A structured, telephone-delivered intervention to reduce methamphetamine use: study protocol for a parallel-group randomised controlled trial

Research suggests that even baseline questioning and brief health education/advice can yield positive short-term changes in AOD use, by prompting reflection, self-regulation of behaviour and treatment-seeking [59‐61]. In addiction treatment research, control groups have comprised no treatment, treatment as usual (e.g. case management) or “minimal input” comparators. In this trial, the information booklet for reducing MA use and related harms, and accessing further support [58], as well as four to six ≤5-min approximately weekly telephone check-ins to control for frequency of contact across treatment arms, is considered to be a minimal input control condition. While this condition may impact positively on participants’ MA use, we expect a more modest benefit relative to the R2C-M intervention condition.

See the timeline in Table 3.

Researcher 2, blind to treatment allocation, will conduct 6 week and 3, 6  and 12 month follow-up assessments by telephone, with a short message service (SMS) sent just prior to each call. Eligibility, baseline and follow-up assessment calls will take approximately 45–60 min. At least five contact attempts will be made per follow-up time point. Participants who cannot be contacted after five contact attempts will be deemed to have missing data for that time point. The research team will attempt contact again at the next follow-up, unless the participant withdraws from the study.

Mixed-methods program evaluation will use applicable elements of the Reach, Effectiveness, Adoption, Implementation and Maintenance (RE-AIM) [62] and Consolidated Framework for Implementation Research (CFIR) [63] implementation research frameworks. A subset of participants randomised to the intervention arm will participate in semi-structured telephone-delivered interviews to understand their experiences of the program, and its implementation. Ready2Change counsellors, program and trial managers and researchers will be interviewed to seek their feedback on program implementation. Program evaluation will also be informed by trial administrative data collected during trial implementation, for example: (i) previous treatment and barriers to help-seeking of people who respond to study advertising; (ii) rate of people randomised to (a) participate in the trial from those who respond to study advertising, (b) participate in an information call and (c) participate in eligibility call; (iii) reasons for ineligibility; (iv) reasons for non-participation and lost to follow-up when available; (v) participant characteristics including health inequity factors.

The study will use retention enhancement strategies utilised in the previous trial of the R2C program for alcohol problems [30], and as suggested in pre-trial focus groups and individual interviews with AOD service clinicians and consumers. Retention strategies include seeking verbal commitment to participate in the program, text message reminders, flexible call schedules and varying reimbursements corresponding to the importance of the data collection time-point.

Participants will be reimbursed with supermarket vouchers as follows: AUD$20 for baseline assessment and AUD$30 per follow-up assessment. Additional reimbursement for follow-up assessments include AUD$10 for completing both 6 week and 3 month follow-ups, AUD$5 for completing 12 month follow-up or AUD$20 for completing all four follow-ups. For participation in additional study tasks, participants will be reimbursed as follows: AUD$20 for completing the CIS cognitive assessment task and AUD$10 for participating in a program feedback interview.

The risks of harm to participants in this study are anticipated to be minor and no compensation for harm is deemed necessary. Although receiving treatment for substance use problems is an exclusion criterion for this study, participants are not restricted from seeking other treatment for MA or other substance use problems after they begin the trial. Throughout the study, participants needing further support will be referred to additional services. The research team will monitor the number of participants who receive or are referred to further AOD services or escalated to the study PI for clinical review.

The trial’s Chief Investigators will perform the function of a Trial Management Committee (TMG), as they have the expertise necessary to oversee all aspects of the conduct of the trial (i.e. capacity to monitor compliance with the protocol, monitor compliance with ethical and clinical governance, provide standardised training and other means of quality control, oversee trial arm fidelity, monitor adverse events and provide leadership to the research team). Data audits will be conducted at periodic intervals (i.e. every 6 months), led by the trial manager. Regular liaison between the principal investigator, study clinician and research team will occur to permit discussion of day-to-day trial progress, participant eligibility and any potential concerns. Any protocol amendments decided by the investigators (e.g. changes to eligibility criteria, outcomes, analyses) will be communicated to the ethics committees and clinical trials register.

R2C-M counsellors will undergo training focusing on competence and adherence to the R2C-M intervention and research procedures. All sessions will be recorded and an independent researcher will rate the fidelity (i.e. adherence and competence) of an intervention session for 20% of the active sample using random start systematic sampling. Adherence to intervention elements will also be monitored by R2C-M counsellors via a standard checklist. Researcher 1 (who conducts control condition check-in calls) will be trained to provide information on further AOD support and use a script to ask about participants’ use of the information booklet (i.e. to ensure that participants in the control condition do not inadvertently receive individualised support). Supervision of Researcher 1 will occur to prevent “drift” (e.g. call duration records will be checked intermittently).

It is recognised that adverse effects can arise from the delivery of psychological interventions in clinical trials [65]. In this trial, adverse events (AEs) and serious adverse events (SAEs) that could be related to methamphetamine use or psychological intervention will be systematically collected during the intervention period and at 6 weeks post-randomisation, for both R2C-M and control conditions. Participants will be asked if they experienced any negative effects following their previous telephone support call and whether they have been hospitalised since their last call. Examples of “negative effects” will be provided if requested and prompts include significantly increased distress, significantly increased MA cravings or the issue for which the participant was looking for help got a lot worse. Unexpected harms will be collected as notes on the participants REDCap file if volunteered by the participant. Participants will also be encouraged to contact the research team if they are concerned about an adverse event. Any AEs or SAEs occurring during the course of this study, whether or not they are deemed to be related to participation in this study, will be followed rigorously, and in conjunction with the participant’s general practitioner as appropriate (general practitioner or practice contact details will be collected at baseline, if the participant has a current general practitioner or practice). All adverse events will be reported to the approving ethics committee and included in subsequent peer-review publications.

If a participant is assessed as being at risk of suicide, referral to appropriate support is immediate. Researchers will undertake suicide intervention skills training, to ensure they are equipped with the skills to respond to suicide risk, particularly when contingencies may be required (e.g. assessing urgency—evaluating the need to keep the participant on the phone; managing the call when risk is immediate). R2C-M counsellors are well-trained and experienced in the management of suicide risk and will follow established clinical risk assessment and management guidelines.

Participants’ right to withdraw from the trial without consequence will be outlined during the consenting process and in the Participant Information Sheet. Participants can withdraw their consent verbally or in written form (i.e. email or text message correspondence), with the option to remove all previously collected data or just remove consent for further data collection. No further contact with the participant will be initiated by the research team upon their withdrawal from the study. In instances where it has been identified that a participant meets exclusion criteria during the study (e.g. active suicidality), and/or that it is not in the best interests of the participant to remain in the study, the Principal Investigator or Study Clinician will decide whether to withdraw the participant from the trial. The reasoning for this will be explained to the participant and they will be offered information on accessing other support. No further data collection will occur, with the exception of the details regarding adverse events.

We aim to randomise between 188 and 204 participants to this study (i.e. total N = 196 ± 8, subjects per study arm = 98 ± 4). This was calculated using the Genstat [66] power procedure. The primary outcome measure (DUDIT score 3 months post-baseline) can range from 0 to 40 and will be analysed via a linear mixed model. Using data from our pilot work [20], we found that the between-subject variance component in DUDIT score was 21, the within-subject variance component was 63 and the estimated improvement (decline) in DUDIT score was 17 (SE = 1.6). We estimate that by 3 months there will be an improvement of at least 16 in the R2C-M arm and that the control arm could exhibit an improvement of up to 10. With 75 evaluable subjects in each treatment arm, the study will have 90% power to detect this difference in improvement using an F-test conducted at the 5% significance level. If these conjectured improvements by 3 months are not durable and, for example, deteriorate by 50% at 6 months, and return, on average, to baseline values by 12 months, then this treatment-by-time interaction scenario will be detected with at least 90% power. The initial target sample size of 188 comprises 75 per arm, inflated to 94 per arm to allow for approximately 20% drop-out, which is based on the attrition rates reported in other treatment [67] and helpline [68] research with AOD cohorts using 12 month endpoints. Provision has been made to increase the recruitment target, based on the 3 and 6 month attrition rates observed after recruitment and data collection commenced (after 12 months of data collection with this complex cohort, 3 and 6 month attrition rates were higher than anticipated). As such, the target sample size was increased from 188 subjects, to up to 204 participants (i.e. an additional 8 subjects based on the 3 month attrition rate; or an additional 16 subjects based on the 6 month attrition rate). Based on our experience with social media advertising for alcohol and other drug treatment trials, it is estimated that 10–12 participants will be recruited per month over 18 months.

Data will be collated, cleaned and validated using programed edit checks, in a database that will be locked prior to the unblinding of the statistician for the primary analysis. The primary analysis will take place after all subjects, not known to have withdrawn or not deemed lost to follow-up, have had their 12 month assessments and will be based on the intention-to-treat principle (i.e. subjects’ data are analysed as randomised and as stratified). A “per-protocol” sensitivity analysis will be restricted to those subjects with at least one follow-up assessment and, for subjects randomised to the R2C-M arm, participation in at least two telephone counselling sessions. Previous research delivering the R2C program to people with alcohol use disorder found exposure of ≥2 sessions yielded a reduction in alcohol use severity compared to a control arm [31]. The first R2C session focuses on a clinical assessment and identifying treatment goals and the second session is when a therapeutic dose is received. As such, exposure to ≥2 sessions is considered “as-treated” for the per-protocol analysis. Additional sensitivity analyses will include a covariate for the number of structured telephone counselling sessions [1 to 6] in which subjects, in the R2C arm, participated. The repeated measurements of the outcome variables will be analysed by fitting linear mixed models using restricted maximum likelihood (REML)—this will allow the most suitable variance-covariance model for the repeated measures to be selected, using Akaike’s Information Criterion, and commonality of nonlinear trends over time to be explored via splines. The F-test will be used to test for an overall group by time interaction and the primary comparison, between groups, of their changes from baseline to 3 month follow-up will be based on a t-test of the corresponding interaction contrast—this t-test will utilise the predicted means and their variance-covariance matrix which are recovered from the fitted mixed model. Diagnostic plots of residuals will be assessed and, if deemed necessary, variance-stabilising transformations such as the empirical logistic transformation will be applied to the outcome variables, and inferences will be based on the analyses conducted on the transformed scale. In a series of exploratory analyses, mixed models with covariates for gender, illicit drug use, extent of exposure to the intervention, differences due to assigned counsellor, exposure to other treatments/programs and baseline levels of MA use, psychological distress, depression, anxiety and stress will be fitted, including their interactions with treatment group, in order to identify moderating factors. The complete list of candidate covariates and details of the analyses will be specified in a Statistical Analysis Plan that will be reviewed and approved by a Study Management Committee prior to database lock. No interim analyses will be conducted and there are no plans to halt data collection before completion. Analyses will be conducted using the most appropriate procedures in GenStat, R and Stata.

The economic evaluation will assess the mean incremental costs and mean incremental benefits of treatment of R2C-M compared to control. Benefits will be measured as quality-adjusted life years (QALYs). Incremental QALYs will be measured by the between-group difference in mean EQ-5D-5L+ score over 12 months. A health system perspective on costs will be taken and will include resource use incurred in the delivery of telephone intervention as well as health services irrespective of payment source. Health care costs will be calculated from the utilisation data and average unit costs for each item. Running costs will be included, but not the costs of training in the primary analysis. In a supplementary analysis, we will model the potential cost-effectiveness using a broader societal perspective and include estimates of the cost of work-related losses using the WHO HPQ28-day, crime and interpersonal related harms associated with MA use from literature sources.

Cost-effectiveness analysis results will be presented as the mean net benefits of treatment across a range of hypothetical money values of QALYs, with 95% CIs and a one-sided p-value calculated using non-parametric bootstrapping. Net benefit estimates will be based on the between-group difference in the means cost and outcome over the 12 months estimated using separate regression analyses controlling for baseline values and the stratification variable. A generalised linear regression model, with an appropriate choice of distribution, will be used to account for any skewness in the cost data. Multiple imputation will be used to address the uncertainty of the estimates due to missing observations. A secondary analysis will estimate a per-protocol cost-effectiveness of the intervention adjusted for non-adherence. Instrumental variable estimation will be used with the randomisation group as the instrument for adherence, defined as at least two telephone counselling sessions [69].

Dissemination of findings to the research community will be via peer-reviewed publications and conference presentations. Chief Investigators will meet near the end of the trial to finalise and implement the research dissemination plan and authorship. There are no plans to engage paid professional writers outside the study team. Participants will be informed they can access the Turning Point website for a summary report of the results at the trial’s end.