This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
Cannabis is the most frequently used illicit substance in the United States resulting in high rates of cannabis use disorders. Current treatments for cannabis use are often met with high rates of lapse/relapse, tied to (1) behavioral health factors that impact cannabis use such as poor sleep, and (2) access, stigma, supply, and cost of receiving a substance use intervention.
This pilot study examined the feasibility, usability, and changes in cannabis use and sleep difficulties following mobile phone–delivered Cognitive Behavioral Therapy for Insomnia (CBT-I) in the context of a cannabis cessation attempt.
Four male veterans with DSM-5 cannabis use disorder and sleep problems were randomized to receive a 2-week intervention: CBT-I Coach mobile app (n=2) or a placebo control (mood-tracking app) (n=2). Cannabis and sleep measures were assessed pre- and post-treatment. Participants also reported use and helpfulness of each app. Changes in sleep and cannabis use were evaluated for each participant individually.
Both participants receiving CBT-I used the app daily over 2 weeks and found the app user-friendly, helpful, and would use it in the future. In addition, they reported decreased cannabis use and improved sleep efficiency; one also reported increased sleep quality. In contrast, one participant in the control group dropped out of the study, and the other used the app minimally and reported increased sleep quality but also increased cannabis use. The mood app was rated as not helpful, and there was low likelihood of future participation.
This pilot study examined the feasibility and initial patient acceptance of mobile phone delivery of CBT-I for cannabis dependence. Positive ratings of the app and preliminary reports of reductions in cannabis use and improvements in sleep are both encouraging and support additional evaluation of this intervention.
Cannabis is the most frequently used illicit substance in the United States [
In light of the relatively poor outcomes associated with standard cannabis use disorder treatments, there has been increasing interest in identifying malleable factors associated with heightened relapse risk among individuals dependent on cannabis, specifically for the purposes of intervention development and refinement [
From these findings, the evaluation of existing sleep treatments as clinical tools to aid in the treatment of cannabis use disorders has been proposed, with the aim of improving sleep and reducing individual “self-medication” with cannabis [
Cognitive Behavioral Therapy for Insomnia (CBT-I) is the leading behavioral treatment for sleep disturbances [
Given identified barriers to substance use disorder care that include access, stigma, supply, and cost [
This pilot study aimed to begin addressing these gaps in the literature by determining the feasibility of providing CBT-I to individuals with a cannabis use disorder engaged in a cessation attempt through a mobile phone app. CBT-I was administered via mobile app instead of in-person sessions as this would allow for greater reach of the intervention (ie, rural areas) and consistent fidelity of the intervention. For the purposes of this feasibility study, we sought to examine mobile app engagement, user-friendliness, and use and future interest in use, in order to provide initial data regarding the feasibility and usability of this technology among veterans with a cannabis use disorder. As a secondary aim, we examined preliminary trends in the impact of app use on cannabis and sleep outcomes. Specifically, we hypothesized that receiving CBT-I through an app, within the context of a self-guided cannabis cessation attempt, would be acceptable to study participants and utilization of the app would correspond with a reduction in cannabis use and an improvement in sleep quality.
In terms of recruitment, 40 individuals contacted the lab expressing interest in the study. Individuals leaving a message were contacted up to four times for the purpose of conducting an initial phone screening. Of the original 40 individuals, 28 veterans who contacted the lab were screened over the phone. Of these, 12 individuals were screened out because they did not report current cannabis use, 6 were not veterans, 5 reported no interest in quitting, and 1 individual was eligible but refused to participate. This resulted in a total sample of 4 individuals.
Study participants (N=4) were adult veterans (18 and over) who reported (1) current sleep problems, defined as a total score of 5 or greater on the Pittsburgh Sleep Quality Index (PSQI) [
The Structured Clinical Interview-Non-Patient Version (SCID I-N/P) for DSM-IV [
A Mobile App Use Measure (MAUM) was developed to gather information about participant use of both the experimental (ie, CBT-I) and control (mood-tracking) apps between baseline and 2-week follow-up, as neither app recorded use data. The MAUM prompted participants to report on (1) frequency of app use, (2) length of use per occasion, (3) components of the app that were used most often, (4) user-friendliness of the app, (5) helpfulness of the app, (6) likelihood of future app use, (7) use of the app for sleep improvement, and (8) use of the app for cannabis cessation. The MAUM was administered at 2-week follow-up.
Sleep quality at baseline and 2-week follow-up was assessed with the 19-item Pittsburgh Sleep Quality Index (PSQI) [
The Timeline Follow
Cognitive behavioral therapy for insomnia (CBT-I) was administered via the CBT-I Coach mobile app for iOS [
Example screenshots from CBT-i Coach.
A self-monitoring mobile app [
Participants were recruited from a VA hospital in the San Francisco Bay Area via flyers placed throughout the hospital campus and by referrals from clinical staff. Recruitment materials were designed to target veterans interested in making a cannabis cessation attempt. Interested individuals contacted the research lab and completed a brief phone screen to assess for initial eligibility. Those meeting initial eligibility criteria were scheduled for a baseline appointment 1 day prior to the date that they would be willing to make an independent, self-guided, cannabis cessation attempt.
On arrival for the baseline appointment, individuals provided written informed consent. Individuals consented to participate in a cannabis cessation attempt and to engage with a behavioral program via mobile app. The inclusion of a sleep intervention was not discussed in recruitment or consenting procedures. Following informed consent, participants were interviewed with the SCID I-N/P and then completed a battery of self-report measures. Following the completion of the baseline assessment, participants were randomly assigned, by coin flip, to receive either the CBT-I (experimental condition) or mood-tracking app (control condition). Following randomization, participants were then loaned an iPod touch with charger, trained on how to use their assigned mobile app (which was pre-installed), and instructed to (1) make a cannabis cessation attempt beginning the following morning, and (2) use their assigned mobile app during the first 2 weeks of their cessation attempt. At the end of the baseline assessment, contact information was obtained for each participant to allow for reminder calls regarding app use and follow-up appointment date. Participants were then scheduled for their 2-week follow-up and compensated US $45 for completion of the baseline appointment.
Two weeks following baseline, participants returned to the research laboratory to complete a self-report assessment battery that was similar to the baseline assessment, with the exception of the SCID I-N/P and addition of the MAUM. Following the 2-week follow-up assessment, participants were instructed to relinquish their iPod touch devices and were compensated US $45. All study procedures were approved by the VA Palo Alto and Stanford Institutional Review Boards.
Based on the small sample size of this pilot study, study outcomes (cannabis use, sleep, and metrics of mobile phone app use) were examined qualitatively for each participant separately.
Information regarding baseline and follow-up scores on study variables can be found in
Participant 1 was a 42-year-old single Caucasian male veteran who met DSM-5 criteria for current cannabis use disorder and alcohol use disorder. Participant 1 was a non-smoker and did not meet criteria for any mood or anxiety disorders. He was a novice with mobile technology and did not own a smartphone. As noted in the Method section, the MSHQ was administered instead of the TLFB for this participant. Based on the number of reported days using cannabis during the prior month, coupled with the quantity reported, we were able to extrapolate this participant’s cannabis use during the 30 days prior to baseline (see
Participant 1 was randomly assigned to receive CBT-I coach. He reported using the mobile app daily over the 2-week course of his cessation attempt. On average, Participant 1 used the app for 5 minutes at a time and typically interacted with the app in the morning. He reported the most helpful aspect of the app as allowing him to track and view his sleep over the 2-week period. He reported the app as user-friendly and indicated interest in using it in the future. Participant 1 also reported that the app helped his sleep because it provided daily accountability to the intervention components. He also noted that it helped with the cessation attempt because it made him more aware of when and how much cannabis he was using. At 2-week follow-up, Participant 1 had self-reported no cannabis use during the 2-week study period. His total PSQI score did not change from baseline (total score=9); however, his sleep efficiency increased from 58% to 80%, bringing it to just below the normal range (85%). At follow-up he reported using medication (prescription or non-prescription) less than once a week for sleep.
Quantity of cannabis use during each day of the 90-day (30-days for participant 1) baseline and 2-week follow-up periods for each participant, as measured by the MSHQ (Participant 1) and TLFB (Participants 2-4). Vertical dashed line indicates the cessation day for each participant. Because Participant 2 did not complete the follow-up assessment, his data are not presented. Horizontal gray solid lines represent mean cannabis use during the represented timeframe (baseline, 2-week follow-up).
Participant 2 was a 54-year-old widowed black/non-Hispanic male veteran who met DSM-5 criteria for current cannabis use disorder and alcohol use disorder. He did not meet criteria for any mood or anxiety disorders. Participant 2 was a tobacco smoker, consuming an average of 4 cigarettes per day. His mean cannabis use over the 90 days prior to the baseline appointment was 0.71 (on the 0-8 visual scale [
Participant 2 was randomly assigned to the control condition (mood-tracking app). This participant dropped out of the study (ie, did not complete his 2-week appointment), thus follow-up data are not available.
Participant 3 was a 65-year-old married black/Hispanic male veteran who met DSM-5 criteria for cannabis use disorder. He did not meet criteria for any mood or anxiety disorders. Participant 3 was also a novice with mobile apps and did not own a smartphone. His mean cannabis use over the 90 days prior to the baseline appointment was 0.18 (on the 0-8 visual scale [
Participant 3 was randomized to the control condition. He reported using the mood-tracking app minimally (less than once per week) over the 2-week cessation period. Participant 3 reported the app as extremely user-friendly, though he indicated that he would not use it in the future and felt that it did not help at all with his cannabis cessation attempt or sleep. On follow-up, Participant 3 reported an increase in cannabis use (mean use 0.57; on the 0-8 visual scale [
Participant 4 was a 27-year-old divorced Caucasian male veteran of recent conflicts in Iraq and Afghanistan. SCID I-N/P assessment indicated that he met criteria for current PTSD and cannabis use disorder; he was a tobacco non-smoker. Participant 4 had the most experience (of the study participants) with mobile technology, and owned an Android smartphone. He reported the highest amount of cannabis use in the 90 days prior to baseline relative to other participants, with a mean use of 4.51 (on the 0-8 visual scale [
Participant 4 was randomly assigned to the CBT-I coach app. He reported using the app daily over the course of the 2-week intervention period. On average, he used the app for 10 minutes during each use, typically in the morning. He noted the app as very user-friendly and indicated that he would use it again in the future. Participant 4 reported that the stimulus control and sleep prescription components of the app were the most helpful, highlighting that the app was useful for sleep, but not for his cannabis cessation attempt. On follow-up, a decrease in mean cannabis use was observed during weeks 1 and 2 (mean use 1.50 on the 0-8 visual scale [
Participant-level cannabis use and sleep data for baseline and follow-up assessments.
Variable | P1: Experimental | P2: Placebo control | P3: Placebo control | P4: Experimental | |||||
Time of assessment | Baseline | Follow-up | Baseline | Follow-up | Baseline | Follow-up | Baseline | Follow-up | |
Cannabis usea | 0.40 | 0.00 | 0.71 | N/A | 0.18 | 0.57 | 4.51 | 1.50 | |
|
|||||||||
|
Quality | 1 | 1 | 3 | N/A | 3 | 3 | 1 | 1 |
|
Latency | 30 min | 10 min | 30 min | N/A | 120 min | 120 min | 30 min | 15 min |
|
Duration | 5 hours | 6 hours | 2 hours | N/A | 2.5 hours | 3.5 hours | 8 hours | 8 hours |
|
Efficiency | 58% | 80% | 28% | N/A | 29% | 41% | 80% | 100% |
|
Disturbances | 2 | 2 | 3 | N/A | 3 | 2 | 1 | 0 |
|
Med use | 0 | 1 | 2 | N/A | 1 | 0 | 0 | 0 |
Daytime dysfunction | 0 | 2 | 3 | N/A | 1 | 0 | 1 | 1 | |
Total PSQIc | 9 | 9 | 19 | N/A | 17 | 14 | 5 | 2 |
aMean cannabis use was based on both frequency of use (ie, number of days) and quantity of use per day (scores range from 0-8). Quantity of cannabis used on each day was indexed by a graphical depiction of joints of varying sizes that ranged from 0-8 [
bSleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbances, daytime dysfunction, and sleep med use represent the 7 components of the PSQI [
cPSQI total scores can range from 0 (excellent sleep) to 21 (very poor sleep).
The current study provided an initial demonstration of using mobile apps for the delivery of psychological interventions among veterans. Feasibility was examined in terms of recruitment, costs, and veteran engagement with the mobile app.
Overall, participants reported interest in using the mobile apps, describing them as easy to use and accessible. Importantly, this feasibility study demonstrated reported app accessibility among veterans without experience with mobile technology. In addition, this study demonstrated the feasibility of using mobile apps among multiple age groups. Indeed, participants ranging from their mid-20s to mid-60s all reported ease in using the mobile apps.
Individuals were compensated up to a total of US $90 for participation. Additional research costs included iPod touch devices (only 1 individual in the study had their own iPhone) and staff costs.
Overall, our sample consisted of novices in terms of mobile technology and mobile app use. Though Participant 4 had experience with using mobile apps, all other participants reported not owning a smartphone or having previous experience with mobile apps. However, all were able to use the apps after a brief explanation. The responses and engagement with the mobile apps was found to vary by treatment condition. Participant 3 (control) reported minimal use of the app (scoring a 1 out of a possible 4). Follow-up assessment of app use found that this individual did not find the mood tracking app helpful for sleep (reporting 0 out of 4) or cannabis cessation (reporting 0 out of 4) and therefore reported little interest in engaging with the app and low likelihood of using it in the future (reporting 0 out of 4). However, for those who received the CBT-I app (Participants 1 and 4), daily use was reported over the 2-week study period, with an average of 5-10 minutes spent with the app per use session.
Participants assigned to the CBT-I app reported that the most helpful portions of the app were the sleep logs and the exercises for improving sleep. Both reported the app to be very user-friendly (scoring 3 out of 4), found that it was very helpful in improving sleep (scoring 3 out of 4), and reported strong interest in using an intervention app in the future (scoring 3 out of 4). All loaned iPod touches were returned to the study team after completion of the study. Although Participant 2 did not complete the 2-week follow-up, he did return to the laboratory to return his iPod touch.
This pilot study served as a feasibility test of mobile-app delivered CBT-I (and mood-tracking control) for both cannabis and sleep outcomes among individuals engaged in a cannabis cessation attempt. This study indicates that individuals who received the CBT-I app reported daily engagement with the app, while individuals in the control condition either dropped out of the study or reported minimal use. Importantly, regardless of condition, this study provided support for loaning out iOS devices for the purposes of treatment, as all devices were returned following study completion.
Findings support the initial feasibility of providing CBT-I via mobile app to improve outcomes among individuals engaged in a cannabis cessation attempt. Coupled with theoretical and empirical work that has highlighted the pivotal role of sleep in the maintenance and treatment of cannabis use disorders [
Although our investigation served as a novel extension of prior empirical and theoretical work, it was not without limitations. First, the small number of participants, all of whom were male, enrolled in this preliminary study limits inference regarding generalizability of findings. Future work would benefit from the replication and extension of this pilot study to a larger and more gender diverse sample. Future work would benefit from focusing recruitment efforts to target female participants. Second, while we screened 40 individuals, due to our inclusion criteria, only four were enrolled. This may have biased the randomness of our sample. Third, given the scope of the study, follow-up data were collected only 2-weeks post-baseline. As the impact of cannabis cessation on sleep can last for over 30-days post-cessation [
This study provides initial evidence for the feasibility of CBT-I for cannabis dependence delivered through a mobile phone app. Findings support the use of technology to enable easy, affordable access to treatment for substance use disorders [
Cognitive Behavioral Therapy for Insomnia
Diagnostic and Statistical Manual of Mental Disorders, 4th edition
Diagnostic and Statistical Manual of Mental Disorders, 5th edition
Mobile App Use Measure
Marijuana Smoking History Questionnaire
Pittsburgh Sleep Quality Index
posttraumatic stress disorder
Structured Clinical Interview–Non-Patient Version for DSM-IV
Timeline Follow-Back Interview
This study was funded, in part, by the following sources: VA Clinical science research and development service funds (Bonn-Miller), VA Health Services Research and Development Service Funds (Babson), and the National Center for PTSD Dissemination and Training Division, VA Palo Alto Health Care System (Bonn-Miller), and a career development award from the National Institute on Drug Abuse (K23 DA032578, Ramo). The views presented in this paper do not necessarily represent those of the Department of Veterans Affairs.
None declared.