Thylstrup 2015.

Study characteristics
Methods	Design: parallel, randomised control trial
Participants	Participants: new and existing patients with AsPD receiving outpatient treatment for drug or alcohol problems Sex: male (n = 124/142, 87%); female (n = 18/142, 13%); data only for the participants who completed the study; no data provided for the 176 randomised patientsa Age: (for sample as a wholea) mean age = 32.2 years (SD = 8.9) Unit of Allocation: cluster; stratified by clinic; randomised permuted blocks or randomly varying size with ratio 1:1 (4 or 6 per block) Number randomised: 176; impulsive lifestyle counselling (ILC) + treatment‐as‐usual (TAU) = 96; TAU = 80 Number completing: 142 patients completed at least 1 follow‐up; ILC + TAU = 78; TAU = 64 Setting: community‐based substance abuse treatment centres across 13 municipalities of Denmark Inclusion criteria: outpatients aged 18 to 65; meet AsPD criteria using Mini International Neuropsychiatric Interview (MINI); able to provide written informed consent; seeking/already in treatment for a substance use disorder Exclusion criteria: patient has plans that would interfere with participation in psychoeducation in next 3 months (e.g. moving away, prison, residential rehab); participating in group treatment with another subject in trial; known to suffer from acute psychosis; severe brain damage; does not speak Danish Ethnicity: no information given Baseline characteristics: Conduct disorder criteria: ILC + TAU group mean = 4.04 (SD = 0.83); TAU group mean = 3.90 (SD = 0.80) Adult antisocial behaviours: ILC + TAU group mean = 5.14 (SD = 0.98); TAU group mean = 4.87 (SD = 1.07) Days of substance use in previous 30‐day period: ILC + TAU group: alcohol use (mean = 4.81, SD = 6.97); alcohol intake + 5 units per day (mean = 3.16, SD = 5.61); heroin (mean = 0.49, SD = 2.75); methadone (mean = 7.73, SD = 12.87); other opioids (mean = 0.83, SD = 3.26); tranquillisers (mean = 7.42, SD = 11.93); cocaine (mean = 1.31, SD = 4.38); amphetamines (mean = 1.43, SD = 3.83); cannabis (mean = 24.26, SD = 13.09); hallucinogens (mean = 0.04, SD = 0.26); inhalants (mean = 0.01, SD = 0.12); buprenorphine (mean = 1.64, SD = 6.70); poly‐substance (mean = 7.57, SD = 11.58) TAU group: alcohol use (mean = 6.46, SD = 8.45); alcohol intake + 5 units per day (mean = 3.44, SD = 5.92); heroin (mean = 1.48, SD = 5.15); methadone (mean = 7.20, SD = 12.83); other opioids (mean = 0.92, SD = 4.27); tranquillisers (mean = 8.00, SD = 12.26); cocaine (mean = 1.31, SD = 3.91); amphetamines (mean = 0.36, SD = 0.83); cannabis (mean = 13.54, SD = 12.86); hallucinogens (mean = 0.14, SD = 0.71); buprenorphine (mean = 2.78, SD = 8.44); poly‐substance (mean = 6.77, SD = 10.52)
Interventions	Two conditionsb: impulsive lifestyle counselling (ILC); or treatment‐as‐usual (TAU) Experimental group: ILC + TAU (n = 96 randomised) Control group = TAU (n = 80 randomised) Details of conditions: Impulsive lifestyle counselling (ILC): treatment was described as "...a brief psycho‐educational intervention..." (quote; p 2); patients randomised to ILC were offered up to six ILC sessions by a specially trained counsellor; the ILC programme is a manualised intervention; sessions cover specific topics and include mandatory questions, printed handouts and worksheets for the patient. Duration of session was not reported. TAU always included: access to opioid substitution treatment (if required); psychosocial support such as casework, counselling, or referral to residential rehabilitation; referral to 'off‐site' psychiatrist for treatment of other psychiatric conditions, such as attention‐deficit/hyperactivity disorder, anxiety or depression. Duration of intervention: 6 sessions, time period not reported Duration of trial: 31 months; January 2012 to July 2014. Length of follow‐up: follow‐up at 3 months and 9 months
Outcomes	Primary outcomes Aggression: trait aggression scores on the 12 item (short‐form) Buss‐Perry Aggression Questionnaire (BPAQ‐SF); scores on the Self‐Report of Aggression and Social Behaviour Measure (SRASBM) Adverse events: number of deaths, incarceration Secondary outcomes Leaving the study early: attrition at 3 and 9 month follow‐up Substance use: scores on Addiction Severity Index (ASI) for drugs and alcohol; days abstinent; % participants reporting complete abstinence; % participants reporting daily use Other outcomes None
Notes	aData analysed for the 142 participants who completed at least one follow‐up (79% of whole group). b36.5% of whole group received opioid substitution treatment at the point of randomisation. Study funding: Trygfonden, and Reckitt‐Benckiser Declaration of interests: None
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The randomization schedules were generated by the trial coordinator and kept secure and confidential at the study coordinating center in Copenhagen. The randomization schedule was constructed using the method of randomized permuted blocks of randomly varying size with a ratio of 1:1 (4 or 6 per block)." (p 3)
Allocation concealment (selection bias)	Unclear risk	Quote:"The trial coordinator informed the referring clinician of the result of randomization immediately after being notified that the patient had been assessed and was found to be eligible for study participation. After this, the clinician informed the patient of the result. In the cases in which patients were randomized to the ILC treatment, the clinician then contacted one of the ILC counsellors at the uptake unit with the participants’ details so that the sessions could be initiated as quickly as possible. Because the randomization had to take place immediately after the assessment interview, the trial coordinator was unable to check whether the baseline assessment was complete before randomizing, and patients with incomplete data at baseline had to be excluded after randomization." (p 3).
Blinding (performance bias and detection bias) of participants	Unclear risk	Comment: Participants were not blind. The trial investigators reported that patients were informed of the result of the randomisation by the clinician undertaking the baseline eligibility assessments.
Blinding (performance bias and detection bias) of personnel	Unclear risk	Quote:"The trial coordinator informed the referring clinician of the result of randomization immediately after being notified that the patient had been assessed and was found to be eligible for study participation... ... In the cases in which patients were randomized to the ILC treatment, the clinician then contacted one of the ILC counsellors at the uptake unit with the participants’ details so that the sessions could be initiated as quickly as possible." (p 3)
Blinding (performance bias and detection bias) of outcome assessors	Unclear risk	Quote: "Research technicians not affiliated with the clinics carried out all assessments at the 3 and 9‐month follow‐up interviews and were blind to treatment group allocation." (p 5). Comment: The review authors noted that there was no reference to the blinding of the data analysts and therefore insufficient information to allow a judgement to be made.
Incomplete outcome data (attrition bias) All outcomes	High risk	Comment: Authors reported it was an ITT analysis but it was not clear that this was the case as 176 were randomised but they only analysed data for subjects who completed baseline data (167) and then completed at least one follow‐up (142).
Selective reporting (reporting bias)	Unclear risk	Comment: Trial information on the ISRCTN register indicated additional secondary outcome measures that are not reported in the 2015 paper: perceived help for antisocial personality disorder (reported in 2017 post hoc secondary analysis only); readiness to change antisocial behaviour measured using the adapted readiness ruler; and staff‐rated improvement in in‐clinic antisocial behaviour, general antisocial behaviour and substance use.
Other bias	Unclear risk	Comment: Vested interest bias (funding and/or author affiliations): although not a pharmacological trial, the study was partially funded by grant from Reckitt‐Benckiser who manufacture an opioid replacement drug. Some subjects in the trial were administered opioid substitution treatment.