Chest Email Content Delivery
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (20)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Diacon, A. H.
Right arrow Articles by Solèr, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Diacon, A. H.
Right arrow Articles by Solèr, M.
(Chest. 2003;123:436-441.)
© 2003 American College of Chest Physicians

Accuracy of Pleural Puncture Sites*

A Prospective Comparison of Clinical Examination With Ultrasound

Andreas H. Diacon, MD; Martin H. Brutsche, MD, PhD and Markus Solèr, MD, FCCP

* From the Pulmonary Division, Department of Internal Medicine, University Hospital, Basel, Switzerland.

Correspondence to: Andreas H. Diacon, MD, Department of Internal Medicine, Tygerberg Hospital, PO Box 19063, 7505 Tygerberg, South Africa; e-mail: ahd{at}sun.ac.za


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Study objective: To assess the value of chest ultrasonography vs clinical examination for planning of diagnostic pleurocentesis (DPC).

Design: Prospective comparative study.

Setting: Pulmonary unit of a tertiary teaching hospital.

Patients and participants: Sixty-seven consecutive patients referred to 30 physicians of varying degrees of experience for DPC.

Interventions: Based on clinical data and examination, physicians determined whether and where a DPC should be performed. Selected puncture sites were evaluated with ultrasound and considered accurate when >= 10 mm fluid perpendicular to the skin were present.

Measurements and results: In 172 of 255 cases (67%), a puncture site was proposed. Twenty-five sites (15%) were found to be inaccurate on ultrasound examination, and a different, accurate site was established in 20 of these cases. Physicians were unable to locate a puncture site in 83 cases (33%). Among these, ultrasound demonstrated an accurate site in 45 cases (54%), while a safe tap was truly impossible in 38 cases (46%). Overall, ultrasound prevented possible accidental organ puncture in 10% of all cases and increased the rate of accurate sites by 26%. The sensitivity and specificity for identifying a proper puncture site with clinical examination compared to ultrasound as the "gold standard" were 76.6% and 60.3% (positive and negative predictive values, 85.5% and 45.8%, respectively). Risk factors associated with inaccurate clinical site selection were as follows: small effusion (p < 0.001), evidence of fluid loculation on chest radiography (p = 0.01; relative risk, 7.8; 95% confidence interval, 1.9 to 32.9), and sharp costodiaphragmatic angle on chest radiography (p < 0.001; relative risk, 7.0; 95% confidence interval, 2.3 to 15.2). Experienced physicians did not perform better than physicians in training.

Conclusions: Puncture site selection with bedside ultrasonography increases the yield of and potentially reduces complication rate in DPC. Physician experience does not predict the accuracy of selected puncture sites.

Key Words: percussion • physical examination • pleural ultrasonography • thoracentesis


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
A variety of disorders can cause pleural effusions, and physicians in many disciplines frequently perform diagnostic pleurocentesis (DPC) in order to obtain fluid for analysis. DPC in patients without effusion or punctures at inappropriate sites can lead to dry punctures, patient discomfort, and potentially to procedure-related complications.1 2 3 Therefore, safety and efficacy of DPC depend on proper patient and puncture site selection.

A pleural effusion is usually suspected on chest radiography (CXR), which indicates the approximate location. The exact puncture site is determined clinically by physical examination of the chest, notably by chest percussion.4 Ultrasound is superior to CXR in identifying pleural fluid collections,5 and the value of chest ultrasonography for thoracocentesis was established more than a decade ago.1 6 7 8 9 However, a reduction in complications or dry taps has not consistently been shown. The routine use of ultrasonography for DPC in unselected patients has not been recommended because of lack of cost-effectiveness data.4 7

Most of the studies on ultrasound for DPC have been done when ultrasound was an expensive and stationary machinery operated only by radiologists. Frequently, patients required transfer for the examination, causing inconvenience and delay, and finally were punctured by a different physician in a potentially different body position than during the ultrasound examination. In recent years, considerable technical development has taken place and the advent of affordable and transportable ultrasound units has made the technique accessible for the physician at the bedside, and on site ultrasonography by the physician performing the puncture is increasingly popular. A recent study using a mobile ultrasound unit showed excellent yield as well as a low complication rate of DPC in patients receiving mechanical ventilation, but data on the routine clinical use of ultrasound for DPC in unselected patients are lacking.10

This prospective study was designed to compare clinical examination with bedside ultrasound for patient and puncture site selection in patients scheduled for DPC. Furthermore, risk indicators associated with inaccurate clinical puncture site selection were investigated.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Subjects, Staff, and Location
The study took place in the respiratory unit of a tertiary teaching hospital. Over a 6-month period, all consecutive patients referred to the unit for the assessment of a presumed pleural effusion were considered for the study if a recent CXR was available and the patient was able to undergo the standard investigational procedure described below. Subjects gave informed consent. Patients receiving mechanical ventilation were not included.

Investigational Procedure
Data Collection: Age and gender were recorded as well as the presumed diagnosis at the time of the referral (infectious, malignant, trauma, other). The effusion was classified as left, right, or bilateral. On CXR, the extent of the effusion was categorized according to predefined illustrations into small (blunting the costodiaphragmatic angle, diaphragm still visible), moderate (diaphragm not visible up to 50% of the hemithorax) or large (>= 50% of the hemithorax). Signs of a loculated effusion were recorded as well as the shape of the costodiaphragmatic angle as blunt (>= 90°) or sharp (< 90°). Decubitus radiographs were not available. The participating physicians were qualified depending on clinical experience as "experts" (respiratory consultants), "experienced" (senior registrars of the Lung Unit), and "inexperienced" (junior medical staff).

Investigational Setting: The patient was positioned sitting upright on the side of a bed or stretcher, facing away from the investigator with arms crossed and resting on a bedside table. The updated patient file, nursing information, and recent CXR were provided. One by one, the participating physicians were invited to examine the patient with the clinical methods deemed appropriate, and then asked for their opinion as to whether and where a DPC could be done safely. The physicians could abstain from indicating a puncture site if they found it unsafe to do so, or if they believed that further testing, eg ultrasonography, was imperative for safety reasons. If a puncture site was indicated by the candidate, the spot was recorded with a coordinate system marked previously on the patients back, so that the following candidate could not be biased.

Ultrasound and Rating of Accuracy of Sites: A chest ultrasound was performed by the main investigator immediately following the clinical assessment (Aloka Echo Camera SSD-650, sonar probe UST-5035-3.5 with 3.5 MHz; Aloka; Zug, Switzerland). At the indicated spots for DPC, the depth of pleural fluid perpendicular to the skin was registered and rated as follows: No fluid (0 mm), 1 to 9 mm of fluid, or >= 10 mm of fluid. The minimal depth of pleural fluid necessary for a safe DPC is not known. We arbitrarily rated a spot with >= 10 mm of pleural fluid as accurate, and a site with no fluid was considered inaccurate. Although a site with 1 to 9 mm of fluid might yield fluid on puncture, such a site was considered unsafe and therefore inaccurate in the sense of this study. After all study-related investigations were done, the main investigator performed the DPC based on the ultrasound findings.

Statistical Evaluation
Sample Size Calculation: We estimated that 20% of the referred patients would not have an effusion suitable for pleurocentesis. Assuming an increase in the accuracy of puncture sites using ultrasound over physical examination of 20%, a sample size of 100 observations would be necessary for statistical significance (power, 0.8; {alpha}, 0.05).

Definitions for Effectiveness of the Methods: If the clinical decision was to proceed with thoracocentesis and the site was rated as accurate on ultrasonography, the attempt was scored as true-positive, and a decision to abstain from tapping when no site with >= 10 mm of fluid was identified on ultrasonography as true-negative; false-positive cases were incorrectly assigned puncture sites with < 10 mm of fluid, and false-negative ratings were given when no puncture was attempted despite a fluid collection of >= 10 mm detectable on ultrasonography.

Statistical Analysis
Statistical analysis was performed with the SPSS/PC software package (version 9.0; SPSS; Chicago, IL). Values were expressed as means ± SD, and a statistical significance level of 0.05 was used. Two-group comparisons were done with two-sided t tests. Risk indicators for inadequate puncture site identification were assessed with Cox proportional hazards regression analysis. Variables significant after univariate Cox regression analysis were considered potential predictors of inadequate puncture site identification, and further tested as covariates in the stepwise multivariate Cox regression analysis to identify independent predictors.


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Patients, Physicians, and Investigations Undertaken
Sixty-seven consecutive patients with pleural effusions of unknown origin participated in the study (Table 1 ). Malignancy was the most frequently suspected cause of effusion, but often no clear clinical suggestion was documented by the referring physicians. Every patient was reviewed by 2 to 6 physicians (mean 3.8); overall, 30 physicians assessed 255 cases for DPC (Table 2 ). In all cases, the physicians used chest percussion with or without other physical investigation techniques, mostly tactile fremitus and auscultation.


View this table:
[in this window]
[in a new window]

  		Table 1.. Patients and Effusions*

 

View this table:
[in this window]
[in a new window]

  		Table 2.. Physicians and Examinations

 
Outcome of Investigations
In 172 of 255 attempts (67%), the investigator proposed a puncture site, all of which were either dorsal or lateral (Fig 1 ). On ultrasonography, 147 sites (85% of indicated sites and 58% overall) were accurate. In 25 attempts in 18 patients (15% of indicated sites and 10% overall), < 10 mm of fluid was found. In 17 of these, a solid structure would have been punctured (liver or spleen [n = 12] and lung [n = 5]). This means that ultrasonography prevented potential complications in 15% of clinically determined puncture sites.



View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1.. Synopsis of outcome of all investigations.

 
Punctures were refused by all investigators in 8 patients, and by at least one investigator in 32 patients. Out of the 83 of 255 rejected attempts (33%), a site with at least 10 mm of fluid could be found in 45 attempts (18%), and in 38 attempts (15%) no fluid collection suitable for puncture was detected. Moreover, in 20 of the 25 inaccurate puncture sites indicated (8%), ultrasonography was able to determine a better, accurate site, while in the remaining 5 sites no effusion was detectable on ultrasonography. This means that ultrasound increased the number of accurate sites detected compared to chest percussion by 26% (p < 0.001, Wilcoxon signed rank test).

A fluid collection of >= 10 mm was present in 212 cases (83%) on ultrasonography. Compared to ultrasound as the "gold standard," clinical examination had a sensitivity and specificity of 76.6% and 60.3%, respectively, to correctly identify a site with fluid depth of at least 10 mm. The corresponding positive and negative predictive values were 85.5% and 45.8%, respectively, compared to ultrasound (100% per definition).

Indicators of Risk
Cases with inaccurate puncture sites were examined for the presence of predefined risk indicators, including the following: age, sex and presumed diagnosis, side of effusion, presence of bilateral effusion, estimated effusion extent on CXR, time lapse between CXR and study day, evidence of loculations on CXR, shape of costodiaphragmatic angle on CXR, and physician experience. Some of these risk indicators have been described in previous reports.9 11 12 Three aspects were associated with inaccuracy of puncture sites in univariate and multivariate analysis: small effusion (p < 0.001), radiologic evidence of fluid loculation (p = 0.01; relative risk, 7.8; 95% confidence interval, 1.9 to 32.9), sharp costodiaphragmatic angle on CXR (p < 0.001; relative risk, 7.0; 95% confidence interval, 2.3 to 15.2). Physician experience was not correlated to outcome.

Table 3 shows all inaccurately chosen puncture sites and the according indicators of risk. No patient featured all three indicators, but two indicators and a single risk factor were present in 7 cases and 11 cases, respectively. Seven inaccurate sites were chosen in cases without any risk indicators (experts [n = 3], experienced [n = 2], inexperienced [n = 2]). Therefore, no constellation suitable for safe clinical puncture site selection can be defined.


View this table:
[in this window]
[in a new window]

  		Table 3.. List of Errors and Risk Indicators

 

   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Clinical assessment with CXR and physical examination of the chest are the recommended standard methods utilized to locate a suitable site for DPC in patients with suspected pleural effusion.4 However, bedside chest ultrasonography has become increasingly popular for this purpose, although its impact on clinical practice had not yet been studied. In this prospective comparative study, we found bedside sonography to increase the number of patients amenable to DPC by 26% compared to clinical assessment including chest percussion. Ultrasonography identified clinically undefinable puncture sites in 18% of cases and allowed to rectify inaccurately placed sites in 8%. Furthermore, ultrasonography avoided punctures on dry chests in 2%. In comparison to ultrasound as the "gold standard," a sensitivity and specificity for clinical assessment of only 76.6% and 60.3%, respectively, were found.

One of the strengths of this study was the large spectrum of patients and staff involved, reflecting the diversity that would normally be encountered on a busy medical service of an academic center. It was an advantage, that pleural procedures in medical patients were almost exclusively carried out by the respiratory unit in our institution. A selection bias of any kind, for example toward more challenging cases, was therefore unlikely to confound our results; however, two possible biases must be discussed. Firstly, the immediate demonstration of the ultrasound findings could have resulted in a learning bias during the study, boosting the experience of less-trained physicians. This teaching effect was, however, one of the incentives of the study. The small number of trials executed per physician in the inexperienced group made a substantial learning bias unlikely. Secondly, the selection of an improper puncture site was not to inflict harm to the subject, and clinical examinations might have been executed less carefully. However, the candidates were highly motivated, and the immediate demonstration of the ultrasound finding was a strong incentive to perform well. On the whole, we are confident that the study results were valid and the findings globally relevant for many medical teaching units.

It is not surprising that a new method based on modern technology was more sensitive than a traditional physical examination technique. The low specificity of clinical assessment with chest percussion, however, is worrying. In almost half of the patients, the physicians were not unanimous in their decision whether or not to puncture. Most of the inaccurate sites were located too low, and in 10% of the puncture sites a solid organ would have been penetrated. These results are in keeping with other published prospective studies on DPC reporting dry taps in 10 to 13%.1 2 3 No ultrasound was used in these studies. Pneumothorax occurred in 11 to 30% of punctures, and hemothorax, subdiaphragmatic hematoma, and lacerations of subdiaphragmatic organs were other serious adverse events. Reported complication rates after DPC vary widely,13 but it seems reasonable to believe that puncture sites with no or very small fluid collections predispose to complications, and avoiding punctures at inaccurate sites using bedside sonography appears to be a promising approach to reducing complications in DPC.

A study by Weingardt et al12 showed that clinical failures of DPC were associated with sonographically small effusions and loculated pleural effusions. Albeit methodologies were different, the predictive value of these indicators was confirmed in our study, where radiologically small effusions and signs of loculation on CXR were associated with inaccurate puncture site placement, as well as a sharp costodiaphragmatic angle. With avoidance of blind punctures in patients with one or more of these risk indicators, the rate of inaccurate puncture sites in our study would theoretically drop from 10 to 4%, but also the rate of possible punctures would be reduced from 67 to 49%. Based on our data and other studies, we do not believe that these figures can be further improved without employing ultrasound routinely.

Physician experience, to our surprise, had no influence on outcome, unlike in the results reported by Bartter et al11 comparing complication rates of trained staff performing DPC to anecdotal control subjects with less-experienced physicians. We excluded relative overexposure of any group of physicians to risk constellations or overproportional refusal of punctures by any group in our study (data not shown). One could argue that the complication rate is less influenced by puncture site selection and rather by instrument handling, which was not assessed in our study. However, in the study of Bartter et al,11 the yield of DPC performed by experienced staff was as much improved as the complication rate, which is only possible with better puncture site selection. Apart from mentioning that chest percussion and tactile fremitus are rather simple clinical skills that can be mastered with little training, we are not able to explain why the clinical circumstances and the radiologic appearance have not allowed the more experienced physicians to recognize hazardous situations better than their less-experienced colleagues.

From a practical point of view, it may not be feasible to use bedside ultrasound every time to locate a site for DPC. As a consequence of this study, we encourage all physicians involved in diagnostic pleural procedures to acquire basic training in ultrasound. We recommend bedside sonography for puncture site selection whenever possible, with the patient in the same body position as for the following puncture. In order to increase yield and safety of DPC, ultrasound assistance is particularly useful in small or loculated effusions, and in effusions with a sharp costodiaphragmatic angle on CXR.


   Acknowledgements
 
Drs. M. Michot, W. Strobel, and E. Ullmer coordinated the study when Dr. Diacon was not available. Professors J.A. Schifferli and A.P. Perruchoud supported the study.


   Footnotes
 
Abbreviations: CXR = chest radiography; DPC = diagnostic pleurocentesis

Received for publication February 27, 2002. Accepted for publication August 27, 2002.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Grogan, DR, Irwin, RS, Channick, R, et al (1990) Complications associated with thoracentesis: a prospective, randomized study comparing three different methods. Arch Intern Med 150,873-877[Abstract]
  2. Collins, TR, Sahn, SA Thoracocentesis: clinical value, complications, technical problems, and patient experience. Chest 1987;91,817-822[Abstract/Free Full Text]
  3. Seneff, MG, Corwin, RW, Gold, LH, et al Complications associated with thoracocentesis. Chest 1986;90,97-100[Abstract/Free Full Text]
  4. Light, RW Thoracocentesis and pleural biopsy. Light, RW eds. Pleural diseases. 2001,358-377 Lippincott, Williams, & Wilkins Philadelphia, PA.
  5. Brandt, WE The thorax. Rumack, MC Wilson, SR Charboneau, JW eds. Diagnostic ultrasound. 1998,575-597 Mosby St. Louis, MO.
  6. O’Moore, PV, Mueller, PR, Simeone, JF, et al Sonographic guidance in diagnostic and therapeutic interventions in the pleural space. AJR Am J Roentgenol 1987;149,1-5[Abstract/Free Full Text]
  7. Kohan, JM, Poe, RH, Israel, RH, et al Value of chest ultrasonography vs decubitus roentgenography for thoracentesis. Am Rev Respir Dis 1986;133,1124-1126[ISI][Medline]
  8. Lipscomb, DJ, Flower, CD, Hadfield, JW Ultrasound of the pleura: an assessment of its clinical value. Clin Radiol 1981;32,289-290[CrossRef][ISI][Medline]
  9. 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[Abstract/Free Full Text]
  10. Lichtenstein, D, Hulot, JS, Rabiller, A, et al Feasibility and safety of ultrasound-aided thoracentesis in mechanically ventilated patients. Intensive Care Med 1999;25,955-958[CrossRef][ISI][Medline]
  11. 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[Free Full Text]
  12. Weingardt, JP, Guico, RR, Nemcek, AA, et al Ultrasound findings following failed, clinically directed thoracenteses. J Clin Ultrasound 1994;22,419-426[ISI][Medline]
  13. Colt, HG, Brewer, N, Barbur, E Evaluation of patient-related and procedure-related factors contributing to pneumothorax following thoracentesis. Chest 1999;116,134-138[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 ChestHome page
J. Heidecker, J. T. Huggins, S. A. Sahn, and P. Doelken
Pathophysiology of pneumothorax following ultrasound-guided thoracentesis.
Chest, October 1, 2006; 130(4): 1173 - 1184.
[Abstract] [Full Text] [PDF]

Home page
 ChestHome page
D. Feller-Kopman
Ultrasound-Guided Thoracentesis
Chest, June 1, 2006; 129(6): 1709 - 1714.
[Abstract] [Full Text] [PDF]

Home page
 ChestHome page
M. H. Brutsche, G.-F. Tassi, S. Gyorik, M. Gokcimen, C. Renard, G. P. Marchetti, and J.-M. Tschopp
Treatment of Sonographically Stratified Multiloculated Thoracic Empyema by Medical Thoracoscopy
Chest, November 1, 2005; 128(5): 3303 - 3309.
[Abstract] [Full Text] [PDF]

Home page
 ChestHome page
M. Piccoli, P. Trambaiolo, A. Salustri, E. Cerquetani, A. Posteraro, G. Pastena, E. Amici, F. Papetti, E. Marincola, S. La Carruba, et al.
Bedside Diagnosis and Follow-up of Patients With Pleural Effusion by a Hand-Carried Ultrasound Device Early After Cardiac Surgery
Chest, November 1, 2005; 128(5): 3413 - 3420.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (20)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Diacon, A. H.
Right arrow Articles by Solèr, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Diacon, A. H.
Right arrow Articles by Solèr, M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS