Patient Feedback on a Mobile Medication Adherence App for Buprenorphine and Naloxone: Closed and Open-Ended Survey on Feasibility and Acceptability

Background Opioid use disorders impact the health and well-being of millions of Americans. Buprenorphine and naloxone (BUP and NAL) can reduce opioid overdose deaths, decrease misuse, and improve quality of life. Unfortunately, poor medication adherence is a primary barrier to the long-term efficacy of BUP and NAL. Objective We aimed to examine patient feedback on current and potential features of a Bluetooth-enabled pill bottle cap and associated mobile app for patients prescribed BUP and NAL for an opioid use disorder, and to solicit recommendations for improvement to effectively and appropriately tailor the technology for people in treatment for opioid use disorder. Methods A convenience sample of patients at an opioid use disorder outpatient clinic were asked about medication adherence, opioid cravings, experience with technology, motivation for treatment, and their existent support system through a brief e-survey. Patients also provided detailed feedback on current features and features being considered for inclusion in a technology designed to increase medication adherence (eg, inclusion of a personal motivational factor, craving and stress tracking, incentives, and web-based coaching). Participants were asked to provide suggestions for improvement and considerations specifically applicable to people in treatment for opioid use disorder with BUP and NAL. Results Twenty people with an opioid use disorder who were prescribed BUP and NAL participated (mean age 34, SD 8.67 years; 65% female; 80% White). Participants selected the most useful, second-most useful, and least useful features presented; 42.1% of them indicated that motivational reminders would be most useful, followed by craving and stress tracking (26.3%) and web-based support forums (21.1%). Every participant indicated that they had at least 1 strong motivating factor for staying in treatment, and half (n=10) indicated children as that factor. All participants indicated that they had, at some point in their lives, the most extreme craving a person could have; however, 42.1% indicated that they had no cravings in the last month. Most respondents (73.7%) stated that tracking cravings would be helpful. Most respondents (84.2%) also indicated that they believed reinforcers or prizes would help them achieve their treatment goals. Additionally, 94.7% of respondents approved of adherence tracking to accommodate this feature using smart packaging, and 78.9% of them approved of selfie videos of them taking their medication. Conclusions Engaging patients taking treatment for opioid use disorder with BUP and NAL allowed us to identify preferences and considerations that are unique to this treatment area. As the technology developer of the pill cap and associated mobile app is able to take into consideration or integrate these preferences and suggestions, the smart cap and associated mobile app will become tailored to this population and more useful for them, which may encourage patient use of the smart cap and associated mobile app.


Background
In 2019, a total of 10.1 million people aged 12 years and older reported misusing opioids, with 1.6 million of those meeting criteria for an opioid use disorder [1]. Intravenous opioid use increases the risk of contracting blood-borne illnesses such as hepatitis B, hepatitis C, and HIV [2]. Buprenorphine and naloxone (BUP and NAL) can reduce opioid overdose deaths by at least 70%, decrease opioid misuse, and improve quality of life [3][4][5][6]. Because of its demonstrated efficacy, the National Institute on Drug Abuse, the Substance Abuse and Mental Health Services Administration, the American Society of Addiction Medicine, and the American Academy of Addiction Psychiatry endorse the use of medications for opioid use disorder in the treatment of opioid use disorder (eg, BUP and NAL and methadone).

Medication Adherence
One of the most substantial barriers to the long-term efficacy of BUP and NAL and poor medication adherence is consistently associated with relapse and overdose risk, as well as several other negative outcomes for patients [7,8]. In a study of 703 patients examining the effect of BUP adherence on opioid relapse, patients who were nonadherent to their prescribed BUP were over 10 times more likely to relapse than patients who were adherent [9]. Poor medication adherence can occur for myriad reasons, as summarized by the World Health Organization's Five Dimensions of Adherence: social and economic, health care system, condition-related, therapy-related, and patient-related [10]. At the patient level, age, perception of treatment, co-occurring psychiatric conditions, and environment (eg, living with other substance users or at a distance from treatment centers) are common causes for treatment discontinuation [8,[11][12][13]. For example, a previous study found that 73% of patients in treatment for an opioid use disorder, with co-occurring symptoms of depression, maintained abstinence at 6 months of treatment. Of those who had combined symptoms of depression and anxiety, only 40% were abstinent in the same time period [14]. Anxiety is a predictive factor for additional psychopathologies, such as insomnia and sleep deprivation, which can, in turn, manifest into symptoms of anxiety and depression [15]. Additionally, a poor therapeutic alliance between patient and provider and lack of prescription monitoring may contribute to patient nonadherence, in part due to limited tools at the level of the prescribing clinician [16].
Nonadherence with BUP and NAL is frequently associated with an increased risk of diversion [17][18][19][20][21]. Given that BUP and NAL is a schedule III opioid medication, improving the management of adherence is critical to decreasing relapse, diversion, and overdose.

Pillsy Technology
Pillsy, which has successfully and significantly increased adherence in other treatment areas, acts like a digital medication coach, providing reminders using a mobile app, text messages, and automated phone calls [22]. The platform automatically tracks doses and timing and sends intelligent reminders to create a unique feedback loop, which allows the developers to constantly optimize incentive or reminder messages to meet user needs and increase adherence. The Pillsy technology was developed to be a simple, inexpensive adjunct to standard treatment. Unlike some other technologies, such as the MySafeRx system [23], the Pillsy system does not require codes or video calls for the patient to access their medication. Still, other technologies, such as reSET-O by Pear Therapeutics, have integrated an intensive behavioral intervention into their platform [24]. Pillsy was designed intentionally without these requirements in an effort not to increase the patient burden, and to begin work with people with opioid use disorder using a simple platform that could be adjusted to specifically include features that were requested by patients to increase the practical usability of the product.
Considering the evidence that tracking medication intakes and maintaining a high motivation with prompts are effective in the intervention phase of adherence up to persistence and considering that Pillsy only nominally increases the cost of oral BUP and NAL treatment and providers can bill for monitoring time (CPT code 99091), it is a potentially attractive solution to patients, providers, and payers. However, there are no studies available that have systematically investigated this technology's impact on medication adherence in this population. The device is under investigation in a small (ClinicalTrials.gov NCT04656899, n=40, 12 weeks) clinical trial that recently completed data collection.

This Study
The intent of this investigation was to examine patient feedback regarding the acceptability and feasibility of an inexpensive Bluetooth-enabled pill bottle cap and associated mobile app (ie, Pillsy, an Optimize Health product) designed to monitor and increase BUP and NAL adherence with the assistance of a designated "helper" (eg, friend and family member). We gathered feedback from patients in BUP and NAL outpatient treatment on existing and potential features of the technology as well to solicit recommendations for improvement in an effort to tailor the technology most effectively and appropriately for people in opioid use disorder treatment.

Methods
This study was conducted in the US states of Idaho and Washington.

Participants
Inclusion criteria were being 18 years of age or older and receiving treatment for a substance use disorder at Ideal Option outpatient addiction medicine clinics. Patients at this clinic are on opioid substitution therapy. Participants may have begun receiving treatment at this clinic due to iatrogenic opioid dependency or heroin or fentanyl or other illicit opioid use prior to treatment induction.

Ethical Considerations
Study procedures were classified as exempt by the Washington State University institutional review board. Participants received a US $25 Walmart gift card (not usable for tobacco, alcohol, or firearms) for participating.

Study Design
Participants were a convenience sample of people receiving medication-assisted treatment that were recruited through 2 Ideal Option clinics. Ideal Option is specialized in medication-assisted treatment of opioid use disorder. Clinic front desk staff invited patients to participate as they arrived at the clinics for their usual care appointments using a short script describing the study. All participants were invited while a study staff member was there. If a study staff member was not present at the clinic, people were not being invited to participate. All research was conducted in 1 appointment at the respective clinic in a private room. Hours for recruitment varied by day based on the researcher's availability, and all data were gathered by 1 researcher. Recruitment was conducted from May 2021 through June 2021. Interested patients met with research staff at the clinic for an eligibility screening, consenting, and study completion in 1 session. Participants were not required to have a smartphone or computer to participate in this study. Researchers provided a laptop for participants to use to view screenshots of the Pillsy system and complete the survey for the study. Survey questions were developed by the research team. Pilot testing on the items was conducted by members of the research lab conducting the study. All surveys were administered by the first author of the manuscript. Participants reported their demographics, medication adherence, diversion, opioid cravings, experience with technology, and support system, and provided suggestions on how to best support people taking BUP and NAL. This information was gathered through a series of closed and open-ended questions administered on the provided laptop, with free text boxes for open-ended questions. Researchers used scripts and screenshots to describe the Pillsy intervention and familiarize patients with it. Potential features under consideration were described verbally. Participants were asked to rate their interest in or opinion on current and potential features of the app, using response options such as "useful, second most useful, and least useful." They were also asked to provide detailed feedback (eg, concerns and suggestions) on examples of product features being considered through open-text areas in the survey. Specific features evaluated included the inclusion of a personal motivational factor identified by the patient, craving tracking, web-based health coach access for the treatment of anxiety or stress, sleep and activity tracking, an online moderated forum, and incentives for optimal medication implementation. Data were collected using the Research Electronic Data Capture, a secure data collection tool. Participation in the study took approximately 30 minutes.

Statistical Analysis
Survey responses were analyzed using descriptive statistics (response frequencies and percentages) to characterize participant behavioral patterns and feedback.

Study Population
A total of 20 participants (N=20) were assessed for eligibility, and all met the inclusion criteria. One participant ended participation after demographics due to a childcare issue that required immediate presence, resulting in n=19 participants for the feedback portion of the study. Participant ages ranged from 20 to 53 (mean 34.15, SD 8.67) years. Participants were primarily White (80%, n=16), female (65%, n=13) and identified as women (65%, n=13). In total, 80% (n=16) of the sample indicated their race as White, and 95% (n=19) indicated non-Hispanic or Latino ethnicity. The mean number of weeks participants had been in treatment at Ideal Option was 1.64 years (SD 86.19 weeks; range 1-261 weeks). The average number of times participants had been in treatment for opioid use disorder was 2.42 (SD 1.95; range 1-7 times).

Medication Adherence, Treatment Motivation, and Diversion
Participants were asked whether they had a strong motivating factor for staying in treatment and what that motivational factor was (see Table 1). Each participant indicated that they had at least 1 strong motivating factor, and 10 indicated children as that factor. When asked how easy or difficult it was to remember to take their medication doses, the majority (79%, n=15) indicated "easy" or "very easy," and 42% (n=8) indicated that they did not miss doses, even occasionally. For those who did occasionally miss doses, the most frequent reason indicated was forgetting (88%, n=7), though people also indicated that they occasionally missed doses due to sleeping through the time when they were supposed to take it, being embarrassed to take it in public, or being in too much of a hurry.

(58) Yes
Have you ever been asked to give your opioid use disorder medication to somebody else? n (%)

(21.1) Yes
Did you provide it to them? (Note: Only asked among those who responded yes to the above question.) n (%)

(50) Yes
Have you ever asked anybody else for some of their opioid use disorder medication? n (%)

(79) Yes
Sleep How many hours per night do you sleep on average? mean (SD) 6.84 (2.46) Hours Do you currently track sleep quality on a device or app on your phone? n (%)

(85) Yes
Would you be interested in using something like that [referring to sleep tracking], if it was available to you? n (%)

(84) Yes
Do you think that understanding your sleep patterns would help you understand if you should be taking your opioid use disorder medication? n (%)

Motivational Reminder Feature Feedback
Participants were asked to rate the helpfulness of a feature that would remind participants of their self-selected motivational factor when it is time to take a dose; 84.2% (n=16) of respondents did not have concerns about this feature. Concerns that were raised by participants included that this may make people feel guilty if they were not reaching their treatment goals and that it may trigger people to use illicit substances if the motivational factor was upsetting. One participant suggested that this may be more helpful for people who were "stronger in their sobriety." Suggestions for improving this feature included diversifying ways to deliver motivation such as adding reinforcing complements when people were doing well, incorporating personalized motivational phrases that users could create, adding a picture of the motivational factor, adding a voice recording of the technology user saying the motivational factor, and including a counseling component such as letting the user know that they are loved and not being judged or bringing in a spiritual component and other positive factors in the person's life.

Technology Use, Craving, Anxiety, and Stress Tracking
All participants had smartphones with unlimited texting and had used Bluetooth connectivity previously. The average length of time they had owned their current smartphone was 16.6 (SD 13.17; range 1-48) months.
Most participants had "a lot" of anxiety or stress in their lives (84.2%, n=16), the majority of which (73.7%, n=14) had received professional treatment for it. All participants were asked if they would be interested in a web-based treatment for anxiety or stress hosted by a real health coach, and 78.9% (n=15) affirmed that this would be helpful. All participants indicated that they had experienced "the most extreme craving a person could have" at the worst point in their life. In response to a question asking participants what their worst cravings were in the past month, 42.1% (n=8) indicated no craving, 57.9% (n=11) indicated that they had some craving, and 0% (n=0) indicated that they had "the most extreme craving a person could have." Most respondents stated that tracking cravings would be helpful (73.7%, n=14) and 68.4% (n=13) did not have concerns with this idea. While there were comments regarding concerns with confidentiality, concerns about tracking cravings primarily focused on triggering a relapse.

Reinforcers
Participants were queried on including reinforcers (ie, "rewards") for consistently taking their opioid use disorder medication. Most participants (84.2%, n=16) believed that reinforcers would help them achieve their treatment goals. When asked about preferences for reinforcers, the majority endorsed a US $25 Walmart card as their top of 3 choices (73.7%, n=14) over a US $25 Amazon card (21.1%, n=4 endorsed) or a month of free cell phone minutes and data (5.3%, n=1 endorsed). After researchers explained that a feature such as this may necessitate adherence tracking (a feature that is currently a primary component of the Pillsy system using pill cap opening data), participants were asked whether they would be open to tracking their doses (94.7%, n=18 responded "yes"), using smart packaging (94.7%, n=18 responded "yes"), and taking a selfie video while taking medication doses (78.9%, n=15 responded "yes").

Feature Preferences
Of the potential features reviewed, participants indicated that motivational reminders would be most useful (42.1%, n=8), followed by craving and stress tracking (26.3%), and web-based support forums (21.1%, n=4; see Table 2 for details). When asked why they selected the features they did, responses for motivational reminders focused on having reminders in general, reminding oneself why they are in treatment, and having reassurance in their choice to continue treatment. Participants who selected craving and stress tracking as a feature preference explained that they would prefer the feature to increase awareness of triggers and make supports aware of a potential need for help.

If you are tracking cravings you can see what is causing it. [Participant 10]
Tracking craving and stress would make people more aware and could cue others to help. [Participant 4] Additional feedback was solicited from participants on how to support people taking BUP and NAL. Although these were all garnered from the same open-ended free-text box, we have organized the feedback into the areas of medical providers and researchers, patients in medications for opioid use disorder programs, and improving treatment or technology. This feedback is listed in Table 3.

Principal Findings
This descriptive study examined feedback from a sample of patients in opioid use disorder treatment to better understand their perceptions and opinions of how a Bluetooth-enabled pill bottle cap and the associated mobile app could be used to improve adherence to opioid use disorder treatment. Poor medication adherence can be the result of discord between patient and provider, lack of resources (availability of transportation to a clinic), and mental health challenges, such as anxiety. Participants indicated that they had strong motivating factors for wanting to stay in treatment. Personalized reminders that target incentives for sobriety, and avoid self-identified triggers for anxiety and cravings, could prove useful in improving the Pillsy technology. Feedback was overwhelmingly positive regarding a feature that would remind them of their self-selected motivational factor when it is time to take a medication dose. A feature providing reinforcers for consistently taking opioid use disorder medications was endorsed at greater than 80%. In addition, over 70% of participants indicated that tracking cravings would be helpful.
The results of this study support modifications to the Pillsy technology that would integrate additional motivational interventions to help promote medication adherence in people with opioid use disorder. First, strong support was present for the integration of personalized motivational factors, follow-up by the research team on suggestions made by participants to include phrases of encouragement and pictures of the family will be considered. Second, the integration of contingency management to provide tangible reinforcers for medication adherence will be considered, particularly, considering the efficacy of contingency management interventions in substance use treatment [25].
Our sample did not suggest additional safeguards such as required video check-ins or unique codes from their provider be added to the technology. This will continue to differentiate the Pillsy technology from other medication access technologies, such as MySafeRx. Our sample also did not request educational materials or behavioral interventions similar to those in the reSET-O technology, therefore continuing to differentiate Pillsy from Pear Therapeutics technology.

Limitations
The study has limitations that should be acknowledged. First, the sample for the study was a small (N=20) convenience sample that was geographically homogeneous (all participants received care within Idaho and Washington states in the United States), primarily White and largely female. Selection bias may be present, as participants were people who had time to engage in the study following their scheduled visit to the treatment clinic. Second, while the participants were given descriptions of how the technology being investigated functioned, they did not have the opportunity to interact with it.

Conclusions
Providing treatment adherence promotion through a digital Bluetooth-based smart pill cap and associated mobile app, like Pillsy, maybe a way to increase adherence and improve treatment outcomes for people in treatment for opioid use disorder using BUP and NAL. Including patient feedback and preferences in the development process is integral to creating a user-friendly, meaningful intervention that patients will use and find helpful. The valuable feedback provided through this study will allow the technology developers to modify the app based on this feedback, with the intention of creating a more patient-centered product.

Data Availability
Data are available upon request, pending approval of the study's primary investigator and external funding sponsor for the project.

Conflicts of Interest
SMM reports that Optimize Health, in the form of a sub-contract to his university, Washington State University, which originated from the National Institutes of Health. He is a Multiple Principal Investigator on the grant that funded this work (R44DA049629, Primary Investigators: SMM, JL). SMM has also received research funding from Orthopedic Specialty Institute, the US Department of Justice and consulted for Consistent Care company. Aside from the noted R44 grant, no other funding or consulting is related to the work reported here. JL is an owner of Optimize Health, the recipient of the NIH award for this study. He also developed the Pillsy app being evaluated. No other authors have disclosures or competing interests.