Effect of a web-based relapse prevention program on abstinence among Japanese drug users: A pilot randomized controlled trial

Journal Pre-proof Effect of a web-based relapse prevention program on abstinence among Japanese drug users: A pilot randomized controlled trial

Ayumi Takano, Yuki Miyamoto, Tomohiro Shinozaki, Toshihiko Matsumoto, Norito Kawakami PII:

S0740-5472(19)30291-0

DOI:

https://doi.org/10.1016/j.jsat.2019.12.001

Reference:

SAT 7952

To appear in:

Journal of Substance Abuse Treatment

Received date:

21 May 2019

Revised date:

8 October 2019

Accepted date:

3 December 2019

Please cite this article as: A. Takano, Y. Miyamoto, T. Shinozaki, et al., Effect of a webbased relapse prevention program on abstinence among Japanese drug users: A pilot randomized controlled trial, Journal of Substance Abuse Treatment(2019), https://doi.org/ 10.1016/j.jsat.2019.12.001

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier.

Journal Pre-proof

Effect of a web-based relapse prevention program on abstinence among Japanese drug users: a pilot randomized controlled trial Ayumi Takanoa, Yuki Miyamotob, Tomohiro Shinozakic, Toshihiko Matsumotod, Norito Kawakamie Department of Mental Health and Psychiatric Nursing, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8510, Japan

b.

Department of Psychiatric Nursing, The University of Tokyo 7-3-1 Hongo, Bunkyo, Tokyo 113-0033, Japan Department of Information and Computer Technology, Faculty of Engineering, Tokyo University of Science

ro

c.

of

a.

-p

re

e.

lP

d.

6-3-1 Sinjuku, Katsushika-ku, Tokyo 125-8585 Department of Drug Dependence Research, National Center of Neurology and Psychiatry 4-1-1 Ogawa-Higashi, Kodaira, Tokyo 187-8553, Japan Department of Mental Health, The University of Tokyo 7-3-1 Hongo, Bunkyo, Tokyo 113-0033, Japan

Jo ur

na

Corresponding author Ayumi Takano Department of Mental Health and Psychiatric Nursing, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8510, Japan TEL: +81 3 5803 5348 [email protected]

１

Journal Pre-proof

Abstract Background: Internet-based intervention could help drug users recover from drug dependence. This study evaluated the effectiveness of a newly developed web-based relapse prevention program (e-SMARPP) for people with a drug problem, including the use of

of

methamphetamine, in Japan.

ro

Methods: The study was a pilot randomized controlled trial comprised of 48 psychiatric

-p

outpatients diagnosed with drug use disorder. The participants were randomly assigned to an

re

eight-week, six-session web-based relapse prevention program (an intervention group) or only

lP

web-based self-monitoring (a control group). The primary outcome was the duration of

na

abstinence from a primary drug during the intervention and relapse risk. Secondary outcomes included motivation to change, self-efficacy, and money spent on drugs. The outcomes, except

Jo ur

for the duration of abstinence during the intervention, were assessed at baseline, 2-, 5-, and 8-months. Program completion rate was also assessed. Results: No significant difference was observed between the intervention and the control groups for the primary and the secondary outcomes. The effect size of the duration of abstinence during the intervention was d = 0.42, which was comparable to previous studies. In the intervention group, about 26% did not complete the entire intervention. Conclusions: e-SMARPP failed to demonstrate efficacy, however, is potentially helpful for

2

Journal Pre-proof

enhancing abstinence. The low attrition rate may suggest the acceptance and feasibility of the program. Further improvement of the program and evaluation in a full-scale trial are needed.

Keywords

of

Web-based intervention; randomized controlled trial; relapse prevention; self-monitoring;

Jo ur

na

lP

re

-p

ro

pilot study; methamphetamine

3

Journal Pre-proof

1.

Introduction Drug use problems are a serious public health concern and drug use is a global burden,

accounting for 1.3% of all disability adjusted life years (Degenhardt et al., 2018). In Asia, the proportion of treatment for the use of amphetamine-type stimulants is higher than in other

of

regions (United Nations Office on Drug and Crime, 2018). In Japan, the most prevalent drug

ro

in the treatment population is methamphetamine, estimated at about 40% of patients who

-p

received any treatment in psychiatry with dependence or related disorders (Matsumoto, 2014).

re

For stimulant users, psychosocial treatment could be a valid and effective approach because

lP

as of yet there is no evidence for the efficacy of pharmacological treatment (Minozzi et al.,

na

2016; National Institute on Drug Abuse, 2016). Despite evidence supporting psychosocial treatment for substance use disorders, there is still a treatment gap (United Nations Office on

Jo ur

Drug and Crime, 2018). Various reasons have been considered as barriers to treatment access: limited availability, concerns about confidentiality and stigmatization, and human-resource limitations for treatment providers (Rooke et al., 2010; Rooke et al., 2013; Sholomskas et al., 2005; Weissman et al., 2006). Above all, stigma toward drug users has been identified as an important barrier to seeking treatment, reducing substance use, and improving mental and physical health (Ahern et al., 2007; Calsyn et al., 2004; Cumminget al., 2016). The national drug policy in some Asian countries is likely to be very strict regarding illicit drug use,

4

Journal Pre-proof

resulting in a strong stigma that is an obstacle to seeking treatment. Therapeutic interventions using information and communication technologies (ICT) have been developed and adapted to illicit drug use and addictive behaviors to address challenges in treatment implementation (Boumparis et al., 2017; Chebli et al., 2016). Various

of

computer-assisted, web-, or mobile-based interventions for drug users developed based on

ro

psychosocial approaches have demonstrated benefit for abstinence, treatment retention and

-p

cost effectiveness with small to moderate effect sizes (Boumparis et al., 2017; Moore et al.,

re

2011; Portnoy et al., 2008; Takano et al., 2015). These interventions were designed to use a

lP

psychosocial approach, such as cognitive behavioral therapy (CBT), community

na

reinforcement approach, and motivational interviewing, as well as for use in face-to-face interventions (Boumparis et al., 2017). In previous studies, the primary outcome was typically

Jo ur

drug use and/or abstinence assessed by self-report, urine test and/or hair analysis (Boumparis et al., 2017; Takano et al., 2015). Treatment retention, relationship with therapists, engagement in the treatment, adverse events, and feasibility have also been assessed as secondary outcomes (Nesvåg and Mckay, 2018; Takano et al., 2015). In the field of substance abuse, psychological factors such as self-efficacy and readiness to change were considered positive predictors of behavioral change (Krebs, Norcross, Nicholson, & Prochaska, 2018; Sliedrecht, de Waart, Witkiewitz, & Roozen, 2019) and some studies used these variables as

5

Journal Pre-proof

treatment outcomes. Additionally, behavioral assessment such as the amount of money lost has been evaluated as a treatment outcome in gambling addiction studies (Goslar, Leibetseder, Muench, Hofmann, & Laireiter, 2019; Peter et al., 2019). Most interventions have been developed in Western countries for specific drugs, such as cocaine, cannabis or opioid users

of

(Carroll et al., 2008, 2009; Kay-Lambkin et al. , 2009; 2011; Ondersma et al., 2007; Rooke et

ro

al., 2013). Along with increasing methamphetamine problems, internet-based programs which

-p

target methamphetamine users have been developed (Reback, Rünger, Fletcher, &

re

Swendeman, 2018; Tait et al., 2012, 2014; Takano et al., 2015). Some internet-based programs

lP

target various types of drug users including methamphetamine users (Boumparis et al., 2017).

na

A meta-analysis reported that internet-based treatment significantly decreased opioid use, however, the effect of such treatment for stimulant users was small and non-significant

Jo ur

(Boumparis et al., 2017). To our knowledge, no empirical study using an internet-based program for Asian drug users, which socioeconomically and culturally differ from western countries, has been conducted.

We developed a flexible and accessible web-based treatment program named “e-learning Serigaya Methamphetamine Relapse Prevention Program (e-SMARPP)” for Japanese drug users (Takano et al., 2016a). The development was based upon an existing face-to-face relapse prevention program, the Serigaya Methamphetamine Relapse Prevention

6

Journal Pre-proof

Program (SMARPP), which was developed based on the Matrix Model for outpatients using stimulants (Rawson et al., 1995). The content of SMARPP was originally based on the Matrix Model, but it was modified to adapt to many types of drugs by using the major modules of the Matrix Model (e.g., function analysis, identification of triggers). SMARPP was implemented

of

in different settings in Japan and its efficacy on drug abstinence, participation in a self-help

ro

group, and motivation to change has been reported (Tanibuchi et al., 2016; Kondo et al., 2014;

-p

Matsumoto et al., 2011). We developed the e-SMARPP content using major and important

re

modules of SMARPP for every drug user regardless of drug type. The e-SMARPP website is

lP

designed to support any device, including personal computers, mobile phones and tablet

na

computers with Internet access. In the feasibility study, the usability and acceptance of the e-SMARPP prototype was assessed using the Web Usability Scale and original questionnaires

Jo ur

(Takano et al. 2016a). The scores of the scale were good and most of the participants felt the program was easy to use and helpful (Takano et al. 2016a). The authors finalized program development after improving the content based on suggestions gathered through the feasibility study (Takano et al., 2016b). The development process and the results of the pilot study have been previously described (Takano et al., 2016a; Takano et al., 2016b). The aim of this study was to obtain a preliminary result regarding the efficacy of the web-based cognitive behavioral relapse prevention program “e-SMARPP” among Japanese

7

Journal Pre-proof

outpatients with drug dependence with a randomized controlled trial (RCT) at 8-month follow-up. The primary hypothesis was that participants assigned to e-SMARPP would maintain a longer duration of consecutive days of abstinence from a primary drug during the intervention and have a reduced relapse risk compared to those who were randomized to only

of

web-based self-monitoring. The secondary hypothesis was that participants in the e-SMARPP

ro

group would report positive changes in motivation to change, self-efficacy, and money spent

lP

Methods

2.1. Trial design

na

2.

re

-p

on drugs.

This study was a two-arm and parallel-group RCT. This study protocol was registered in University

Hospital

Medical

Jo ur

the

Information

Network

clinical

trial

registry

(UMIN000016075). Details of the study protocol were described elsewhere (Takano et al., 2016c). The RCT protocol was changed twice after study commencement. First, the inclusion criteria were extended from those who used a primary abused drug in the past month to those who used a primary abused drug in the past year because it was difficult to recruit outpatients who currently used drugs. Most patients had already quit using drugs for several weeks or months before their first visit. Second, two trial sites were added to recruit more patients. The

8

Journal Pre-proof

Ethics Committee of the University of Tokyo and each trial site approved this study protocol. 2.2. Participants and setting The participants were recruited at six psychiatric hospitals from January 2015 to March 2016 (National Center of Neurology and Psychiatry, Saitama Psychiatric Medical Center,

of

Kanagawa Psychiatric Medical Center, Okayama Psychiatric Medical Center, Tokyo

ro

Metropolitan Matsuzawa Hospital and APARI clinic). The hospitals provide specialized

-p

treatment for people with substance use disorders. Posters and flyers with a brief outline of

re

the study (e.g., introduction of e-SMARPP, participant eligibility criteria, duration, and study

lP

tasks and activities) were distributed by the hospital staff. The inclusion criteria were: 1)

na

outpatients who were diagnosed with substance use disorder assessed by DSM-IV or 5 (psychoactive substances other than alcohol and tobacco) by a psychiatrist, 2) those who used

Jo ur

a primary abused drug in the past year, and 3) those with access to the Internet via PC, smartphone or tablet computer and could exchange e-mail. Patients with the following conditions were judged ineligible to participate in the study by a psychiatrist and were excluded: 1) patients with severe physical diseases, 2) patients with high suicide risk, 3) patients with severe symptoms of a substance-induced psychotic disorder, and 4) patients with impaired cognitive function. 2.3. Sample size

9

Journal Pre-proof

A minimal sample size was calculated for the two primary outcome variables (the duration of abstinence and relapse risk) to detect a medium effect size of d = 0.4 based on previous studies for drug users. As for the duration of abstinence, the effect size between the intervention group and control group after the intervention was reported as d = 0.45 in a study

of

conducted for computer-assisted cognitive behavioral therapy (Carroll et al., 2008). For

ro

relapse risk, the effect size between pre and post intervention was d = 0.39 in a study

-p

conducting a relapse prevention program in Japan (Morita, 2013). We estimated a sample size

re

of 100 per group (total 200), assuming α = 0.05 and a power (1 - β) = 0.8.

lP

2.4. Randomization and blinding

na

Outpatients were recruited by distributing flyers and posters. After baseline assessment, eligible participants were randomly assigned to either the intervention group (e-SMARPP

Jo ur

group) or the control group (self-monitoring group), using the method of permuted block with a random block size of four. Randomization was stratified by the trial site. The computer-generated allocation list was made by an independent researcher (YM) and concealed to other researchers and participants until the time of assignment. Enrollment was done by a researcher (AT) and the intervention started immediately. The researcher (AT) managed study progress and sent e-mail reminders to participants who did not respond to the follow-up assessments (a maximum of two times in the subsequent week). Psychiatrists and

10

Journal Pre-proof

health professionals who worked for the trial sites were blinded. An independent researcher (YM) downloaded data from the e-SMARPP database and an independent research assistant masked the group variable before analysis, then researchers (AT, TS) analyzed the final data. 2.5. Intervention

of

2.5.1. Web-based relapse prevention program: e-SMARPP

ro

The e-SMARPP website included five modules: 1) a six-session relapse prevention

-p

course, 2) self-monitoring, 3) information (downloadable PDF information and website links

re

to drug addiction support services), 4) user guide (how to use the system, frequently asked

lP

questions, and contact form); and 5) survey (the baseline and the follow-up assessments). The

na

main intervention modules were the relapse prevention session and self-monitoring. Videos and homework were provided in each session (Supplementary Table 1). Users submitted their

Jo ur

homework through an Internet text form after watching the videos. The content of the relapse prevention session was versatile and can be used for various drug problems. The content was intended to be user-friendly with minimal text and limited use of difficult Kanji characters referencing specialized medical terminology. Personalized feedback comments from a trained web-therapist who was a registered nurse with skills specific to the relapse prevention program were provided through the e-SMARPP website after the homework was submitted. The web-therapist was like a facilitator of the relapse prevention program. The feedback

11

Journal Pre-proof

messages were empathic and gave supportive advice that utilized motivational interviewing techniques. Self-monitoring was done in a calendar format like the Timeline Followback (TLFB) method (Fals-Stewart et al., 2000; Norberg et al., 2012; Sobell et al., 1979; 1996). During the intervention, the participants were expected to record daily drug use by the weekly

of

deadline. This self-monitoring was also used to retrospectively record substance use at the

ro

follow-up surveys. Additionally, e-SMARPP had some automated functions, including

-p

tracking progress for users and a notification email function for the users and the

re

web-therapists.

lP

---Supplementary Table 1---

na

2.5.2. Intervention group: e-SMARPP group

Participants who were assigned to the e-SMARPP group were provided access to all

Jo ur

e-SMARPP content. They were expected to complete each relapse prevention session over a week period in a sequential order by each deadline. For an 8-week intervention period, they were expected to complete a total of six sessions, but they had a 2-week grace period and were allowed to progress at their own pace. If they did not complete a session, the session was carried over to the next week. After the 8-week study period, the program was closed and thereafter the participant could not use the program. Participants were expected to record their daily drug use situation on the web-based self-monitoring calendar by each deadline. If they

12

Journal Pre-proof

did not go through an expected session and/or self-monitoring by each deadline, the web-therapist sent an e-mail reminder (a maximum of two times in the subsequent week). Participants continued to receive outpatient treatment as usual, such as medication and face-to-face group/individual psychosocial treatment. Even if participants stopped receiving

of

outpatient treatment or changed hospitals, the web-based intervention continued.

ro

2.5.3. Control group: self-monitoring group

-p

Participants who were assigned to the self-monitoring group were provided access to a

re

limited set of e-SMARPP content, including the web-based self-monitoring and information

lP

content. The self-monitoring group participants did not have access to the relapse prevention

na

program. Similar to the e-SMARPP group, they were expected to record their daily drug use situation for eight weeks. E-mail reminders and outpatient treatment were continued in a

Jo ur

similar way of the intervention group. After the study period, the relapse prevention sessions were provided if requested. 2.6. Measures

2.6.1. Primary outcomes One of the primary outcomes was the longest duration of consecutive days of abstinence from the primary drug during an eight-week intervention (56 days) which counted using the self-monitoring calendar, according to previous studies (Carroll et al., 2008, 2014)

13

Journal Pre-proof

(Supplementary table 2). Another primary outcome was relapse risk assessed using the Stimulant Relapse Risk Scale (SRRS) at follow-up assessments at 2, 5 and 8 months after the randomization (Ogai et al., 2007). The total score ranges from 30 to 90. Higher scores indicate a higher relapse risk.

of

---Supplementary table 2---

ro

2.6.2. Secondary outcomes

-p

We assessed psychological factors related to behavioral change and money spent on

re

drugs at the 2-, 5- and 8-month follow-up assessments. Motivation to change was measured

lP

with the Stage of Change Readiness and Treatment Eagerness Scale-8 version for Drug Use

na

(SOCRATES) (Miller and Tonigan, 1996; Kobayashi et al., 2010). The total score ranges from 19 to 95. Higher scores indicate a higher motivation to change. Self-efficacy in handling drug

Jo ur

craving was measured with the Self-efficacy Scale for Drug Dependence (SSDD) (Morita et al., 2007). The total score ranged from 16 to 102. A higher score means higher self-efficacy in handling a drug craving. The total money spent on drugs (yen) in the past month was also assessed. The money spent on drugs depended on type of drug, amount of drug, and the relationship between a drug user and a dealer. In Japan, methamphetamine is more expensive than other drugs. An outlier was defined as the cost variable being over 100,000 yen, even though the participant did not use a drug in the past month.

14

Journal Pre-proof

2.6.3. Other variables We collected data on participant characteristics and information related to drug use at the baseline (presented in Table 1). We assessed the severity of drug use problems using the Japanese version of the Drug Abuse Screening Test (DAST-20) (Skinner, 1982; Shimane et al.,

of

2015). Total score ranges from 0 to 20 and a high score represents a severe condition. The

ro

cutoff scores for low (1-5), intermediate (6-10), substantial (11-15), severe (more than 16)

-p

were used based on the recommendations in previous studies (Cocco and Carey, 1998; Gavin

re

et al., 1989; Yudko et al., 2007).

lP

The intervention completion rate of each group was assessed for program evaluation.

na

We defined dropouts as those who did not complete the entire intervention. In the e-SMARPP group, the participants’ characteristics at the baseline were compared between those who

Jo ur

completed the intervention and the dropouts. Adverse effects, hospitalization, arrest, and death during the intervention, were also assessed. Additionally, the participants were asked about harmful effects such as drug craving while using e-SMARPP after the intervention. 2.7. Statistical analysis Abstinent days from the primary drug were compared between the intervention and control groups using t-test when there was complete information about drug use/abstinence during the intervention. The effect size (Cohen’s d) was calculated. Values of 0.2, 0.5 and 0.8

15

Journal Pre-proof

were considered as small, medium and large effect, respectively (Cohen, 1992). Analysis for the repeated measures was on an intention-to-treat basis for all available data, using mixed-effect models. All participants at the baseline were included in the primary analysis. We fitted two linear

mixed-effect

models

for each outcome:

random-intercept

of

“analysis-of-variance” models that treated an assessment time-point as a categorical (nominal)

ro

covariate and random-intercept and random-slope “growth-curve” models that treated

-p

assessment time-point as continuous covariates (coded as 0, 2, 5, 8 [months]). Both models

re

included group, time-point, and their product term as covariates and assumed first-order

lP

autoregressive correlation structure for outcome-residuals within the same individual. If

na

convergence was not achieved for random-intercept and random-slope models, these random effects were forced to be independent, or random-slope was deleted. To help in interpretation,

Jo ur

the effect sizes were calculated at each assessment point by using estimated means based on the random-intercept analysis-of-variance models. At each follow-up assessment, the effect size was obtained by dividing the model-based estimated mean-difference by the pooled standard deviation (SD) calculated from available raw data. The intervention completion rate by the intervention group was described by calculating the progress of each session and weekly self-monitoring. The baseline characteristics between those who completed the intervention and the dropouts were compared using t-test, chi-squared test, or Fisher's exact

16

Journal Pre-proof

test. Analyses were conducted with a level of 5% in the two-sided test, using SPSS Statistics Ver. 23.

3.

Results

of

3.1. Participants

ro

Figure 1 is the participant flow diagram. In total, 48 outpatients were recruited and

-p

randomly assigned into either the e-SMARPP group (n = 23) or the self-monitoring group (n

re

= 25). All participants received the intervention, however, six (26.1%) in the e-SMARPP

lP

group did not complete the intervention. In the e-SMARPP group, the number of participants

group.

na

who did not respond to the follow-up assessments was more than in the self-monitoring

Jo ur

Table 1 shows characteristics of the participants by group. The majority were smartphone users and used the Internet every day. About half of the participants mainly used methamphetamine and 18.7% had a severe drug dependence. About half of them had received outpatient treatment for more than three years and a face-to-face relapse prevention program in the past. There was no significant difference in the characteristics between the groups. ---Figure 1-----Table 1---

17

Journal Pre-proof

3.2. Efficacy of e-SMARPP The longest period of abstinence days from the primary drug during the intervention was 48.8 (SD = 14.7) in the e-SMARPP group and 41.2 (SD = 20.3) in the self-monitoring group (Table 2). The e-SMARPP group maintained a longer abstinence duration than the

of

self-monitoring group during the follow-up with a moderate effect size (d = 0.42), while there

ro

was no significant difference between the groups (t = 1.45, p = 0.16).

-p

Table 2 shows the raw scores of relapse risk, motivation to change, self-efficacy, and

re

money spent on drugs at baseline, 2-, 5-, and 8-month follow-up assessment by the group.

lP

Table 3 shows the estimated efficacy of the e-SMARPP on the outcomes based on the mixed

na

model analyses (i.e., estimates of product terms of the group and assessment time). For the relapse risk, the SRRS scores had no significant difference for the interaction of group and

Jo ur

time (t = -0.19, p = 0.85 from the growth-curve model). At 2-, 5-, and 8-month follow-up assessments, the effect size were very small. Although the effects size of money spent on drugs at the 5-month was medium (d = -0.54), the effect disappeared at other follow-up assessments. ---Table 2-----Table 3--3.3. Program evaluation

18

Journal Pre-proof

As shown in Table 4, about 70% of the participants in the e-SMARPP group completed all relapse prevention sessions. All the participants completed at least two sessions. The completion rate of the self-monitoring was over 80% in both groups. Table 5 shows differences for the baseline characteristics between the intervention completers and dropouts

of

in the e-SMARPP group. The dropouts were significantly likely to use drugs at an earlier age,

ro

had a more severe drug dependence and relapse risk, and had lower self-efficacy.

-p

There were no adverse effects such as hospitalization, arrest and death during the

re

intervention. However, four participants (9.3%) reported that they felt drug craving and

4.

Discussion

na

---Table 5---

Jo ur

---Table 4---

lP

negative feelings when they used the e-SMARPP program.

We evaluated the efficacy of e-SMARPP with an RCT design at 8-month follow-up. To the best of our knowledge, this is the first RCT assessing the efficacy of a web-based treatment for Asian drug users that includes methamphetamine users. No significant difference was observed between the e-SMARPP group and the self-monitoring groups. Also, other outcomes including relapse risk, motivation to change, self-efficacy, and money spent on drugs were not significantly improved in the e-SMARPP group compared to the

19

Journal Pre-proof

self-monitoring group. Although we tried to recruit as many patients as possible, only about 25% of the expected participants were finally involved in the study. The participants’ characteristics in this study were almost the same as those of patients with drug dependence who received treatment in

of

Japan (Matsumoto, 2014). This study population may be representative of Japanese patients

ro

with drug dependence. In a comparison with study participants in other countries, the

-p

participants in this study were likely to receive outpatient treatment for several years and

re

maintain a long abstinence prior to the intervention. Similar previous RCTs excluded drug

lP

users who were currently receiving any treatment for substance use disorders or were

na

abstinent from drugs in the past month (Carroll et al., 2008, 2014; Kay-Lambkin et al., 2009; Rooke et al., 2013; Tait, 2014). For instance, the average days of primary drug use in the last

Jo ur

28 days at the baseline were 3.9 in this study (Table 1). These drug use days were much less than the drug use days in previous studies in other countries (Carroll et al., 2008, 2014; Rooke et al., 2013; Tait, 2014). In Japan, it was difficult to recruit outpatients with active drug use using exclusion criteria like previous studies. These differences might cause a ceiling effect and lead to an attenuation of the intervention effect for abstinence, relapse risk, and other psychological outcomes compared to previous studies. Other recruitment strategies may be needed to involve many active drug users. For example, some studies used online recruitment

20

Journal Pre-proof

methods using online portal sites, advertisements on popular search engines or social media, and links on affiliated websites of an academic society or recruitment institute, etc. (Arnaud et al., 2016; Rooke et al., 2013). In a future study, we will try to develop these methods to recruit more potential participants who seek treatment.

of

The efficacy of e-SMARPP on abstinence was not significant. This might be because of

ro

the small sample size. However, the effect size of abstinence during the intervention in the

-p

study (d = 0.42) was moderate and comparable to previous studies (Boumparis et al., 2017;

re

Carroll et al., 2008; Portnoy et al., 2008; Rooke et al., 2013). Further study with a large

lP

sample is needed to evaluate the efficacy of e-SMARPP on abstinence. The control group

na

condition might have affected the results. This study used an active control group. The control group was provided self-monitoring which is one of the effective methods for maximized

Jo ur

self-regulatory capacity and skills (Michie et al., 2012). The control group may have also received some benefit from the self-monitoring program. On the other hand, the condition of the control group in the previous study which assessed the effectiveness of the computer-assisted/web-based program was a non-active control, such as with provision of information (Carroll et al., 2008; Rooke et al., 2013). A previous meta-analysis reported the same results that studies employing active treatment as a comparison group demonstrated mostly an effect close to zero (Rooke et al., 2010). The methods using the active control

21

Journal Pre-proof

group may have attenuated the intervention effect. The efficacy of e-SMARPP might be underestimated when compared to previous studies. A further study with a large sample size and non-active control group will be needed to assess the exact efficacy of the web-based relapse prevention program.

of

All scales used to assess psychological factors related to behavioral change were not

ro

significantly improved. The effect sizes of motivation to change at 2- and 8-months were

-p

close to medium. Meanwhile, the effect sizes of self-efficacy were negative. This suggested

re

that e-SMARPP might have the potential to improve motivation to change, however, the

lP

efficacy was unstable. Motivation to change and self-efficacy among drug users have been

na

reported to be important predictors of abstinence, however, that may have a curvilinear relation with drug use behavior rather than a linear one (Crouch et al., 2015; Kobayashi et al.,

Jo ur

2007; Kondo et al., 2014). Self-efficacy is a more complex predictor. It is unclear whether higher self-efficacy is always better among substance users, and an optimal level of self-efficacy for abstinence is unknown (Kadden and Litt, 2011). Some patients may be overconfident in their ability to abstain (Burling et al., 1989). Moreover, the motivational component, which was included in e-SMARPP, has been considered to increase the participant’s ambivalence toward behavioral change (Portnoy et al., 2008). It is necessary to carefully test the motivation to change and self-efficacy by using detailed subscales or a more

22

Journal Pre-proof

longitudinal study. Regarding reduction of drug cost, money spent on drugs in the past month at the 5-month follow-up tended to be lower in the e-SMARPP group. This was on average a reduction of about 15,000 yen (about 135 USD) in the e-SMARPP group compared to the baseline.

of

Although no previous study has examined the efficacy of a web-based program on drug cost,

ro

this outcome may be useful because drug users can more easily notice a direct benefit of the

-p

program. However, there was some concerns regarding data reliability. The data collection

re

methods should be refined to obtain an accurate drug cost because drug cost varies for each

lP

drug and drug market and it might be difficult to remember the money spent on drugs. We

na

could calculate the total money spent on the primary drug if we revise the self-monitoring module to collect the frequency of drug use per day and the general cost of each drug.

Jo ur

There were no serious adverse effects observed although some participants felt drug cravings while they were going through the relapse prevention session. The program completion rate was better or comparable to previous studies (Carroll et al., 2008, 2014; Rooke et al., 2013; White et al., 2010). There were more dropouts in the e-SMARPP group. An excessive amount of modules may have caused an additional strain on the participants. There are other possible limitations. First, the sample size was small, and the statistical power was limited. Additionally, most participants had already quit using beforehand, which

23

Journal Pre-proof

makes it even more difficult to detect a significant effect between groups. Further research with a larger sample including active drug users is required to identify the efficacy of e-SMARPP. Second, the follow-up term was relatively short. Recovery is a long process, and as such, future studies need a longer follow-up to evaluate long-term effects. Third,

of

generalization of the findings was limited because the participants were recruited from

ro

large-scale psychiatric hospitals in areas with a large population. Although participant

-p

characteristics in this study were almost the same as for Japanese outpatients receiving

re

treatment for drug dependence, the efficacy of e-SMARPP for drug users who do not receive

lP

treatment was not confirmed. Future study should be conducted among different populations

na

in various areas including rural areas where treatment resources are limited. Fourth, the content of e-SMARPP was originally developed using an abstinent-oriented Matrix Model

Jo ur

which is geared toward stimulant users. This means that e-SMARPP may not be suitable for some specific populations such as prescribed benzodiazepine users. When we provide e-SMARPP for various types of drug users, the content should be based on their various needs and there is still room for improvement in terms of content. Also, the target population should be considered when implementing e-SMARPP and further studies. Lastly, the reliability of the collected data was uncertain because all the variables were self-reported. Abstinence and motivation to change might be influenced by social desirability bias and may be an

24

Journal Pre-proof

overestimation.

5.

Conclusions This study assessed the efficacy of a web-based relapse prevention program for

of

outpatients with drug dependence in Japan. The program failed to demonstrate significant

ro

improvements in any outcomes, however, the dropout rate from the intervention was lower

-p

than other internet-based programs. The effect size for the duration of abstinence from the

re

primary drug during the intervention was moderate and comparable to those reported

lP

previously in other countries, however, further study with a large sample is needed to evaluate

na

the efficacy of the program. The e-SMARPP program may be promising as safe and feasible relapse prevention program. We still lack evidence of what kind of internet-based

Jo ur

interventions can best contribute to recovery from drug dependence, especially for methamphetamine users in Asian countries. Further studies are needed to evaluate the efficacy of various interventions and tools in this population.

Acknowledgement We would like to acknowledge Dr. Nobuya Naruse, Dr. Ohoji Kobayashi, Dr. Nozomu Hashimoto, Dr. Takashi Sunami, Dr. Satoshi Sakakibara, Dr. Arisa Kadowaki, Dr. Sachiko

25

Journal Pre-proof

Yamada, and Dr. Asuka Hida, for helping us to recruit patients for this study.

Funding The study was supported by the Pfizer Research Foundation and JSPS KAKENHI Grant

ro

of

Number JP16K20813.

-p

Declaration of interests

na

References

lP

re

None.

Ahern, J., Stuber, J., Galea, S. (2007). Stigma, discrimination and the health of illicit drug

Jo ur

users. Drug Alcohol Depend, 88(2), 188–196. Arnaud, N., Baldus, C., Elgán, T. H., De Paepe, N., Tønnesen, H., Csémy, L., & Thomasius, R. (2016). Effectiveness of a Web-Based Screening and Fully Automated Brief Motivational Intervention for Adolescent Substance Use: A Randomized Controlled Trial. Journal of Medical Internet Research, 18(5), e103. Boumparis, N., Karyotaki, E., Schaub, M. P., Cuijpers, P., Riper, H. (2017). Internet interventions for adult illicit substance users: a meta-analysis. Addiction, 112(9), 1521–

26

Journal Pre-proof

32. Burling, T. A., Reilly, P. M., Moltzen, J. O., Ziff, D. C. (1989). Self-efficacy and relapse among inpatient drug and alcohol abusers: a predictor of outcome. J Stud Alcohol, 50(4), 354–60.

of

Calsyn, R. J., Klinkenberg, W. D., Morse, G. A., Miller, J., Cruthis, R., HIV/AIDS Treatment

ro

Adherence, Health Outcomes and Cost Study Group (2004). Recruitment, engagement,

-p

and retention of people living with HIV and co-occurring mental health and substance

re

use disorders. AIDS Care, 16 Suppl 1, S56-70.

lP

Carroll, K. M., Ball, S. A, Martino, S., Nich, C., Babuscio, T. A., Rounsaville, B. J. (2009).

na

Enduring effects of a computer-assisted training program for cognitive behavioral therapy: a 6-month follow-up of CBT4CBT. Drug Alcohol Depend, 100(1–2), 178–81.

Portnoy,

Jo ur

Carroll, K. M., Ball, S. A., Martino, S., Nich, C., Babuscio, A., Nuro, K. F., Gordon, M. A., G.

A.,

Rounsaville,

B.

J.

(2008).

Computer-assisted

delivery of

cognitive-behavioral therapy for addiction: a randomized trial of CBT4CBT. Am J Psychiatry, 165(7), 881–8. Carroll, K. M., Kiluk, B. D., Nich, C., Gordon, M. A., Portnoy, G. A., Marino, D. R., Ball, S. A. (2014). Computer-assisted delivery of cognitive-behavioral therapy: efficacy and durability of CBT4CBT among cocaine-dependent individuals maintained on methadone.

27

Journal Pre-proof

Am J Psychiatry, 171(4), 436–44. Chebli, J. L., Blaszczynski, A., Gainsbury, S. M. (2016). Internet-based interventions for addictive behaviours: a systematic review. J Gambl Stud, 32(4),1279-304. Cocco, K. M., Carey, K. B. (1998). Psychometric properties of the Drug Abuse Screening

of

Test in psychiatric outpatients. Addict Behav, 10(4), 408–14.

ro

Cohen, J., (1992). Quantitative methods in psychology. Psychol Bull, 112(1), 155–9.

-p

Crouch, T. B., DiClemente, C. C., Pitts, S. C. (2015). End-of-treatment abstinence

re

self-efficacy, behavioral processes of change, and posttreatment drinking outcomes in

lP

Project MATCH. Psychol Addict Behav, 29(3), 706–15.

na

Cumming, C., Troeung, L., Young, J. T., Kelty, E., Preen, D. B. (2016). Barriers to accessing methamphetamine treatment: A systematic review and meta-analysis. Drug Alcohol

Jo ur

Depend, 168, 263–73.

Degenhardt, L., Charlson, F., Ferrari, A., Santomauro, D., Erskine, H., Mantilla-Herrara, A., … Vos, T. (2018). The global burden of disease attributable to alcohol and drug use in 195 countries and territories, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. The Lancet Psychiatry, 5(12), 987–1012. Degenhardt, L., Whiteford, H. A., Ferrari, A. J., Baxter, A. J., Charlson, F. J., Hall, W. D., Freedman, G., Burstein, R., Johns, N., Engell, R. E., Flaxman, A., Murray, C. J., Vos, T.

28

Journal Pre-proof

(2013). Global burden of disease attributable to illicit drug use and dependence: findings from the Global Burden of Disease Study 2010. Lancet, 382(9904), 1564–74. Fals-Stewart, W., O’Farrell, T. J., Freitas, T. T., McFarlin, S. K., Rutigliano, P. (2000). The timeline followback reports of psychoactive substance use by drug-abusing patients:

of

psychometric properties. J Consult Clin Psychol, 68(1), 134–44.

ro

Gavin, D. R., Ross, H. E., Skinner, H. A. (1989). Diagnostic validity of the drug abuse

-p

screening test in the assessment of DSM-III drug disorders. Br J Addict, 84(3), 301–7.

re

Goslar, M., Leibetseder, M., Muench, H. M., Hofmann, S. G., & Laireiter, A.-R. (2019).

lP

Pharmacological Treatments for Disordered Gambling: A Meta-analysis. Journal of

na

Gambling Studies, 35(2), 415–445.

Kadden, R. M., Litt, M. D. (2011). The role of self-efficacy in the treatment of substance use

Jo ur

disorders. Addict Behav, 36(12), 1120–6. Kay-Lambkin, F. J., Baker, A. L., Lewin, T. J., Carr, V. J. (2009). Computer-based psychological treatment for comorbid depression and problematic alcohol and/or cannabis use: a randomized controlled trial of clinical efficacy. Addiction, 104(3), 378– 88. Kay-lambkin, F. J., Baker, A. L., Kelly, B., Lewin, T. J. (2011). Clinician-assisted computerised versus therapist-delivered treatment for depressive and addictive disorders:

29

Journal Pre-proof

a randomised controlled trial, Med J Aust, 195(3), S44–S50. Kobayashi, O., Matsumoto, T., Chiba, Y., Imamura, F., Morita, N., Wada, K. (2010). [Factor structure and validity of the Japanese version of SOCRATES (Stages of Change Readiness and Treatment Eagerness Scale) in adolescent delinquents placed in detention

of

home]. Nihon Arukoru Yakubutsu Igakkai Zasshi, 45(5), 437–51.

ro

Kobayashi, O., Matsumoto, T., Otsuki, M., Endo, K., Okudaira, K., Harai, H., Wada, K.

-p

(2007). [A preliminary study on outpatient relapse prevention program for

re

methamphetamine dependent patients: Serigaya Methamphetamine Relapse Prevention

lP

Program (SMARPP)]. Nihon Arukoru Yakubutsu Igakkai Zasshi, 42(5), 507–21.

na

Kondo, A., Ide, M., Takahashi, I., Taniai, T., Miura, K., Yamaguchi, A., Yotsuji, N., Matsumoto, T. (2014). [Evaluation of the relapse prevention program for substance

Jo ur

abusers called "TAMARPP" at mental health and welfare center]. Nihon Arukoru Yakubutsu Igakkai Zasshi, 49(2), 119–35. Krebs, P., Norcross, J. C., Nicholson, J. M., & Prochaska, J. O. (2018). Stages of change and psychotherapy outcomes: A review and meta-analysis. Journal of Clinical Psychology, 74(11), 1964–1979. Matsumoto, T. (2014). [The Nationwide Mental Hospital Survey on Drug-related Psychiatric Disorders].

Retrieved

December

30

02,

2018

from:

Journal Pre-proof

http://www.ncnp.go.jp/nimh/yakubutsu/drugtop/data/researchJHS2014.pdf Matsumoto, T., Chiba, Y., Imamura, F., Kobayashi, O., Wada, K. (2011). Possible effectiveness of intervention using a self-teaching workbook in adolescent drug abusers detained in a juvenile classification home. Psychiatry Clin Neurosci, 65(6), 576–83.

of

Matsumoto, T., Tachimori, H., Tanibuchi, Y., Takano, A., Wada, K. (2014). Clinical features

ro

of patients with designer-drug-related disorder in Japan: a comparison with patients with

-p

methamphetamine- and hypnotic/anxiolytic-related disorders. Psychiatry Clin Neurosci,

re

68(5), 374–82.

lP

Michie, S., Whittington, C., Hamoudi, Z., Zarnani, F., Tober, G., & West, R. (2012).

na

Identification of behaviour change techniques to reduce excessive alcohol consumption. Addiction, 107(8), 1431–1440.

Jo ur

Miller, W. R., Tonigan, J. S. (1996). Assessing drinkers’ motivation for change: The Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES). Psychol Addict Behav, 10(2), 81–9.

Minozzi, S., Saulle, R., De Crescenzo, F., Amato, L. (2016). Psychosocial interventions for psychostimulant misuse. Cochrane Database Syst Rev. 9. Moore, B. A., Fazzino, T., Garnet, B., Cutter, C. J., Barry, D. T. (2011). Computer-based interventions for drug use disorders: a systematic review. J Subst Abuse Treat, 40(3),

31

Journal Pre-proof

215–23. Morita, N. (2013). [Development and efficacy of psychoeducation therapy for drug dependence with comorbid psychiatric disorder]. [Development and efficacy of cognitive behavioral therapy for drug dependence], 89-119. Retrieved December 02, 2018 from:

of

http://www.ncnp.go.jp/nimh/yakubutsu/drug-top/data/research_SMARPP2012sogo.pdf

ro

Morita, N., Suetsugu, S., Shimane, T., Okasaka, Y., Kiyoshige, T., Iwai, K. (2007).

-p

[Development of a manualized cognitive behavioral therapy program for Japanese drug

re

addicts and a study of the efficacy of the program]. Nihon Arukoru Yakubutsu Igakkai

lP

Zasshi, 42(5), 487–505.

Addiction.

na

National Institute on Drug Abuse (2016). DrugFacts: Treatment Approaches for Drug Retrieved

December

02,

2018

from:

Jo ur

https://www.drugabuse.gov/publications/drugfacts/treatment-approaches-drug-addiction Nesvåg, S., McKay, J. R. (2018). Feasibility and effects of digital interventions to support people in recovery from substance use disorders: Systematic review. J Med Internet Res, 20(8):e255. Norberg, M. M., Mackenzie, J., Copeland, J. (2012). Quantifying cannabis use with the timeline followback approach: a psychometric evaluation. Drug Alcohol Depend, 121(3), 247–52.

32

Journal Pre-proof

Ogai, Y., Haraguchi, A., Kondo, A., Ishibashi, Y., Umeno, M., Kikumoto, H., Hori, T., Komiyama, T., Kato, R., Aso, K., Asukai, N., Senoo, E., Ikeda, K. (2007). Development and validation of the Stimulant Relapse Risk Scale for drug abusers in Japan. Drug Alcohol Depend, 88(2–3), 174–81.

of

Ondersma, S. J., Svikis, D. S., Schuster, C. R. (2007). Computer-based brief intervention a

ro

randomized trial with postpartum women. Am J Prev Med, 32(3), 231–8.

-p

Peter, S. C., Brett, E. I., Suda, M. T., Leavens, E. L. S., Miller, M. B., Leffingwell, T. R., …

re

Meyers, A. W. (2019). A Meta-analysis of Brief Personalized Feedback Interventions for

lP

Problematic Gambling. Journal of Gambling Studies, 35(2), 447–464.

na

Portnoy, D. B., Scott-Sheldon, L. A., Johnson, B. T., Carey, M. P. (2008). Computer-delivered interventions for health promotion and behavioral risk reduction: a

Jo ur

meta-analysis of 75 randomized controlled trials, 1988-2007. Prev Med, 47(1), 3–16. Rawson, R. a, Shoptaw, S. J., Obert, J. L., McCann, M. J., Hasson, A. L., Marinelli-Casey, P. J., … Ling, W. (1995). An intensive outpatient approach for cocaine abuse treatment. The Matrix model. Journal of Substance Abuse Treatment, 12(2), 117–127. Reback, C. J., Rünger, D., Fletcher, J. B., & Swendeman, D. (2018). Ecological momentary assessments for self-monitoring and counseling to optimize methamphetamine treatment and sexual risk reduction outcomes among gay and bisexual men. Journal of Substance

33

Journal Pre-proof

Abuse Treatment, 92,17–26. Rooke, S., Copeland, J., Norberg, M., Hine, D., McCambridge, J. (2013). Effectiveness of a self-guided web-based cannabis treatment program: randomized controlled trial. J Med Internet Res, 15(2), e26.

of

Rooke, S., Thorsteinsson, E., Karpin, A., Copeland, J., Allsop, D. (2010). Computer-delivered

ro

interventions for alcohol and tobacco use: a meta-analysis. Addiction, 105(8), 1381–90.

-p

Shimane, T., Imamura, A., Ikeda, K., Yamamoto, M., Tsuji, M., Nagayo, Y., Ohkubo, T.,

re

Ohta, M., Kanda, H., Okazaki, S., Ohe, M., Matsumoto, T. (2015). [Reliability and

na

Zasshi, 50(6), 310–24.

lP

validity of the Japanese version of the DAST-20]. Nihon Arukoru Yakubutsu Igakkai

Sholomskas, D. E., Syracuse-Siewert, G., Rounsaville, B. J., Ball, S. A., Nuro, K. F., Carroll,

Jo ur

K. M. (2005). We don’t train in vain: a dissemination trial of three strategies of training clinicians in cognitive-behavioral therapy. J Consult Clin Psychol, 73(1), 106–15. Skinner, H. A. (1982). The drug abuse screening test. Addict Behav, 7(4), 363–71. Sliedrecht, W., de Waart, R., Witkiewitz, K., & Roozen, H. G. (2019). Alcohol use disorder relapse factors: A systematic review. Psychiatry Research, 278, 97-115. Sobell, L. C., Brown, J., Leo, G. I., Sobell, M. B. (1996). The reliability of the Alcohol Timeline Followback when administered by telephone and by computer. Drug Alcohol

34

Journal Pre-proof

Depend, 42(1), 49–54. Sobell, L. C., Maisto, S. A., Sobell, M. B., Cooper, A. M. (1979). Reliability of alcohol abusers’ self-reports of drinking behavior. Behav Res Ther, 17(2), 157–60. Tait, R. J., McKetin, R., Kay-Lambkin, F., Bennett, K., Tam, A., Bennett, A., Geddes, J.,

controlled

trial

of

an

internet-based

intervention

addressing

ro

randomised

of

Garrick, A., Christensen, H., Griffiths, K. M. (2012). Breakingtheice: a protocol for a

-p

amphetamine-type stimulant use. BMC Psychiatry, 12(1), 67.

re

Tait, R. J., McKetin, R., Kay-Lambkin, F., Carron-Arthur, B., Bennett, A., Bennett, K.,

lP

Christensen, H., Griffiths, K. M. (2014). A web-based intervention for users of

Ment Health, 1(1), e1.

na

amphetamine-type stimulants: 3-month outcomes of a randomized controlled trial. JMIR

Jo ur

Takano, A., Miyamoto, Y., Kawakami, N., Matsumoto, T. (2016a). Web-based cognitive behavioral relapse prevention program with tailored feedback for people with methamphetamine and other drug use problems: development and usability study. JMIR Ment Health, 3(1), e1. Takano, A., Miyamoto, Y., Kawakami, N., Matsumoto, T. (2016b). [The potential of e-Health for treatment and prevention for drug dependence: process of development and revision of web-based relapse prevention program “e-SMARPP”]. Nihon Arukoru Yakubutsu

35

Journal Pre-proof

Igakkai Zasshi, 51(6):382-92. Takano, A., Miyamoto, Y., Kawakami, N., Matsumoto, T., Shinozaki, T., Sugimoto, T. (2016c). Web-based cognitive behavioral relapse prevention program with tailored feedback for people with methamphetamine and other drug use problems: protocol for a

of

multicenter randomized controlled trial in Japan. BMC Psychiatry, 16(1), 87.

ro

Takano, A., Miyamoto, Y., Matsumoto, T. (2015). [A review about new approaches using the

re

Yakubutsu Igakkai Zasshi, 50(1), 19–34.

-p

Internet and computer technology for people with drug use disorder]. Nihon Arukoru

lP

Tanibuchi, Y., Matsumoto, T., Imamura, F., Wakabayashi, A., Kawachi, H., Hikitsuchi, E.,

na

Takano, A., Yonezawa, M., Kato, T., Yamada, M., Wachi, A., Aboshi, M., Wada, K. (2016). [Efficacy of the Serigaya Methamphetamine Relapse Prevention Program

Jo ur

(SMARPP): for patients with drug use disorder: A study on factors influencing 1-year follow-up outcomes]. Nihon Arukoru Yakubutsu Igakkai Zasshi, 51(1), 38–54. United Nations Office on Drugs and Crime (2018). World Drug Repot 2018. Retrieved December

02,

2018

from:

https://www.unodc.org/wdr2018/prelaunch/WDR18_Booklet_2_GLOBAL.pdf Weissman, M. M., Verdeli, H., Gameroff, M. J., Bledsoe, S. E., Betts, K., Mufson, L., Fitterling, H., Wickramaratne, P. (2006). National survey of psychotherapy training in

36

Journal Pre-proof

psychiatry, psychology, and social work. Arch Gen Psychiatry, 63, 925–34. White, A., Kavanagh, D., Stallman, H., Klein, B., Kay-Lambkin, F., Proudfoot, J., Drennan, J., Connor, J., Baker, A., Hines, E., Young, R. (2010). Online Alcohol Interventions: A Systematic Review. J Med Internet Res. 12(5), e62.

of

Yudko, E., Lozhkina, O., Fouts, A. (2007). A comprehensive review of the psychometric

Jo ur

na

lP

re

-p

ro

properties of the Drug Abuse Screening Test. J Subst Abuse Treat, 32(2), 189–98.

37

Journal Pre-proof

Table 1 Participants’ characteristics and Internet use at baseline assessment

Male Currently married Never married Divorced Single Middle school High school Some college College or higher Full-time Part-time Unemployed Sick leave Housewife/other Every day 2 hours or more/day Smartphone Personal computer Tablet/mobile phone

Cohabiter Education

-p

ro

Job

of

Age Sex Marital status

Intervention Control (n=25) (n=23) n/mean %/(SD) n/mean %/(SD) 37.0 (7.3) 39.5 (7.5) 14 60.9 19 76.0 4 17.4 5 20.0 15 65.2 17 68.0 4 17.4 3 12.0 4 17.4 7 28.0 2 8.7 4 16.0 9 39.1 4 16.0 6 26.1 7 28.0 6 26.1 10 40.0 4 17.4 3 14.6 5 21.7 2 8.0 12 52.2 14 56.0 0 0 2 8.0 2 8.7 4 16.0 19 82.6 21 84.0 15 65.2 18 72.0 18 78.3 17 68.0 4 17.4 7 28.0 1 4.3 1 4.0

re

Internet use

lP

Internet device (most use)

Primary drug

na

Methamphetamine NPS MDMA

Jo ur

Benzodiazepine

Cough medicine Heroine Inhalant Poly drug

Duration of consecutive abstinence from primary drug in past 28 days Duration of consecutive abstinence from all substances in past 28 days Total abstinent days from primary drug in past 28 days Total abstinent days from all substances in past 28 days Age of first drug use

Arrest in past Jail in past Drug dependence severity (DAST-20)

Total score Low (1-5) Intermediate (6-10) Substantial (11-15)

38

13 1 3 1 2 0 1

56.5 4.3 13.0 4.3 8.7 0 4.3

11 5 2 3 2 2 0

44.0 20.0 8.0 12.0 8.0 8.0 0

2

8.7

0

0

23.4

(9.2)

21.0

(10.7)

20.3

(10.3)

14.8

(11.2)

25.2

(7.0)

23.0

(9.3)

23.7

(7.9)

19.8

(10.1)

21.3

(7.6)

21.5

(5.6)

16 4 13.2 1 4

69.6 17.4 (3.6) 4.3 17.4

16 6 11.7 3 4

64.0 24.0 (3.9) 12.0 16.0

14

60.9

13

52.0

Journal Pre-proof

Severe (16-20) < 1 year 1-3 years > 3 years

Outpatient treatment period

4 8 2 13 2.9 11 6

Number of hospitalizations Face-to-face relapse prevention Self-help group NPS: New psychoactive substances MDMA: 3,4-methylenedioxymethamphetamine

Jo ur

na

lP

re

-p

ro

of

DAST-20: Drug Abuse Screening Test

39

17.4 34.8 8.7 56.5 (6.4) 47.8 26.1

5 7 5 13 2.2 13 8

20.0 28.0 20.0 52.0 (6.3) 52.0 32.0

Journal Pre-proof Table 2 Average abstinence days and scores of primary and secondary outcomes by treatment condition Abstinence duration in intervention period (56 days)

e-SMARPP mean 48.8

e-SMARPP Baseline (n=23) mean SD

SD 14.7

Median 56

2-months (n=19) mean SD

Relapse risk 72.4 13.1 68.0 (SRRS) Motivation to change 76.2 7.5 80.9 (SOCRATES ) Self-efficacy 56.4 21.2 62.6 (SSDD) Money spent (n=22) (n=19) on drugs 18772. 36564. 8368. 7 3 4 (yen) a SRRS: Stimulant Relapse Risk Scale

Self-monitoring mean 41.2

(n=19)

5-months (n=15) mean SD

f o

8-months (n=13) mean SD

Self-monitoring Baseline 2-months (n=25) (n=24) mean SD mean SD

o r p

13.7

64.2

11.5

64.4

8.1

79.2

7.1

80.4

7.4

r P

e

71.2

18.8

66.9

(n=15) 21294. 8

l a

n r u

21.7

3333. 3

o J

18.0

10465. 4

15.0

(n=13) 12615. 4

SOCRATES: Stage of Change Readiness and Treatment Eagerness Scale SSDD: Self-efficacy Scale for Drug Dependence a: Sample size varies because outliers were excluded.

40

median 56

5-months (n=21) mean SD

(n=25) t=1.45 p=0.16 d=0.42 8-months (n=20) mean SD

66.3

12.0

63.5

10.9

63.7

10.5

62.7

12.9

79.8

9.9

81.1

10.1

82.2

9.0

80.3

11.0

62.9

18.1

69.5

15.9

68.5

20.3

69.0

19.6

(n=24) 29250. 5

SD 20.3

15667. 1

(n=23) 25068. 2

8239. 1

(n=20) 17618. 3

13670. 0

(n=19) 26081. 0

9157. 9

25257. 0

Journal Pre-proof Table 3 Efficacy of e-SMARPP on relapse risk, motivation to change, self-efficacy, and money spent on drugs (intervention: n=23, control: n=25) By assessment point

Relapse risk (SRRS)

Motivation to (SOCRATES)

Mixed-effect model estimate of group-mean difference (ref: Self-monitoring) 2-months a 0.21

(95% CI)

t

(-5.42 to 5.83)

5-months a

-1.04

(-7.64 to 5.55)

0.20

(-6.83 to 7.23)

-0.09

(-1.02 to 0.85)

2.51

(-0.97 to 6.00)

Self-efficacy (SSDD)

f o

0.94

8.41

0.02

-0.31

0.75

9.06

-0.12

0.06

0.95

11.90

0.02

-0.19

0.85

1.43

0.16

5.97

0.42

(-4.49 to 4.84)

0.08

0.94

7.30

0.02

(-2.22 to 8.62)

1.19

0.24

8.56

0.37

(-0.41 to 1.01)

0.85

0.40

-4.16

(-13.23 to 4.91)

-0.91

0.36

15.60

-0.27

-4.24

(-13.75 to 5.28)

-0.88

0.38

15.46

-0.27

-2.14

(-12.01 to 7.72)

-0.43

0.67

14.77

-0.15

-0.27

(-1.48 to 0.95)

-0.44

0.66

-1348.25

(-15789.40 to 13092.89)

-0.19

0.85

25658.81

-0.05

-14115.37

(-29403.89 to 1173.16)

-1.84

0.07

26346.53

-0.54

-1179.78

(-17036.34 to 14676.79)

-0.15

0.88

22341.46

-0.05

-698.44

(-2596.75 to 1199.86)

-0.74

0.46

5-months a

0.18

a

3.20

5-months

rn

u o

a

8-months Expanding rate per assessment point b Money spent on drugs 2-months a (yen) 5-months a

J

8-months a Expanding rate per assessment point c SRRS: Stimulant Relapse Risk Scale

l a 0.30

a

Effect size

0.07

8-months a Expanding rate per assessment point b change 2-months a 8-months Expanding rate per assessment point b 2-months a

Pooled SD p

o r p

e

r P

SOCRATES: Stage of Change Readiness and Treatment Eagerness Scale 41

Journal Pre-proof SSDD: Self-efficacy Scale for Drug Dependence a: Mixed-effect analysis-of-variance models including assessment time-point as a categorical (nominal) covariate and random-intercepts. b: Mixed-effect growth-curve models including assessment time-point as a continuous covariate (coded as 0, 2, 5, 8 [months]) and random effects for intercepts and time-point. These random effects were allowed to be dependent (SRRS, SOCRATES) or forced to be independent (SSDD). c: Mixed-effect growth-curve models including assessment time-point as a continuous covariate (coded as 0 to 3 for baseline, 2-, 5-, 8-months,

f o

respectively) and random effects.

l a

e

o r p

r P

n r u

o J

42

Journal Pre-proof

% 100.0 100.0 100.0 91.3 91.3 87.0 87.0 82.6 % 100.0 100.0 95.7 91.3 87.0 73.9

ro

n 23 23 23 21 21 20 20 19 n 23 23 22 21 20 17

-p

Self-monitoring (week) 1 2 3 4 5 6 7 8 Relapse prevention session 1 2 3 4 5 6

of

Table 4 Intervention completion rate in the e-SMARPP group (n=23)

Jo ur

na

lP

re

All participants in the control group (n=25) completed every weekly self-monitoring.

43

Journal Pre-proof

Table 5 Comparison of participants’ characteristics at the baseline between those who completed the intervention and the dropouts in the e-SMARPP group

Age Sex Primary drug

b

ro

-p

re

lP

Jo ur

Number of hospitalizations Face-to-face relapse prevention Self-help group Relapse risk (SRRS) Motivation to change (SOCRATES) Self-efficacy (SSDD) Money spent on drugs (yen)

Total score Low (1-5) Intermediate (6-10) Substantial (11-15) Severe (16-20)

na

Age of first drug use Arrest in past Jail in past Drug dependence severity (DAST-20)

of

Male Methamphetamine NPS MDMA Benzodiazepine Cough medicine Heroine Inhalant Poly drug

Complete (n=17) Dropout (n=6) n/mean %/(SD) n/mean %/(SD) 38.7 (6.2) 32.3 (8.6) 12 70.6 2 33.3 11 64.7 2 33.3 1 5.9 0 0.0 2 11.8 1 16.7 1 5.9 0 0.0 1 5.9 1 16.7 0 0.0 0 0.0 0 0.0 1 16.7 1 5.9 1 16.7 23.1 (8.0) 16.2 (1.7) 13 76.5 3 50.0 3 17.6 1 16.7 12.2 (3.5) 16.2 (1.8) 1 5.9 0 0.0 4 23.5 0 0.0 10 58.8 4 66.7 2 11.8 2 33.3 3.2 (7.4) 2.0 (2.1)

pa 0.06 0.16 0.47

0.05 0.32 0.23 0.02 0.39

0.70

8

47.1

3

50.0 1.00

6 68.8

35.3 (13.1)

0 82.3

0.0 0.14 (6.5) 0.03

77.1

(7.6)

73.5

(7.1) 0.32

61.7

(20.3)

41.7

(17.2) 0.04

11937.5 (32766.8) 37000.0 (42965.1) 0.16

a: t-test, chi-squared test, or Fisher's exact test b: Sample size varies because of excluding outliers. NPS: New psychoactive substances

MDMA: 3,4-methylenedioxymethamphetamine DAST-20: Drug Abuse Screening Test SRRS: Stimulant Relapse Risk Scale SOCRATES: Stage of Change Readiness and Treatment Eagerness Scale SSDD: Self-efficacy Scale for Drug Dependence

44

Journal Pre-proof

Supplementary table 1 Content for relapse prevention session of e-SMARPP

6.

Jo ur

5.

na

lP

4.

re

-p

3.

of

2.

What is drug dependence? Video  Mental and physical consequences caused by drug use (11’ 02”)  Change in the brain (11’ 39”)  How to stop drug craving (7’ 43”) Homework  Think about your pros and cons for drug use and quitting drug.  Define your drug use situation: when, where, who, why, what and emotion. Triggers of drug use Video  Process of craving and drug use (5’ 27”)  Various internal and external triggers of drug craving (11’ 00”)  Anchors keeping you from drug use (5’ 01”) Homework  Define your internal and external triggers.  Who and what are your anchors? Recovery process; “Just for today” Video  Process and stage of recovery (12’ 38”)  Safe lifestyle and signs of relapse (10’ 19”)  How to plan a safe daily life (9’ 27”) Homework  Think of your signs of relapse and barriers to recovery.  Plan a safe daily life schedule without drugs. Features of dependence symptoms Video  Typical features of dependence (9’ 05”)  Typical thoughts and behaviors when people fall for drugs (12’ 32”)  Justification for relapse (9’ 21") Homework  Think of your patterns of thinking and behavior during drug use  Think of your possible justification for relapse Supporters for recovery Video  Typical internal triggers: “HALT” (hungry, angry, lonely and tired) (10’ 05”)  To trust and be honest to yourself and others (5’ 41”)  Support from peers and professionals (13’ 39”) Homework  Think of ways to handle internal triggers.  Think of your supporters. Who? How to find? No need to be strong, be smart and practiced Video  Tips for recovery (6’ 04”)  Review of skills to handle triggers and relapse (12’ 21”)  To accept the way you are, messages from peers (4’ 32”) Homework  Think of crisis plans when you relapse into drug use.  Think of your future when you recover from drug addiction.

ro

1.

Each session additionally has a weekly diary activity. Parentheses indicate minutes and seconds of each video. Reference: Takano A, Miyamoto Y, Kawakami N, Matsumoto T, Shinozaki T, Sugimoto T. Web-based cognitive behavioral relapse prevention program with tailored feedback for people with methamphetamine and other drug use problems: protocol for a multicenter randomized controlled trial in Japan. BMC Psychiatry. 2016, 16:87.

45

Journal Pre-proof Supplementary table 2 Assessment schedule of primary and secondary outcomes Measurement

Follow-up 2-month 5-month

8-month

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

f oo pr e-

on

Pr

Self-efficacy Money spent drugs

Stage of Change Readiness and Treatment Eagerness Scale Self-efficacy Scale for Drug Dependence Cost of drug in the last month (yen)

al

4 5

Longest consecutive abstinent days during the intervention Stimulant Relapse Risk Scale

rn

Primary outcome 1 Duration of abstinence 2 Relapse risk Secondary outcome 3 Motivation to change

Baseline

Jo u

Outcome

Journal Pre-proof

Figure 1. Participant flow diagram.

Enrollment

Assessed for eligibility Baseline assessment Randomized (n=48)

Allocation Allocated to e-SMARPP (n=23) Received allocated intervention (n= 23) Did not receive allocated intervention (n=0)

Allocated to web-based self-monitoring (n= 25) Received allocated intervention (n= 25) Did not receive allocated intervention (n=0)

oo

f

Follow-Up Lost to follow-up (unable to contact)

Lost to follow-up (unable to contact) 2-month: questionnaires (n=1, 4%), TLFB (n=0, 0%) 5-month: questionnaires (n=4, 16%), TLFB (n=5, 20%) 8-month: questionnaires (n=5, 20%), TLFB (n=6, 24%)

e-

Analysis

pr

2-month: questionnaires (n=4, 17.4%), TLFB (n=4, 17.4%) 5-month: questionnaires (n=8, 34.9%), TLFB (n=7, 30.4%) 8-month: questionnaires (n=10, 43.5%), TLFB (n=10, 43.5%)

Analysed (n=23)

Analysed (n=25)

Pr

TLFB: Timeline method to assess participants’ drug use condition Excluded from analysis ExcludedFollowback from analysis a: In the intervention group, 4 participants awere excluded from the analysis because of missing data.

Duration of abstinence (n=0) Other repeated outcomes (n=0)

Jo u

rn

al

Duration of abstinence (n=4)

Journal Pre-proof

Highlights

rn

al

Pr

e-

pr

oo

f

Internet-based treatment could be a promising approach for drug use problems. The effects of such treatment in Asia where stimulant drugs are common was unclear. A web-based relapse prevention program was assessed using a RCT in Japan. A significant effect on abstinence was not demonstrated due to a small sample size. Feasibility was confirmed because attrition from the intervention was low.

Jo u

    

Figure 1