Text Messaging Intervention to Improve HIV Adherence in Rural Drug Users (TEXTR34)

The investigators will develop and test the feasibility and promise of a combined HIV adherence assessment and intervention application for rural drug users using an available, familiar technology whose reach will grow exponentially: text messaging via mobile phones. By 2007, 84% of U.S. residents had mobile phones, with near 100% mobile phone penetration projected by 2013. While technology adoption is often slower in under-served communities, the trend is different with mobile phone technology. African-Americans are using more mobile phone minutes per capita and increasing their use at a higher rate compared with other ethnic groups. This technology has great potential to reduce health disparities. In this project, the investigators will develop and test the feasibility and promise of a text messaging application and system using Ecological Momentary Assessment methods to detect nonadherence and drug use and immediately intervene to improve HIV treatment adherence in drug users living with HIV/AIDS who reside outside major metropolitan areas. This R34 is a Stage 1b/2a project in the Stage Model of Behavioral Therapy Development that will develop novel interventions and methods, and generate preliminary estimates of effect sizes that will determine whether a larger clinical trial with extended follow-up and cost-effectiveness evaluation is warranted. The specific aims of this project are: To identify assessment and intervention features relevant to the adherence barriers and drug use patterns of rural and non-urban HIV+ drug users using formative methods including: structured interviews and focus groups to identify specific barriers to adherence and engagement in care and needs related to drug craving and drug use that should be addressed by the intervention iterative usability testing of components and drafts of the intervention To create a text messaging mobile phone application and system (Treatment Extension by Texting, Text) to assess and improve HIV treatment adherence and drug use in real time --Text will be built upon a piloted unidirectional personalized text phone application and system, STeM, and will include pre- and post-programming usability testing To test the feasibility and promise of the assessment and intervention tool in a randomized pilot trial of rural HIV+ drug users with detectable viral load (VL) comparing Text to usual care Feasibility: Identify recruitment rates, consent rates, participant flow, completion rates, and variance of key covariates and outcome variables Promise: determine point estimates and the precision of effects for primary and secondary adherence outcomes including pharmacy refills and unannounced pill counts (medication adherence), missed visit percentage (treatment engagement), VL, and drug craving and drug use.

Active illicit drug use, regardless of diagnostic category (i.e., abuse or dependence), along with medication non-adherence, are the problematic behaviors targeted for change in this study. The investigators will develop and test an innovative mobile-phone delivered text messaging intervention designed to reduce drug use and increase HIV medication adherence. In the first phase of our work (quarters 1-2), the investigators will use qualitative, formative methods with user input to develop the assessment and intervention contents, while our technology partner develops the mobile phone application shell with which the investigators will conduct initial pre-coding usability testing. Focus groups will be used to inform the final design of the text messaging application, and participant input will provide important information to help us determine the number and rate of messages. The investigators plan to recruit 1-2 focus groups from three participating clinics and will spend a day to two days in each location to conduct the groups and to conduct individual interviews with participants who prefer a private interview. In the past, the investigators have succeeded in recruiting groups of 5-8 participants by using flyers, word of mouth, and reminders from providers to prompt participants to call our study staff to arrange participation. The investigators also have typically scheduled groups with several weeks' notice, in both daytime and evening hours, to accommodate the schedules of employed and unemployed participants. The investigators plan to provide lunch or snacks, and to compensate participants for their time. The purpose of the focus groups and individual interviews is to develop the content of the phone application, and to address their adherence barriers and drug use patterns, so that the investigators can program messages (queries) that are the most appropriate for this population. Other topics to be covered in focus groups and individual interviews will include preferred rate of messaging, types of messages desired, use of language and terms that participants understand for random and scheduled queries, followed by iterative user testing of the pre-programming shell of the program to determine whether they understand the flow, understand all terms and symbols, etc. The investigators will gather quantitative (count) data as well as qualitative (free response) data, which the investigators will analyze for rates and themes, respectively. Themes will be derived by analysis of transcripts of interviews and focus group sessions, then grouped into related categories to determine common and unusual concerns of participants. The investigators will ensure that the program addresses all common issues, and the investigators will decide on a case by case basis whether and how to incorporate unusual issues. However, unusual issues may be important ones, representing likely crisis intervention topics to plan for. The count data will provide raw numbers in which to anchor the thematic data. Given that people who text may send and receive 50 or more messages per day routinely, it does not seem overly burdensome to request that they respond to the system 5 times per day, but this issue will be addressed directly in the focus groups. This formative work will be used to generate appropriate content and structure for the intervention that best fits the needs of this unique rural, HIV+, drug using population. The revision of our existing text messaging tool, Stem, will take approximately 6 months, proceeding iteratively. The investigators will iteratively test the acceptability and usability of components of the tool with the target population before finalizing its programming by conducting guided sessions with participants who will review the program via screenshots. The resulting product will be the novel mobile phone application, Text, which the investigators will test with users on the phones in their homes for a 2 week period in quarter 2 to identify any remaining programming needs to finalize the system for the pilot trial. There are three distinct types of messaging events planned Random queries will be sent at a rate of 4 per day during normal waking hours for that participant, as recommended by the leading Ecological Momentary Assessment (EMA) researchers. The second type is a medication dose reminder, which for most patients will occur once or twice per day with current ART medication regimens. The third type is an event-contingent message, which the PARTICIPANT initiates, for events like drug craving, drug use, etc. Therefore, most participants will receive 5 or maybe 6 messages per day from the system, and may send 1-2 messages confirming taking their medications, or may send us more messages if they report a critical event. The current Stem system is programmed on a web-based application. A staff member interviews the participant to determine which kinds of messages they want to have sent with emphasis on personal meaning and concerns for confidentiality. The staff member then types the message into the web-based program while the patient is present. The investigators have an archive of previously used personal messages that new participants could review and reuse or create their own messages. For the enhanced TEXT application proposed here, the investigators will use the same procedure to develop messages. Staff members will also assist the participants in programming the cell phone with appropriate numbers and reviewing its relevant features. Therefore, staff members will do all of the programming, with the participant's guidance about what messages will be helpful to them. When literacy is an issue, the investigators will use a combination of simple textable icons along with a few words to assist the participant in understanding the message. The finalized intervention will be compared to usual care in a small pilot randomized trial that will demonstrate the feasibility of the approach, and yield evidence of promise (estimates of effect sizes) as recommended by recent publications on conducting pilot work. In this phase, the investigators will test the new Text intervention in the field with real participants who will be assessed at baseline, then randomized to receive the phones and Text intervention or usual care over a 12 week period (quarters 3-6). The investigators will assess participants again at 3 and 6 months as a post-treatment follow-up, and complete data cleaning, analyses, and reporting during the final months of the project (quarters 5-8). The investigators plan to begin enrollment of participants in the pilot trial in quarter 3, enrolling 8-12 participants per month until the investigators have enrolled 70 people, which will occur during quarter 5 or 6. This will permit a 6 month follow-up of all participants. Hypotheses: Based on the rates of change the investigators saw in the two pilot studies, the investigators expect that: 1).The intervention will result in a 20% decrease in Missed Visit Proportion (MVP) compared to usual care. 2).The intervention will result in a 20% increase in adherence by pharmacy refill data compared to usual care. 3).The intervention will result in a 25% decrease in drug use over 30 days compared to usual care. Lastly, while not a formal hypothesis, the investigators expect that the intervention will be feasible and acceptable for participants, and the investigators will measure this using an exit interview.

Participants randomized to this arm will receive a mobile phone if needed, and the 12 week personalized text messaging/EMA intervention

Participants randomized to this arm will receive treatment as usual in the ID clinics and in the communities for nonadherence and drug use problems

Personalized, tailored text messaging to prompt medication adherence, with EMA random queries about nonadherence and drug use, and two-way messaging for participants to report critical events

usual care in one of the three participating ID clinics, along with usual care (if any) in the community for problematic drug and alcohol use

The primary outcome measure for ADHERENCE is missed visit proportion (MVP) across a 6 month follow-up period, used to derive the sample size for this stage 1 RCT. The primary outcome measure for DRUG USE is a Timeline Follow-back to gather 90 day retrospective reports of daily drug use.

Secondary outcome measures for adherence will be unannounced telephone pill counts and pharmacy refill data; these will be assessed and used to plan a larger, more definitive study, if the intervention shows promise. Pharmacy compliance can be measured directly in the large subset of patients who fill their prescription through the Virginia ADAP Program. Unannounced pill counts conducted by phone can be assessed for all participants and used to measure medication adherence.

We will have daily data from participants' text messages about their adherence behavior. Comparing texts sent by participants about adherence to TLFB will allow us to test the level of agreement about adherence by two distinct measures, one (TLFB) with known properties. Differences in the magnitude and direction of the effect of the intervention for different outcomes will help us choose an outcome for the larger trial, and may shed light on the particular aspects of behavior that the text messaging tool can influence.

Inclusion Criteria: 18 or older has a detectable viral load reports use of at least one illicit drug in last 30 days requires ART able to receive mobile phone service at home address Exclusion Criteria: lack of English proficiency planning to move out of area in the next six months inability to provide consent. Note regarding "inability to provide consent": Traditionally, some patients with HIV whom we evaluate for potential participation in research studies are not able to provide informed consent, either due to significant cognitive decline, or their designation of someone else as their legal proxy. When this occurs, we do not offer them study participation because legally, they are unable to consent, even if currently, their cognitive capacities seem intact. We have generally assessed this through a series of branching questions to determine their legal and cognitive capacity for consent.

Ingersoll K, Dillingham R, Reynolds G, Hettema J, Freeman J, Hosseinbor S, Winstead-Derlega C. Development of a personalized bidirectional text messaging tool for HIV adherence assessment and intervention among substance abusers. J Subst Abuse Treat. 2014 Jan;46(1):66-73. doi: 10.1016/j.jsat.2013.08.002. Epub 2013 Sep 9.

Ingersoll KS, Dillingham RA, Hettema JE, Conaway M, Freeman J, Reynolds G, Hosseinbor S. Pilot RCT of bidirectional text messaging for ART adherence among nonurban substance users with HIV. Health Psychol. 2015 Dec;34S(0):1305-15. doi: 10.1037/hea0000295.