0
Transparency in Health Care |

Reducing Iatrogenic Risk in Thoracentesis: Establishing Best Practice Via Experiential Training in a Zero-Risk Environment FREE TO VIEW

Darlene R. Duncan, MD; Timothy I. Morgenthaler, MD, FCCP; Jay H. Ryu, MD, FCCP; Craig E. Daniels, MD, FCCP
Author and Funding Information

*From the Department of Internal Medicine (Dr. Duncan) and the Division of Pulmonary and Critical Care Medicine (Drs. Morgenthaler, Ryu, and Daniels), Mayo Clinic, Rochester, MN.

Correspondence to: Craig E. Daniels, MD, FCCP, Division of Pulmonary and Critical Care Medicine, Gonda South 18, Mayo Clinic, 200 First St SW, Rochester, MN 55905; e-mail: daniels.craig@mayo.edu


This research was supported by Mayo Clinic Institutional Funds.

The authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal.org/site/misc/reprints.xhtml).

For editorial comment see page 1120


© 2009 American College of Chest Physicians


Chest. 2009;135(5):1315-1320. doi:10.1378/chest.08-1227
Text Size: A A A
Published online

Background:  We studied the reasons why patients undergoing thoracenteses performed in our outpatient pulmonary clinic had a higher frequency of iatrogenic pneumothorax compared to that in the concurrent radiology practice in our institution, which utilizes ultrasound guidance. We reviewed our practice model and implemented a unique experiential training paradigm in a zero-risk simulation environment to improve efficacy, timeliness, service orientation, and safety.

Methods:  We retrospectively determined the rate of clinically significant pneumothoraces in our practice (phase I, July 1, 2001, to June 30, 2002). The training system redesign included the following: (1) a designated group of pulmonologist instructors dedicated to treating pleural disease and reducing the number of iatrogenic complications; (2) the use of ultrasound image guidance for all thoracenteses; and (3) structured proficiency and competency standards for proceduralists. Postintervention (phase II) data were prospectively collected (January 2005 to December 2006) and compared with our baseline data.

Results:  The baseline rate of pneumothorax was 8.6% (5 of 58 patients) in our pulmonary practice. Following intervention (phase II), the rate of pneumothorax declined to 1.1% (p = 0.0034). During phase II, the number of thoracenteses performed increased (186 vs 58 per year, respectively; p < 0.05). The iatrogenic pneumothorax rate was stable in the 2 years following intervention (2005, 0.7% [1 of 137 pneumothoraces]; 2006, 1.3% [3 of 226 pneumothoraces]; p > 0.9). Postintervention complications included procedure-related pain (n = 19), cough (n = 4), and hypotension (n = 10).

Conclusions:  An improvement program that included simulation, ultrasound guidance, competency testing, and performance feedback reduced iatrogenic risk to patients. We recommend application of this process to procedural practices.

Figures in this Article

The Mayo Clinic Pulmonary and Critical Care Medicine division is a group of 44 pulmonary and critical care specialists engaged in a large, integrated, multispecialty group practice. We have performed thoracentesis in our outpatient clinic for nearly 3 decades while providing training opportunities for our physicians who are in training (ie, fellows and internal medicine residents who are on clinical rotation). Recently, we examined closely the quality of our thoracentesis procedural practice with emphasis on the review of iatrogenic complications.

Through a retrospective investigation of our institutional thoracentesis practice (phase I), which was performed as part of a previously published study, we learned that the rate of pneumothorax was higher for thoracenteses performed in our outpatient clinic (8.6%) than for those performed concurrently elsewhere in our medical facility by our procedural radiologists utilizing ultrasound guidance (1.9%; p = 0.099).1 While our rate of iatrogenic pneumothorax following thoracentesis was acceptable when compared with other published data, with a pneumothorax rate ranging between 3% and 30%,25 we were concerned with the discrepancy between our iatrogenic pneumothorax rates compared to those of our radiology colleagues. This prompted a careful evaluation of our entire thoracentesis practice. We noted that our overall procedural volume had been decreasing while referral for thoracentesis to the radiology practice was increasing (Fig 1). As a result, our experience in performing and teaching the thoracentesis procedure was declining and was accompanied by a decrease in procedural learning opportunities for our physicians-in-training. Further, we recognized that the formal procedural instruction of our physicians-in-training was not consistent and needlessly exposed patients to learning curve risk. In addition to safety, we identified several other areas of system waste,6 including excessive and/or incomplete analysis of pleural fluid, delay in procedural scheduling, patient inconvenience, and miscommunication.7

Figure Jump LinkFigure 1 Yearly thoracentesis volume: outpatient pulmonary clinic compared with radiology. Total number of yearly thoracentesis procedures (1998 to 2003) performed using sonographic guidance by a radiologist (solid line with diamonds) vs pulmonary physicians without the use of sonography (dashed line with dots).Grahic Jump Location

Using the measures of quality proposed by the Institute for Healthcare Improvement7 and the Institute of Medicine,8 we developed and implemented a comprehensive quality improvement project with the stated goals of reducing iatrogenic thoracentesis complications in our outpatient pulmonary practice while improving efficiency, patient satisfaction, and procedural instruction for our physicians-in-training (Fig 2). In this article, we report the results of our efforts thus far.

Figure Jump LinkFigure 2 Thoracentesis quality improvement initiative. A graphic representation of the components of our project with the goal of reducing complications and improving patient care. US = ultrasound. Refer to online supplement for detailed explanation.Grahic Jump Location

All data (except the initial baseline complication data [see introductory comments]) was collected as a quality improvement initiative, but in order to report the results here, we subsequently obtained approval from the Mayo Clinic Institutional Review Board. Our methodology followed the “design, measure, analyze, improve, control” (or DMAIC) framework.9

Phase I of the study included a review of the records of all patients with pleural fluid obtained from a thoracentesis procedure that had been analyzed from July 2001 through June 2002 (n = 538), which was conducted via the Mayo Clinic-Rochester laboratory database search. We then determined the number of patients who had undergone thoracentesis that was performed in our outpatient pulmonary practice (n = 58) and their complications. Retrospective data collection for this group was previously performed and has been partially published by Barnes et al,1 along with methodology, following peer review. The procedural data and complication rate from this period of our practice served as the baseline against which we compared future interventions. Available complication data were limited to clinically significant pneumothoraces and tube thoracostomy as prospective collection of other complications (eg, cough, hypotension, and pain) were not performed. A chest radiograph to exclude postthoracentesis pneumothorax is not routinely performed as part of our standard practice.10 Therefore, the presence of pneumothorax is defined as “clinically meaningful” as was previously published by Barnes et al1 and can be summarized as follows. All chest imaging (radiographs and CT scans) performed within 30 days of the procedure were evaluated. In addition, postprocedural instruction for all patients included contact information and instruction to report any postprocedural complication resulting in care elsewhere. Medical records and physician correspondence were reviewed to determine patient-reported complications. The requirement of tube thoracostomy to treat iatrogenic pneumothorax was a clinical decision made at the discretion of the treating physician.

We analyzed our practice and the available literature to identify systematic errors, suboptimal performance characteristics, and potential evidence-based interventions. From our analysis of our clinic practice, we believed that we would need to improve the following: procedural expertise; standardization of thoracentesis equipment; availability of and competency with ultrasound guidance; and appointment availability. We then designed and implemented a comprehensive improvement project that was centered on establishing proceduralist competency and proficiency in an experiential, zero-risk environment.

Experiential Training Methodology and Quality Improvement Initiatives
Expertise:

We determined that those physicians performing outpatient thoracentesis would need to achieve expertise, as defined in the following ways:

  1. A need to demonstrate competency. All proceduralists were required to complete a half-day thoracentesis workshop. The key components of experiential training included the following: use of bedside ultrasound imaging utilizing cadaveric and inanimate models; and observed performance of thoracentesis with ultrasound imaging. Following this establishment of competency, all proceduralists performed their initial 10 thoracenteses on human subjects while under the direct supervision of an experienced proceduralist (C.E.D.) in a clinical setting. Feedback was given in real time, and if competency was not demonstrated by the tenth observed procedure, the proceduralist was not given privileges to perform a thoracentesis (all proceduralists obtained competency).

  2. A need to maintain proficiency. We required that all proceduralists complete a yearly minimum of 10 thoracenteses.

  3. The formation of a focused thoracentesis group. To maintain the highest level of proficiency, we limited thoracentesis privileges, in our large pulmonary group (n = 44), to providers who completed criteria 1 and 2 (n = 5).

The standardization of equipment and procedure included the following: (1) the use of real-time ultrasound marking and/or guidance for all thoracenteses; (2) the standardization of procedure and equipment; (3) the requirement of preprocedural verification of the correct patient by two identifiers; and (4) the requirement of preprocedural verification of the correct procedure and the correct side on which the procedure will be conducted (ie, right vs left). Improvement of appointment availability and convenience of patient scheduling included the following components: we changed our scheduling template and developed a dedicated pleural appointment calendar; and we equipped a procedural room and provided support staff (nurse or assistant) to assist with procedures as well as to provide preprocedural and postprocedural education.

With these quality improvement plans in place (January 1, 2005), we prospectively collected predefined quality metrics in all patients undergoing outpatient thoracentesis in our division. We analyzed each procedure as a separate event with a unique potential for complication regardless of whether it was an initial or repeat thoracentesis for an individual patient. A prospective collection of procedural complications (eg, cough, hypotension, syncope, respiratory distress, desaturation, and pain) was recorded by the clinician performing the procedure at the end of each procedure. Provider-specific performance data and complication data were reported to the division as part of a yearly quality improvement conference. Changes were incorporated into the procedural practice based on feedback from data review and a quality improvement conference.

For the purpose of assessing improvements, data are described as the median (range) for continuous variables, and a percentage for categorical variables. Comparisons of pneumothorax were performed using the Fisher exact test for categorical variables and the two-sample rank sum test for continuous variables. In all cases, two-tailed p values of ≤ 0.05 were considered to be statistically significant.

Demographic Data

We report the results of 363 of 367 patients (4 patients were excluded from the study due to lack of research consent) who underwent thoracentesis in our outpatient clinic in the 2 years following our quality intervention compared to the 58 preintervention patients (Table 1). More thoracenteses were performed on the right side and on men. Therapeutic thoracentesis performed solely for symptom relief was less common (18%).

Table Graphic Jump Location
Table 1 Patient Demographics*

*Values are given as the median (range) or No. (%), unless otherwise indicated.

Rate of Pneumothorax and Tube Thoracostomy

Prior to the institution of our quality improvement intervention, our rate of pneumothorax was 8.7% (5 of 58 patients), with 3 of these patients (6%) requiring tube thoracostomy drainage (Table 2). We did not identify a significant difference in the rate of iatrogenic pneumothorax between staff physicians (7.4%; 2 of 27 physicians) and physicians-in-training (9.6%; 3 of 31 physicians-in-training; p > 0.9). Following the introduction of our quality improvement intervention, our pneumothorax rate declined to 1.1% (4 of 362 patients; p = 0.0034). None of these patients required tube thoracostomy drainage (p = 0.0025). One patient was hospitalized for observation of a modest-sized pneumothorax that did not require tube thoracostomy drainage.

Table Graphic Jump Location
Table 2 Complication Rates Preintervention (2001 to 2002) and Postintervention (2005 to 2006)*

*Values are given as No. (%), unless otherwise indicated.

Rate of Other Complications

Following the quality improvement intervention, the proceduralist prospectively recorded other clinically significant complications such as pain, cough, vasovagal events, syncope, hypotension, and respiratory distress (Table 3). Our overall complication rates for the 2 years following the intervention were 9.5% and 13.3%. Thoracentesis-induced respiratory distress or new desaturation events did not occur. Physician-recorded patient chest pain was the most frequent complication (5.2%). A vasovagal event or hypotension was reported 2.7% of the time; however, all episodes were treated conservatively without the administration of IV fluids, drugs, or subsequent hospitalization. Other complications included nausea, cough, transient hypoxemia, and bloody fluid return. One patient died within 24 h after thoracentesis. This event was investigated extensively, and the coroner concluded that the patient died of “natural causes.” No thoracentesis procedural complications caused or contributed to death.

Table Graphic Jump Location
Table 3 Other Complications Postintervention*

*Values are given as No. (%).

†Cough, nausea, or transient hypoxia.

Number of Thoracenteses Performed (Preintervention and Postintervention)

After the institution of our quality initiative, the number of thoracenteses performed in our pulmonary outpatient practice increased. In the 6 years preceding our project (1998 to 2003), there was a decline in the total number of thoracenteses performed per year, reaching a low of 39 outpatient thoracenteses in 2003, while institution-wide thoracentesis procedures were continually increasing (Fig 1). After the institution of our quality improvement intervention, the mean yearly number of thoracenteses we performed significantly increased (p < 0.05).

The culture of procedural training in medicine has, for decades, been focused on experiential training using human subjects with the often-heard educational motto “see one, do one, teach one.”11 As a result, the adoption of a zero-risk environment for initial thoracentesis instruction was not previously used at our institution. The basis of our (phase II) methodology to train physicians with the use of ultrasound and to assure procedural competency included the creation of a zero-risk experiential training environment using simulated cadaveric and anatomic models. This resulted in a marked reduction in the number of iatrogenic pneumothoraces without an identifiable learning curve during the introduction of phase II. We have since required all physicians-in-training to complete similar zero-risk training for thoracentesis and to demonstrate competency in this simulation environment prior to performing thoracentesis on human subjects. We believe that this represents a best-practice model for thoracentesis training.

Prior reports of thoracentesis complications have investigated the effect of single interventions (ie, ultrasound imaging or operator experience) on the rate of thoracentesis complications. Patient factors such as body habitus, effusion size, loculated fluid, and repeat thoracentesis may all contribute.2,3,12 Operator skill and experience as well as the use of a vacuum bottle to withdraw fluid may all be independent factors contributing to thoracentesis complications.5,13 The use of sonographic guidance has been shown1,12,14 to reduce the rates of pneumothorax but has not been well studied independent of other factors. Current guidelines15 advise the routine use of ultrasound guidance when effusions are small or loculated. It remains uncertain whether the utilization of ultrasound truly reduces the rate of pneumothorax or perhaps plays a smaller role than the skill and experience of the operator.1619 While controversy exists as to whether the operator, the availability of ultrasound, or the presence of a learner is the most important contributor to complications, we suspected that all of these factors affect the risk for iatrogenic complications.

Overall, our iatrogenic complication rate, postintervention, compares favorably with the current “gold standard” of interventional radiologists practicing at a tertiary referral center.12 The comprehensive nature of our approach limits our ability to determine the contribution of individual components to the overall “success” of this project. It is likely that the introduction of ultrasound image guidance will be viewed as one of the key components in the reduction of iatrogenic complications. Another potential key in the reduction of iatrogenic complications may have been the focusing of our procedural group. We limited our thoracentesis practice from a large group (n = 44) to a focused group of pulmonologists with a pleural interest (n = 5). The net effect was that we revoked the privileges of 39 board-certified pulmonary physicians and concentrated the performance of thoracentesis procedures on the 5 focused proceduralists with training and privileges. In addition, the impact of experiential training in this procedural group, utilizing simulated thoracentesis and cadaveric models in a zero-risk environment, likely explains the immediate impact of the intervention (phase II) to reduce iatrogenic pneumothoraces without evidence for a training effect.

We view our efforts as a success. A critical component of the success of our project was the ability of our health-care team to be transparent with one another. There was an initial inclination among our group to regard our unusually high pneumothorax rate with skepticism, but, fortunately, this initial skepticism quickly gave way to genuine concern for our patients and a determination to fix the problem. Along the way, we learned to view the “problem” as one outcome variable of an overarching process, and we clarified our goals to improve the process of training and patient care, not just to reduce pneumothoraces.

Our study is not a rigorous clinical trial. We recognize that multiple interventions were implemented simultaneously, and this prevents our ability to differentiate the impact of specific interventions. Complication data from our retrospective control group were compared against prospectively collected data from our quality improvement project, and this limited our analysis. Further, we acknowledge the potential for bias introduced by physician self-reporting of complications other than iatrogenic pneumothorax. While our main outcomes were statistically significant, some may argue that the pneumothorax rate was not egregious to begin with. We are striving for rates below the “usual complication rates,” and for the highest possible safety of thoracentesis, with complication rates far < 1%.

Implementation of a multifaceted quality improvement project centered on the demonstration of competence in an experiential, zero-risk training environment coupled with the mandated use of bedside ultrasound decreased our rate of iatrogenic pneumothorax and the need for tube thoracostomy. In addition, the number of thoracenteses performed in our outpatient practice increased. We believe that this represents a best-practice model for procedural training in performing thoracentesis and that this comprehensive approach reduces the iatrogenic risk to patients.

Barnes TW, Morgenthaler TI, Olson EJ, et al. Sonographically guided thoracentesis and rate of pneumothorax. J Clin Ultrasound. 2005;33:442-446. [PubMed] [CrossRef]
 
Raptopoulos V, Davis LM, Lee G, et al. Factors affecting the development of pneumothorax associated with thoracentesis. AJR Am J Roentgenol. 1991;156:917-920. [PubMed]
 
Colt HG, Brewer N, Barbur E. Evaluation of patient-related and procedure-related factors contributing to pneumothorax following thoracentesis. Chest. 1999;116:134-138. [PubMed]
 
Seneff MG, Corwin RW, Gold LH, et al. Complications associated with thoracocentesis. Chest. 1986;90:97-100. [PubMed]
 
Bartter T, Mayo PD, Pratter MR, et al. Lower risk and higher yield for thoracentesis when performed by experienced operators. Chest. 1993;103:1873-1876. [PubMed]
 
Chalice R, Chalice R. American Society for Quality. Improving healthcare using Toyota lean production methods: 46 steps for improvement. 2007;2nd ed Milwaukee, WI ASQ Quality Press
 
Institute for Healthcare Improvement Improvement tip: find muda and root it out. 2008;Accessed November 26, 2008 Available at:www.ihi.org.
 
Committee on Quality of Health Care in America, Institute of Medicine Crossing the quality chasm: a new health system for the 21st century. 2001; Washington, DC National Academies Press
 
George Ml, Rolands D, Price M, et al. The lean six sigma pocket toolbook. 2005;1st ed New York, NY McGraw Hill
 
Capizzi SA, Prakash UB. Chest roentgenography after outpatient thoracentesis. Mayo Clin Proc. 1998;73:948-950. [PubMed]
 
Dunn WF. Education theory applied to critical care: Dewey versus tradition; it really does matter. Am J Respir Crit Care Med. 2003;167:4-5. [PubMed]
 
Jones PW, Moyers JP, Rogers JT, et al. Ultrasound-guided thoracentesis: is it a safer method? Chest. 2003;123:418-423. [PubMed]
 
Petersen WG, Zimmerman R. Limited utility of chest radiograph after thoracentesis. Chest. 2000;117:1038-1042. [PubMed]
 
Grogan DR, Irwin RS, Channick R, et al. Complications associated with thoracentesis: a prospective, randomized study comparing three different methods. Arch Intern Med. 1990;150:873-877. [PubMed]
 
Light RJ. Clinical practice: pleural effusion. Thorax. 2003;346:ii8-ii17
 
Jain M, Miller L, Belt D, et al. Decline in ICU adverse events, nosocomial infections and cost through a quality improvement initiative focusing on teamwork and culture change. Qual Saf Health Care. 2006;15:235-239. [PubMed]
 
Misset B, Timsit JF, Dumay MF, et al. A continuous quality-improvement program reduces nosocomial infection rates in the ICU. Intensive Care Med. 2004;30:395-400. [PubMed]
 
Resar R, Pronovost P, Haraden C, et al. Using a bundle approach to improve ventilator care processes and reduce ventilator-associated pneumonia. Jt Comm J Qual Patient Saf. 2005;31:243-248. [PubMed]
 
Young MP, Gooder VJ, Oltermann MH, et al. The impact of a multidisciplinary approach on caring for ventilator-dependent patients. Int J Qual Health Care. 1998;10:15-26. [PubMed]
 

Figures

Figure Jump LinkFigure 1 Yearly thoracentesis volume: outpatient pulmonary clinic compared with radiology. Total number of yearly thoracentesis procedures (1998 to 2003) performed using sonographic guidance by a radiologist (solid line with diamonds) vs pulmonary physicians without the use of sonography (dashed line with dots).Grahic Jump Location
Figure Jump LinkFigure 2 Thoracentesis quality improvement initiative. A graphic representation of the components of our project with the goal of reducing complications and improving patient care. US = ultrasound. Refer to online supplement for detailed explanation.Grahic Jump Location

Tables

Table Graphic Jump Location
Table 1 Patient Demographics*

*Values are given as the median (range) or No. (%), unless otherwise indicated.

Table Graphic Jump Location
Table 2 Complication Rates Preintervention (2001 to 2002) and Postintervention (2005 to 2006)*

*Values are given as No. (%), unless otherwise indicated.

Table Graphic Jump Location
Table 3 Other Complications Postintervention*

*Values are given as No. (%).

†Cough, nausea, or transient hypoxia.

References

Barnes TW, Morgenthaler TI, Olson EJ, et al. Sonographically guided thoracentesis and rate of pneumothorax. J Clin Ultrasound. 2005;33:442-446. [PubMed] [CrossRef]
 
Raptopoulos V, Davis LM, Lee G, et al. Factors affecting the development of pneumothorax associated with thoracentesis. AJR Am J Roentgenol. 1991;156:917-920. [PubMed]
 
Colt HG, Brewer N, Barbur E. Evaluation of patient-related and procedure-related factors contributing to pneumothorax following thoracentesis. Chest. 1999;116:134-138. [PubMed]
 
Seneff MG, Corwin RW, Gold LH, et al. Complications associated with thoracocentesis. Chest. 1986;90:97-100. [PubMed]
 
Bartter T, Mayo PD, Pratter MR, et al. Lower risk and higher yield for thoracentesis when performed by experienced operators. Chest. 1993;103:1873-1876. [PubMed]
 
Chalice R, Chalice R. American Society for Quality. Improving healthcare using Toyota lean production methods: 46 steps for improvement. 2007;2nd ed Milwaukee, WI ASQ Quality Press
 
Institute for Healthcare Improvement Improvement tip: find muda and root it out. 2008;Accessed November 26, 2008 Available at:www.ihi.org.
 
Committee on Quality of Health Care in America, Institute of Medicine Crossing the quality chasm: a new health system for the 21st century. 2001; Washington, DC National Academies Press
 
George Ml, Rolands D, Price M, et al. The lean six sigma pocket toolbook. 2005;1st ed New York, NY McGraw Hill
 
Capizzi SA, Prakash UB. Chest roentgenography after outpatient thoracentesis. Mayo Clin Proc. 1998;73:948-950. [PubMed]
 
Dunn WF. Education theory applied to critical care: Dewey versus tradition; it really does matter. Am J Respir Crit Care Med. 2003;167:4-5. [PubMed]
 
Jones PW, Moyers JP, Rogers JT, et al. Ultrasound-guided thoracentesis: is it a safer method? Chest. 2003;123:418-423. [PubMed]
 
Petersen WG, Zimmerman R. Limited utility of chest radiograph after thoracentesis. Chest. 2000;117:1038-1042. [PubMed]
 
Grogan DR, Irwin RS, Channick R, et al. Complications associated with thoracentesis: a prospective, randomized study comparing three different methods. Arch Intern Med. 1990;150:873-877. [PubMed]
 
Light RJ. Clinical practice: pleural effusion. Thorax. 2003;346:ii8-ii17
 
Jain M, Miller L, Belt D, et al. Decline in ICU adverse events, nosocomial infections and cost through a quality improvement initiative focusing on teamwork and culture change. Qual Saf Health Care. 2006;15:235-239. [PubMed]
 
Misset B, Timsit JF, Dumay MF, et al. A continuous quality-improvement program reduces nosocomial infection rates in the ICU. Intensive Care Med. 2004;30:395-400. [PubMed]
 
Resar R, Pronovost P, Haraden C, et al. Using a bundle approach to improve ventilator care processes and reduce ventilator-associated pneumonia. Jt Comm J Qual Patient Saf. 2005;31:243-248. [PubMed]
 
Young MP, Gooder VJ, Oltermann MH, et al. The impact of a multidisciplinary approach on caring for ventilator-dependent patients. Int J Qual Health Care. 1998;10:15-26. [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).
Supporting Data
Data Supplement

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
Ultrasound-Guided Thoracentesis*: Is It a Safer Method?
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543