Liu 2016.

Study characteristics
Methods	RCT design: 2‐arm parallel‐group trial Total N randomised: 149 Length of follow‐up: no follow‐up Analysis: unclear (27 dropouts in the treatment group, 28 dropouts in the control group)
Participants	Location: China Number of study centres and setting: single centre, Cardiology of Harbin Medical University Hospital Department CAD criteria: MI in the past month (diagnosis of acute MI according to the Braunwald standard of cardiology); MI patients must comply with symptomatic myocardial ischaemia, and the ECG appears to ischaemic ST‐segment decline or the ST segment elevation or the new pathological Q wave, myocardial enzyme changing observed, such as elevated serum creatine kinase CK‐MB, increased lactate dehydrogenase (LDH1), increased CK (CPK) Depression criteria: depression score HAM‐D ≥ 18 Other entry criteria: none Exclusion criteria: (1) Those suffering from cardiovascular disease: (1.1) acute MI caused by non‐atherosclerotic disease; (1.2) uncontrollable hypertension (systolic BP > 180 mmHg or diastolic BP > 100 mmHg); (1.3) less than 3 months after arterial bypass surgery; (1.4) suffering from arrhythmia; (1.5) suffering from non‐arterial sclerosis (such as anaemia). (2) Those suffering from other somatic diseases: (2.1) an obvious laboratory examination exception; (2.2) distinct hepatic and renal dysfunction; (2.3) with a history of allergies to sertraline, hyperforin, and acanthopanax. (3) Those suffering from mental illness: (3.1) with alcohol or other substance abuse in the last 6 months; (3.2) with psychotic symptoms, psychiatric history or suffering from bipolar disorder, organic mental disorder, dementia, and other diseases; (3.3) frequently receiving benzodiazepines; (3.4) receiving psychological treatment in the last 3 months; (3.5) with suicide attempts; (3.6) receiving monoamine oxidase inhibitor in the last 4 weeks; (3.7) if the patient has a suicide attempt or the depressive symptoms are aggravated in the course of treatment, the psychiatrist would advise the patient to quit the experiment; (3.8) with a history of depression and the HAM‐D24 score higher than 35 before MI. Treatment N: 73 (43.4% female, mean age: 53.4 (SD: 10.3)) Control N: 76 (41.1% female, mean age: 54.1 (SD: 10.8)) Comparability of groups: no significant baseline differences
Interventions	Treatment: sertraline + placebo of Shugan Jieyu capsule treatment Control: Shugan Jieyu capsule + sertraline placebo Duration of treatment: 24 weeks
Outcomes	Review outcomes: depression symptoms (HAM‐D), all‐cause mortality, cardiovascular vital signs (BP, HR), ECG waves, pharmacological side effects Other outcomes: clinical global impression, number needed to treat for non‐inferiority
Funding	None
Notes	Inconsistency in HAM‐D change scores. No significant differences between trial arms were reported, but the data suggest otherwise (likely standard error instead of standard deviation reported). No contact possible in order to clarify results.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "...patients were numbered chronologically and then labeled by a random number." (p.535) Comment: Group allocation via odd or even number
Allocation concealment (selection bias)	Unclear risk	Comment: No sufficient information provided
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Comment: No sufficient information provided Quote: "We used the double‐blind experiment..." (pg. 537)
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: No sufficient information provided
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: data‐analysis paragraph is missing, insufficient information on data‐analysis (ITT/per protocol) and missing data handling procedure
Selective reporting (reporting bias)	Unclear risk	Comment: No study protocol available; assessment points described (see below) do not correspond with results
Other bias	High risk	Comment: data on assessments at 4, 8 and 24 weeks are missing; available results are based on week 12 (middle of the treatment). End of treatment (week 24) not reported