Perng 1997.

Methods	Randomized multi‐centerphase III trial Eligible patients were Histologically or cytologically confirmed NSCLC Inoperable stage IIIA, IIIB, or IV 18 to 75 years old Measurable disease No previous chemotherapy, immunotherapy, or radiotherapy regimens PS: 0 to 2 on Zubrod Scale Adequate bone marrow, hepatic, and renal function Exclusion criteria Prior diagnosis of malignancy excluded, with the exception of in situ carcinoma of the cervix or adequately treated basal cell carcinoma of the skin
Participants	G arm: 27 participants (ITT population) ‐ median age (range): 63 (36 to 75) years/number of elderly participants not reported EP arm: 26 participants (ITT population) ‐ median age (range): 60 (35 to 75) years/number of elderly participants not reported
Interventions	G arm: gemcitabine 1250 mg/m2 over 30‐minute i.v. infusion on days 1, 8, and 15, every 4 weeks. Dexamethasone and metoclopramide were given before gemcitabine infusion as antiemetic prophylaxis EP arm: cisplatin 80 mg/m2 over 1‐hour i.v. infusion on day 1 and etoposide 80 mg/m2 over 60‐minute i.v. infusion on days 1, 2, and 3 of each 28‐day cycle. In EP arm, chemotherapy administered after hospitalization. Granisetron, dexamethasone, metoclopramide, and lorazepam given before cisplatin as antiemetic prophylaxis. Dexamethasone and metoclopramide used on days 2 and 3 as antiemetics
Outcomes	Neither planned primary nor secondary outcomes reported
Notes	No information on inclusion of elderly patients nor specific subgroup analysis was obtained, despite multiple attempts to contact study authors
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Eligible participants randomly assigned to GEM and EP regimens by a statistical office not involved in the trial, using a computer‐ generated list of random numbers
Allocation concealment (selection bias)	Unclear risk	No information on allocation concealment
Blinding of outcome assessment (detection bias) OS and 1y OS rate outcome	Unclear risk	No information on blinding
Blinding of outcome assessment (detection bias) Other outcomes	High risk	No information on blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	"Two patients in the EP arm were excluded from analysis due to protocol violation. One patient suffered grade 3 hearing impairment before entering the trial and the other patient was found to have brain metastases on the second day of the first cycle of treatment"
Selective reporting (reporting bias)	Low risk	No evidence of reporting bias
Other bias	Unclear risk	No separate elderly subgroup analysis