Eguchi 2015.

Study characteristics
Methods	Prospective, randomised, double‐blind, placebo‐controlled trial Multicentre study with 22 institutions Country: Japan Study duration: 8 days
Participants	Number of participants: 34 (18 intervention/16 control) adult cancer patients (35 allocated) Inclusion criteria: age > 18 yrs; histological diagnosis of advanced cancer; expected life expectancy of > 4 months; no future plans for chemotherapy, radiotherapy, or surgical treatment; CRF refractory to other treatments; ability to take oral medications, inpatient treatment; ALT ≤ 300 U/mL; AST ≤ 300 U/mL; creatinine ≤ 3.0 mg/dL; bilirubin ≤ 3.0 mg/dL Exclusion criteria: severe heart disease; diabetes mellitus; active gastrointestinal ulcers, viral hepatitis, infectious disease, or tuberculosis; radiotherapy or chemotherapy in the previous 4 weeks; surgery for their cancer in the prior 2 weeks; history of corticosteroid allergy; corticosteroids use in the last 2 weeks; inadequate understanding of their condition
Interventions	Participants were randomly assigned to 2 groups in 1:1 ratio  Intervention group: methylprednisolone 16 mg twice daily, orally for 7 days Control group: placebo twice daily for 7 days
Outcomes	Primary outcome Change in VAS score for fatigue from day 0 to 8 Secondary outcomes  Change in VAS score for appetite from day 0 to 8. Change in QoL on QoL–ACD score on days 0, 3, and 8 Laboratory examinations were taken on days 0 and 8 Adverse events were recorded daily and the level of severity was recorded on a 5‐point Likert‐type scale
Notes	The study was underpowered, so was unable to prove the efficacy of methylprednisolone in improving CRF.   The was evidence of a difference in the secondary endpoints. The study was partially funded by the Epidemiological and Clinical Research Information Network; no reference made to potential COIs
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"dynamic randomization method"
Allocation concealment (selection bias)	Low risk	Central randomisation at data centre
Blinding of participants and personnel (performance bias) Participant‐reported fatigue relief	Low risk	Double‐blind
Blinding of outcome assessment (detection bias) Participant‐reported fatigue relief	Unclear risk	No clear statement of how assessors were blinded
Incomplete outcome data (attrition bias) Participant‐reported fatigue relief	Low risk	< 10% of participants did not complete the study
Selective reporting (reporting bias)	Low risk	None detected