Medical thoracoscopy (MT) has gained widespread acceptance in the diagnosis of exudative pleural effusion (EPE), as it allows direct visualization of the pleura and biopsy from abnormal sites.1 The diagnostic yield of thoracentesis is about 60% and 90% for malignant pleural effusions (MPE) and tubercular pleural effusions (TPE), respectively.2,3 Closed-blind pleural biopsy (CBPB) has a yield of 80% to 90% for TPE and 50% to 60% for MPE, provided an optimal number of biopsies are obtained.2–5 CBPB in combination with thoracentesis, increases the diagnostic yield of the latter by up to 10% in patients with MPE.3,6 In contrast, MT has a diagnostic yield of about 91% to 95% for MPE, and almost 100% for TPE.2,7–9 Thus, in many centers, including ours, thoracoscopic pleural biopsy has replaced CBPB as the procedure of choice in EPE, when thoracentesis fails to provide a conclusive diagnosis.7 MT can be performed either with rigid or the semirigid thoracoscope.10,11 Studies conducted this far have shown comparable diagnostic yield with these instruments.12–14
In developing countries, including India, thoracoscopy is not widely available. Moreover, few pulmonary physicians are trained to perform this procedure. Hence, in many centers, CBPB continues to be performed in the workup of undiagnosed EPE.6,15–17 In this retrospective study, we describe our experience with MT and also compare the outcome of thoracoscopy (both rigid and semirigid) with that of CBPB.
PATIENTS AND METHODS
This was a retrospective observational study conducted at our institute between August 2004 and May 2014. The study was approved by the Ethics Review Committee and a written informed consent was obtained from all patients. Some patients in this study have been included in the previous study from our group.12,18,19 We had been performing CBPB in the evaluation of undiagnosed EPE till June 2009. Rigid thoracoscopy (RT) and semirigid thoracoscopy (SRT) were started in November 2006 and March 2011, respectively. A routine point-of-care ultrasonography in the bronchoscopy suite was started in January 2013. From January 2013 onward, RT was generally chosen if there were loculations on computed tomography (CT) chest or multiple septations on thoracic ultrasonography or if a therapeutic procedure such as adhesiolysis was additionally required.
We included patients with undiagnosed EPE who underwent either MT or CBPB. An EPE (based on Light’s criteria20) was considered as undiagnosed, if the etiology remained unclear despite diagnostic thoracentesis and routine pleural fluid (PF) investigations including adenosine deaminase (ADA), cytology, and microbiology. Analysis of PF for interferon-γ and nucleic acid amplification tests were not routinely performed. Patients with PF ADA>70 U/L were treated with empiric antituberculosis therapy (ATT); thoracoscopy was considered only when there was clinicoradiologic failure after 4 to 8 weeks of ATT or the clinical picture suggested an alternative diagnosis (active extrathoracic malignancy, pleural nodules on CT of the thorax, and others). All subjects underwent a complete blood count, coagulation profile, and CT scan of the chest before the procedure. The effusion was categorized as mild (<1/3 of the hemithorax), moderate (1/3 to 2/3 of the hemithorax), or massive (occupying >2/3 of the hemithorax).
Patients underwent CBPB (between August 2004 and June 2009) either with an Abrams or a trucut biopsy needle under local anesthesia, as previously described.21 Briefly, after a 5 mm skin incision, the Abrams needle was inserted. Once PF was aspirated, the needle was withdrawn, and placed parallel to the ribs with the orifice of the needle facing the operator. The biopsy was taken after twisting the needle to advance the cutting edge. Pleural biopsy with the trucut needle was also performed after a 5 mm skin incision. After aspirating PF with a syringe, the trucut needle was inserted into the pleural space, completely angled toward the skin and the stylet advanced along the chest wall. Subsequently, the outer sheath was advanced over the stylet, and the needle was then removed. At least 4 biopsies were obtained in all patients.
Patients underwent either RT (Richard Wolf, Knittlingen, Germany) or SRT (Olympus Medical Systems, Tokyo, Japan) under local anesthesia and conscious sedation, as previously described.12 Briefly, after positioning the patient in the lateral decubitus position with the affected side upward, a chest ultrasound was performed (January 2013 onward) to determine the site of entry in the lateral chest wall. Under strict aseptic precautions, a skin incision (1.5 to 2 cm length for SRT and 2 to 2.5 cm for RT) was made and blunt dissection carried out with a curved artery forceps till the pleura was breached. The trocar supplied with the respective thoracoscopes was inserted followed by introduction of the thoracoscope. PF was completely aspirated while air was allowed to enter the pleural space. The pleural surfaces were then inspected and parietal pleural biopsies (6 to 8 and 2 to 4 with semirigid and rigid thoracoscope, respectively) were obtained in all the patients with the biopsy forceps of the respective thoracoscopes. A 24 to 28 F intercostal tube drainage was then placed in the pleural cavity.
Demographic characteristics of the patient including age, sex, clinical diagnosis, presence of comorbidities (CKD, extrathoracic malignancy), PF analysis (total and differential count, protein and glucose values, ADA levels, microbiology, and cytology), findings on CT chest, and thoracoscopic pleural findings were recorded. We also noted the complications encountered during the procedure. Patients were followed for up to 3 to 6 months. The procedure was defined as successful if a biopsy representative of the pleura could be obtained. Pleural effusions were categorized as TPE if granulomas were demonstrated in pleural tissue or Ziehl-Neelsen stain showed acid-fast bacilli. A MPE was said to be present if biopsy was positive for malignant cells.
Data are presented in a descriptive manner. The differences between continuous and categorical variables were compared using Kruskal-Wallis ANOVA and χ2 tests, respectively. The procedural yield of MT was compared with that of CBPB. A P-value <0.05 was considered as statistically significant.
During the study period, 348 patients [mean age (SD): 49.8 (15.7) y] underwent MT or CBPB. Of these, 264 patients underwent thoracoscopy and 84 patients underwent CBPB. RT was performed in 198 patients, whereas 66 patients were subjected to SRT. Pleural biopsy was performed using Abrams and trucut needle in 50 and 34 patients, respectively. The primary indication for MT was undiagnosed EPE (n=248) followed by pleurodesis (n=12) and adhesiolysis (n=4). The baseline characteristics of the subjects are shown in Table 1. There was no significant difference between the groups except PF ADA and glucose levels, which were higher in the CBPB and the SRT groups, respectively. The most common preprocedure clinical diagnosis was malignancy (n=142, 42.8%) followed by tuberculosis (n=76, 22.9%); the diagnosis was unclear in 106 patients (31.9%). Forty-one (12.3%) and 18 (5.4%) patients had CKD and active extrathoracic malignancy, respectively. The effusion was massive in 132 (39.7%) and bilateral in 28 (8.4%) patients.
The procedural yield was significantly higher (P=0.02) in the thoracoscopy group compared with CBPB (Table 2). Of the 84 patients who underwent CBPB, a representative pleural biopsy was obtained in 71 (84.5%) patients; whereas in 13 patients, histologic examination revealed skeletal muscle and tissue from other internal organ. A pleural biopsy could be obtained in 231 of the 248 patients who underwent MT [RT: 170/182 (93.4%); SRT: 61/66 (92.4%)]. The most common reason for failure to obtain a pleural biopsy in MT group was the presence of dense adhesions (n=15); however, all the biopsies obtained were representative of the pleura. Of the 15 procedures that failed due to adhesions, 14 were performed without the use of point-of-care ultrasound. After introduction of chest ultrasound, the incidence of failed thoracoscopies decreased significantly (1/77, 1.3%). The procedural yield was significantly higher when thoracoscopy was performed along with chest ultrasonography than those without (98.7% vs. 90.6%, P=0.04). The most common histopathologic diagnosis obtained was malignancy followed by tuberculosis (Table 3). The pleural biopsies showed nonspecific inflammation in 72 (23.9%) patients. Twenty of these patients had CKD and the effusion responded to dialysis. The rest of the patients were managed with antibiotics alone and the effusion was classified as either idiopathic or parapneumonic.
We correlated the preprocedure clinical diagnosis with the final histopathologic diagnosis. In patients with clinical suspicion of tuberculosis, only 58.5% (38/65) were finally diagnosed as tuberculosis (Table 4). Similarly in those with clinical suspicion of MPE, only 62.5% (85/136) had histopathologic evidence of malignancy. In patients with inconclusive clinical picture, a final histopathologic diagnosis of either tuberculosis or malignancy could be achieved in 47.3% (44/93) of the patients. A clinical diagnosis of tuberculosis had a positive predictive value (PPV) of 58.5% and a negative predictive value (NPV) of 83.9% for the final diagnosis of tuberculosis. Similarly, a clinical diagnosis of malignancy had a PPV and NPV of 62.5% and 89.1%, respectively, for confirmed MPE.
Pleural effusions were classified as having high ADA (>70 U/L), borderline ADA (41 to 70 U/L), and low ADA (≤40 U/L) (Table 5). A total of 61.2% and 70.5% of patients with borderline and high ADA, respectively, had tuberculosis, whereas only 14.4% of the patients with low ADA were diagnosed as tuberculosis. This difference was statistically significant (P=0.0001). In addition, malignancy and other nontubercular diagnoses were significantly more common in the subgroup with low ADA as compared with those with borderline and high ADA.
The clinical and radiologic findings, PF analysis, and thoracoscopic appearances of pleura in patients with a final histopathologic diagnosis of either tuberculosis or malignancy are shown in Table 6. Patients with malignancy had a higher age and were more likely to have a massive effusion, mediastinal shift, nodular pleural thickening on CT chest, large-sized/variable-sized pleural nodules visualized during thoracoscopy, and lower PF ADA levels. Patients with tuberculosis had higher PF ADA and protein levels and more often had small pleural nodules visualized during thoracoscopy.
Of the 248 patients who underwent thoracoscopic pleural biopsy, major complications occurred in 15 patients (Table 7). Three patients died following thoracoscopy. One patient with malignancy developed acute kidney injury on the second day of thoracoscopy and succumbed to the illness after 2 days. The second patient also had malignancy and developed reexpansion pulmonary edema and respiratory failure, which required noninvasive ventilation. Although the respiratory failure improved, his general condition deteriorated and the patient expired in the second week after thoracoscopy. The third patient had CKD, developed empyema following thoracoscopy followed by sepsis, and expired in the second week after thoracoscopy. All patients who died had poor baseline performance status. Death directly due to the procedure, however, was attributed only in the patient who developed empyema and sepsis. The most common major complication following thoracoscopy was empyema (n=10, 3.9%) followed by reexpansion pulmonary edema (n=4, 1.6%). Seven (8.3%) complications were encountered with CBPB. Most of the complications were with the trucut needle (5 instances of hemothorax), whereas 1 patient each developed subcutaneous hematoma and syncope, respectively, with the Abrams needle. There was no mortality with CBPB.
The results of this study suggest that MT is a useful and reasonably safe tool in the workup of undiagnosed EPE, and is superior to CBPB. However, if MT is unavailable, CBPB provides a good diagnostic yield, superior to clinical judgment. Undiagnosed EPE is a common problem encountered by pulmonary physicians. In countries where tuberculosis is endemic, differentiating tuberculosis from malignancy, the 2 most common causes of EPE, is of utmost importance as the treatment and prognosis vary. It is a common practice in resource-constrained settings to initiate empiric ATT in all undiagnosed EPE, and consider an alternate diagnosis only if the effusion fails to respond to therapy. However, treating MPE as tuberculosis would not only lead to an unnecessary delay in the initiation of appropriate therapy but also subject them to risk of adverse reactions of ATT.
In resource-constrained settings, CBPB continues to be performed. In fact, a combination of PF ADA, lymphocyte/neutrophil ratio of >0.75, and CBPB has been shown to have diagnostic accuracy similar to MT in diagnosis of TPE.2 In another study from tuberculosis endemic country in patients younger than 50 years, CBPB was found to have diagnostic accuracy approaching thoracoscopy.22 Thus, the debate on the role of CBPB as compared with MT as the initial diagnostic modality for undiagnosed EPE continues. Some pulmonary physicians favor CBPB as the initial investigation, with thoracoscopy only if the effusion still remains undiagnosed.7 In fact, if pleural biopsy is performed under image guidance, the diagnostic yield could be as high as 87%, albeit lower than MT.23,24
In our study, the success rate of a representative pleural biopsy was significantly higher with thoracoscopic biopsy. A recent randomized trial has also suggested the superiority of MT over CBPB with lower yield and higher risk of complications with the latter.25 Blind pleural biopsy was nonrepresentative in 15.5% patients who underwent CBPB, and similar rates of nonrepresentative samples with blind pleural biopsies have been reported in other studies also.3,15 Pleural biopsies could not be obtained in 6.8% of the patients undergoing MT in this study, and is marginally higher than the earlier reported rates.13,14 This is probably due to the delayed presentation of patients to our center, which led to the formation of multiple pleural adhesions.12 Moreover, chest ultrasound was not used routinely to confirm the site of entry before 2013. The success rate of MT significantly improved when a point-of-care ultrasound was used to guide the choice of a thoracoscope and determine the chest wall entry point. It has been demonstrated previously that ultrasound reliably identifies entry sites for trocar placement during MT, even in the presence of pleural adhesions.26,27
Our study reemphasizes that a clinicoradiologic diagnosis of tuberculosis or malignancy has a poor PPV and was accurate in only 60% of the patients. Similar results were also seen in the study by Kannan et al.28 These findings suggest that pleural biopsy should be performed in all patients with EPE that remain undiagnosed after thoracentesis. Initiating empiric ATT without a histopathologic confirmation, a practice common in resource-constrained settings, is not justified. The 2 most common diagnoses in our study were malignancy and tuberculosis. The rate of malignancy (34.4%) is lower in our study compared with western series (>50%),29–31 but comparable with the reported rates from the developing world.18,32,33 In addition, the rate of nonspecific inflammation in our study (23.9%) is similar to that in other large series of thoracoscopy,34 and is due to inclusion of large number of patients with uremia and parapneumonic pleural effusions.
The visual findings during thoracoscopy have been correlated with histopathologic diagnosis in earlier studies as well. In the study by Mohan et al,29 presence of pleural nodules, pleural infiltration, and hemorrhagic fluid were clinical predictors of malignancy. In our previous study, the presence of uniform nodules and variable-sized nodules were suggestive of tuberculosis and malignancy, respectively.12 Even in this study, the only thoracoscopic finding of clinical significance was the size of pleural nodules. Uniformly small-sized nodules were suggestive of tuberculosis and the presence of variable-sized or large-sized nodules/masses suggested malignancy.
PF ADA level has been extensively studied in diagnosis of TPE.35,36 In a meta-analysis of 63 studies, the sensitivity and specificity of PF ADA in diagnosing TPE was 92% and 90%, respectively.37 The optimal cutoff for ADA in diagnosis of TPE remains unclear with different studies reporting cutoffs ranging from 35 to 55 U/L.35,38–44 An ADA level of >100 U/L is virtually diagnostic of tuberculosis,43 whereas a value <16 U/L virtually excludes tuberculosis.41 In our study, none of the patients with a final diagnosis of MPE had PF ADA value >70 U/L. Thus, this cutoff seems reasonable for initiating ATT, whereas a cutoff of 40 U/L (suggested in many studies) does not seem optimal as only 60% of patients with ADA value between 40 and 70 U/L had a histopathologic diagnosis of tuberculosis.
Several studies have shown MT to be a relatively safe procedure.45,46 In our study, thoracoscopy was associated with an attributable mortality rate of 0.37% and major complication rate of 5.6%, the most common being empyema (3.7%). Similar mortality (0.28%) and empyema (5%) rates have also been shown in a large series of 355 patients who underwent MT.31 In our study, pleural biopsy had a complication rate of 8.3% similar to previous reports.47
Finally, our study is not without limitations. It is a retrospective study from a single center. We do not have the recommended long-term follow-up of 1 year in those with nonspecific diagnosis on pleural biopsy. In addition, the results are applicable to CBPB and not image-guided CBPB.23 The strength of our study is the large sample size and the fact that we have compared the commonly used techniques for obtaining pleural biopsy.
In conclusion, MT is the procedure of choice in the diagnosis of undiagnosed EPE, given its better success rate and an acceptable safety profile, compared with CBPB. In centers where thoracoscopy is not feasible, we recommend CBPB (image guided, wherever feasible) over initiating empiric treatment (Fig. 1). Both RT and SRT are equally effective and the choice of the thoracoscope would depend on the availability, extent of pleural septations, and the need for additional procedures such as adhesiolysis.
1. Lee P, Colt HG. Pleuroscopy
in 2013. Clin Chest Med. 2013;34:81–91.
2. Diacon AH, Van de Wal BW, Wyser C, et al.. Diagnostic tools in tuberculous pleurisy: a direct comparative study. Eur Respir J. 2003;22:589–591.
3. Prakash UB, Reiman HM. Comparison of needle biopsy with cytologic analysis for the evaluation of pleural effusion
: analysis of 414 cases. Mayo Clin Proc. 1985;60:158–164.
4. Poe RH, Israel RH, Utell MJ, et al.. Sensitivity, specificity, and predictive values of closed pleural biopsy. Arch Intern Med. 1984;144:325–328.
5. Escudero Bueno C, Garcia Clemente M, Cuesta Castro B, et al.. Cytologic and bacteriologic analysis of fluid and pleural biopsy specimens with Cope’s needle. Study of 414 patients. Arch Intern Med. 1990;150:1190–1194.
6. Bhattacharya S, Bairagya TD, Das A, et al.. Closed pleural biopsy is still useful in the evaluation of malignant pleural effusion
. J Lab Physicians. 2012;4:35–38.
7. Koegelenberg CF, Diacon AH. Pleural controversy: close needle pleural biopsy or thoracoscopy
-which first? Respirology. 2011;16:738–746.
8. Lee P, Hsu A, Lo C, et al.. Prospective evaluation of flex-rigid pleuroscopy
for indeterminate pleural effusion
: accuracy, safety and outcome. Respirology. 2007;12:881–886.
9. Loddenkemper R. Thoracoscopy
—state of the art. Eur Respir J. 1998;11:213–221.
10. Lee P, Colt HG. Rigid and semirigid pleuroscopy
: the future is bright. Respirology. 2005;10:418–425.
11. Agarwal R, Aggarwal AN, Gupta D. Diagnostic accuracy and safety of semirigid thoracoscopy
in exudative pleural effusions: a meta-analysis. Chest. 2013;144:1857–1867.
12. Dhooria S, Singh N, Aggarwal AN, et al.. A randomized trial comparing the diagnostic yield of rigid and semirigid thoracoscopy
in undiagnosed pleural effusions. Respir Care. 2014;59:756–764.
13. Khan MA, Ambalavanan S, Thomson D, et al.. A comparison of the diagnostic yield of rigid and semirigid thoracoscopes. J Bronchology Interv Pulmonol. 2012;19:98–101.
14. Rozman A, Camlek L, Marc-Malovrh M, et al.. Rigid versus semi-rigid thoracoscopy
for the diagnosis of pleural disease: a randomized pilot study. Respirology. 2013;18:704–710.
15. Hira HS, Ranjan R. Role of percutaneous closed needle pleural biopsy among patients of undiagnosed exudative pleural effusion
. Lung India. 2011;28:101–104.
16. James P, Gupta R, Christopher DJ, et al.. Evaluation of the diagnostic yield and safety of closed pleural biopsy in the diagnosis of pleural effusion
. Indian J Tuberc. 2010;57:19–24.
17. Pandit S, Chaudhuri AD, Datta SB, et al.. Role of pleural biopsy in etiological diagnosis of pleural effusion
. Lung India. 2010;27:202–204.
18. Mootha VK, Agarwal R, Singh N, et al.. Medical thoracoscopy
for undiagnosed pleural effusions: experience from a tertiary care hospital in North India. Indian J Chest Dis Allied Sci. 2011;53:21–24.
19. Kumar S, Agarwal R, Bal A, et al.. Utility of ADA, PCR and thoracoscopy
in differentiating tuberculous and non-tuberculous pleural effusion
complicating chronic kidney disease. Indian J Med Res. 2015(In press).
20. Light RW, Macgregor MI, Luchsinger PC, et al.. Pleural effusions: the diagnostic separation of transudates and exudates. Ann Intern Med. 1972;77:507–513.
21. McLeod DT, Ternouth I, Nkanza N. Comparison of the Tru-cut biopsy needle with the Abrams punch for pleural biopsy. Thorax. 1989;44:794–796.
22. Emad A, Rezaian GR. Diagnostic value of closed percutaneous pleural biopsy vs pleuroscopy
in suspected malignant pleural effusion
or tuberculous pleurisy in a region with a high incidence of tuberculosis
: a comparative, age-dependent study. Respir Med. 1998;92:488–492.
23. Koegelenberg CF, Diacon AH. Image-guided pleural biopsy. Curr Opin Pulm Med. 2013;19:368–373.
24. Metintas M, Ak G, Dundar E, et al.. Medical thoracoscopy
vs CT scan-guided Abrams pleural needle biopsy for diagnosis of patients with pleural effusions: a randomized, controlled trial. Chest. 2010;137:1362–1368.
25. Haridas N, Suraj KP, Rajagopal TP, et al.. Medical thoracoscopy
vs closed pleural biopsy in pleural effusions: a randomized controlled study. J Clin Diag Res. 2014;8:Mc01–Mc04.
26. Hersh CP, Feller-Kopman D, Wahidi M, et al.. Ultrasound guidance for medical thoracoscopy
: a novel approach. Respiration. 2003;70:299–301.
27. Medford AR, Agrawal S, Bennett JA, et al.. Thoracic ultrasound prior to medical thoracoscopy
improves pleural access and predicts fibrous septation. Respirology. 2010;15:804–808.
28. Kannan SK, Lin WJ, Teck TS, et al.. Pleuroscopy
: early experience in an East Malaysian state with high tuberculosis
prevalence. J Bronchology Interv Pulmonol. 2009;16:250–253.
29. Mohan A, Naik S, Naseer R, et al.. Performance characteristics of semirigid thoracoscopy
in pleural effusions of undetermined etiology. J Bronchology Interv Pulmonol. 2010;17:289–294.
30. Rozman A, Camlek L, Kern I, et al.. Semirigid thoracoscopy
: an effective method for diagnosing pleural malignancies. Radiol Oncol. 2014;48:67–71.
31. Metintas M, Ak G, Yildirim H, et al.. The safety of medical thoracoscopy
in a group at high risk for complications. J Bronchology Interv Pulmonol. 2013;20:224–231.
32. Prabhu VG, Narasimhan R. The role of pleuroscopy
in undiagnosed exudative pleural effusion
. Lung India. 2012;29:128–130.
33. Thangakunam B, Christopher DJ, James P, et al.. Semi-rigid thoracoscopy
: initial experience from a tertiary care hospital. Indian J Chest Dis Allied Sci. 2010;52:25–27.
34. Janssen JP, Ramlal S, Mravunac M. The long-term follow up of exudative pleural effusion
after nondiagnostic thoracoscopy
. J Bronchol. 2004;11:169–174.
35. Krenke R, Korczynski P. Use of pleural fluid levels of adenosine deaminase and interferon gamma in the diagnosis of tuberculous pleuritis. Curr Opin Pulm Med. 2010;16:367–375.
36. McGrath EE, Anderson PB. Diagnostic tests for tuberculous pleural effusion
. Eur J Clin Microbiol Infect Dis. 2010;29:1187–1193.
37. Liang QL, Shi HZ, Wang K, et al.. Diagnostic accuracy of adenosine deaminase in tuberculous pleurisy: a meta-analysis. Respir Med. 2008;102:744–754.
38. Yildiz PB, Yazar EE, Gorgun D, et al.. Predictive role of adenosine deaminase for differential diagnosis of tuberculosis
and malignant pleural effusion
in Turkey. Asian Pac J Cancer Prev. 2011;12:419–423.
39. Tay TR, Tee A. Factors affecting pleural fluid adenosine deaminase level and the implication on the diagnosis of tuberculous pleural effusion
: a retrospective cohort study. BMC Infect Dis. 2013;13:546.
40. Porcel JM, Esquerda A, Bielsa S. Diagnostic performance of adenosine deaminase activity in pleural fluid: a single-center experience with over 2100 consecutive patients. Eur J Intern Med. 2010;21:419–423.
41. Zaric B, Kuruc V, Milovancev A, et al.. Differential diagnosis of tuberculous and malignant pleural effusions: what is the role of adenosine deaminase? Lung. 2008;186:233–240.
42. Mehta AA, Gupta AS, Ahmed S, et al.. Diagnostic utility of adenosine deaminase in exudative pleural effusions. Lung India. 2014;31:142–144.
43. Verma SK, Dubey AL, Singh PA, et al.. Adenosine deaminase (ADA) level in tubercular pleural effusion
. Lung India. 2008;25:109–110.
44. Kelam MA, Ganie FA, Shah BA, et al.. The diagnostic efficacy of adenosine deaminase in tubercular effusion. Oman Med J. 2013;28:417–421.
45. Rahman NM, Ali NJ, Brown G, et al.. Local anaesthetic thoracoscopy
: British Thoracic Society Pleural Disease Guideline 2010. Thorax. 2010;65suppl 2ii54–ii60.
46. Shujaat A, Bajwa AA, Usman F, et al.. Safety and accuracy of semirigid pleuroscopy
performed by pulmonary fellows at a major university hospital: our initial experience. J Bronchology Interv Pulmonol. 2013;20:213–223.
47. Pereyra MF, San-Jose E, Ferreiro L, et al.. Role of blind closed pleural biopsy in the managment of pleural exudates. Can Respir J. 2013;20:362–366.