Pre-Written Text Messages
The formative research activities were instrumental in developing specific text messages that were deemed culturally appropriate for this target population. The text messages were developed within the theoretical framework of Social Support Theory, Health Belief Model and Social Cognitive Theory. According to Social Support Theory, social support encompasses instrumental, emotional, and informational assistance provided by members of one’s social network [32
]. Health Belief Model asserts that individuals’ beliefs regarding threats to their health and their beliefs that specific health behaviors can reduce these threats predict their likelihood of engaging in protective health behaviors [34
]. Social Cognitive Theory posits interactive causal relationships among personal determinants, behavior and environmental influences [36
]. Effective HIV prevention interventions applying this theory increases individuals’ knowledge and awareness of their health risks, guides them in developing self-regulation skills, offers practice and feedback opportunities, and engages social support resources to maintain prevention behavior.
Focus group and pre-test participants were especially helpful in pointing out specific nuances in the messages that would make them “speak” to the population, keep participants engaged in a text-messaging dialogue, and provide health education, risk reduction and/or social support. At the conclusion of the focus groups and intervention pre-test, 400 pre-written text messages were developed and categorized accordingly: Welcome = 14 messages; Social Support/Informational Support = 70 messages; Social Support/Emotional Support = 42 messages; Social Support/Instrumental Support = 18 messages; Health Belief Model/Health Threat = 50 messages; Health Belief Model/Health Behaviors to Reduce Threat = 65 messages; Social Cognitive Theory/Awareness of Health Risks = 52 messages; Social Cognitive Theory/Self-regulation Skills = 44 messages; Social Cognitive Theory/Self-efficacy = 28 messages; Post Intervention Follow-up = 17 messages (see Table for sample messages).
Sample of Pre-written Text Message by Behavioral Change Theory
Although some pre-written messages could reflect more than one behavioral change theory, messages were categorized under the dominant theory. For example, in the message, “Get tested 2nite on the Blvd. We’re here 4 U” both Informational Support (Get tested 2nite on the Blvd.) and Emotional Support (We’re here 4 U) are provided. However, in this instance, the main objective is to convey information regarding a testing site and, therefore, the message is categorized under Informational Support. Constructing messages within the guidelines of behavioral chance theory enabled the intervention to be guided by evidence-based concepts and, when the full IT intervention is completed, will provide valuable outcome findings on the acceptability of the different theoretically based messages. These messages were categorized by participant profile/behavioral characteristics (e.g., HIV positive, injector) and inputted into an Excel document before the intervention so the research staff could easily cut and paste an appropriate message into instant messaging text boxes on a computer for instant delivery to the text-messaging device.
The research staff were trained on how to identify and deliver the appropriate pre-written messages as well as how to respond and provide social support and health education messages for any related requests for information received from a participant (e.g., hepatitis information, where to receive free dental care, the location of a food line or soup kitchen). Research staff were instructed to memorize the pre-written texts so these messages could be seamlessly incorporated into text conversations.
To measure and assess behavioral change among the participants, the pre-written text messages were categorized according to key constructs in behavioral change theory (i.e., social support theory, health belief model, social cognitive theory) to provide education, social support, and referrals for healthier, prosocial changes regarding methamphetamine- and sexual-risk behaviors. In addition, text messages were coded by prospective participants’ profiles such as HIV status and risk activities (e.g., online hook-up or patron of commercial sex venues; receptive, insertive or versatile sexual partner) to facilitate appropriate messaging. Therefore, the IT intervention assesses certain factors of behavioral change such as perceived benefits of engaging in protective health behaviors, and is specifically designed to address barriers to practicing such behaviors, and enhancing self-efficacy. Finally, a complete message log was developed, which will tally the number and type of messages used in the intervention to facilitate a comprehensive content analysis of the text conversations in the data analysis phase of this project.1
During the formative stage of the study, the text-messaging devices changed three times. Given that cellular telephones and cellular phone plans vary tremendously it was determined that in order to ensure uniformity in text-messaging capabilities and messaging service and to ensure that each phone was password protected, each participant would receive a study-provided device, which he could keep after his participation was completed.
The Motorola T900 two-way paging device originally considered for the intervention was the most economical option at $20 per unit; however, testing of the device showed multiple deficiencies. The first three pre-test participants indicated that the pager was unacceptable and, particularly given the rapid change in innovative smart-phone technology, they would not be willing to carry this device. A pager also had the following weaknesses as a platform for an IT intervention: 1) The menus were not intuitive, were difficult to understand and involved many steps to access the screens to enter and delete messages; 2) Message transmission was slow, with some delays up to three hours; 3) There were no rechargeable batteries, therefore, disposable batteries were provided by the study and needed to be replaced often, approximately every two days; 4) There was very little memory in the device, which required participants to constantly erase incoming and outgoing messages; and 5) the device was bulky and outdated.
A more advanced device, The first generation T-Mobile Sidekick, at $40 per unit, included a flip color screen display, full keyboard, easy to use menus, rechargeable battery, and a fast Internet connection. Still, this device was an early generation product and also showed deficiencies. The device had limited memory capacity, and messages needed to be deleted regularly. The battery life was limited and participants were required to recharge the device nightly. In addition, each device had to be set-up separately with a Subscriber Identity Module (SIM) card and loaded with referrals for participants, with texting service, but no phone service, and without capability to download games, ringtones, or other extras. A few devices had some hardware issues such as loose screens, broken side doors and bay covers, and additional signs of wear and tear such as paint rubbed off the individual keyboard letters.
Finally, a slightly newer generation of the T-Mobile Sidekick that addressed the hardware and internal battery issues were purchased, at $85 per unit. There were no hardware, software, or battery issues observed with these devices.
Software to Manage Message Transmission
The selection of a messaging system software to manage text transmission between the study staff and participants proved to be a challenge. The original system was deemed to be inadequate during the pre-test intervention as the software resided on one computer and all text messages were sent through email messages only. The system was originally chosen as it had the capability of programming the pre-written messages into pull-down menus. However, the system was lacking in its ability to track and tally messages, ensure instant messaging (there was significant delay between transmission and receipt), and allow for multiple staff to manage messaging to and from participants.
Following the pre-test intervention an online system was selected. This system was accessible from any computer with an Internet connection, was economical, allowed for the Study Coordinator to login to the system at any time and monitor conversations, and could easily manage simultaneous conversations from multiple participants at one time.2
Finally, the online instant messaging system (AIM) offered by America Online was selected. This system was compatible with the text-messaging devices, was offered for free, could be set-up on any computer with an Internet connection, offered good logging of messages with exact timestamp, and provided an easy instant messaging window that was acceptable and familiar to the staff. However, the staff were still required to cut and paste the pre-written texts and keep a daily tally of which messages were transmitted in an Excel document.
Service lines were purchased separately, each with its own one-year contract; however, the study had to use devices that were compatible with that service only. The study purchased 35 lines at $30 each per month and chose a “data only” service contract with unlimited text messaging and Internet service, but without a “voice” plan or phone minutes.3
Additional procedural refinements were made following the intervention pre-test. Given the learning curve for creating texts, responding to texts, and regularly erasing messages a one-page fact sheet with instructions was designed to give to study participants with their device. The pre-test participants also helped to determine the optimal days and times during which the staff would be online to communicate with participants. Finally, it was determined that, to adequately participate in the intervention, all participants needed to have the capability to charge their device daily for eight consecutive hours.
Pre-test participants would either initiate contact or respond to research staff text messages. To maintain consistency between text-messaging correspondences, the upper limit on each “text messaging conversations” was 20 messages sent per party (i.e., 20 messages from the research staff to the participant and 20 messages from the participant to the research staff), or 40 messages total per conversation. The research staff engaged in a maximum of four “text messaging conversations” per day, per participant. Several markers of message acceptability during the pre-test were observed. These included the fact that pre-test participants remarked they could not differentiate between the pre-written messages and the extemporaneous messages. Another marker of acceptability came when a participant confirmed, through the text conversation, that the message(s) written were relevant and important to their current situation. For example, when staff wrote, “U can njoy & protect urself at the same time,” and the participants responded, “Decide not to smk T went to a friends 4 the nite thnx 4 makn me think bout it.” A message was also considered acceptable when a participant continued a text conversation that staff initiated; such as staff wrote, “H20 is UR friend, drink it down,” and the participant responded, “no T 2nite just weed n didn’t have sex with my friend, so I’m safe, thnx.”