Pleural disease

Pleural effusions

Treat exudative effusions as tuberculous in areas of high prevalence if, and only if, they are no discordant features. A rapid response to therapy serves to confirm the diagnosis. A small proportion of patients with exudative effusions that would be classified as idiopathic by criteria that are more rigorous will be treated as tuberculous by these criteria. Lack of response, or discordant features, require further testing.

Discordant features:

• Elderly
• Clubbing
• Any features suggestive of SLE (especially arthritis or skin rash)
• Empyema (polymorphs rather than lymphocytes on diff count of the fluid)

Pleural exudate versus transudate

Light’s criteria¹ are the most widely recognised: an exudate is more like serum than a transudate. (Pleural fluid protein/serum protein > 0.5; pleural fluid LDH/serum LDH >0.6.) Alternatively, use the absolute value of pleural fluid LDH >200 IU/l (this last has been modified to an LDH > greater than two-thirds the upper limit of normal for serum LDH).

Blood stained fluid is common in tuberculosis and its presence on its own does not make the diagnosis of TB less likely. The probability that someone presenting with a pleural effusion has tuberculosis is dictated by the population prevalence and the clinical setting, and investigation should be tailored accordingly. An elderly cyanosed man with cor pulmonale, severe fluid overload, and an effusion is likely to have a transudate. A 20 year old presenting with unilateral chest discomfort, fever, and a moderate effusion in an environment of very high TB prevalence probably has an exudative effusion, and in the absence of palpable or radiological adenopathy suggesting lymphoma, this scenario has a high probability of being due to tuberculosis, so further tests will only change that figure slightly. The absolute level of ADA (if done) is useful – using a cut-off of 60 U/l results in a specificity of 96% and a PPV of 96%. (NPV = 95%.)²

Pleural biopsy and cytology

Pleural fluid cytology is indicated in situations of clinical equipoise – e.g. an elderly smoker with clubbing and an effusion. Pleural biopsy is indicated when cytology doesn’t give an answer, and in patients who appeared to have typical tuberculous effusions but have failed to respond to treatment.

Therapeutic aspiration of pleural effusions

Tapping effusions therapeutically is appropriate if there is clinical evidence of respiratory embarrassment. Remove enough fluid to make the patient comfortable, but be wary of removing more than about 500 ml at a time in patients with ‘chronic’ effusions to avoid re-expansion pulmonary oedema. The simplest system is an intravenous cannula and a giving set draining into a receiver, but it is important to remember not to leave the needle in the cannula as it may lacerate the re-expanding lung. Once the needle is removed, the cannula often kinks irrevocably unless it is adequately secured and monitored. It is important to remain in the vicinity of the patient while performing the procedure. Don’t leave it up to the nursing staff to decide when sufficient fluid has been removed.

Steroids – there is currently insufficient evidence, to know whether steroids are of any benefit in the management of tuberculous pleural effusions, with weak evidence of a reduction in time to resolution but little useful information on effect on long term lung function. ³

Perspective – exudates and transudates

In Light’s original study, about two thirds of the patients had exudates. Light’s criteria are ‘or’ criteria – if any one of the three was true, the fluid was classified as an exudate. This use of the criteria in ‘parallel’ enhances sensitivity at the expense of specificity. It is also worth noting that two of the criteria (absolute LDH and LDH/serum ratio) often demonstrate multicollinearity and so shouldn’t be combined in a diagnostic model. The differentiation of transudate from exudate is usually clear just from a glance at the figures. Where there is doubt it may be worth using the actual values to derive a multilevel likelihood ratio⁴ to derive an ‘exact’ post-test probability.

Proceed as follows:

1. Read off the number where the two values intersect – this is the combined LR.
2. Use this value in the Bayes table to derive a post-test probability.

e.g. If you guess before doing the test that the fluid is an exudate with a probability of 50%, and the protein concentration of the fluid is 26 with a serum protein of 71 (transudate using protein ratio) with a fluid LDH of 590 and an upper limit of normal for your laboratory of 600 (exudate using LDH ratio, which is 590/600 = 0.98), then the combined LR will be at the intersection of 26-30 and 0.9- 1.00, which is 5.2, and the post-test probability will be 83%.

Combined likelihood ratios: pleural fluid protein concentration and ratio pleural fluid LDH to upper limit of normal reference range of serum LDH for laboratory:

(‘Protein’ = pleural fluid protein concentration in g/l. ‘LDH’ = ratio of the measured pleural fluid LDH concentration to the value of the upper limit of the normal range of serum LDH for the laboratory. * Single LRs: the column in bold below this represents the LRs for the various protein concentrations to the left without knowledge of the LDH; the similar bold row represents LDH ratio LRs without protein concentration information. )

Perspective – differentiating exudates from transudates using urinary dip sticks

The classification of pleural fluid as an exudate or a transudate based on protein and LDH according to Light’s criteria is covered in the section on clinical epidemiology; in practice there is often a desire to know the chemistry results promptly, and this can be difficult in the absence of ready access to laboratory facilities. One possibility⁵ is to take 1 ml of pleural fluid and dilute it to 10 ml with saline.(Tap water may work although this has not been studied.) Then place a drop of this fluid onto a normal urine dip stick and read off the protein concentration. The presence of 3+ (5.0g/l) had a sensitivity of 78% and specificity of 93% (LR+ 14.0, LR- 0.6) for the fluid being subsequently found to be an exudate.

Perspective – influence of diuretics on protein and LDH concentrations

A small study⁶ looking at fluid chemistry after treating cardiac failure found that both protein and LDH increased by between 50% and 90%, resulting in potential misclassification of 30-70%, depending on the exact criterion used.  

Perspective – ADA in pleural fluid

Adenosine deaminase measurements on pleural fluid are most accurate when least needed. Many of the studies exploring its validity have been done in patients readily identifiable by other criteria as having or not having tuberculous pleural effusions. There are quite literally hundreds of studies looking at this topic, with a wide range of results. Of interest, two meta-analyses reached nearly identical conclusions. The cut-off chosen is important:

View the ADA in the context of your pre-test probability. In an area where more than 80% of effusions are tuberculous, the finding of an ADA of >100 is diagnostic and an ADA of >30 is highly suggestive, probably sufficiently so to warrant treatment without further testing. On the other hand, in a group of patients suspected to have low prevalence of TB, an ADA of >30 is not diagnostic, although an ADA of < 16 excludes the diagnosis.

Perspective – diagnosing pleural tuberculosis using PCR

Using a composite reference standard, a recent systematic review⁷ found excellent (as expected) specificity, and a considerable improvement in sensitivity with the next version GeneXPert assay (Ultra). So in practice it is diagnostic if positive, but still not adequate to exclude the diagnosis if negative:

In patients with a low CD4, testing for a urinary LAM, or combining a LAM and the PCR, may make the most sense.

Bronchopleural fistula

This usually arises after a chest drain is inserted for pneumothorax in cavitating tuberculosis or old fibrocavitatory disease. Drains may carry on bubbling for some time, and surgeons are reluctant to intervene in the presence of active tuberculosis due to frequent wound breakdown and difficult surgery. Such patients thus tend to remain in hospital on drainage until the tuberculosis comes under control. There is a common tendency to want to put the drain onto continuous suction. There is little evidence to support this practice, and some data from the surgical literature⁸ to suggest that it may do harm, possibly by keeping the fistula open by entraining air.

Pleurodesis.

In most cases, the decision about pleurodesis should be made in consultation with a thoracic surgeon, but the reason for including this guideline is the difficulty accessing the information when it when its needed. Consider it in:

• Recurrent malignant pleural effusion
• Recurrent pneumothorax
• Persistent ‘benign’ pleural effusion compromising lung function.

Perspective – success rates for treating malignant effusions

A review of studies looking at the success of pleurodesis as defined by the absence of any re-accumulation of fluid⁹ found that doxycycline 500mg worked in 72% (40% got pain), methylprednislone 160 mg worked in 30%, talc (2.5 to 10 mg) worked in 93% and bleomycin 15 units worked in 54%.

Surgical pleurodesis is probably still the preferred method for a fit person with good respiratory function, but consider bleomycin in patients with a malignant effusion (fluorouracil and mitomycin have also been used.). In patients without malignancy, the choices are more limited. Sterile medicinal quality talc is challenging to obtain, as is the parenteral form of doxycycline. An alternative which has been proposed by a group in Mexico¹⁰ is 20 ml of 10% povidone iodine in 80 ml saline, but this is clearly an ‘off label’ use and should only be considered with great circumspection.

Performing a pleurodesis

• Check on CXR that the lung is fully expanded, and insert venous access.
• ‘Pre-med’ the patient with a full dose of non-steroidal – e.g. 800 mg of ibuprofen orally or 100 mg of indomethacin rectally.
• Under sterile conditions, remove the white aspiration port from a urine drainage bag tube, and place one end into the end of the chest drain tube, and connect the other end to the tube to the suction bottle, to give yourself a ‘injection port’ on the chest drain.
• With the patient supine, inject, in aliquots of 5 ml, 20 ml of 1% lignocaine (less if the patient is very thin and frail) into the drain, using your injection port. Remember that the pleura is a very large absorptive surface, so giving the lignocaine rapidly is the same as giving an intravenous bolus. Flush with 2 aliquots of 20 ml of saline
• Wait a full ten minutes for it to take effect.
• Inject the sclerosant, also in aliquots.
• Flush again with 2 aliquots of 20 ml of saline.
• Be prepared to deal with sudden severe pain if it occurs in spite of your precautions – you may need to use morphine.
• Leave the drain clamped for 2 hours.
• Thereafter either simply unclamp, or unclamp and put on low-pressure suction.
• Leave the drain in for 24 hours.

---

1. Light RW, MacGregor ML, Luchsinger DC, et al. Pleural effusions: the diagnostic separation of transudates and exudates. Ann Intern Med 1972;77:507-13 ↩

2. Riantawan P, Chaowalit P, Wongsangiem, M et al. Diagnostic value of pleural fluid adenosine deaminase I tuberculous pleuritis with reference to HIV infection and a Bayesian analysis. Chest 1999;116:97-103 ↩

3. Ryan H, Yoo J, Darsini P. Corticosteroids for tuberculous pleurisy. Cochrane Database of Systematic Reviews 2017, Issue 3. Art. No.: CD001876. DOI: 10.1002/14651858.CD001876.pub3. Accessed 16 February 2023 ↩

4. Heffner JE, Sahn SA, Brown LK. Multilevel likelihood ratios for identifying exudative pleural effusions. Chest. 2002;121:1916-20 ↩

5. Burgess LJ, Taljaard JJ, Maritz FJ. A rapid screening test for pleural exudates. S Afr Med J. 1999;89:14 ↩

6. Romero-Candeira S, Fernandez C, Martin C, et al. Influence of diuretics on the concentration of proteins and other components of pleural transudates in patients with heart failure. Am J Med. 2001;110:681-6 ↩

7. Aggarwal AN, Agarwal R, Dhooria S, Prasad KT, Sehgal IS, Muthu V. Xpert MTB/RIF Ultra versus Xpert MTB/RIF for diagnosis of tuberculous pleural effusion: A systematic review and comparative meta-analysis. PLoS One. 2022 Jul 11;17(7):e0268483. doi: 10.1371/journal.pone.0268483. PMID: 35816471; PMCID: PMC9273090. ↩

8. Marshall MB, Deeb ME, Bleier JI, et al. Suction vs. water seal after pulmonary resection: a randomised prospective study. Chest 2002;121:831-5 ↩

9. Walker-Renard PB, Vaughan LM, Sahn SA. Chemical pleurodesis for malignant pleural effusions. Ann Intern Med 1994;120:56-64 ↩

10. Olivares-Torres CA, Laniado-Laborín R, Reyes-Escamilla, A, et al.  Iodopovidone pleurodesis for recurrent pleural effusions. Chest. 2002;122;581-3 ↩

Was this article helpful?

0 out Of 5 Stars

5 Stars		0%
4 Stars		0%
3 Stars		0%
2 Stars		0%
1 Stars		0%

5 Submit

Send Feedback

How can we improve this article?

Please submit the reason for your vote so that we can improve the article.

Submit