Soler 2005.

Study characteristics
Methods	12‐week trial with 2 arms: olanzapine + DBT placebo + DBT Duration: 12 weeks (after a 4‐week selection phase during which the pre‐intervention baseline was established but no therapeutic intervention was given) Country: Spain Setting: outpatient
Participants	Method of recruitment of participants: Patients were referred from clinical services. Overall sample size: 60 Diagnosis of borderline personality disorder: DSM‐IV Means of assessment: SCID‐II and DIB‐R Mean age: 40.74 years (SD = no information; range = no information) Sex: 86.57% women Comorbidity: no information; however, patients with unstable comorbid axis I disorders were excluded. Inclusion criteria Meeting DSM‐IV diagnostic criteria for borderline personality disorder assessed by SCID‐II and the DIB‐R Aged 18–45 years CGI‐S of illness score ≥ 4 Exclusion criteria Comorbid unstable axis I disorder Women that did not use medically accepted contraception Patients receiving psychotherapy
Interventions	Experimental group Treatment name: olanzapine Number randomised to group: 30 Duration: 12 weeks Control/comparison group Comparison name: placebo (no further details) Number randomised to group: 30 Duration: 12 weeks Both groups Concomitant psychotherapy: All patients received DBT (weekly, 150‐minute skills training group sessions; phone calls). Concomitant pharmacotherapy: Participants could continue treatment with benzodiazepines, antidepressants and mood stabilisers, but doses could not be modified. Proportions of participants taking standing medication during trial observation period: Participants were taking other medications before or during the treatment in both groups. In the olanzapine group, 73.3% of participants were taking benzodiazepine, 80% were taking antidepressants and 33.3% were taking mood stabilisers. In the placebo group, 60% of participants were taking benzodiazepine, 70% were taking antidepressants and 16.7% were taking mood stabilisers.
Outcomes	Primary outcomes Suicidal behaviour/self‐mutilating behaviour, measured by behavioural, biweekly reports of episodes of self‐injuring behaviour/suicide attempts. Assessed at baseline and every 2 weeks for 12 weeks (EOT) Mental health status (functioning), measured by CGI‐S. Assessed at baseline and every 2 weeks for 12 weeks (EOT) Secondary outcomes Impulsivity, measured by behavioural, biweekly reports of episodes of impulsivity/aggressive behaviour. Assessed at baseline and every 2 weeks for 12 weeks (EOT) Depression, measured by Ham‐D. Assessed at baseline and every 2 weeks for 12 weeks (EOT) Attrition Adverse effects, measured by patients' reports and scales assessing extrapyramidal side effects, weight, cholesterol levels. Assessed biweekly for 12 weeks
Notes	Sample calculation: no information Ethics approval: no information Funding source: funded or partially funded by pharmaceutical industry Conflicts of interest: No conflicts of interest were reported besides partial funding from the pharmaceutical industry. Comments from trial authors (limitations) "results cannot be fully extrapolated to inpatients, to patients with active comorbid axis I disorders, or to those with less clinically severe disorders". (Soler 2005, p. 1223) "We are unaware of any possible drug‐drug interactions, the 'masking' effect of psychotherapy, or whether olanzapine is useful as maintenance treatment." (Soler 2005, p. 1223)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: insufficient information on methods used to generate random sequence to permit a judgement of low or high risk of bias
Allocation concealment (selection bias)	Unclear risk	Comment: No information given on allocation concealment to permit a judgement of low or high risk of bias
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Comment: The trial was referred to as being double‐blind, however, insufficient information was given on how blinding of participants and personnel was carried out (packaging of trial medication etc.) and maintained, to permit a judgement of low or high risk of bias.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: The trial was referred to as being double‐blind, however, insufficient information was given on how blinding of outcome assessors was carried out and maintained, to permit a judgement of low or high risk of bias.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Quote: "All analyses were conducted on an intent‐to‐treat basis. The endpoint was based on a last‐observation‐carried‐forward strategy. Patients were included in the analyses only if they had a baseline measure and at least one post‐baseline measure." (Soler 2005 p. 1222) Quote: "Sixty subjects were randomly assigned to dialectical behaviour therapy plus olanzapine or placebo and started the experimental phase; 42 subjects (70%) completed the study.There were no between‐group differences regarding demographic variables or concomitant treatments at baseline. Neither dialectical behaviour therapy intervention time nor dropout rates differed significantly between the two groups (eight of the 30 patients who received olanzapine versus 10 of the 30 who received placebo dropped out before the end of the study." (Soler 2005 p. 1222 et seq.) Comment: reasons for dropouts given; numbers balanced across groups Continuous outcomes based on ITT (LOCF) Of the 60 patients enrolled, 42 completed treatment (22 in active drug group, 20 in placebo group) Reasons for early termination: No reasons given
Selective reporting (reporting bias)	Low risk	Comment: The trial protocol was available and all of the prespecified primary and secondary outcomes that were of interest in the review were reported in the way they were prespecified.
Vested Interest (funding and/or author affiliations)	High risk	Quote: "Supported by grants from the Fondo de Investigación Sanitaria (Ministry of Health, Spain) and from Eli Lilly and Co. Madrid." (Soler 2005 p. 1223)
Other bias	Low risk	Comment: No indication of other bias