Pleural effusion: tuberculosis: no clear role for steroids.

Clinical bottom line (level 1b-)

  1. Steroids did not clearly reduce symptoms or prevent pleural thickening in patients with tuberculous pleurisy and effusions.
Wyser et al: Chest 1996; 110 (2): 333-338
Expires October 2003

The study

Double-blinded concealed randomised trial with intention-to-treat
Setting: university hospital, South Africa

74 patients (aged mean 33 years, 61% male) exudative pleural effusions and tuberculous pleurisy proven on thoracoscopy or bronchoscopy (presence of caseating granuloma on biopsy and/or positive M. tuberculosis culture.)

Excluded if
  • malignancy or pneumonia causing effusion
  • contraindication to corticosteroids (diabetes mellitus, uncontrolled hypertension, peptic ulcer disease, empyema)
  • HIV positive
  • neoplastic disease
  • noncompliance with treatment


    • Control Group: (n = , 36 analysed): placebo
    Experimental Group: (n = , 34 analysed): prednisone 0.75 mg/kg/day. After 2 to 4 weeks, depending on the effect this was reduced to 5 mg/day in all patients.
    All patients received rifampicin 10 mg/kg/day, isoniazid 8 mg/kg/day and pyrazinamide 25 mg/kg/day, and pyridoxine 25 mg/kg/day. Patients had pleural effusions completely drained.
    95% followed for 24 weeks

    The evidence

    Outcome Time to outcome CEREERRRR
    (95% CI)    		ARR
    (95% CI)    		NNT
    (95% CI)
    residual pleural thickening on CT 6 months 21
    (58.3%)    		 17
    (50.0%)    		 14%
    (-32% to 45%)    		 8.33%
    (-14.9% to 31.6%)    		 12
    (NNT = 3 to infinity;
    NNH = 7 to infinity)

    Outcome Control Group
    (SD)    		 Experimental Group
    (SD)    		 Mean Difference
    (95% CI)
    morbidity on visual analogue score (mm): baseline 434
    ()     		449
    ()     		NS
    ( to )
    morbidity on visual analogue score: 8 weeks 20
    ()     		3
    ()     		NS
    ( to )

  • morbidity: (dyspnoea, cough, night sweats, tiredness, appetite, pleuritic chest pain, general well-being) on a visual analogue scale
  • At no time was there a difference in the VAS scores between the two groups.
  • No patient had recurrent pleural effusion (95% CI: 0.0% to 4.2%)

    • Comments

    1. The study is too small to show any clear difference between steroids and placebo.

    Citation

    1. Wyser C, Walzl G, Smedema JP, et al: corticosteroids in the treatment of tuberculous pleurisy: a double-blind, placebo-controlled, randomized study. Chest 1996; 110 (2): 333-338
    Search Terms: pleural effusion in Cochrane
    Contributor: Chris Ball and Clare Wotton, October 2000
    Reviewer:

    Clinical Question.
    Patient tuberculous pleurisy and effusions
    Intervention or Exposure steroids
    Comparison placebo
    Outcome symptoms