Simpson 2011.

Methods	RCT with wait‐list control, cross‐over design.
Participants	Seventeen patients recruited from the Liverpool (Australia) Hospital brain injury community team caseload. Inclusion: severe TBI (PTA > 1 day), aged 18 years or older, moderate to severe levels of hopelessness, suicidal ideation, or both. Exclusion: severe neuropsychological impairments in cognitive or language functions, extremely challenging behaviour that would preclude compliance with the study protocol, and non‐fluency in English.
Interventions	Cognitive‐behavioural therapy delivered via a 20‐hour manualised group‐based programme, delivered in 10 weekly 2‐hour sessions.
Outcomes	Primary outcome measures: Beck Hopelessness Scale (BHS) Beck Scale for Suicide Ideation (BSS) Hospital Anxiety and Depression Scale (HADS) Secondary outcome measures: Herth Hope Index Rosenberg Self‐Esteem Scale Social Problem Solving Inventory‐Revised (SPSI‐R) Timepoints measured: Baseline At completion of treatment (10 weeks after baseline) 3 months after completion of treatment
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Block randomisation: groups of 4 participants allocated to an intervention, using a computer‐generated set of random numbers.
Allocation concealment (selection bias)	Low risk	Allocation to intervention conducted off‐site and allocation was concealed.
Blinding of participants and personnel (performance bias) All outcomes	High risk	The study interventions were either an active treatment or a wait‐list control, and therefore, blinding of participants and personnel was not possible.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Assessments at completion of treatment and at 3‐month follow‐up were conducted by an independent assessor who was blind to the intervention group. Participants were asked not to disclose their intervention group to the assessor. The independent assessor was asked to record any inadvertent disclosure of the participants' intervention group. However, the primary outcome measures were self‐report scales and therefore, subject to bias since the participants were aware of the intervention group to which they were assigned.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Seventeen participants' were randomised to groups. Only one subject withdrew prior to the final assessment time point.
Selective reporting (reporting bias)	Low risk	Primary author provided the study protocol, which showed that all outcomes collected were reported.
Other bias	High risk	Small sample size (intervention group, N = 8 and control group, N = 9).