F, and 1 to 36F. There were
No. two other instances of a second TI placement. One
91
patient developed a recurrent pneumothorax imme-
51:!::2 diately after a TT was removed, while another
required a second TT to drain a loculated effusion
67/91 (74) because the initial TT was not completely effective.
67/126 (53)
55/126 (44) DISCUSSION
67/126 (53)
23/91 (25) Our data suggest that chest tubes can be placed
safely by pulmonologists for a variety of causes.
Clinical Investigations
 N 103

CIT] Barotrauma-induced Pneumothorax

E:::::l Iatrogenic Pneumothorax
• Symptomatic or Malignant Pleural Effusion
!{i{i:::l Spontaneous Pneumothorax
mmJ Empyema/Parapneumonic Effusion
D Miscellaneous

FIGURE 1. The original indication for placement of each tube thoracostomy. Pe rcentages reflect the
number of Tis placed for that indication divided by the total number of Tis placed.

Table 2-Size of Chest Tube and Initial Indication for Placement (n =103)
Chest Tube Size

Indication 7-8 14 16 20 24 28 32 36 Unknown

Effusion
Malignant 10 7 3 1
Parapneumonic 1 4
Empyema 2
Symptomatic* l l
Hemothorax l
Chylothorax l
Pneumothorax
Iatrogenic 1 2 2 2 ll 6 2
Barotrauma ll 5 8
Spontaneous 2 6 2 l 2
Hydropneumothorax
Total 3 4 4 2 41 22 22 4
*Nonmalignant causes of symptomatic effusion included pancreatitis and congestive heart failure.
'Iatrogenic causes followed placement of central lines, transbronchial biopsies, thoracenteses, nerve blocks, and transthoracic needle biopsies.

Table 3-Cause of Death and Indication for Placement of TT

Cause of Death

Respiratory Failure End-Stage

TI Indication Sepsis::+::ARDS Cancer (Excluding ARDS ) Liver Disease TIP*

Iatrogenic PTX 2 2 l
Malignant effusion 3
Parapneumonic 3
effusion/empyema
Barotrauma PTX 16 2
Spontaneous PTX 2
*TIP=thrombotic thrombocytopenic purpura; PTX=pneumothorax

Indications for TT include the following: pneumo-
thorax (spontaneous, related to barotrauma, iatro-
genic); hemothorax or hemopneumothorax; pleural
effusions (malignant, complicated parapneumonic,
symptomatic, chylothorax); and following trauma or
thoracic surgery. 3 Our TTs were not placed for
trauma or surgical-related indications, but usually for
treatment of pneumothorax or pleural effusions. The
pneumothoraces frequently occurred in the ICU
setting secondary to mechanical ventilation-related
barotrauma or as complications of procedures such
as central line placement and bronchoscopy. Spon-

CHEST I 112 I 3 I SEPTEMBER, 1997 711

 Table 4-Problems Associated With TT (n=14)
Problem Early* Late'
Not placed in the pleural space 1 0
Nonfunctional rrt 2 8
Laceration of trapped lung l 0
Site infection 0 l
Leak around TT 0 l
*Early: wi thi n 24 h of TI placement.
1
Late: >24 h following TI place ment.
1 Due to cotting,
l kinking, and/or dislodgment of TI.
taneous pneumothorax was a less frequent indication
for tube thoracostomy, comprising only 14% ofTTs.
Chest tube size was chosen at the discretion of the
treating physician. The most commonly used size
was 24F. Small-caliber tubes (sizes 7 , 8,and 14F) in
our series had a higher complication rate (36%)
compared to large-caliber tubes (9%) .The differ-
ence in outcomes with the two tubes could be due to
a selection bias, as some of the tubes were used for
different indications and the study was not a random-
ized, controlled trial. Smaller tubes have become
more popular recently and have been used success-
fully for a v ariety of indications, including the follow-
ing: spontaneous and iatrogenic pneumothoraces;4-7
pneumothorax secondary to barotrauma; 8 and pleu-
ral effusions ,9·10 including empyemas. 11 - 13 These
catheters have been touted as easier to insert and
less painful than larger-bore tubes; however, to our
knowledge, there has not been a controlled study
comparing the two. Success rates \vith small-caliber
tubes range from 50 to 100% for pneumothorax5-7
and 72 to 95% for pleural effusions and empy-
ema. 9·10·14 The difficulties encountered in the
present series \'lith these tubes could be related to
operator inexperience or to inappropriate patient
selection. Catheter obstruction has been d escribed
as occurring in these small-caliber chest tubes/ ·1.5
this problem occurred in 2 of 11 tubes in our series.
Flushing the catheter h as been suggested to prevent
obstruction; however, this is not a routine practice in
our institution. Other reported complications in-
Table 5- Complications of TT
Insertional
Esophageal perforation Horner's syndrome
Partial aortic obstmction Arteriovenous fistul a
Laceration of internal organs a nd vessels Reexpansion pulmonary edema
Acute diaphragmatic paralysis Contralateral pn eumothorax
Avulsion injury to s tomach Cardiogenic shock from displaceme nt of right ventricle
Chylothorax
Lun g perforati on
Diaphragmati c perforation
Subcutaneous p lacement
712
elude catheter displacement and site infections.16 A
disadvantage of small-caliber Tis is the inability to
do a finger exploration prior to insertion, and this
likely played a role in one of our early TT problems
when th e TT did not enter the pleural space.
Pleurodesis was performed in 23% (21/91 ) of
patients. Early in the study, the pleurodesis agent
used was doxycycline or minocycline; whereas in
later pleurodesis attempts, talc slurry was used.
There were no clinically important complications
related to pleurodesis.
The patient population requiring Tis in this series
had severe illness. Unfortunately, we did not have
any APACHE (acute physiology and chronic health
evaluation ) or similar s coring system in place at the
time of this study to compare severity of illness.
Although TT was not directly a acuse of any deaths,
35% of patients (32191 ) died \'lith a chest tube in
place, suggesting a patient population \'lith severe
disease. Barotrauma-related pneumothoraces com-
prised 25% 26/103)
( of Tis in our series. There is a
3 to 15% incidence of pneumothorax in patients
undergoing mechanical ventilation P- 19 Many of our
patients required more than one chest tube, a phe-
nomenon that has been previously described in
mechanically ventilated patients .20 As seen in our
series, the size of the TT selected for a pneumotho-
rax seems to have little bearing on whether subse-
quent TTs are needed. In the series of Heffner and
colleagues20 of 47 pneumothoraces, 16 patients re-
quired a second chest tube for recurrence on the
ipsilateral side. The most common disease process
present in their patients requiring more than one TT
was ARDS; in our series, 80% of patients requiring a
second chest tube for barotrauma-related pneumo-
thorax had ARDS .
Our complication rate is similar to that of a s eries
involving trauma patients .1 However, as previously
stated, our patient population likely differed from
those involved in trauma. A s we hadno comparison
group, such as Tis placed b ysurgical personnel, it is
difficult to directly compare complication rates. Ad-
ditionally, one weakness of our study is that several
Positional Miscellaneous
Pul monaty infarction
Empyema
Subcutaneous emphysema
Necrotizing fasciitis
Osteomye litis
Site infection
Clinical Investigations
individuals were placing the TT and reporting sub-
sequent proble ms, instead of one individual assess-
ing all TT placements for complications; therefore,
our complication rate may be understated.
Complications of TT include those related to
inse rtion,21-2 7 position of the tube, 28- 35 and other
miscellaneous causes36 ·37 (Table 5). All of the com-
plications in the present series have been d escribed
previously. The most common complication in our
se1ies was r elated to the tube not functioning due to
clotting, kinking, or dislodgment Although most of
our tubes w ere sec ured b y a purse string suture and
tape, movement and kinking are inevitable compli-
cations, particularly in critically ill patients who are
frequently moved for procedures.
In conclusion, TT can be performed with an
acceptable complication rate and high efficacy by
pulmonologists despite a severely ill patient popula-
tion. Further study is needed to assess the efficacy
and complication rate of small- vs large-caliber
tubes .
REFERENCES
Millikan J , Moore E, Steiner E , et al . Complications of tube
thoracostomy for acute trauma. Am J Surg 1980; 140:738-41
2 American M edical Association. Directory of graduate medical
edu cation programs. Chicago: American M edical Association,
1992--1993; 59
3 Miller KS , Sahn SA. Chest tubes: indications, technique,
management and complications. Chest 1987; 91:258-64
4 Bernstein A, Waqaruddin M, Shah M. Management of
spontaneous pneumoth orax using a Heimlich flutter valve.
Thorax 1973; 28:386-89
5 Casola B,van S onnenberg E, Keig htley A, e t la. Pneumotho-
rax: r adiologic treatment with small catheters. Radiology
1988; 166:89-91
6 Perlmutt L, BraunS, Newman G, et al. Transthoracic needle
aspiration: use of a small chest tube to treat pneumothorax.
AJR Am J Roentgenol 1987; 148:849-51
7 Pete rs J, Kubitschek K. Clinical evaluation of a percutaneous
pneumothorax catheter. Chest 1984; 86:714-17
8 Conces DJ, Tarver RD , G ray WC, et al. Treatment of
pneumothoraces utilizing s mall caliber chest tubes. Chest
1988; 94:55-57
9 Wood R, Tzakis A, Shaw B, et la. A simplified technique for
the treatment of simple pleural effusions. Surg Gynecol
Obstet 1987; 164:283-84
10 Fuhrman B , La ndrum B, Ferrara T, et al. Pleural drainage
using modified pigtail catheters. Crit Care Med 1986; 14:
575-77
11 Westcott J. Percutaneous cathe te r drainage of pleural effu-
sion and empyema. AJR Am J Roentgenol1985; 144:1189-93
12 Silverman S, Mueller P, Sai ni S, et al. Thoracic empyema:
management with image-guided catheter drainage. Radiology
1988; 169:5-9
13 van S onnenberg E , NakamotoS, Mueller P, et al. CT- and
ultrasound-guided catheter drainage of empyemas after chest
tube fai lure. Radiology 1984; 151:349-53
14 So S, Yu D. Catheter drainage of spontaneous pneumothorax:
suction or no suction, early o r late re moval . Thorax 1982;
37:46-48
15 Morrison M, Mueller P, Lee M, e t !a. Sclerotherapy of
malignant pleural effusion through sonographically placed
small-bore catheters. AJR Am J Roentgenol 1992; 158:41-43
16 Shand J, McCreath G, Monie R. The use o ffin e bore Silastic
catheters to drain carcinomatous pleural effusions. Br J Dis
Chest 1988; 82:394-97
17 Zwillich C, Pierson D , Creagh C, et l.a Complications of
assisted ventilation: a prospective study of 354 consecutive
cases. Am J Med 1 974; 567:161-68
18 Peterson G, Baier H . Incidence of pulmonary barotrauma in
a medical IC U. Crit Care Med 1983; 11:67-69
19 Flemming W, Bowen J , HatcherC. Early complications o f
long-term respiratory support. J Thorac Cardiovasc Surg
1972; 64:729-32
20 Heffner JE, McDonald J, Barbie ri C. Recurrent pneumotho-
races in ventilated patients despite ipsilateral chest tubes.
Chest 1995; 108:1053-58
21 Foresti V, Villa A, Casati 0 , et al. Abdominal placement of
tube thoracostomy due to ack l of re cognition of paralysis of
hemidiaphragm. Chest 1992; 102:292-93
22 Shapira OM , Aldea GS, Kupferschmid J, et al. Delayed
perforation of the esophagus by a closed thoracostomy tube.
Chest 1 993; 104:1897-98
23 Resnick DK. Delayed pulmonary perforation: a rare compli-
cation of tube thoracostomy. Chest 1993; 103:311-13
24 Fraser R. Lung perforation complicatin g tube thoracostomy:
pathologic description of three cases . Hum Pathol 1988;
19:518-23
25 Gooding C, Kerlan R, Brasch H.. Partial aortic obstruction
produced by a thoracostomy tube. J Pedi atr 1981; 98:471-73
26 Kumer S, Belik J. Chylothorax: a complication of chest tube
placement in a n eonate. Crit Care Med 1984; 12:411-12
27 Meisel S, Ham Z, Friel I, et al. Another complication of
thoracostomy-performation of the right atrium . C hest 1 990;
98:772-73
28 Cox P, Keshishian J, Blades B. Traumatic arteriovenous
fistula o f the c hest wall and lung secondary to insertion of an
intercostal cathete r. J Thorac Cardiovasc Surg 1966; 54:
110-12
29 Fein A, Godwin J, Moore A, et la. Systemic artery-to-
pulmonary vascular shunt: a complication of closed tube
thoracostomy. AJR Am J Roentgenol 1983; 140:917-19
30 Fleishman J, Bullock J, Rosset J, et al. Iatrogenic Horner's
syndrome secondary to chest tube thoracostomy. J Clin
Neuroophthalmol 1983; 3:205-10
31 Campbell P, Neil T, Wake P. Horne r's syndrome caused b y
an intercostal drain. Thorax 1989; 44:305-06
32 Bertino R, Wesbey G, Johnson R. Horner syndrome occur-
ring as a complication of chest tube placement. H.adiology
1987; 164:745
33 Gerard PS, Kaldawi E, Litani V, et al . Right-sided pneumo-
thorax as a result of a left-sided chest tube. Chest 1993;
103: 1602-03
34 Kollef MH, Dothager DW. H.eversible cardiogenic shock due
to c hest tube compression of th e right ventricle. Chest 1991;
99:976-80
35 Timby J, R eed C , Zeilender S, et al . ' Mechanical' causes of
pulmonary edema. Chest 1990; 98:973-79
36 Stahly T, Tench W. Lung entrapment and infarction by chest
tube suction. Radiology 1977; 122:307-09
37 Chen Y, Wu M, Lee P, et al. Necrotizing fasciitis: is it a fatal
complication of tube thoracostomy? Report of three cases.
Respir Med 1 992; 86:249-51
CHEST I 112 I 3 I SEPTEMBER, 1997 713