ENRICHD 2003.

Study characteristics
Methods	RCT design: 2‐arm parallel‐group trial Total N randomised: 2481 Length of follow‐up: evaluations after 6 months and annually thereafter (follow‐up duration 18 to 54 months) Analysis: ITT for mortality and cardiac events (93 participants in the intervention arm did not receive the intervention)
Participants	Location: USA Number of study centres and setting: Outpatients from 73 hospitals affiliated with 8 clinical centres CAD criteria: acute MI with elevation in 1 or more biomarkers as well as MI‐compatible symptoms or characteristic ECG ST‐T changes or new Q waves; randomisation within 28 days after MI Depression criteria: major depression or dysthymia diagnosis based on the DISH according to modified DSM‐IV criteria Other entry criteria: low perceived social support assessed through the ENRICHD Social Support Instrument (ESSI) Exclusion criteria: patients with acute MI following PCI or CABG, receiving psychotherapy or taking an antidepressant for longer than 14 days but remained depressed, non‐cardiac conditions likely to be fatal within 1 year, too ill to participate, participating in another trial, major psychiatric disorder (including schizophrenia, bipolar disorder, severe dementia, or active substance abuse), at risk for suicide, refusal of participation or physician disallowed participation, could not be enrolled within 28 days, inaccessible for intervention or follow‐up Treatment N: 1238 (43% female, mean age: 61 (SD: 12.6)) Control N: 1243 (44% female, mean age: 61 (SD: 12.5)) Comparability of groups: no significant baseline differences except for the use of ACE inhibitors
Interventions	Treatment: individual (at least six 1‐hour sessions weekly) and group (weekly 2‐hour sessions) CBT by Beck supplemented by techniques based on social learning theory for participants with low perceived social support; participants with scores > 24 on the HAM‐D or those with less than 50% reduction in BDI score after 5 weeks were referred to study psychiatrist for consideration of pharmacotherapy with sertraline (50 to 200 mg/d) Control: usual care Duration of treatment: individual behavioural intervention up to 6 months with additional 12 weeks for group therapy, adjunctive pharmacotherapy up to 12 months
Outcomes	Review outcomes: depression symptoms (HAM‐D), all‐cause mortality, cardiovascular mortality, cardiac events, healthcare and resource utilisation (cardiovascular hospitalisation), quality of life (12‐item Short Form Health survey) Other outcomes: social support and social networks, life satisfaction, change in cardiac risk factor profile, perceived stress, self‐efficacy
Funding	National Heart, Lung, and Blood Institute Other support: Pfizer Inc provided sertraline for the study
Notes	Mixed study sample (patients with depression and/or low perceived social support were enrolled)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Comment: Automated telephone randomization system using permuted blocks with varying sizes, stratified by clinical center; test for selection bias potentially resulting from unmasking of previous assignments (participants and interventionists were unblinded) with nonsignificant results
Allocation concealment (selection bias)	Low risk	Comment: Allocation obtained by an automated telephone randomization system
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Comment: Participants and interventionists unmasked Quote: "Psychosocial interventions including those used in ENRICHD cannot be fully blinded" (protocol paper, pg. 4)
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: Conflicting information reported for blinded outcome assessment. Quote: "Staff who collected, verified, or classified end point data or follow‐up assessments were masked as much as possible" (Berkman, 2003). Quote: "End point data collection, verification and classification, and follow‐up psychosocial assessments are conducted by staff who are blinded to treatment assignment." (protocol paper, 2000)
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: Depression outcomes analysed per protocol, all other reported outcomes ITT
Selective reporting (reporting bias)	High risk	Comment: results of main outcomes reported as described in the design papers of the trial. Secondary outcomes (change in cardiac risk factor profile, perceived stress and self‐efficacy) were not reported per protocol or ITT
Other bias	Unclear risk	Comment: Therapy quality and adherence to treatment protocol were monitored by an external organisation (the Beck Institute) Comment: QoL was not assessed at baseline and it thus remains unclear whether or not QoL was balanced in the two groups at baseline