Weinberg 2006.

Study characteristics
Methods	design: randomised controlled trial
Participants	sex: 30/30 females (100%) age: 28.2 years on average location: USA setting: outpatient exclusions: comorbid psychotic disorders, bipolar I disorder, substance dependence, elevated suicide risk level of functioning/severity of illness: baseline frequency of self‐harming behaviour: group 1: 9.33, SD = 14.78; group 2: 8.20, SD = 10.46 BPD diagnosis according to: DSM‐IV means of assessment: SCID, DIB‐R (both sets of criteria had to be met for inclusion)
Interventions	group 1 (EG): Manual‐assisted cognitive treatment (MACT; i.e. adjunctive intervention to ongoing treatments of the participants; six session therapy incorporating elements of DBT, cognitive behavioural treatment, and bibliotherapy) group 2 (CG): Treatment as usual (TAU); all subjects took part in additional treatments not further specified duration: 6 weekly sessions concomitant psychotherapy: both EG and CG participants received treatment as usual; all participants took part in additional treatments not further specified concomitant pharmacotherapy: both EG and CG participants received treatment as usual; no further details on amounts or types of medications used
Outcomes	outcomes considered in this review self‐rated: ‐ observer‐rated: suicidality (SBQ), parasuicidality (Parasuicide History Interview (PHI) ‐ deliberate self‐harm frequency), time‐points used here: 2 months (i.e. post‐treatment)
Notes	analyses: ITT
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"randomly assigned" (Weinberg 2006, p. 485) No further details.
Allocation concealment (selection bias)	Unclear risk	No further details. After screening of 60 referrals by phone, 37 were invited for further assessments. Reasons for exclusion of N = 7 are given. N = 30 were included in the final sample. N = 15 were assigned to the EG and N = 15 to the CG.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"The baseline assessment and administration of the MACT [i.e. the treatment under test] were performed by the primary investigator." (Weinberg 2006, p.486) "Interviewers were randomly assigned for following assessments. The interviewers were blind to baseline ratings and to participants' group allocation" (Weinberg 2006, p. 487)
Selective reporting (reporting bias)	Low risk	No indications for selective reporting.
Treatment adherence?	High risk	"This study did not monitor adherence and competence." (Weinberg 2006, p. 489)
Allegiance effect improbable?	Low risk	No indication for allegiance effect.
Attention bias: equal amounts of attention to all groups (obligatory treatment components)?	High risk	More attention spent to EG participants.