Abstract
Background
Pragmatic innovations are needed to optimize clinical outcomes among people who use opioids (PWUO) initiating buprenorphine. This pilot randomized controlled trial assessed the feasibility of integrating text messaging in a low threshold tele-buprenorphine bridge program for PWUO during the COVID-19 pandemic.
Methods
Eligible adult patients with opioid use disorder (OUD) inducted on buprenorphine (N=128) in the NYC Health+Hospitals Virtual Buprenorphine Clinic between May and November 2020 were randomized to an automated texting intervention based on the medical management model versus treatment as usual. A participant feedback survey was administered at 8-weeks (n=18). Primary outcomes consisted of acceptability (e.g., study enrollment, engagement with the intervention) and feasibility (e.g., lack of phone number and/or mobile phone ownership) of integrating texting in clinical care. A secondary outcome included retention in treatment at week 8 (i.e., active buprenorphine prescription within the prior 7 days).
Results
Nearly all eligible patients consented to enroll in the study (90.8%) and few were excluded due to lack of mobile phone ownership (n=27, 14.6%). Requests to discontinue receipt of texts (n=6, 9.4%) was attributed to excessive message frequency, perceived lack of relevancy, and reduced interest in the intervention. Respondents completing the follow-up feedback survey were generally satisfied with the frequency of software-generated messages (14/18, 77.8%) and half shared text content with peers (9/18, 50%). There were no perceived issues with privacy, intrusiveness, or ease of use. Retention did not differ between participants randomized to the texting (M=5.23 weeks, SD=3.41) and TAU groups (M=4.98 weeks, SD=3.34) at week 8 (p= 0.676).
Conclusion
This pilot randomized controlled trial confirms high acceptability and feasibility of integrating an automated texting tool in a tele-buprenorphine bridge program. Future studies should assess if text messaging may be efficacious when combined with staff contact and content addressing social determinants of health.
INTRODUCTION
The opioid epidemic has worsened since the emergence of coronavirus 2019 (COVID-19)1. Foremost among opioid overdose prevention strategies during COVID-19 are easing of federal regulations permitting video and telephone visits (telemedicine) to prescribe buprenorphine for people who use opioids (PWUOs)2. The provision of buprenorphine via telemedicine offers a safe and evidence-based approach for expanding access to care for PWUOs3,4.
Current protocols for telebuprenorphine are approximated on the medical management model offering patient education, instructions on unobserved or ‘home’ induction on buprenorphine, encouragement of self-help group and counseling participation, linkage to specialty care, and between-visit telephone support to troubleshoot administrative (e.g., prior authorization requests, appointment scheduling) and clinical concerns (e.g., cravings, withdrawal symptoms)5. Some challenges to expanding telebuprenorphine include insufficient reimbursements, telecommunications equipment and support, and staff to offer between-visit clinical support (e.g., managing complicated inductions, patient-provider communication)4,6,7. Further, studies of telebuprenorphine have described differential rates of retention 3 months post-induction on buprenorphine (50–80%) and presses the need for patient-centered strategies to reduce loss to follow-up8–11.
Mobile health (mHealth) interventions offer a low-cost and longitudinal approach to potentially improving retention among patients in telebuprenorphine while partially offsetting clinician-level barriers to increasing buprenorphine prescribing (e.g., insufficient administrative and clinical support, between-visit patient support). Text messaging is the most commonly used feature of mobile phones nationally and among PWUOs in opioid use disorder (OUD) treatment.12–14 Prior texting studies have established their acceptability and clinical impact in reducing alcohol, tobacco, and illicit substance use,15,16 and could enhance retention in telebuprenorphine by reinforcing core elements of the medical management model.
This article describes a pilot randomized clinical trial of a text messaging-based medical management system (“TeMeS” hereafter) to enhance retention in a low threshold telebuprenorphine bridge program established during COVID-19 for NYC residents with OUD. This tool is built upon the medical management model to offer appointment reminders; medication adherence support during induction, stabilization, and maintenance; cognitive behavioral therapy and motivational enhancement therapy content; and details supporting engagement with self-help groups, counseling, specialty care (e.g., behavioral health, HIV and Hepatitis C services), and harm reduction (e.g., opioid overdose education, Naloxone, pre- and post-exposure prophylaxes for HIV).17
METHODS
Study procedures and participants
This single-site study was an unmasked, proof-of-concept, pilot randomized clinical trial that utilized convenience sampling to enroll adult patients (≥18 years old) completing an initial visit in the NYC Health+Hospitals Virtual Buprenorphine Clinic between May and November 2020. Details pertaining to the clinic structure, workflow, and patient population of this low threshold tele-buprenorphine bridge program launched during COVID-19 have been described previously6. Telephone or video visits completed by addiction medicine or addiction psychiatry clinicians were centered on the medical management model and facilitated home-induction on buprenorphine using the “teach-back” method. Patients were given instructions to contact the program in the event of an unanticipated clinical issue and were followed up weekly or twice-monthly until stabilized on buprenorphine. The study received approval from the New York University School of Medicine Institutional Review Board.
Eligible patients included adults with OUD (per medical records) within their first 7 days of induction on sublingual buprenorphine, a planned treatment episode of at least 8 weeks, mobile phone ownership at the time of enrollment, ability to read and understand English, and willingness to be randomized. Patients were excluded from participating if they had any acute psychiatric or medical condition at the time of enrollment, were already receiving a mHealth intervention for OUD, had an auditory or visual disability preventing mobile phone use, or were unable to provide informed consent. The study team contacted potential patient participants by phone on the same day following completion of their initial buprenorphine visit, informed of the study details, and were enrolled after providing verbal consent.
Patient participants were randomized in a 1:1 allocation ratio to the texting intervention (intervention) or treatment as usual (TAU; passive clinician telephone support to address any unanticipated administrative or clinic issues) using an opaque, sealed envelope to reveal the random assignment generated prior to study initiation using a publicly available web-based randomization program18. Participants randomized to the TeMeS group had their phone numbers entered in a HIPAA-compliant text messaging software (Apptoto©), encouraged to utilize phone privacy safeguards (e.g., mobile phone password protection, regularly deleting text messages), and given instructions on TeMeS, including that it was fully automated and not intended to substitute for patient calls to the clinic to address any unanticipated administrative or clinic needs. A smaller sample of participants were contacted at 8-weeks to assess satisfaction with the texting tool using open-ended items and mailed a $20 transportation voucher. The study received approval from the New York University School of Medicine Institutional Review Board and Bellevue Hospital Department of Research.
Intervention development
Details pertaining to intervention development (text content, frequency of text messages delivered during induction, stabilization, and maintenance) have been reported previously in detail.12,13,19,20 Prior usability testing of TeMeS identified design specifications requiring additional modifications, and incorporated key stakeholder feedback (i.e., patients, clinicians, administrators) to ensure seamless integration in clinical care.21,22 In preparation for this pilot study, message content was revised to include the Virtual Buprenorphine Clinic program phone number, operating hours, and telephone hotline and online resources for self-help groups, specialty care, and counseling services. The intervention consists of over 400 archived messages that were delivered as follows: a) patient-provider communication (i.e., clinic contact information and location; sent x1–3/week with a maximum of 1 message per day); b) adherence to buprenorphine (queries pertaining to cravings and withdrawal symptoms and unobserved home induction instructions, daily medication reminders; x1–3 messages per day during induction/stabilization that were reduced to once daily messages during maintenance); c) self-management (informational content on OUD, buprenorphine, opioid overdose prevention, specialty care; once daily); d) goal of opioid abstinence per motivational enhancement and cognitive behavior therapy (once daily); and e) self-help group meeting and psychosocial counseling participation (x1–3/week with a maximum of 1 message per day). Patients received approximately 6–8 messages per day during induction/stabilization and was reduced to 3–5 messages daily during maintenance. Archived messages in each category were then adapted and approximated for delivery during induction, stabilization, and maintenance phases of treatment to address potential clinical queries identified in our prior feasibility studies.21,22Texts primarily comprised of informative or supportive content (90%) with less than 10% of texts prompting participants regarding perceived cravings or withdrawal symptoms using numerical responses (“reply 1 if yes, reply 2 if no”) that generated automated texts reinforcing home induction instructions and to contact their provider if symptoms persisted. Message content, scheduled delivery, and operational sequences were programmed in a HIPAA-compliant texting software (Apptoto©) and allowed the study team to track participant use (i.e., replies to message queries) over time. Once daily texts were sent during pre-scheduled times either at 9:00am or 12:00pm. Texts sent during induction were scheduled for delivery at 9:00am, 1:00pm, and 5:00pm. Texts did not relay or solicit for any patient health information. Lastly, participants could not “push” or elicit responses from the texting software to address any clinical or administrative queries unless they received a text that offered response choices. Instances in which participants responded to text queries, a subsequent text response directed participants to clinic phone numbers to speak directly with clinic or administrative staff.
TAU participants were offered routine clinical care consisting of pro-active contact with clinical staff and regular visits with buprenorphine prescribers. Although participants randomized to receive TeMeS were informed that the intervention was automated lacking input from clinical staff, the study team checked the software daily to report any participant issues to clinic staff.
Measures
Baseline demographic (e.g., gender, self-reported race/ethnicity, age, housing, and insurance status) and clinical characteristics (self-reported past 30-day opioid and drug/alcohol use, any medical and/or psychiatric history) as well as prescription records at 8-weeks were captured using the electronic health records. Electronic medical record abstraction of baseline demographic and clinical characteristics were based on review forms refined in prior pilot studies that identified active diagnosis for any medical and/or psychiatric conditions.21,22 The 8-item, open-ended, participant feedback survey was administered at 8-weeks among a convenience sample of patient participants that were contacted in chronological order based on their study enrollment date to capture perceptions and experiences with the texting tool based on the Technology Acceptance Model.23 More specifically, participants were probed to share details regarding: 1) the ease of use in accessing content disseminated in delivered texts, navigating two-way response trees (e.g., queries regarding cravings and withdrawal symptoms that triggered personalized home-induction buprenorphine dosing instructions), multimedia and web-content hyperlinks shared in texts pertaining to OUD, HIV, and HCV informational, prevention and treatment resources; 2) perceived usefulness of the tool to improve adherence to buprenorphine dosing instructions, patient-physician communication, and self-management?; 3) factors that would increase or limit patients’ intention to use the tool over time following induction on buprenorphine, and factors that influenced the respondent to utilize the tool during enrollment; 4) perceived enjoyment and/or annoyance with this tool, particularly with the frequency and timing of scheduled message queries, and if they could recall any “disappointing” experiences utilizing the tool; and 5) an open-ended item encouraging respondents to delve further regarding the aforementioned domains and identify approaches to improving the intervention.
Recruitment extended to a sub-sample of n=18 participants to achieve data saturation. Process measures were gathered from the TeMeS software and included the frequency of participant replies to scheduled message queries and requests to discontinue receipt of messages. Clinical outcomes were based on a structured audit of the electronic medical records and consisted of retention in treatment at week 8 defined as an active buprenorphine prescription within 7 days following enrollment at 8 weeks.
Analytic plan
We estimated that a sample size of 128 participants would provide 80% power to detect a 20% between-group difference in retention at 8 weeks and alpha of 10%. The sample size was aligned with prior texting pilot randomized controlled trials (RCTs), proposed standards for text messaging studies, and sized for feasibility testing and ‘de-bugging’ of the texting prototype16,21. The intended target sample size of N=128 would allow for initial effect size estimates to support the design of an adequately powered efficacy study in the future. Analysis relied on descriptive estimates of baseline participant characteristics and follow-up survey responses, and frequencies/percentages to describe feasibility measures (e.g., retention, participant responses to text queries).
RESULTS
A total of 128 participants were randomly assigned to receive TeMeS (n=64) or treatment as usual (n=64). Nearly all eligible patients consented to enroll in the study (90.8%) and few were excluded due to lack of mobile phone ownership (n=27, 14.6%) or limited interest in the intervention (n=13, 9.2%; see CONSORT flow diagram Figure 1).
Figure 1.
CONSORT flow diagram
Most participants were male (n=108, 84.4%), aged 44.7 years (range 22, 73), and Medicaid-insured (n=94, 73.4%). Approximately 34.4% of participants were unstably housed (n=41, 32%) and reported prior criminal justice involvement (n=41, 32%; see Table 1).
Table 1.
Demographic and clinical characteristics of PWUO randomized to TeMeS vs. TAU in a telebuprenorphine program during COVID-19 (N=128)
| Measures | Distribution, n (%) | ||
|---|---|---|---|
| Intervention group (n=64) | Treatment as usual (n=64) | Overall (=128) | |
| Age (range) | 42.4 (22, 73) ±11.5 | 46.8 (23, 69) ±11.6 | 44.7 (22, 73) ±11.7 |
| Gender | |||
| Male | 50 (78.1%) | 58 (90.6%) | 108 (84.4%) |
| Female | 14 (21.9%) | 6 (9.4%) | 20 (15.6%) |
| Race-ethnicity | |||
| African-American | 15 (23.4%) | 16 (25.0%) | 31 (24.2%) |
| Hispanic/Latinx | 28 (43.8%) | 26 (40.6%) | 53 (41.4%) |
| White | 21 (32.8%) | 22 (34.4%) | 43 (33.6%) |
| Insurance status | |||
| Medicaid | 50 (78.1%) | 44 (68.8%) | 94 (73.4%) |
| Medicare | 6 (9.4%) | 3 (4.7%) | 9 (7.0%) |
| Privately-insured | 6 (9.4%) | 6 (9.4%) | 12 (9.4%) |
| Uninsured | 2 (3.1%) | 11 (17.2%) | 13 (10.2%) |
| Housing status | |||
| Own apartment/housing | 43 (67.2%) | 31 (48.4%) | 74 (57.8%) |
| Shelter/street homeless | 16 (25.0%) | 25 (39.1%) | 41 (32.0%) |
| Temporarily residing with friend/family | 5 (7.8%) | 8 (12.5%) | 13 (10.2%) |
| Past criminal justice involvement | 16 (25.0%) | 25 (39.1%) | 41 (32.0%) |
| Clinical characteristics | |||
| Psychiatric comorbidities | 32 (50.0%) | 33 (51.6%) | 65 (50.8%) |
| Medical comorbidities | 38 (59.4%) | 34 (53.1%) | 72 (56.3%) |
| Past 30-day substance use (self-reported) | |||
| Injection drug use | 16 (25.0%) | 21 (32.8%) | 37 (28.9%) |
| Heroin | 55 (85.9%) | 57 (89.1%) | 112 (87.5%) |
| Prescription opioids | 27 (42.2%) | 27 (42.2%) | 54 (42.2%) |
| Alcohol | 21 (32.8%) | 24 (37.5%) | 45 (35.2%) |
| Crack/cocaine | 10 (15.6%) | 11 (17.2%) | 21 (16.4%) |
| Cocaine | 26 (40.6%) | 24 (37.5%) | 50 (39.1%) |
| Benzodiazepine | 13 (20.3%) | 13 (20.3%) | 26 (20.3%) |
| Cannabis | 33 (51.6%) | 31 (48.4%) | 64 (50.0%) |
| Tobacco | 49 (76.6%) | 46 (71.9%) | 95 (74.2%) |
The intervention software sent 4613 text messages and archived 362 participant responses. Requests to discontinue receipt of text messages (n=6, 9.4%) was attributed to excessive message frequency, perceived lack of relevancy, and reduced interest in the intervention.
At 8-weeks, a sub-sample of 18 participants completed individual interviews that assessed perceptions and experiences with the texting tool components based on the Technology Acceptance Model. Positive feedback included the usefulness of the tool in linking with self-help groups, reducing illicit opioid use, mitigating cravings, and improving adherence to scheduled telebuprenorphine visits. Most respondents were satisfied with the frequency of software-generated messages (14/18, 77.8%) and half of respondents shared text content with peers (9/18, 50%). Suggestions for improving the intervention included improving access to mental health services, clinic locations offering in-person support in the event of any adverse events (e.g., persistent withdrawal symptoms), reinforcing between-visit support post-induction in the event of worsening cravings or illicit opioid reuse, linking patients with health insurance or social services, and using multimedia content to enhance patient education (e.g., diagrams, video links). There were no perceived issues with privacy, intrusiveness, or ease of use.
Clinical outcomes
At 8-weeks, retention in the Virtual Buprenorphine Clinic among the 128 participants receiving the texting intervention (mean=5.23 weeks, SD=3.41) was comparable to TAU (mean=4.98 weeks, SD=3.34). Retention did not differ between the texting and TAU groups (p= 0.676 at week 8).
DISCUSSION
Findings from this pilot randomized controlled trial study in a low threshold tele-buprenorphine bridge program demonstrate the feasibility of integrating text messaging. The study was sized for feasibility testing and effect size estimates to inform a future efficacy study and was not intended to provide a definitive test of intervention efficacy on 8-week treatment retention. However, design modifications that may enhance the positive effect of future iterations of the intervention on retention include personalizing message content, frequency, and delivery schedules using baseline patient demographic and clinical characteristics (e.g., age, injection drug use, poly-substance use) and post-induction queries for cravings and withdrawal symptoms that are predictive of illicit opioid reuse risk16,24. Combining automated text messaging with human contact (e.g., clinicians, peers) may also enhance intervention efficacy by streamlining real-time troubleshooting to address unanticipated clinical and administrative needs during regular business hours to reduce illicit opioid reuse and loss to follow-up16. For instance, 8-week follow-up interviews and analysis of archived text messages revealed participant preferences for real-time staff contact to promptly address unanticipated administrative (e.g., prescription errors, prior authorization requests, appointment rescheduling) and clinical issues (persistent cravings, withdrawal symptoms) that place patients at elevated risk of opioid reuse and loss-to-follow-up.
Lastly, participants enrolled in the study experienced high percentages of unstable housing (42.2%), criminal justice involvement (32%), and lack of mobile phone ownership (14.6%). The provision of mobile phones, subsidized mobile phone plans, and content addressing social determinants of health (e.g., access to health insurance, emergency housing, food pantries, legal aid) may mitigate barriers to linkage and retention in tele-buprenorphine treatment among vulnerable patient populations.
With the expansion of telebuprenorphine services nationally, remote technologies such as text messaging are uniquely positioned to enhance clinical outcomes25. Text messaging, though in its early stages, offers a practical, low-cost, and effective approach to enhancing chronic disease management. Texting is the most utilized cellular phone feature among patients enrolled in OUD treatment and findings from this study demonstrate its acceptability in real-world clinical settings to foster adherence to federal and consensus practice guidelines to buprenorphine treatment (e.g., patient-provider communication and self-management)13,26.
Limitations
Limitations to this study include the lack of generalizability due to the small sample size of patients enrolled in an urban tele-buprenorphine bridge program and reliance on convenience sampling among participants that had working mobile phones to conduct follow-up interviews at 8 weeks. Restrictions to conducting in-person study visits and lack of sufficient research staff during the early months of the pandemic limited participant contact to telephone calls during enrollment and follow-up study visits. Although the texting software sent and received messages, it is unknown how many texts were actually read, comprehended, and acted upon by participants.
Conclusions
This pilot randomized controlled trial demonstrates the feasibility of rapidly integrating an automated text messaging intervention in a tele-buprenorphine bridge program. Future studies should assess the efficacy of text messaging to enhance linkage and retention among underserved patient populations in tele-buprenorphine settings.
Funding:
The project was sponsored by the National Institute on Drug Abuse (K23DA042140-01A1 and the Bellevue Hospital Department of Psychiatry.
COI/Disclosures:
Dr. Lee has received in-kind study drug for recent and current NIDA-funded trials from Alkermes Inc and Indivior PLC. Dr. Lee has received a recent investigator sponsored study grant from Indivior PLC.
References:
- 1.Becker WC, Fiellin DA. When epidemics collide: coronavirus disease 2019 (COVID-19) and the opioid crisis. American College of Physicians; 2020. p. 59–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Nunes EV, Levin FR, Reilly MP, El-Bassel N. Medication treatment for opioid use disorder in the age of COVID-19: Can new regulations modify the opioid cascade? Journal of substance abuse treatment. 2021;122:108196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lin LA, Fortney JC, Bohnert AS, Coughlin LN, Zhang L, Piette JD. Comparing telemedicine to in-person buprenorphine treatment in US veterans with opioid use disorder. Journal of Substance Abuse Treatment. 2022;133:108492. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lin LA, Casteel D, Shigekawa E, Weyrich MS, Roby DH, McMenamin SB. Telemedicine-delivered treatment interventions for substance use disorders: A systematic review. Journal of substance abuse treatment. 2019;101:38–49. [DOI] [PubMed] [Google Scholar]
- 5.Lee JD, Grossman E, DiRocco D, Gourevitch MN. Home buprenorphine/naloxone induction in primary care. Journal of General Internal Medicine. 2009;24(2):226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tofighi B, McNeely J, Walzer D, et al. A telemedicine buprenorphine clinic to serve New York City: initial evaluation of the NYC public hospital system’s initiative to expand treatment access during the COVID-19 pandemic. Journal of Addiction Medicine. 2022;16(1):e40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Uscher-Pines L, Sousa J, Raja P, Mehrotra A, Barnett M, Huskamp HA. Treatment of opioid use disorder during COVID-19: Experiences of clinicians transitioning to telemedicine. Journal of Substance Abuse Treatment. 2020;118:108124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Eibl JK, Gauthier G, Pellegrini D, et al. The effectiveness of telemedicine-delivered opioid agonist therapy in a supervised clinical setting. Drug and Alcohol Dependence. 2017;176:133–138. [DOI] [PubMed] [Google Scholar]
- 9.Brunet N, Moore DT, Lendvai Wischik D, Mattocks KM, Rosen MI. Increasing buprenorphine access for veterans with opioid use disorder in rural clinics using telemedicine. Substance Abuse. 2022;43(1):39–46. [DOI] [PubMed] [Google Scholar]
- 10.Weintraub E, Greenblatt AD, Chang J, Himelhoch S, Welsh C. Expanding access to buprenorphine treatment in rural areas with the use of telemedicine. The American journal on addictions. 2018;27(8):612–617. [DOI] [PubMed] [Google Scholar]
- 11.Zheng W, Nickasch M, Lander L, et al. Treatment outcome comparison between telepsychiatry and face-to-face buprenorphine medication-assisted treatment (MAT) for opioid use disorder: A 2-year retrospective data analysis. Journal of addiction medicine. 2017;11(2):138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Tofighi B, Grossman E, Buirkle E, McNeely J, Gourevitch M, Lee JD. Mobile phone use patterns and preferences in safety net office-based buprenorphine patients. Journal of addiction medicine. 2015;9(3):217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tofighi B, Leonard N, Greco P, Hadavand A, Acosta MC, Lee JD. Technology use patterns among patients enrolled in inpatient detoxification treatment. Journal of addiction medicine. 2019;13(4):279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ashford RD, Lynch K, Curtis B. Technology and social media use among patients enrolled in outpatient addiction treatment programs: cross-sectional survey study. Journal of medical Internet research. 2018;20(3):e84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Whittaker R, McRobbie H, Bullen C, Rodgers A, Gu Y. Mobile phone‐based interventions for smoking cessation. Cochrane database of systematic reviews. 2016;(4) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Tofighi B, Nicholson JM, McNeely J, Muench F, Lee JD. Mobile phone messaging for illicit drug and alcohol dependence: a systematic review of the literature. Drug and alcohol review. 2017;36(4):477–491. [DOI] [PubMed] [Google Scholar]
- 17.Tofighi B, Grossman E, Bereket S, D. Lee J. Text message content preferences to improve buprenorphine maintenance treatment in primary care. Journal of Addictive Diseases. 2016;35(2):92–100. [DOI] [PubMed] [Google Scholar]
- 18.Urbaniak G, Plous S. Research Randomizer (Version 4.0)[Computer software]. Retrieved on Jan 06, 2020. 2013.
- 19.Cole-Lewis H, Kershaw T. Text messaging as a tool for behavior change in disease prevention and management. Epidemiologic reviews. 2010;32(1):56–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tofighi B, Williams AR, Chemi C, Suhail-Sindhu S, Dickson V, Lee JD. Patient barriers and facilitators to medications for opioid use disorder in primary care. Substance Use & Misuse. 2019;54(14):2409–2419. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tofighi B, Grazioli F, Bereket S, Grossman E, Aphinyanaphongs Y, Lee JD. Text message reminders for improving patient appointment adherence in an office‐based buprenorphine program: A feasibility study. The American Journal on Addictions. 2017;26(6):581–586. [DOI] [PubMed] [Google Scholar]
- 22.Tofighi B, Grossman E, Bereket S, Aphinyanaphongs Y, Lee JD. Integrating text messaging in a safety-net office-based buprenorphine program: A feasibility study. Drug and Alcohol Dependence. 2015;100(156):e223. [Google Scholar]
- 23.Venkatesh V, Davis FD. A theoretical extension of the technology acceptance model: Four longitudinal field studies. Management science. 2000;46(2):186–204. [Google Scholar]
- 24.Konova AB, Lopez-Guzman S, Urmanche A, et al. Computational markers of risky decision-making for identification of temporal windows of vulnerability to opioid use in a real-world clinical setting. JAMA psychiatry. 2020;77(4):368–377. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Tofighi B, Abrantes A, Stein MD. The role of technology-based interventions for substance use disorders in primary care: a review of the literature. Medical Clinics. 2018;102(4):715–731. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Milward J, Day E, Wadsworth E, Strang J, Lynskey M. Mobile phone ownership, usage and readiness to use by patients in drug treatment. Drug and alcohol dependence. 2015;146:111–115. [DOI] [PubMed] [Google Scholar]