Sculier 2001.

Methods	Randomisation 1:1:1 ratio Standard chemotherapy arm (6 courses of EVI (epirubicin 90 mg/m² , vindesine 3 mg/m² and ifosfamide 5 g/m² ; all drugs given IV on day 1); no infection prophylaxis given (therefore not evaluated in this review) Intervention arm: accelerated chemotherapy (the same as above, given every 14 days) and GM‐CSF for prevention of infection Control arm: accelerated chemotherapy (the same as above, given every 14 days) and cotrimoxazole prophylaxis Recruitment Period April 1993 to April 2000 Median follow‐up time Not reported
Participants	243 patients randomised, 233 eligible 78 standard arm (not evaluated in this review) 78 patients GM‐CSF 77 patients antibiotics Inclusion criteria Patients with small‐cell lung cancer and extensive disease (with metastases or as a locoregional disease that could not be locally treated in a single radiotherapy field) Age ≤ 75 years Patients should not have had prior therapy (radiotherapy, chemotherapy, surgery) Mean age in years GM‐CSF arm: 64 years (range: 35 to 74) Antibiotics arm: 61 years (range: 37 to 74) Stage GM‐CSF arm: Stage III: 7 patients Stage IV: 71 patients Antibiotics arm Stage III: 7 patients Stage IV: 70 patients Brain metastases GM‐CSF arm: 14 patients Antibiotics arm 17 patients Countries Several countries in Europe
Interventions	All patients 6 courses of EVI (epirubicin 90 mg/m² , vindesine 3 mg/m² and ifosfamide 5 g/m² ; all drugs given IV on day 1, in the accelerated arms every 14 days GM‐CSF arm GM‐CSF was given, as a daily subcutaneous dose of 5 µg/kg, from day 3 through day 13, or until neutrophil count reached ≥ 4000 mm³ after nadir. Antibiotics arm Cotrimoxazole (160 mg trimethoprim plus 800 mg sulfamethoxazole) was administered orally every 12 hours from day 3 until the end of the course of chemotherapy
Outcomes	Overall survival Tumour response Absolute and relative dose intensity Incidence of infections and severe infection Adverse events
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"eligible patients were randomised"
Allocation concealment (selection bias)	Unclear risk	not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Open label trial (subcutaneous injection of GM‐CSF versus oral antibiotics)
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Blinding of outcome assessor not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	"In the 233 eligible patients, 14 were nonassessable for response (2 in arm A, 6 in arm B, and 6 in arm C) for the following reasons: too long delay between 2 courses of chemotherapy (1), early death unrelated to cancer or treatment complications (9), protocol violation (2), death prior to starting treatment (1), no work‐up at evaluation (1)"
Selective reporting (reporting bias)	Unclear risk	No study protocol identified, therefore unclear if all the planned outcomes are reported
Other bias	Unclear risk	Not reported