Trial Design
The study design is reported in detail elsewhere [20]. This is a parallel, two-arm, feasibility RCT comparing the clinical and cost-effectiveness of SPADe and usual care, against usual care only, in maintaining alcohol abstinence in adults with alcohol dependence receiving care in the community. The trial follows the guidelines of phase 2 of the Medical Research Council (MRC) for complex interventions [21] and includes: adaptation of the intervention, feasibility study including health economics, and process evaluation. No changes of the protocol trial design has taken place.
Assessments
Participant assessments were conducted at baseline and at 3, 6 and 12 months by the research assistants (RAs) who were trained in assessing capacity and obtaining consent, promoting the study, administering the questionnaires, and interviewing participants. A window of two weeks at either side of the assessment due points was endorsed in order to maximise engagement with follow up.
Participants
Inclusion criteria
-
1. Presentation to either of the two alcohol services in London (Hounslow, site 1 and Camden, site 2) seeking abstinence from alcohol.
-
2. Alcohol dependence (moderate to severe), scoring 16 and above on Severity of Alcohol Dependence Questionnaire (SADQ) (see outcome measures below). This level of dependence indicates that it would be clinically appropriate to receive a medically assisted detoxification (4).
-
3. Stated intention to stay in the area within the time period of the intervention.
-
4. Willingness to be part of a group intervention if randomised to receive it.
Exclusion criteria
-
1. Age less than 18 (as not usually treated by specialist alcohol services).
-
2. Pregnancy: pregnant women need urgent intervention to withdraw from alcohol, due to the effect of alcohol on the foetus.
-
3. Known terminal illness with life expectancy of less than 6 months.
-
4. Severe medical condition that requires urgent medical admission, which would lead to an unplanned medically assisted withdrawal.
-
5. Severe cognitive impairment that compromises capacity and /or ability to participate in a group intervention.
-
6. Acute stage of severe and enduring mental illness (schizophrenia, Bipolar Affective Disorder, recurrent depressive disorder: current episode severe), when acute symptomatology compromises service user’s ability to participate in a group intervention.
Study Settings
Participants were recruited from specialist alcohol community services offering recovery orientated treatment for people with Alcohol Use Disorders (AUD). Recruitment took place in two sites in London, both offering the intervention and usual care, in order to explore challenges associated with implementation of the intervention across two different set up of services. Both sites were run by a partnership between a third sector organisation and the National Health Service (NHS), a common funding model in the UK. The two sites had different integration levels; site 1 was fully integrated with a single management and clinical governance system; site 2 had two parallel systems. Furthermore site 1 had been offering the intervention for several years as part of the standard treatment pathway, whereas in site 2 the intervention had to be added to the clinical pathway.
Interventions
SPADe
The intervention provides structured group preparation (additional to usual care) with the aim of helping participants (i) regain partial control over their drinking prior to detoxification and (ii) initiation of lifestyle changes for the individual and the immediate family environment. These changes are linked with developing new coping skills and enhanced self-efficacy [22].
The six sessions are numbered and offered weekly in a given order. Stabilisation of the amount and pattern of drinking is a common theme across all the sessions. To that effect each session can act as an entry point (i.e. an open rolling programme group), despite the special theme covered in depth during the second part of the session. Each session has two facilitators, lasts for one hour and is divided into three parts:
-
• In the first part (15 minutes) group rules are established, new members are introduced, aims of the intervention. In-between sessions practice allocated in previous session where individual targets were set are reviewed.
-
• The second part (30 minutes, main part) explores the following themes depending on session number: 1 – Understanding habit, addiction and alcohol dependence; 2 – Stabilise and control your drinking; 3 – Lifestyle changes for you and the people around you; 4 – Reduction of your drinking; 5 – Achieving abstinence; 6 – Relapse prevention strategies.
-
• In the third part (15 minutes), the group summarises the main learning points and agrees in-between sessions practice and targets to be achieved before the next session. A group work folder provided enabling notes and worksheets to be kept together.
The number of participants per group at any point is between two and eight. This is considered appropriate for theory-based treatment groups, to (i) reach a balance between the educational and treatment elements of the intervention, (ii) promote interactions between group members (group therapeutic effect) and (iii) secure the facilitator’s attention to each member individually [23]. The duration of 1 hour as suggested in Patient and Public Involvement meetings, reduces the risk of withdrawal symptoms and help participants to maintain concentration.
Usual Care
Usual Care includes: planning for detoxification; detoxification delivery; and aftercare. Participants entered detoxification at the first available opportunity (expected to be within 4 weeks from presentation). Whilst waiting for detoxification, they meet their keyworker (one-to-one) on 3–4 occasions to maintain motivation and plan aftercare. Participants allocated to the intervention arm entered detoxification at the first available opportunity following completion of the intervention.
Detoxification was medically assisted in the community as an outpatient (both sites), or inpatient (only for site 1), as clinically indicated. The choice depends on health risk factors and availability of social support during detoxification [4]. Furthermore participants allocated to Intervention could achieve abstinence by Guided Self-Detox. Guided Self-Detox refers to gradual reduction of drinking guided by the group facilitator during the intervention period. As mentioned above aftercare (following detoxification) included a relapse prevention group, a small number of individual key worker sessions, pharmacological interventions as appropriate peer support groups such as Self-Management and Recovery Training (SMART) Recovery or Alcoholics Anonymous (AA) or more comprehensive aftercare group programme either as an outpatient (both sites) or inpatient (only site 1).
Participants in both intervention and control groups received all the above elements of usual care available in the recruiting service. At any time during the trial, change of clinical needs or risks were monitored by the clinical team and participant’s treatment plan was modified accordingly, e.g. urgent hospital admission. These incidents were monitored and reported as per standard ethical recommendations for safety monitoring. Each participant’s care pathway was recorded and analysed for variability within and between sites as part of the economic evaluation (see below), to ensure these are equivalent across trial arms.
Outcomes
Feasibility measures
Feasibility outcomes were: (I) Recruitment and retention rates: monthly monitoring of the number of alcohol dependent clients accessing services during the recruitment period of the study; how many meet the eligibility criteria; and how many were invited and accepted into the study and retained in each group for the full 12 months. (II) Compliance with treatment: number of SPADe sessions attended (for the intervention arm) using the facilitator’s record of attendance. (III) Data collection and completeness: attendance for assessments and completeness of instruments; loss to follow up and missing data for all outcomes were analysed.
A variety of possible outcomes were used at different time points (3, 6 and 12 months following randomisation), so the primary outcome for the main trial could be identified and sample size calculations to be conducted, including: duration of continuous abstinence with no incidents of lapse or relapse; percentage of days of abstinence (PDA) (both self-report using Time Line Follow Back method) (TLFB) [24]; time to relapse (from stopping alcohol to first day of alcohol use; also as defined by self-report).
Secondary outcomes
Secondary outcomes were measured using validated instruments wherever possible at 3, 6 and 12 months:
-
1. Severity of Alcohol Dependence Questionnaire (SADQ), a 20-item self-completion questionnaire, scores range 0 to 60, (16 to 29 indicates moderate severity, above 30 severe dependence) [25].
-
2. Alcohol Urge Questionnaire (AUQ), an eight-item self-completion questionnaire containing three domains of drinking urges: desire for a drink; expectation of positive effect from drinking; inability to avoid drinking [26].
-
3. Incentive Conflict Task (ICT). This task assesses inherent conflict in abstaining drinkers, between the intention to abstain from drinking and the desire to drink, that may contribute to relapse [9]. Alcohol-dependent individuals, as they experience successive detoxifications, become increasingly impaired in performing the ICT [27].
-
4. Euro-QoL (EQ-5D-5L); English version [28], which is a short questionnaire assessing general healthcare used in economic evaluations for the calculation of Quality Adjusted Life Years.
-
5. Self-reported participation in aftercare activities, using a specifically developed log, measuring type and frequency of activity attended, during the period prior the follow up interview.
Process evaluation – treatment fidelity
Observation of twenty five percent of the intervention sessions, offered across both recruitment sites, during the whole duration of the studywas planned using the Yale Adherence and Competence Scale (YACS II) (2005) [29]. Furthermore, 10% of the sessions was planned to be rated by an additional independent rater using YACS II. Group facilitators were asked to complete a self-assessment form following each session to reflect on their fidelity to the intervention manual.
Audio recording and rating of twenty five percent of the key-working sessions offered to the usual care group was plannedin order to assess possible contamination between the study arms, using a specifically developed form based on YACS II items and the main objectives of the SPADe group intervention with low score indicating no presence of specific CBT content and no contamination.
Sample Size
No formal calculation is required for a feasibility study since measuring effectiveness is not a key objective. The sample size of 50 was considered appropriate for assessing key objectives related to recruitment, retention, randomisation, data capture, performance of outcome measures and acceptability.
Randomization And Masking
Once consented, participants were randomised using a third-party web-based randomisation system which ensured concealed allocation. Participants were stratified according to number of previous detoxifications (> 2 vs ≤ 2) and site. Randomisation had a random block size (2–4). Research Assistants (RAs) were blind to the randomisation of the opposite study site. It was planned to for them to conduct the follow-up interviews of the opposite site. This initial plan for cross site follow up was abandoned early during the 3 months follow up period as it was proven administratively challenging and was considered a barrier for successful follow up efforts.
Statistical And Health Economic Analysis
The main analysis was planned to be based on the intention-to-treat principle considering all randomised clients according to the arm they were allocated.
The feasibility outcomes were summarised using descriptive statistics where appropriate. The potential primary and secondary outcomes were summarised by arm, as well as completion rates estimated for each outcome measure. Duration of continuous abstinence, as measured from randomisation, and the time to relapse, as measured from the end of the detoxification wereanalysed using a Kaplan-Meier curve and log-rank test. The percentage of Days of Abstinence (PDA) were analysed using a regression model. Formal hypothesis tests and confidence intervals were conducted although due to the nature of this trial they were treated cautiously and the main focus was on the completion rate of outcomes, and the estimation of parameters required for a sample size calculation for the main trial.
Economic evaluation in the feasibility study tested the collection of data on costs and outcomes that would be required in any future definitive trial assessing the cost effectiveness of the SPADe intervention. Clinic contacts with key workers and doctors for delivering the intervention and usual care were obtained retrospectively from clinic records by research assistants. To assess potential service use offsets, data on other service utilisation by participants was gathered by self report using the Client Service Receipt Inventory. The CSRI was administered to participants by researchers in the clinic at 3, 6 and 12 months. The questionnaire covers all forms of health and social care, and includes contacts with the police and justice services [30].
Qualitative Interviews
A purposive sample of participants across both trial groups and sites (approximately 20) were planned to be interviewed at three months to establish experiences of randomisation, recruitment and initial trial procedures (wave 1) and follow up interviews were planned to take place at 9 months (completion of the study, wave 2) to give specific feedback on retention issues and treatment conditions. Due to difficulties interviewing participants on two time points, interviews were conducted at any time between 6 and 9 months since recruitment into the study. A selected sample of staff involved with the intervention (both directly and in-directly) were interviewed on study completion.