Keywords
ultrasound - thoracentesis - pneumothorax - chest radiograph - chest tube placement
- drainage catheter placement - interventional radiology
Introduction
Pleural effusions are commonly treated with thoracentesis as part of diagnostic and
therapeutic management strategies. In the United States, a pleural effusion is diagnosed
in 1.5 million patients annually.[1] Thoracentesis involves percutaneous aspiration of pleural fluid through the chest
wall. Common adverse events include pneumothorax (4–39%), cough (24%), dyspnea (15%),
chest pain (5%), and vasovagal reactions (3%).[2]
[3] Less common complications include inadvertent liver or splenic laceration, hemorrhage,
infection, and re-expansion pulmonary edema.[2]
[3]
Ultrasound guidance is increasingly utilized for thoracentesis as it has been shown
to reduce the risk of pneumothoraces (< 3% overall with a 90% risk reduction compared
with nonultrasound-guided procedures) and has been associated with lower total hospital
costs. Despite the reduction in pneumothoraces, it is still common practice to obtain
routine post-procedure chest radiographs after thoracentesis. There are no consensus
or societal guidelines that recommend or refute the utility of chest radiographs after
thoracentesis.
The purpose of this study was to report the clinical utility of chest radiography
following interventional radiology-performed ultrasound-guided thoracentesis.
Materials and Methods
Study Design
This study was conducted with Institutional Review Board approval (HUM00151864 and
18–3202) at two institutions and complied with the Health Insurance Portability and
Accountability Act. Informed consent was not required for this retrospective study.
This study was assessed using the Strengthening the Reporting of Observational studies
in Epidemiology (STROBE) guidelines.[4] Patients with pleural effusions identified on chest radiography or computed tomography
and who underwent thoracentesis between 2003 and 2018 were identified.
Inclusion and Exclusion Criteria
A total of 3,998 patients underwent thoracentesis. A total of 3,022 (75.6%) patients
were older than 18 years old, underwent interventional radiology-performed ultrasound-guided
thoracentesis, and had same-day (< 24 hours) post-procedure chest radiograph evaluation.
Patients younger than 18 years old, those who underwent thoracentesis by noninterventional
radiology specialties, or those without same-day chest radiograph evaluation were
excluded (n = 976). The decision to obtain a radiograph was based on the individual provider
and patient’s clinical status.
Thoracentesis and Pleural Drainage Catheter Placement Techniques
Thoracentesis procedures have been previously described.[5]
[6] All procedures were performed by an interventional radiologist using ultrasound
guidance. Ultrasound-guided thoracentesis was performed using standard operating procedure
and a generic 5-French catheter (end-hole, nonvalved) with syringe or vacuum container
at the operator’s discretion. Procedures were terminated when 1.5 L of fluid was removed;
patient experienced discomfort or excessive coughing, or at operator discretion. Pleural
drainage catheters were placed to treat the pneumothorax, under fluoroscopic or computed
tomography-guidance, based on operator’s discretion. Catheter size was based on operator’s
discretion.
Variables and Definitions
Patient age (years), laterality of thoracentesis (left, right, or bilateral), procedural
technical success, volume of fluid removed (mL), method of post-procedure chest imaging
(anteroposterior [AP or portable], posteroanterior [PA or one view], PA and lateral
[two view], or PA, lateral, and left lateral decubitus [three view] chest radiography),
absence or presence of pneumothorax, pneumothorax size (mm), pneumothorax management
measures, and post-procedure clinical outcomes were recorded. Technical success was
defined as successful aspiration of pleural fluid. Method of post-procedure chest
imaging was determined by operator preference. Pneumothorax was defined as any abnormal
air in the pleural cavity. Pneumothorax size was recorded based on the longest distance
from the chest wall to the lung.[7]
[8] Pneumothorax management measures included repeat chest imaging or placement of a
pleural drainage catheter. The pleural drainage catheter size was recorded in French.
Post-procedure clinical outcomes included new patient-perceived dyspnea and hypoxia
(oxygen saturations < 90% on room air).
Cost Estimation
Current Procedural Terminology (CPT) codes for one-, two-, and three-view chest radiographs
were obtained from publicly available government reimbursement schedules. CPT codes
included 71,045 (chest one view), 71,046 (chest two views), and 71,047 (chest three
views). Mean costs associated with chest radiographs after thoracentesis were estimated
using Medicare and Medicaid fee schedules.[9]
[10] Medicare reimbursement of one-, two-, and three-view radiographs were $23.78, $30.19,
and $37.95, respectively, with weighted mean of $29.84.[9] Medicaid reimbursements for these three examinations were $13.87, $17.63, and $22.19,
with weighted mean of $17.42.[10]
Statistical Analyses
All statistical analyses, including means, percentages, and standard deviations, were
calculated using R software version 3.2.2 (R Core Team).
Results
A total of 3,022 (75.6%) patients were older than 18 years old, underwent interventional
radiology-performed ultrasound-guided thoracentesis, and had same-day post-procedure
chest radiograph evaluation. Mean patient age was 56.7 ± 15.5 years. Laterality of
thoracentesis, volume of fluid removed, method of post-procedure chest imaging, absence
or presence of pneumothorax, and pneumothorax size are shown in [Table 1]
. Of the total 3,022 procedures, 1,531 (50.7%; 1,531/3,022), 1,477 (48.9%; 1,477/3,022),
and 14 (0.5%; 14/3,022) were performed on left, right, or bilaterally in a single
session. Technical success was 100% (n = 3,022). A mean volume of 940 ± 550 mL fluid was removed in a single session. Post-procedure
imaging was performed in the form of PA (2.6%; 78/3,022), AP (17.0%; 513/3,022), PA
and lateral (77.9%; 2,355/3,022), or PA, lateral, and left lateral decubitus (2.5%;
76/3,022) chest radiographs. Post-procedural pneumothorax was identified in 21 patients
with a total incidence of 0.69% (n = 21/3,022). Mean, pneumothorax size, measured on chest radiograph as the longest
distance from the chest wall to the pleural reflection, was 18.8 ± 10.2 mm (range:
5.0–35.0 mm).
Table 1
Thoracenteses and post-procedural chest radiograph results
|
Variable
|
Outcome
|
|
Numerical measures are summarized as mean (standard deviation) and categorical variables
are shown as number of cases (percentage of 3022).
|
|
Laterality of effusion
|
Left
|
1,531 (50.7%)
|
|
Right
|
1,477 (48.9%)
|
|
Bilateral
|
14 (0.5%)
|
|
Mean volume removed
|
940 mL (550 mL)
|
|
Chest radiograph technique
|
Portable
|
78 (2.6%)
|
|
One view
|
513 (17.0%)
|
|
Two views
|
2,355 (77.9%)
|
|
Three views
|
76 (2.5%)
|
|
Post-procedural pneumothorax
|
21 (0.69%)
|
|
Mean pneumothorax size
|
18.8 (10.2 mm)
|
Pneumothorax management and clinical outcomes are shown in [Table 2]. Of the 21 pneumothoraces, seven (33.3%) were asymptomatic and resolved spontaneously
and had a mean size of 6.4 ± 2.4 mm. Fourteen pneumothoraces, of mean size 25.0 ±
5.8 mm, required management with a pleural drainage catheter (66.6%; 14/21). Nine
(42.8%; 9/21) were managed with 10-French pigtail placement and five (23.8%; 5/21)
were managed with larger chest tube placement, based on operator discretion. Two chest
tubes were 20-French, two were 21-French, and one was 22-French. The overall incidence
of pneumothorax requiring pleural drainage catheter placement following interventional
radiology-performed ultrasound-guided thoracentesis was 0.46% (14/3,022). Of the patients
requiring drainage catheter placement, 12 (85.7%) and 13 (92.9) had dyspnea and hypoxia,
respectively.
Table 2
Outcomes of pneumothoraces
|
Variable
|
Outcome
|
|
Numerical measures are summarized as mean (standard deviation) and categorical variables
are shown as number of cases (percentage of 21).
|
|
Method of pneumothorax resolution
|
Spontaneous
|
7 (33.3%)
|
|
10-French pigtail catheter
|
9 (42.8%)
|
|
Chest tube
|
5 (23.8%)
|
|
Chest tube size
|
20-French
|
2 (9.5%)
|
|
21-French
|
2 (9.5%)
|
|
22-French
|
1 (4.8%)
|
|
Mean pneumothorax size
|
Not requiring intervention
|
6.4 mm (2.4 mm)
|
|
Requiring intervention
|
25.0 mm (5.8 mm)
|
There were 937 patients older than 65 years; thus, the cost to the Medicare population
was $27,547. The proportion of adults in Medicaid was 20.1% at the time of this manuscript,
equating to 607 patients in this study; thus, the cost to Medicaid was $10,581.
Discussion
With the advent of ultrasound-guided thoracentesis, the overall incidence and risk
of post-procedural complications have been significantly reduced, especially pneumothorax
rates.[11] In this study, the overall incidence of iatrogenic pneumothorax following interventional
radiology-performed ultrasound-guided thoracentesis was 0.69%. Previous studies have
reported the incidence of thoracentesis-related pneumothorax ranging from 4 to 39%
without the use of ultrasound[2]
[3] and more recent, smaller investigations have found the pneumothorax rates ranging
from 0 to 3% with the use of ultrasound.[5]
[6]
[12]
[13]Overall, ultrasound guidance has led to a 90% reduction in the incidence of iatrogenic
pneumothorax compared with nonimage-guided thoracentesis.
Recently, two large studies of 9,320 and 3,067 patients reported a 0.61 and 0.62%
incidence of thoracentesis-related pneumothorax, respectively, which is similar to
the 0.69% reported in this study.[14]
[15] Ault et al included thoracenteses that were performed, or supervised, by a single
internist, whereas this study included thoracenteses performed by interventional radiologists.[14] Cho et al described thoracenteses performed in the emergency department by emergency
medicine physicians, internists, and radiologists.[15] Taken as a whole, considering the various clinical situations and operators, the
similar low rate of pneumothorax suggests the marked efficacy of ultrasound and resultant
consistency across operator specialties.
When rare post-procedural pneumothoraces did occur, 33.3% (7/21) were asymptomatic
and resolved spontaneously. About 0.46% (14/3,022) of patients who underwent interventional
radiology-performed ultrasound-guided thoracentesis required pleural drainage catheter
placement. Moreover, of the patients requiring drainage catheter placement, the majority
of patients was symptomatic with 12/14 (85.7%) and 13/14 (92.9%) experiencing new
dyspnea and hypoxia, respectively. Cho et al reported a similarly low rate of chest
tube insertion in 36.8% of those patients (0.2% overall) for the management of iatrogenic
pneumothorax in the emergency department.[15] Other studies reported chest tube insertion following 1.7% of all thoracentesis
procedures to evacuate symptomatic pneumothoraces.[6] Pneumothorax identification, however, does not lead to patient benefit in all circumstances
and may increase costs related to unnecessary observation or treatment. In cases of
asymptomatic pneumothorax ex vacuo, or pneumothorax resulting from failed re-expansion
of noncompliant lung rather than disruption of the visceral pleura, drain placement
provides little benefit and may expose patients to iatrogenic harm.[16]
[17]
[18]
Despite the low incidence of iatrogenic pneumothorax following ultrasound-guided thoracentesis
(0.69%) and even lower requirement rates of pleural drainage catheter placement (0.46%),
it is still common clinical practice to obtain a routine post-thoracentesis chest
radiograph regardless of patient symptoms. Because follow-up chest radiography protocol
is not standardized, various combinations of portable, single-, or multiple-view radiographs
without or with expiration are performed. In this study, chest radiograph was performed
in the form of portable (78; 2.6%), one view (513; 17.0%), two views (2,355; 77.9%),
and three views (76; 2.5%). A single-view chest radiograph costs $23.78, while a two-view
chest radiograph costs $30.19 to Medicare.[9]
[10] Thoracentesis procedures are performed in 173,000 patients annually with most patients
receiving some form of post-procedural follow-up chest radiograph.[19]
[20]
[21]
[22]
[23] If the same proportion of patients in the United States qualified for Medicare and
Medicaid as those in this study, there would be 53,640 and 34,600 thoracentesis patients
receiving Medicare and Medicaid benefits, respectively. At a minimum, if all patients
received a single-view chest radiograph at $23.78, the total cost to Medicare would
be $1.28 million per year. Total cost to Medicaid at $13.87 would equate to $480,000
per year. While computed tomography remains the gold standard for the diagnosis of
pneumothorax, the utilization of point-of-care ultrasound may provide a cost-effective
alternative to screen for pneumothorax in symptomatic patients following thoracentesis.[24]
The routine practice of obtaining immediate post-procedure follow-up chest radiograph
comes with little tangible patient benefit and at significant health care costs. In
asymptomatic patients, without dyspnea or hypoxia, routine chest radiographs are of
limited clinical utility immediately after interventional radiology-performed ultrasound-guided
thoracentesis.
Limitations of the present study relate in part to the retrospective study design.
Mechanically ventilated patients are at an increased risk of pneumothorax following
thoracentesis compared with nonventilated patients and, as a result, the complications
within the present analysis may have occurred disproportionally among the ventilated
population.[12] More broadly, limited data on the indication for thoracentesis and the etiology
of the pleural effusions in each patient inhibits identification of clinical scenarios
or patient subpopulations where routine post-procedure radiography may be value-added.
Chest radiograph follow-up technique was not standardized with some patients receiving
supine portable or single-view radiographs that may have limited detection of pneumothorax.
Furthermore, since all thoracentesis procedures were performed by interventional radiologists,
with expertise in image-guided interventions, the generalizability of the results
may be limited among other medical specialties. Further, the level of training and
expertise of the interventional radiologists performing the procedures was not controlled
for and may have been correlated with procedural complications. While presumed to
be significant, practice costs related to routine post-thoracentesis radiography,
including potentially unnecessary monitoring and treatment of asymptomatic patients,
are beyond the scope of the economic analysis presented.
Conclusion
The incidence of clinically significant pneumothorax requiring catheter drainage following
interventional radiology-operated ultrasound-guided thoracentesis is incredibly low
(0.46%), and routine post-procedure chest radiographs in asymptomatic patients provide
little value. Reserving post-procedure chest radiographs for patients with new post-procedure
dyspnea or hypoxia will result in more efficient resource utilization and health care
cost savings.
