Sartori 2004.

Methods	Single centre RCT evaluating intrapleural bleomycin vs interferon alfa‐2b in the palliative treatment of malignant effusion (Italy)
Participants	Inclusion: cytologically proven MPE requiring at least two thoracenteses in preceding four weeks; at least 3 L drained in the preceding four weeks; adequate pulmonary re‐expansion on CXR after thoracentesis; last systemic treatment administered at least six weeks prior to enrolment; no further chemotherapy options; Karnofsky performance score > 40 Exclusion: none 160 participants randomised
Interventions	All patients underwent a 9 Fr intercostal drain insertion under ultrasound scan (USS) guidance. Fluid was drained via a 3‐way‐tap until USS revealed no residual effusion. Study drug administered IP via the chest tube. Tube was then clamped for two hours and participants changed position every 15 minutes. Tube removed 24 ‐ 48 hours after the last dose Bleomycin group: 0.75 mg/kg bleomycin in 50 ml saline. A repeated dose was given if > 100 ml/day output three days after the first dose IFN alfa‐2b group: 1 million units/10 kg in 200 ml saline. Six doses given every four days
Outcomes	Treatment response at 30 days (complete response (no fluid re‐accumulation), partial response (asymptomatic fluid recurrence < 50% of the original effusion, not requiring thoracentesis), no response (fluid recurrence > 50% of the original effusion, requiring thoracentesis)) Time to progression Number of thoracenteses until death
Notes	Deaths included in the analysis as failures (as presented in the paper as ITT analysis) People with trapped lung excluded from trial entry Not included in network meta‐analysis
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated randomisation
Allocation concealment (selection bias)	Low risk	Computer‐generated randomisation
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not stated explicitly but two drugs were given as different volumes
Blinding of outcome assessment (detection bias) All outcomes	High risk	Synptom recurrence and adverse event reporting may be biased by lack of blinding. Mortality not biased by lack of blinding. Not stated if CXR interpretation was blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	ITT analysis performed
Selective reporting (reporting bias)	Low risk	All outcomes reported
Other bias	Low risk	No other biases identified