Kessinger 1987.

Methods	Single centre RCT comparing intrapleural (IP) bleomycin and tetracycline in MPE (USA)
Participants	Inclusion: histologically proven malignancy; symptomatic pleural effusion with either > 3 g/dl protein or malignant cells on cytology Exclusion: allergy to either study drug 42 procedures randomised in 34 participants
Interventions	All participants underwent chest tube drainage Tetracycline arm: 500 mg tetracycline in 50 ml saline IP. 1 dose Bleomycin arm: 89 units in 50 ml saline IP. 1 dose For both arms, drain clamped for eight hours after instillation and participant moved positions. Thereafter, tube opened and suction applied. Drain removed when < 40 ml/24hours drained (or on day 7 if ongoing high output)
Outcomes	Treatment response at one month ('Complete response' (no re‐accumulation of the effusion); 'Partial response' (asymptomatic re‐accumulation of the effusion developed that was < 50% of its original volume); 'no response') Side effects Length of time chest tube in place following pleurodesis
Notes	Bilateral disease included. Some participants randomised to the trial more than once People with trapped lung eligible for trial entry Included in network meta‐analysis for pleurodesis efficacy, fever and pain
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Toss of coin"
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	No mention of blinding. Both drugs administered in 50 ml saline
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not stated. No mention of whether CXR interpretation was performed by a blinded individual
Incomplete outcome data (attrition bias) All outcomes	High risk	11/34 (32%) participants non‐evaluable for pleurodesis outcome (3 in bleomycin group and 8 in tetracycline group)
Selective reporting (reporting bias)	Low risk	All stated outcomes reported
Other bias	High risk	Unclear whether participants who were given both agents because the first agent failed were included in the analysis