A review of pleural fluid TB culture at a large acute UK hospital

Abstract

Background

The British Thoracic Society (BTS) recommend pleural fluid TB culture (TBC) should only be requested where there is clinical suspicion of TB pleuritis. Many clinicians request TBC when investigating pleural effusions regardless of clinical suspicion.

Objectives

To examine the sensitivity of pleural fluid TBC in TB patients.

To audit the appropriateness of pleural fluid TBC requests.

To determine if enhanced clinical information increased the utility of TBC.

Method

We assessed pleural fluid from TB patients and clinical information accompanying all pleural fluid TBC requests at our hospital between Jan 2008 and Dec 2012. Clinical information suspecting TB on the request form was considered a high level of suspicion. Absence of information was considered a low level of suspicion.

Results

The prevalence of pleural effusion in TB was low 6/120 (5%). The sensitivity of pleural fluid TBC in cases of TB presenting with a pleural effusion was 33%.

Of 918 pleural fluid samples, 67 (7.3%) had a high clinical suspicion. The positive predictive value (PPV) of pleural fluid TBC with high clinical suspicion was 3% (table 1). The remaining 851 had low clinical suspicion with a negative predictive value of 100%.

Discussion

The prevalence of TB diagnosed by pleural fluid TBC is low (2/918). Clinical suspicion of TB did not significantly improve the PPV. No patients with a low clinical suspicion cultured TB. Our data supports BTS recommendations that pleural fluid TBC should only be requested when TB pleuritis is suspected.