Kawashima 2016.

Study characteristics
Methods	Randomised double‐blinded RCT with cross‐over design: participants randomised to coenzyme Q10 or placebo at initiation of study and received therapy for 3 months. They underwent 1‐month wash‐out period and then participants were crossed over to receive the other therapy (coenzyme Q10 or placebo).
Participants	20 adult participants with chronic heart failure and ejection fraction ≤ 40% as documented with the modified Simpson method via echocardiogram, and who had received standard heart failure treatment for at least 1 month. Excluded: people with malignant disease, severe infectious disease, trauma, people undergoing haemodialysis, and who were receiving other supplements. Some people also excluded based on the decision of the attending doctor.
Interventions	Intervention with dose: oral ubiquinol 200 mg twice daily (400 mg/day) Control: placebo
Outcomes	Used in this review (only from the first phase, at 3 months): Levels of coenzyme Q10, Brain natriuretic peptide (BNP) measured by blood test. Exercise capacity measure by exercise duration and METs during treadmill exercise test. Left ventricular ejection fraction measured by echocardiography. Not used in this review (not outcomes of interest): weight, systolic and diastolic blood pressure, heart rate, white blood cell, haemoglobin, creatinine, eGFR, cystatin‐C, total cholesterol, LDL cholesterol, triglycerides, high‐sensitivity troponin I, C‐reactive protein, urinary 8‐OHdG/Cr, CoQ: reduced form, oxidized form, ratio for reduced form to total CoQ10, ratio for reduced form to oxidized form Data about NYHA clinical status from the first phase was not provided by the authors and thus, was not used.
Notes	5 participants dropped out of the treatment group and 1 participant dropped out of the control group. Only 14 out of 20 participants completed the study and were included in the analysis.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Study was mentioned to be randomised, however, random sequence generation method was not mentioned.
Allocation concealment (selection bias)	Unclear risk	Details about allocation concealment were not mentioned.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double blinded
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No information about blinding the assessors given.
Incomplete outcome data (attrition bias) All outcomes	High risk	Only 14 out of 20 participants completed the study and were included in the analysis. 5 participants dropped out of the treatment group and 1 participant dropped out of the control group.
Selective reporting (reporting bias)	Low risk	All outcomes listed in the protocol were reported in the manuscript.
Other bias	Low risk	No other bias could be found.