Work package 4: feasibility and acceptability studies of the ADVANCE programme to reduce intimate partner violence perpetration by men in substance use treatment

Background

Work package 4(i) assessed the feasibility of conducting an evaluation trial of ADVANCE in substance use treatment. Findings informed an efficacy trial in WP5 that was suspended due to COVID-19. ADVANCE was then adapted for digitally supported delivery (ADVANCE-D) [WP3(ii)], and the feasibility of delivering ADVANCE-D was evaluated [WP4(ii)]. These WPs will be presented chronologically: WP4(i), WP5, WP3(ii), WP4(ii).

Work package 4(i): feasibility and acceptability randomised controlled trial of the ADVANCE group intervention

For more information, see Gilchrist et al.,3 Gilchrist et al.7 and Dheensa et al.9 Report Supplementary Material 1 describes the statistical analysis plan (SAP), and Appendix 2 describes the health economics report for WP4(i). Some text in this chapter has been reproduced with permission from Gilchrist et al.7 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.

Aims

To assess the feasibility and acceptability of conducting an efficacy and cost-effectiveness trial to evaluate ADVANCE, including the acceptability of delivering the intervention in substance use treatment and the feasibility of outcome measure collection from male perpetrators and their female (ex)-partners.

Methods

A multicentre, parallel-group, individually randomised controlled feasibility trial with a nested formative evaluation was conducted, comparing ADVANCE plus substance use TAU to TAU only, for men in substance use treatment. Five pre-specified criteria were established to determine progression to an effectiveness trial: ≥ 60% of eligible male participants recruited; ADVANCE acceptable to staff and male participants; ≥ 70% of participants followed up and no increase in the level of substance use or IPV perpetrated by men in the intervention arm, 16 weeks post randomisation.

Recruitment, setting and participants

Potential male participants were approached by researchers in waiting rooms or in treatment groups or were identified by substance use treatment staff in six community services in England (London, the West Midlands and the Southwest). Flyers and posters in services reached more men. Men first consented to be screened for eligibility by a researcher and for the researcher to discuss this with their keyworker. If eligible after screening, men then consented to trial participation.

Men were eligible if they:

1. had perpetrated IPV towards an (ex)-partner in the last 12 months assessed using the adapted Revised Abusive Behaviour Inventory (ABI-R)
2. had face-to-face, phone, e-mail or social media contact with their (ex)-female Partner/s at least once in the past 12 months
3. planned to stay in the current location for the next 6 months
4. agreed to provide contact details of their (ex)-partner/s and
5. could understand and communicate in English.

Men were excluded if they:

1. had a current restraining order prohibiting them or anyone on their behalf from contacting their (ex)-partner
2. had pending IPV court cases
3. had pending child protection hearings or
4. were attending an IPV perpetrator programme.

Researchers administered the baseline interview with eligible men.

Randomisation and blinding

Researchers randomised participants to ADVANCE + TAU (intervention) versus TAU only (control) immediately following baseline using an online randomisation system (1 : 1), managed by the UK-registered King’s Clinical Trials Unit, using randomly varying block sizes, stratified by a combination of sites and temporal cycles. Up to 18 men per treatment service were randomised. Men and keyworkers were then informed of allocation. Researchers were not blind to treatment allocation. Both statisticians were subgroup blind until database lock, with the senior statistician remaining blind during analysis.

Recruitment of (ex)-partners

Following randomisation, (ex)-partners were offered support from an ISS and provided more information about the trial. Researchers called women interested in taking part. Following consent, interviews with women took place in the substance use treatment service or the ISS where possible. Several interviews took place in women’s homes or a library with two researchers present.

Intervention

The Template for Intervention Description and Replication (TIDieR) checklist140 describes the ADVANCE programme components (Table 4).

Control

Male participants in both treatment arms received substance use TAU.

Outcomes

Confounding variables and potential outcome measures for use in a future trial were assessed at baseline and after treatment (16 weeks post randomisation) (Table 5).141–163

Assessment of trial feasibility

Eligibility, recruitment, randomisation and follow-up rates for men by site and group allocation, and recruitment and follow-up rates for (ex)-partners assessed feasibility.

Assessment of intervention acceptability

Process variables assessed the intervention’s acceptability: rate of intervention session attendance; days between randomisation and attending the first individual session and group intervention starting. The Working Alliance Inventory-ShortRevised (WAI-SR)162 and the California Psychotherapy Alliance Scale – Short Form163 measured therapeutic and psychotherapy alliance, respectively, for men allocated to ADVANCE.

Procedure

Data were collected during July 2018–July 2019. Three temporal cycles of ADVANCE were delivered. Men received a £5 voucher (for a chosen shop/service) for each session attended, awarded at sessions 6 and 12, with a £10 ‘bonus’ for attending all 12 sessions. Travel was reimbursed, and refreshments were provided.

Sample size

The target sample size to estimate the parameters for a definitive trial (WP5) was 76 (ex)-partners and approximately 108 perpetrators (6 sets of 18 men) based on a retention rate of 80% of survivors post intervention.

Data analyses

Feasibility parameters were estimated with 95% CI precision. The difference in mean outcomes was estimated between participants randomised to ADVANCE + TAU and TAU only by intention to treat at 16 weeks post randomisation. No formal significance tests were performed. These effects were also standardised by dividing the estimated mean difference by the respective (pooled group) SD at baseline. Linear regression models were used. Fixed effects included baseline measures of the outcome, trial arm and randomisation stratifiers, site and cycle. The number of (ex)-partners with follow-up data was too low (n = 17) for statistical analyses.

Formative evaluation

Focus groups or semistructured interviews with substance use treatment staff, ISS workers and men who attended at least one ADVANCE session were digitally recorded and transcribed verbatim. Data were organised and coded by multiple coders using NVivo. Framework analysis164 was used to explore patterns in themes across different participants and groups of participants.

Economic analysis

The costs of training staff to deliver ADVANCE and the delivery costs of ADVANCE based on attendance were estimated.

Key findings

Screening, recruitment and follow-up

The screening, recruitment and follow-up of men and their (ex)-partners is described in Figures 11 and 12: 104 men were randomised to the intervention (n = 54) or control (n = 50) arm: 39 from London, 25 from the West Midlands and 40 from the South West.

FIGURE 11 The CONSORT diagram for male participants in the ADVANCE feasibility RCT. a Not currently listed in protocol. Reproduced with permission from Gilchrist et al.7 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)

FIGURE 12 The CONSORT diagram for female participants in the ADVANCE feasibility RCT. Reproduced with permission from Gilchrist et al.7 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)

Participant characteristics

Table 6 presents participants’ baseline characteristics.

Feasibility parameters

Table 7 describes the feasibility parameters for male participants and (ex)-partners.

Patient-centred outcome measures

Estimated treatment differences at follow-up were calculated (Table 8). The sign of the estimated group difference indicated improvement on all scales, except for Primary Care PTSD Screen for DSM-5 (PC-PTSD-5), the anger subscale from the Propensity for Abusiveness Scale (PAS) and the Brief Self-Control Scale (BSCS) (Figure 13). Standardised estimated effect sizes are in the small range (< 0.2), and all associated CIs cross the line of no difference, zero.

FIGURE 13 Standardised effect sizes at 16-week follow-up for male participants in the ADVANCE feasibility study. BIDR-SF, Balanced Inventory of Desirable Responding – Short Form. Reproduced with permission from Gilchrist et al.7 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)

Attendance

Sets of up to 18 men per treatment service were randomised. Three sets (one in each region) were completed (cycle 1) and a formative evaluation undertaken to inform the implementation of cycle 2 (a further 3 sets). Unfortunately, in London, it was not possible to deliver ADVANCE during cycle 1, as no men allocated to ADVANCE were available when the intervention began. A third cycle was therefore undertaken in London.

On average, 4.4 (SD 4.5) of the 14 compulsory sessions were attended for the 47 men offered ADVANCE. Attendance improved by cycle 3 (Table 9). Overall, 66% of men attended at least one intervention session, rising to 85.7% in cycle 3. For those 34 men who attended at least one session, the mean number sessions attended was 6.6 (SD 3.9). Men who attended no intervention sessions were slightly younger (39 vs. 44 years), and a greater proportion were living in a hostel/supported accommodation (37.5% vs. 25.8%) and met criteria for moderate to severe depression (68.8% vs. 51.6%), anxiety (56.3% vs. 38.7%) or PTSD (75.0% vs. 58.6%).

Intervention acceptability

For men allocated to ADVANCE and followed up, therapeutic alliance was high. Seven focus groups and 7 interviews with 31 facilitators and substance use keyworkers, and a focus group with 5 men who attended at least one session of the intervention (cycle 3) assessed acceptability.

Across all cycles, keyworkers and facilitators supported the need for ADVANCE and felt it could be easily integrated into their services, but facilitators required protected time to prepare for intervention delivery; otherwise, it affected existing workload. Fortnightly integrity support from the ADVANCE team helped clarify issues about session content: ‘once I’d spoken to (ADVANCE integrity support) a few times and I understood it, I realised how much it flowed’ (facilitator, London, cycle 3). Safeguarding procedures, including the case management meetings and weekly follow-ups with the ISS workers, ensured risk was managed or mitigated, and the importance of multiagency working was stressed: ‘We had one incident of risk and we referred to social services. We worked together and spoke to each other’ (facilitator, London, cycle 3).

Men attending ADVANCE were motivated to improve their relationships: ‘I wanted to address some issues that I had’ (P030032), ‘I had assaulted her, I just wanted it to try and make myself better’ (P020041). Men found ADVANCE relevant: ‘I’ve learnt things from this course which I can put into practice’ (P010099), ‘That crisis plan, yes. it was a very good thing to learn, for me personally’ (P020041). One facilitator reported that ‘we did exactly what it said in the manual and we let everybody go round the table and say two last words. One of them, his words were “life changing”’ (facilitator, London, cycle 3).

Across the cycles, men benefited from being in a group of men with shared experiences and appreciated the support they received from other group members. In all three cycles, men were impressed by the skills, enthusiasm and sensitivity of the facilitators, which motivated men to continue attending.

Intervention safety

Two serious adverse events were recorded and assessed by the DMEC chair and NHS Ethics Committee as unrelated to the trial.

Economic evaluation

Overall, the total group session duration was 126.25 hours, and the total cost was £8585 across six sets of intervention delivery. Among male participants, the mean costs in the control arm were consistently higher than in the intervention arm.

Limitations

Delays between randomisation and the group starting resulted in cycle 1 in London not being undertaken. Attendance was low overall, although it improved by cycle 3. A low proportion of men’s (ex)-partners were recruited.

Discussion

It was possible for trained staff to safely deliver ADVANCE to male IPV perpetrators in substance use treatment services, and for risk to their (ex)-partners to be effectively managed and mitigated through case management and integrated support. Challenges encountered were similar to other evaluations of complex interventions: staffing and contextual issues.165

The reduction across cycles in the number of days from randomisation to the individual sessions (46.7 to 17.2 days) and the first group session (from 46.6 to 39.1 days) starting could have impacted on attendance as the median rate of intervention session attendance increased to 64.3% by cycle 3. Improvements in attendance and retention by cycle 3 may also be related to researchers and staff being more confident with study procedures.

The low recruitment of (ex)-partners confirmed it would not be feasible to conduct a future trial of ADVANCE using (ex)-partners’ reports of IPV victimisation as the primary outcome. Alternative ways of assessing outcomes should be considered, including using record linkage.

Feasibility of conducting a definitive trial

Funders supported progression to a definitive trial: ≥ 60% of eligible male participants recruited (met); ADVANCE acceptable to staff and male participants (partially met); ≥ 70% of participants followed up (not met) and no increase in the level of substance use or IPV perpetrated by men in the intervention arm 16 weeks post randomisation (met; data not presented).