Mortensen 2014.

Study characteristics
Methods	Randomized, double‐blinded RCT with parallel design with short‐term (16 weeks) and long‐term (106 weeks) follow‐up
Participants	420 adult participants with heart failure were randomised. Inclusion criteria: chronic heart failure in NYHA functional class III or IV with stable HF therapy Exclusion criteria: myocardial infarction, unstable angina pectoris, percutaneous coronary intervention, cardiac resynchronisation device, cardiac surgery, or stroke, all within 6 weeks of enrolment; HF from congenital heart disease; uncorrected valvular heart disease or planned valve surgery; planned other cardiac surgery or resynchronisation therapy; on urgent waiting list for heart transplantation (status 1 person); implanted mechanical assist device; on continuous inotropic support for HF; restrictive or hypertrophic cardiomyopathy; alcoholic heart disease; acute inflammatory myocarditis; severe non‐cardiac disease including malignancy with life expectancy < 1 year; psychosocial instability or anticipated problems with compliance; women of childbearing potential and lactating women; allergy to the constituents of the test medication; supplementary Coenzyme Q10 intake within the last month before run‐in; 6‐minute walk distance > 450 metres at run‐in; participation in another controlled trial 202 participants randomised to coenzyme Q10 group and 218 participants to the placebo group
Interventions	Intervention with dose: oral coenzyme Q10 100 mg 3 times a day with standard HF therapy Control: placebo with standard HF therapy.
Outcomes	Used in this review: all‐cause mortality (at 106 weeks); cardiovascular mortality (at 106 weeks); major adverse cardiovascular events (at 106 weeks); myocardial infarction (at 106 weeks); stroke (at 106 weeks); revascularisation procedures percutaneous intervention and coronary artery bypass grafting (at 106 weeks); hospitalisation due to heart failure (at 106 weeks); left ventricular ejection fraction (at 16 and 106 weeks); symptoms improvement: NYHA, visual analogue scale score, 6‐minutes walk test (at 16 and 106 weeks); N‐terminal pro‐brain natriuretic peptide (NT‐proBNP; at 16 and 106 weeks); serum levels of coenzyme Q10 (at 16 and 106 weeks); adverse events (at 106 weeks). Not used in this review (not outcomes of interest): heart rate, systolic and diastolic blood pressure, left ventricular end systolic and end diastolic pressure (at 16 and 106 weeks)
Notes	36 participants (22 in the coenzyme Q10 group and 14 in the placebo group) withdrew from the study; however, were included in the intention‐to‐treat analysis. 8 participants (4 in each group) were lost to follow‐up at the end of the study period. Intention‐to‐treat analysis was used for all‐cause mortality, cardiovascular mortality, hospitalisations, and adverse events (including revascularisation procedures, strokes, and MI). Study was partially funded by pharmaceutical companies that manufacture and distribute coenzyme Q10 supplements.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Study was randomised. Random sequence generation method was detailed in the methods section.
Allocation concealment (selection bias)	Low risk	Random numbers generated for randomisation were kept in sealed envelopes. Randomisation code was unavailable to investigators, participants, or statisticians.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Study was double‐blinded.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	The randomisation code was unavailable to investigators, participants, or statisticians at any time during the study until all data and materials had been collected, all blood samples had been analysed, and statistical analyses had been performed.
Incomplete outcome data (attrition bias) All outcomes	Low risk	36 participants (22 in the coenzyme Q10 group and 14 in the placebo group) withdrew from the study; however, were included in the intention‐to‐treat analysis. Analysing withdrawal reasons did not show any significant differences between the groups. 8 participants (4 in each group) were lost to follow‐up at the end of the study period.
Selective reporting (reporting bias)	Low risk	All outcomes listed in the protocol were reported in the manuscript.
Other bias	Low risk	No other risk of bias was found.