U‐CARE 2018.
| Study characteristics | ||
| Methods | RCT design: 2‐arm parallel design Total N randomised: 3928 Length of follow‐up: 14 weeks Analysis: ITT with multiple imputation by chained equations for depression, ITT without missing data for cardiovascular mortality and cardiac events (28 dropouts intervention group, 10 dropouts control group at 12 months) |
|
| Participants | Location: Sweden Number of study centres and settings: 25 Swedish hospitals CAD criteria: recent MI < 3 months Depression entry criteria: > 7 on 1 or both of the 2 HADS subscales Other entry criteria: < 75 years old Exclusion criteria: scheduled for coronary artery bypass surgery; unable to use computer, internet, email, or mobile phone; unable to read Swedish; expected to live < 1 year; anticipated to show poor compliance (e.g. substance abuse or not showing up to the cardiac nurse visit); self‐reported severe depression or suicidal ideation (MARDS‐S total score > 34 or MARDS‐S item 9 > 3); participating in another behavioural intervention trial Treatment: 117 participants (44% women, mean age: 58.4 (SD: 9)) Control: 122 participants (36% women, mean age: 60.8 (SD: 7.8)) Comparability of groups: no significant baseline differences |
|
| Interventions | Treatment: therapist‐guided internet CBT treatment Control: usual treatment Duration of treatment: 14 weeks |
|
| Outcomes | Review outcomes: depression symptoms (Montgomery–Åsberg Depression Rating Scale, also HADS), cardiovascular mortality, cardiac events Other outcomes: anxiety symptoms (HADS Anxiety), behavioural activation (Behavioral Activation for Depression Scale), cardiac anxiety (Cardiac Anxiety Questionnaire), adherence to treatment |
|
| Funding | Swedish Research Council | |
| Notes | Mixed study sample (patients with symptoms of depression or anxiety were recruited) | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Comment: sequence generation with stratification by clinical centre Quote "1:1 allocation, using a computer‐generated code" |
| Allocation concealment (selection bias) | Unclear risk | Comment: Randomization occurred automatically in the internet‐based portal |
| Blinding of participants and personnel (performance bias) All outcomes | Unclear risk | Comment: technical and telephone support staff blind to allocation |
| Blinding of outcome assessment (detection bias) All outcomes | Unclear risk | Comment: no details on outcome assessment blinding; primary outcomes self‐reported |
| Incomplete outcome data (attrition bias) All outcomes | High risk | Comment: Intention‐to‐treat as the main analysis. Reasons for drop‐out not provided in Norlund or Humphries |
| Selective reporting (reporting bias) | Low risk | Comment: primary outcomes reported (depression and anxiety) Comment: secondary endpoints in protocol not yet reported ‐ quality of life, stress behaviors, fatigue, sleep pattern, posttraumatic stress, posttraumatic growth, health economy aspects, cost‐effectiveness of the intervention, major adverse cardiac events |
| Other bias | High risk | Comment: There was a change in the inclusion criteria during the study due to low recruitment numbers. The HADS threshold was lowered from ≧10 to >7 on any subscale of the HADS. The recruitment target in the protocol was 500 (eventual recruitment n = 239) Comment: manualised treatment (online) and therapist support was provided for the online modules in patients allocated to treatment; no information on therapists adherence or quality Comment: email and telephone prompts provided; unclear if to intervention and control groups |