Are participants concerned about privacy and security when using short message service to report product adherence in a rectal microbicide trial?

Are participants concerned about privacy and security when using short message service to report... Abstract Objective During a Phase 2 rectal microbicide trial, men who have sex with men and transgender women (n = 187) in 4 countries (Peru, South Africa, Thailand, United States) reported product use daily via short message service (SMS). To prevent disclosure of study participation, the SMS system program included privacy and security features. We evaluated participants’ perceptions of privacy while using the system and acceptability of privacy/security features. Materials and Methods To protect privacy, the SMS system: (1) confirmed participant availability before sending the study questions, (2) required a password, and (3) did not reveal product name or study participation. To ensure security, the system reminded participants to lock phone/delete messages. A computer-assisted self-interview (CASI), administered at the final visit, measured burden of privacy and security features and SMS privacy concerns. A subsample of 33 participants underwent an in-depth interview (IDI). Results Based on CASI, 85% had no privacy concerns; only 5% were very concerned. Most were not bothered by the need for a password (73%) or instructions to delete messages (82%). Based on IDI, reasons for low privacy concerns included sending SMS in private or feeling that texting would not draw attention. A few IDI participants found the password unnecessary and more than half did not delete messages. Discussion Most participants were not concerned that the SMS system would compromise their confidentiality. SMS privacy and security features were effective and not burdensome. Conclusion Short ID-related passwords, ambiguous language, and reminders to implement privacy and security-enhancing behaviors are recommended for SMS systems. text messaging, privacy, HIV, LGBT persons BACKGROUND AND SIGNIFICANCE Protecting participants’ privacy and security is key to conducting ethical research. Privacy refers to a participant’s right to control when, how, and to what extent personal information is shared with others. Security refers to protection against unauthorized access to data or devices.1,2 Researchers are obligated to protect participants from potential harm resulting from a privacy or security breach, including disclosure of study participation or personal health information, especially on sensitive topics such as HIV, sexual behavior, sexual orientation, and gender identity. Disclosure of sexual orientation by gay or bisexual men can increase exposure to negative responses and stressors, and men who have same-sex sexual behavior may elect not to disclose it to avoid stigmatizing reactions.3,4 Data collection via short message service (SMS, or text messaging) is convenient but also presents new challenges to protecting participants’ privacy and security. SMS is a useful and practical data collection method that is widely accessible, with 95% of the global population living in areas covered by mobile-cellular networks.5 Messages can be received silently and even when the phone is turned off.6 However, SMS messages leave a record on the phone, which can have relevant implications for the protection of privacy and security,6 particularly in cases where a cell phone is shared by 2 or more people.7,8 Institutional Review Boards (IRBs) at research institutions charged with ensuring the protection of participants’ privacy and security often raise concerns regarding the use of SMS in research. These concerns focus on perceptions that text messages are not secure and could breach participant confidentiality or expose personal health information. Given that only 30% of IRBs have had training regarding new media technology in research,9 board members must take into account a plurality of views from information technology specialists, technology developers, study investigators, ethicists, and legal officials when evaluating the use of SMS in research. In addition, research on participants’ perceptions of privacy and security during SMS-based mHealth interventions should also inform IRB discussions. Participants’ perceptions may differ from researchers’ and health care workers’ expectations. For example, in a study of an SMS reminder intervention for HIV and tuberculosis patients, a majority of health care workers perceived a risk of unintentional disclosure of patients’ health status, while a vast majority of patients did not perceive similar risks.7 Therefore, although participants’ perceptions of privacy may be context-dependent and malleable,10 as end users their perspectives are valuable in informing researchers and ethical boards of the potential for study-related social harms when using SMS in research. Current literature on participants’ views on privacy and security when receiving health information via SMS has demonstrated that concerns exist. In research exploring the use of SMS-based reminders for antiretroviral treatment adherence, some participants reported a fear of unintentional disclosure of health status, with the potential for resulting stigma and discrimination.7,8,11,12 In one such study, participants reported deleting SMS messages immediately after receipt to prevent others from viewing them.12 However, in another study, participants did not consider the reminders intrusive to their privacy and did not report negative consequences to receiving messages, even when others inadvertently received them.13 Finally, for some, social norms around SMS privacy dictate an unspoken etiquette by which people do not read others’ messages.14 To date, most of the studies that have examined privacy in the context of SMS focused on systems that provide reminders or health information but are not interactive, as ours was.7,12,15 We designed an interactive SMS-based adherence reminder and data collection system for MTN-017, an international Phase 2 safety and acceptability study comparing an oral tablet (emtricitabine/tenofovir disoproxil fumarate, or FTC/TDF) and a gel (reduced glycerin tenofovir, or TFV-RG 1% gel) applied rectally for HIV prevention among men who have sex with men and transgender women. To protect participants’ privacy/security and mitigate the potential for related concerns, we implemented several privacy and security design features. We evaluated participants’ perceptions of privacy while using the system and the acceptability of the system’s privacy features. Our report can contribute to the development of future SMS-based data collection systems and inform ethical review of research involving SMS for data collection. MATERIALS AND METHODS Study design The data for this study come from an international Phase 2 safety and acceptability trial comparing 2 HIV-prevention products (TFV-RG 1% gel and FTC/TDF tablets) and 3 study regimens: (1) daily use of the gel applied rectally with an applicator, (2) gel applied before and after receptive anal intercourse (RAI) with an applicator, and (3) daily oral tablet. All participants completed an 8-week study period for each regimen (daily oral tablet, daily rectal gel, RAI rectal gel) in a randomly assigned order, with a 1-week washout period between each study period.16 A secondary aim of the study was to evaluate adherence to product use. Sample Participants included men and transgender women ≥18 years of age, non-HIV infected, in good health, with a history of RAI at least once in the past 3 months. The study took place at 8 sites in 4 countries: Cape Town, South Africa; Bangkok and Chiang Mai, Thailand; Lima, Peru; San Juan, Puerto Rico (US territory); and San Francisco, Pittsburgh, and Boston, continental United States. Participants were recruited from health clinics and community-based organizations. The study was reviewed and approved by the IRBs at all participating institutions, and participants provided written informed consent. Procedures This paper describes a substudy focusing on use of an SMS system for adherence monitoring among participants in the larger study comparing safety and acceptability of 3 HIV prevention regimens. At enrollment, participants were randomized to 1 of 6 study-product sequences and given study product, either a bottle of oral FTC/TDF tablets to take daily or 30 white plastic applicators prefilled with RG-TFV 1% gel to insert rectally either daily or before and after RAI. Participants were given a study-owned mobile phone if they did not have their own and were trained to use the SMS system to report product use daily during the trial. We created alphanumeric passwords, made up of 2-letter abbreviations for each study site plus 1- or 2-digit numbers that corresponded to the participant identification numbers. For example, participant 1 in Cape Town was assigned the password CT1. During the SMS training, participants chose their preferred reminder time, programmed the SMS system mobile number into their mobile phones, and joined the system. Joining the system entailed sending the word “Join,” the participant’s password, and the preferred reminder time (using the 24-h clock for clarity and brevity) to the system (eg, “Join CT1 2200”). After choosing a reminder time and joining the system, participants completed their first SMS session at the clinic under staff supervision. They responded to a reminder by sending their passwords, reported the number of times they used the study product, and then deleted the messages from their phones. They were given an SMS instruction card, which included the SMS system mobile number; their password; and a reminder to lock their phone, protect it with a password, and delete the study text messages to safeguard their privacy. During each study period, participants were asked to respond to daily text messages from the system. They were compensated for 1 completed text message session per day (defined as responding to the reminder with a password and sending the number of times the product was used). Each study site determined a locally appropriate compensation structure, ranging from US$0.15 to US$2.00 per session. Participants were informed that they would receive a bonus at the end of the month, ranging from US$1.00 to US$10.00, for completing at least 6 sessions per week. SMS system description The SMS script consisted of a generic daily reminder, programmed to arrive at the chosen reminder time, requesting that participants send their password to initiate the session. Upon sending the password, participants received a text message asking them to report the number of times they had used the product since their prior report. After sending a number, participants received a final daily text reminding them to delete the study messages and lock their phone (see Figure 1). Messages were programmed in the regional language of the study site, either English, Spanish, Thai, Xhosa, or Afrikaans. Additional system specifications, challenges to and solutions for implementation, and evaluations of system acceptability and feasibility are detailed elsewhere.17,18 Figure 1. View largeDownload slide MTN-017 SMS question tree. Figure 1. View largeDownload slide MTN-017 SMS question tree. CASI questionnaires At enrollment, participants underwent a baseline assessment via computer-assisted self-interview (CASI), which captured demographic data, among other items. At the final clinic visit, after completing all study product use, they completed a final CASI assessment, which asked about experiences using the SMS system. Participants were asked to indicate how much they were bothered by having to use a password and by having to erase the SMS session from their phone (answer choices: not at all, a little, a lot); they were also asked, “How concerned were you about privacy while using the SMS system?” (not concerned at all, a little concerned, very concerned). Questionnaire items were developed by our team, based on questions tested in previous studies. Based on prior research in this area,19 we anticipated that there would not be adequate dispersion if the scale asked about positive feelings associated with these features, hence items measured how much participants were bothered by the SMS system protocol for privacy and security. In-depth interviews Upon returning to the clinic after the first 8-week study period, a subsample of 33 participants, evenly distributed by country and study regimen, underwent in-depth interviews (IDIs) to address the secondary study aim of evaluating adherence to the study product. The IDIs were conducted by phone by a trained interviewer based at Columbia University in New York. Stratified purposeful sampling was used to select participants for the interviews based on having excellent (100%) or poor (<80%) adherence to study product use.20 During the IDIs, participants discussed their experiences using the SMS system during the study, including about privacy issues with the question “How did you feel about privacy while texting?” and the probes “Were you worried other people would see?” and “Did you erase the questions after answering them?,” as well as “How did you feel about having a password for the system?” To better understand participants’ perceptions of privacy and disclosure of study participation and product use, and also to add context to their responses about privacy while texting, participants were also asked, “Who knew you were using the product?,” “What was their reaction?,” “Who didn't you tell that you were using the product?,” and “How did you feel about the possibility of them finding out about it?” The interview guide was developed by our team, based on items tested in previous studies. Interviews were audio-recorded, transcribed, and translated into English when necessary (Spanish interviews were not translated, since many of the research staff are bilingual). Data analysis We used a convergent parallel design for our mixed-methods analyses.21 For our quantitative analyses, we generated descriptive statistics for demographic variables, how much people were bothered by having to send a password and erase the text messages from their phones, and how concerned they were about privacy while using the SMS system. In addition, we tested for differences between countries with respect to reaction to the SMS system using analysis of variance. SPSS v.23 software was used for all quantitative analyses. We used the qualitative data to elucidate the findings noted in the survey. To analyze data from qualitative interviews, a codebook was developed jointly by the interviewers, the project manager, and the project’s principal investigator22 on the basis of prior research and inductively. It incorporated categories and themes from the interview guide and included definitions, inclusion and exclusion criteria, and examples of passages for inclusion. To validate and finalize the codebook, 3 researchers coded an initial set of 3 transcripts independently and then compared the codes to assess concordance. Any discrepancies were discussed until consensus was reached, and the codebook was modified where necessary, refining inclusion and exclusion criteria and adding new codes. Researchers coded the remaining transcripts independently using NVivo 10.0 software. Every transcript was coded in pairs to ensure intercoder agreement. Spot checks for intercoder agreement consistently yielded >90% agreement. Coding reports were generated for the codes titled “Privacy while texting” and “Privacy of product use.” The coding reports were analyzed using content analysis,20 then summarized and discussed by team members. We further analyzed coding reports by country to identify any country-level differences in attitudes toward the privacy and security features or perceptions of privacy while using SMS. Finally, we selected quotes that contributed to understanding the participants’ experiences and, when necessary, translated them from Spanish into English. RESULTS Demographics Participants included 187 men and transgender women, of whom 180 responded to the CASI section evaluating the SMS system. Demographic information for this subsample is included in Table 1. Participants had a mean age of 31 years, and ranged in age from 18 to 64 years. Eleven percent considered themselves transgender women and 89% identified as gay or homosexual. Almost half of the participants (43%) were from the United States (including Puerto Rico), close to one-third (29%) from Thailand, one-fifth from Peru, and the rest (8%) from South Africa. Table 1. Demographics of MTN-017 participants (N = 180) Demographic characteristics  Mean (SD)  Range  Age  31.5 (9.3)  18–64  Education (years)  12.3 (1.9)  0–13    Na (%)    Currently working full or part time  140 (80)    Currently in school full or part time  51 (29)    Country of residence       Peru  36 (20)     South Africa  15 (8)     Thailand  52 (29)     United States (including Puerto Rico)  77 (43)    Gender       Male  159 (89)     Transgender female  20 (11)    Sexual identity       Gay/homosexual  157 (89)     Bisexual  13 (7)     Straight/heterosexual  3 (2)    Demographic characteristics  Mean (SD)  Range  Age  31.5 (9.3)  18–64  Education (years)  12.3 (1.9)  0–13    Na (%)    Currently working full or part time  140 (80)    Currently in school full or part time  51 (29)    Country of residence       Peru  36 (20)     South Africa  15 (8)     Thailand  52 (29)     United States (including Puerto Rico)  77 (43)    Gender       Male  159 (89)     Transgender female  20 (11)    Sexual identity       Gay/homosexual  157 (89)     Bisexual  13 (7)     Straight/heterosexual  3 (2)    aNs may not sum to 180 due to missing data; percentages are of those with non-missing data. SMS system privacy and security In the final follow-up questionnaire, we assessed 3 components of privacy and security while using the SMS system: use of a password to access the system, instructions to erase all SMS messages from the system after each report, and overall privacy concerns while using the system (see Table 2). Table 2. MTN-017 participant reactions to privacy while texting and privacy-enhancing features of SMS system (N = 180) Participant reactions  Na (%)  Bothered by having to enter password     Not at all  128 (73)   A little  34 (19)   A lot  14 (8)  Bothered by having to erase SMS session     Not at all  143 (82)   A little  24 (14)   A lot  7 (4)  Concerned about privacy while using system     Not at all  151 (85)   A little  18 (10)   A lot  9 (5)  Participant reactions  Na (%)  Bothered by having to enter password     Not at all  128 (73)   A little  34 (19)   A lot  14 (8)  Bothered by having to erase SMS session     Not at all  143 (82)   A little  24 (14)   A lot  7 (4)  Concerned about privacy while using system     Not at all  151 (85)   A little  18 (10)   A lot  9 (5)  aNs do not sum to 180 due to missing data; percentages are of those with non-missing data. Reactions to using a password to access the SMS system The majority of participants (73%) reported not being bothered at all by having to enter a password, while 19% were bothered a little and 8% were bothered a lot. There were no significant differences between countries. In the IDIs, participants elaborated on this topic (see Table 3). Table 3. Qualitative responses on SMS privacy and security Age, study site  Reactions to using a password to access the SMS system  The password was useful  a  42, Bangkok  I think it’s something that is needed to identify that I am the person who is responding to the message.  b  36, Cape Town  I tried to send an SMS using my friend’s phone because mine, I had no airtime. I tried to send an SMS using my friend’s phone and when I am done I texted “one.” So after that I phoned [the study coordinator] and asked her what is going to happen if I reply using a friend’s phone. Because I did enter my password, and after that I texted “one.”  c  44, Lima  Passwords are part of our lives, right, because you are always authorizing them or changing them, but this one, as I said, since it is so short – it’s four digits – it’s easier to remember, so …  The password was unnecessary  d  32, Boston  I mean, I can see the need for confidentiality in the system, but I didn’t feel personally a need for that, so in some ways the, like, doing your password and then doing your answer seemed a little bit more than I needed.  d  26, San Francisco  [Entering the password] was a little bit precarious. You know, especially since it’s both letters and numbers, you know? Or, I would think that since it’s so easy to track where a message is coming from by phone number, that it’s a little obsolete to have the passcode.  Age, study site  Reactions to using a password to access the SMS system  The password was useful  a  42, Bangkok  I think it’s something that is needed to identify that I am the person who is responding to the message.  b  36, Cape Town  I tried to send an SMS using my friend’s phone because mine, I had no airtime. I tried to send an SMS using my friend’s phone and when I am done I texted “one.” So after that I phoned [the study coordinator] and asked her what is going to happen if I reply using a friend’s phone. Because I did enter my password, and after that I texted “one.”  c  44, Lima  Passwords are part of our lives, right, because you are always authorizing them or changing them, but this one, as I said, since it is so short – it’s four digits – it’s easier to remember, so …  The password was unnecessary  d  32, Boston  I mean, I can see the need for confidentiality in the system, but I didn’t feel personally a need for that, so in some ways the, like, doing your password and then doing your answer seemed a little bit more than I needed.  d  26, San Francisco  [Entering the password] was a little bit precarious. You know, especially since it’s both letters and numbers, you know? Or, I would think that since it’s so easy to track where a message is coming from by phone number, that it’s a little obsolete to have the passcode.  Age, study site  Reactions to instructions to erase messages from cell phones  Participants were not concerned about deleting study-related messages  e  30, Boston  It's funny, I delete – delete messages, things you don’t want people to see or whatever. In this case, it’s one of those that I have not even touched, I just let it – you know, it’s there, there’s no reason to erase it … there’s nothing there that’s incriminating or weird, or stands out, so I just haven’t even touched it.  Participants deleted messages after showing study staff  f  21, Cape Town  Sometimes just for proof I was keeping [the messages on my phone] to show them when I go to the clinic. I was showing them so they won’t criticize me for not texting, to show them it says no response sometimes.  Age, study site  Reactions to instructions to erase messages from cell phones  Participants were not concerned about deleting study-related messages  e  30, Boston  It's funny, I delete – delete messages, things you don’t want people to see or whatever. In this case, it’s one of those that I have not even touched, I just let it – you know, it’s there, there’s no reason to erase it … there’s nothing there that’s incriminating or weird, or stands out, so I just haven’t even touched it.  Participants deleted messages after showing study staff  f  21, Cape Town  Sometimes just for proof I was keeping [the messages on my phone] to show them when I go to the clinic. I was showing them so they won’t criticize me for not texting, to show them it says no response sometimes.  Age, study site  Privacy concerns  Participants felt others would not be suspicious of study text messages  Upon arrival of reminder message  g  38, Lima  [The daily reminder] is a quick message, and it’s anonymous because it just asks for a password, so, there’s no problem. Whether it be your partner or your mother, there is no reason for them to feel, I don’t know, to mistrust you because it is just asking for a password.  While texting  h  28, Chiang Mai  Other people might look at me and say that I was just using the Internet, Facebook, or sending messages online with my friends. … I am not concerned that other people might know what I am sending. I am not concerned that other people will sneak in and read the SMS.  h  36, Cape Town  I was not worried about [other people seeing what I was texting]. I was doing it even in front of them. They just didn't know what I was doing.  After texting  i  22, Cape Town  I don’t worry about that because nobody noses around my phone because I lock it.  Participants felt message content would not reveal study participation  j  22, Cape Town  It’s pretty much private itself because you only text your password, it doesn’t have any name or anything, and doesn’t say what it is all about. And it reminds you to delete everything when you are done.  j  51, Lima  I wouldn’t want my family to find out right now [that I am gay]. … So, I am assuming that all of this is private and that my name doesn’t appear. The text messages, well, it is something that is much more private because there are no names, no last names, you don't enter any of that information … and no one else sees my phone.  Participants had suggestions for how to make the system more private  k  43, San Francisco  Have something really, really vague where only the study participant and you guys know what the real question is, so that if they got the text across, nobody would be really suspicious, they’d be like, oh, what’s the weather like today.    Age, study site  Privacy concerns  Participants felt others would not be suspicious of study text messages  Upon arrival of reminder message  g  38, Lima  [The daily reminder] is a quick message, and it’s anonymous because it just asks for a password, so, there’s no problem. Whether it be your partner or your mother, there is no reason for them to feel, I don’t know, to mistrust you because it is just asking for a password.  While texting  h  28, Chiang Mai  Other people might look at me and say that I was just using the Internet, Facebook, or sending messages online with my friends. … I am not concerned that other people might know what I am sending. I am not concerned that other people will sneak in and read the SMS.  h  36, Cape Town  I was not worried about [other people seeing what I was texting]. I was doing it even in front of them. They just didn't know what I was doing.  After texting  i  22, Cape Town  I don’t worry about that because nobody noses around my phone because I lock it.  Participants felt message content would not reveal study participation  j  22, Cape Town  It’s pretty much private itself because you only text your password, it doesn’t have any name or anything, and doesn’t say what it is all about. And it reminds you to delete everything when you are done.  j  51, Lima  I wouldn’t want my family to find out right now [that I am gay]. … So, I am assuming that all of this is private and that my name doesn’t appear. The text messages, well, it is something that is much more private because there are no names, no last names, you don't enter any of that information … and no one else sees my phone.  Participants had suggestions for how to make the system more private  k  43, San Francisco  Have something really, really vague where only the study participant and you guys know what the real question is, so that if they got the text across, nobody would be really suspicious, they’d be like, oh, what’s the weather like today.    Note: a–k quotations correspond to in-text descriptions. Five people felt the password was useful.a This was especially salient in cases where participants had to use a phone that was not registered with the system; the password allowed them to be identified.b Five others had positive comments: 2 mentioned the password was good protection, while another said it was a nice thing to have. Two compared it to passwords for other systems. One said it was simple and convenient, and another said it was short and easy to remember.c Ten people had a neutral stance, stating it was fine and did not cause them any problems. Four participants had trouble remembering their password at first, but most were able to memorize it quickly, except for one who said he could not remember it. Finally, 3 people felt that the password was unnecessary, whether due to lack of privacy concerns or perceived inconvenience.d Reactions to instructions to erase messages from cell phones In the quantitative assessment, participants were also asked how much they were bothered by having to erase the SMS sessions from their cell phones. Again, the majority (82%) said they were not bothered at all, a few (14%) were bothered a little, and 4% were bothered a lot. There were no significant differences by country. We explored with participants in the IDIs whether they deleted the messages from their phones following each report, as instructed. A little over half of them (18 participants) said they did not delete the messages, and some specifically mentioned that they were not concerned about others seeing them.e Given that participants received a financial incentive when they returned for their clinic visit based on the number of SMS reports sent, many mentioned keeping the messages on their phone to show them to clinic staff as proof of sent messages. Some participants deleted the messages after receiving that month’s incentive. This was the case particularly at the Cape Town site, where issues with cellular service caused some messages not to be received by the system.f Those who deleted the messages often had reasons other than privacy, like keeping their message inbox from getting too full. Others deleted only the incoming messages from the SMS system, but kept the messages they had sent on their phones. Privacy concerns Finally, participants were asked via CASI how concerned they were about privacy while using the SMS system during the study. Most (85%) were not concerned at all about privacy, 10% were a little concerned, and only 5% reported being very concerned. Participants in South Africa showed a significantly higher level of concern about privacy while using the SMS system than participants in other countries, with a mean rating of 2.00 on a 3-point scale (1 = not concerned at all, 2 = a little concerned, 3 = very concerned), while mean scores for other countries ranged from 1.08 to 1.21 (P < .001). During the IDIs, participants did not report concerns about the privacy of study text messages at any point in the cycle of interaction with the system (upon receiving the first message, while texting the system, or after completing the text report), and no differences between countries were identified. One participant stated that the initial reminder only asked for a password and could not arouse the suspicion of others who might see it.g Participants were not worried that others might wonder with whom they were texting while responding to the system, stating that no one could know what they were doing by simply observing their texting behavior.h Finally, participants were unconcerned about others viewing their messages after they completed their responses.i Others made reference to the harmless content of the messages, which were worded carefully to prevent revealing study participation. This gave them assurance that their privacy was protected within the SMS system.j Finally, one had suggestions for making the first question more private so that no one would be suspicious if it flashed on the screen.k Overall, the great majority of participants did not have concerns about privacy while using the SMS system. For those who had concerns about revealing study participation to others, the privacy elements worked well to protect them. Privacy of HIV prevention product use, study participation, and sexual identity In addition to participants’ perceptions of SMS system privacy, we analyzed IDI data on the privacy of product use (see Table 4). Themes regarding disclosure of HIV prevention product use, study participation, and sexual identity emerged, allowing us to contextualize responses on SMS privacy within overall study-related privacy concerns. Table 4. Qualitative responses on privacy of sexual identity, HIV prevention product use, and study participation Age, study site  Overall privacy concerns  Participants were concerned about others’ assumptions upon disclosing study product use  l  30, Boston  I’ve told a couple of people, and they’d look at me like, like is he telling me that he’s HIV positive?  Participants did not want family members to learn of study participation for fear of disclosure of sexual identity  m  51, Lima  Well, I would also have to tell them, “So, this is how I am, I am gay, and so, my health is fine, I am taking [the pills] for a study” and all that, right, and my family, whom I live with currently, is a little – I could say – gossipy, right … and I have family members on my father’s side who are very religious. I think that, well it would be better to avoid all of this, by not telling them.  m  31, Cape Town  Ehh … it’s my mom. I never told her. Because she still hasn’t accepted the fact that I am gay. So things like this study I did not want to talk to her about it because of that.  Age, study site  Overall privacy concerns  Participants were concerned about others’ assumptions upon disclosing study product use  l  30, Boston  I’ve told a couple of people, and they’d look at me like, like is he telling me that he’s HIV positive?  Participants did not want family members to learn of study participation for fear of disclosure of sexual identity  m  51, Lima  Well, I would also have to tell them, “So, this is how I am, I am gay, and so, my health is fine, I am taking [the pills] for a study” and all that, right, and my family, whom I live with currently, is a little – I could say – gossipy, right … and I have family members on my father’s side who are very religious. I think that, well it would be better to avoid all of this, by not telling them.  m  31, Cape Town  Ehh … it’s my mom. I never told her. Because she still hasn’t accepted the fact that I am gay. So things like this study I did not want to talk to her about it because of that.  Note: l–m quotations correspond to in-text descriptions. Out of 33 participants interviewed, less than half discussed product use with friends (n = 16) and/or their partners or boyfriends (n = 14). Only 9 (27%) discussed it with family members, and 5 did not tell anyone they were using the study gel or tablets. Some felt comfortable talking about the study and their involvement in it. However, others expressed concern that people could assume they were HIV-infected if they disclosed product use, and some even had such an experience during the study.l In countries other than the United States, many participants lived with their parents and did not tell them about the study to avoid having to disclose their sexual identity. Four people specifically mentioned that their families either did not know or did not accept that they were gay, and they hid their study participation and study product use from family members for this reason.m Overall, many people in our sample did not feel comfortable discussing study participation or product use with others, indicating the relevance of exploring privacy issues around the SMS system. DISCUSSION Our study of participants’ perceptions of privacy while using SMS to report adherence to HIV-prevention product use during a Phase 2 study showed that most participants were not highly concerned about privacy while texting. Although some participants reported concerns about disclosure of study participation, sexual identity, or HIV-prevention product use, they were not concerned that the SMS system would compromise their confidentiality. Whether they considered texting behavior to be private or the text message language sufficiently ambiguous to protect their privacy, most did not have privacy concerns when responding to the SMS system. The use of SMS, a ubiquitous technology, did not exacerbate privacy concerns for participants, given that texting behavior is very common, and therefore did not arouse suspicion. Nevertheless, a subsample of 9 participants, 6 of whom were from South Africa and most of whom were unemployed, reported being very concerned about privacy while using the SMS system. Unfortunately, the subsample participants interviewed were not selected based on responses to CASI questions on SMS privacy, and none who reported SMS privacy concerns were interviewed. Given that phone sharing in South Africa is common23,24 and that these participants may have had fewer economic resources and an increased likelihood of sharing phones, this could be the reason for higher privacy concerns among this group. Prior research in African settings has found that participants with the lowest incomes who may be more likely to share mobile phones reported the greatest concerns with SMS privacy.7 Design features were effective in protecting participants’ privacy and mitigating any concerns. First, participants were able to select the time of day they preferred to receive the study messages. Participants in other studies also emphasized self-selection of time and/or day of SMS messages to protect privacy.12 Second, the first message, asking if they were “available to text,” was instrumental in preparing the participants for answering study questions. It also allowed them to physically position themselves in a place that could give them greater privacy while texting, if necessary. Third, the password had 2 functions. It was effective in preventing access to sensitive study questions without authorization, and it was linked to the participant ID, which allowed for participant tracking if a different phone number was used to report. This was critical in the data analysis phase to quickly link reports sent from multiple phones by a single participant. While a previous study found that having to enter a personal identification number to access the study SMS resulted in poorer access to the messages,15 in our study most participants were not bothered by having to enter a password, and some felt it provided an increased sense of privacy. Fourth, the security feature reminding participants to lock their cell phone and delete the study-related messages was not burdensome. While in other studies participants preferred to delete messages upon receipt to prevent others from viewing them,11,12 in our study many participants found it unnecessary and deleted the messages only after showing them to study coordinators and receiving compensation or when their message inbox became too full. Finally, the carefully constructed language of our SMS system prevented disclosure of study participation or the type of product being used. In prior research, participants suggested that language used in messages and how much it revealed about the nature of the study could either bolster or hinder confidentiality,25 and in this study the ambiguous language was key. Research has shown that participants may prefer coded messages, in which only they can understand the meaning,11 which can be as effective as direct messages about health information.15 Nevertheless, there is the potential for confusion with coded language, and participants prefer messages that clearly convey meaning to minimize uncertainty.11 We achieved this by designing a clear message that was carefully worded to preclude revealing study participation or the nature of the study product. Limitations While the qualitative data provide a rich description of SMS privacy and security concerns in this population, a larger sample size could allow for greater generalizability and comparative analyses. Furthermore, the IDI subsample was selected to explore adherence to the study products; therefore, we could not purposefully select those who reported fewer or greater SMS privacy concerns. In addition, our sample included different countries and cultural contexts, so we cannot make generalizations, given that small numbers of participants from each context were interviewed and others may have different perspectives. Variations in participants’ technological literacy or proficiency may have influenced their privacy concerns or lack thereof; however, these data were not available to supplement the analysis. CONCLUSION Privacy and security are concerns for researchers charged with the protection of research participants. Studies like ours can inform both investigators and IRBs on effective SMS practices to safeguard privacy and provide data security. Based on our evaluation of privacy and security features for an interactive SMS data collection system, the most useful features for health informatics practitioners who are developing SMS-based interventions include: (1) reminders that reinforce privacy and security-enhancing behaviors (ie, lock your phone, delete study-related messages), (2) ambiguous language that can be understood by study participants without revealing confidential information to others, and (3) the use of short, participant ID–related passwords that allow for participant tracking. These last 2 features are innovative and have yet to be highlighted in in the published SMS literature in combined use. COMPETING INTERESTS The authors have no competing interests to declare. CONTRIBUTORS RG, WB, CD, TH, AS, and MI collected and cleaned the data. RG, WB, and CD analyzed and interpreted the data. IB, WB, CD, JL, IM, RC, and ACD contributed to the study design. RG, WB, CD, IB, AS, and ACD drafted the report. TH, MI, JL, IM, and RC provided critical revision of the article. All authors gave input to the final version and provided final approval of the version to be published. FUNDING MTN-017 was sponsored by the US National Institutes of Health (NIH) and CONRAD (Contraception Research and Development). The study was designed and implemented by the Microbicide Trials Network (MTN). The MTN is funded by the National Institute of Allergy and Infectious Diseases (UM1AI068633, UM1AI068615, UM1AI106707), with co-funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Institute of Mental Health (NIMH), all components of the US National Institutes of Health. The study products used in this study were supplied by the Investigational New Drug sponsor, CONRAD, and the pharmaceutical collaborator, Gilead. This work was also supported by a center grant from the NIMH to the HIV Center for Clinical and Behavioral Studies at New York State Psychiatric Institute and Columbia University (P30-MH43520). WB was supported by the National Library of Medicine (R01-LM012355-01A1 and T15-LM007079) and the NIMH/National Institute on Drug Abuse (R03-MH103957). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. ACKNOWLEDGMENTS We are grateful to the study participants for their participation and dedication. We also thank the study team members at the research sites, who worked tirelessly to assist in implementing the SMS system and to train study participants to use it. We thank the protocol management team and the Microbicide Trials Network leadership operations center for their contributions to the study. References 1 Andriole KP. Security of electronic medical information and patient privacy: what you need to know. J Am Coll Radiol.  2014; 11: 1212– 16. Google Scholar CrossRef Search ADS PubMed  2 Tavani H. Privacy and security. Internet Ethics  .   London :  Macmillan; 2000: 65– 95. 3 White Y, Sandfort T, Morgan Ket al.  , Family relationships and sexual orientation disclosure to family by gay and bisexual men in Jamaica. Int J Sex Heal.  2016; 28: 306– 17. Google Scholar CrossRef Search ADS   4 Schrimshaw EW, Downing MJ, Cohn DJ. Reasons for non-disclosure of sexual orientation among behaviorally bisexual men: non-disclosure as stigma management. Arch Sex Behav.  2016; 1– 15. [Epub ahead of print]. 5 Sanou B. ICT Facts and Figures 2016 . Published online first: 2016. http://www.itu.int/en/ITU-D/Statistics/Documents/facts/ICTFactsFigures2016.pdf. Accessed July 29, 2016. 6 Kaplan WA. Can the ubiquitous power of mobile phones be used to improve health outcomes in developing countries? Global Health.  2006; 2: 9. Google Scholar CrossRef Search ADS PubMed  7 Nhavoto JA, Grönlund Å, Klein GO. Mobile health treatment support intervention for HIV and tuberculosis in Mozambique: Perspectives of patients and healthcare workers. PLoS One.  2017; 12: e0176051. Google Scholar CrossRef Search ADS PubMed  8 Crankshaw T, Corless IB, Giddy Jet al.  , Exploring the patterns of use and the feasibility of using cellular phones for clinic appointment reminders and adherence messages in an antiretroviral treatment clinic, Durban, South Africa. AIDS Patient Care STDs.  2010; 24: 729– 34. Google Scholar CrossRef Search ADS PubMed  9 Allison S, Bauermeister JA, Bull Set al.  , The intersection of youth, technology, and new media with sexual health: moving the research agenda forward. J Adolesc Heal.  2017; 51: 207– 12. Google Scholar CrossRef Search ADS   10 Acquisti A, Brandimarte L, Loewenstein G. Privacy and human behavior in the age of information. Science (80–).  2015; 347: 509 LP– 514. Google Scholar CrossRef Search ADS   11 Curioso WH, Alex Quistberg D, Cabello Ret al.  , “It’s time for your life”: How should we remind patients to take medicines using short text messages? AMIA Annu Symp Proc.  2009; 2009: 129– 33. Google Scholar PubMed  12 Rodrigues R, Poongulali S, Balaji Ket al.  , “The phone reminder is important, but will others get to know about my illness?” Patient perceptions of an mHealth antiretroviral treatment support intervention in the HIVIND trial in South India. BMJ Open.  2015; 5:e007574. 13 Sidney K, Antony J, Rodrigues Ret al.  , Supporting patient adherence to antiretrovirals using mobile phone reminders: patient responses from South India. AIDS Care.  2012; 24: 612– 17. Google Scholar CrossRef Search ADS PubMed  14 Hakkila J, Chatfield C. It’s like if you opened someone else’s letter: user perceived privacy and social practices with SMS communication. In: Proceedings of the 7th International Conference on Human Computer Interaction with Mobile Devices and Services.  2005; 219– 22. 15 Siedner JM, Santorino D, Haberer EJet al.  , Know your audience: predictors of success for a patient-centered texting app to augment linkage to HIV care in rural Uganda. J Med Internet Res.  2015; 17: e78. CrossRef Search ADS PubMed  16 Cranston RD, Lama JR, Richardson BAet al.  , MTN-017: a rectal phase 2 extended safety and acceptability study of tenofovir reduced-glycerin 1% gel. Clin Infect Dis.  2017; 64: 614– 20. Google Scholar PubMed  17 Brown III W, Giguere R, Ibitoye Met al.  , Successfully addressing challenges to implemeting a multinational SMS-based reminder and data collection system in a biomedical HIV prevention trial. AIDS Res Hum Retroviruses.  2014; 30: A87. Google Scholar CrossRef Search ADS   18 Brown III W, Giguere R, Sheinfil Aet al.  , Feasibility and acceptability of an international SMS text message–based adherence and survey system in a biomedical HIV prevention study (MTN-017). AIDS Res Hum Retroviruses.  2016; 32: 386. Google Scholar CrossRef Search ADS PubMed  19 Brown III W. Instrumentation and mobile technology for monitoring sexual behavior and product use in clinical trials. In: Microbicide Trials Network Annual Meeting . 2013. http://www.mtnstopshiv.org/sites/default/files/attachments/BROWN-MTN 017 SMS meeting presentation.pdf. Accessed July 24, 2017. 20 Patton MQ, Patton MQ. Qualitative Research and Evaluation Methods . 3rd ed. Thousand Oaks, CA: SAGE Publications; 2002. 21 John W. Creswell, Vicki L. Plano Clark. Chapter 3. Choosing a mixed methods research design. In: Designing and Conducting Mixed Methods Research . Thousand Oaks, CA: SAGE Publications; 2011: 53– 106. 22 MacQueen KM, McLellan E, Kay Ket al.  , Codebook development for team-based qualitative analysis. Cult Anthropol Methods.  1998; 10: 31– 36. 23 James J, Versteeg M. Mobile phones in Africa: how much do we really know? Soc Indic Res.  2007; 84: 117– 26. Google Scholar CrossRef Search ADS PubMed  24 Kreutzer T. Generation mobile: online and digital media usage on mobile phones among low-income urban youths in South Africa. Centre for Film and Media Studies, University of Cape Town, Cape Town; 2009. www.tinokreutzer.org/mobile. Accessed July 24, 2017. 25 Baranoski AS, Meuser E, Hardy Het al.  , Patient and provider perspectives on cellular phone-based technology to improve HIV treatment adherence. AIDS Care.  2014; 26: 26– 32. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the American Medical Informatics Association. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of the American Medical Informatics Association Oxford University Press

Are participants concerned about privacy and security when using short message service to report product adherence in a rectal microbicide trial?

Loading next page...
 
/lp/ou_press/are-participants-concerned-about-privacy-and-security-when-using-short-dzCgCcoRUp
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of the American Medical Informatics Association. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
1067-5027
eISSN
1527-974X
D.O.I.
10.1093/jamia/ocx081
Publisher site
See Article on Publisher Site

Abstract

Abstract Objective During a Phase 2 rectal microbicide trial, men who have sex with men and transgender women (n = 187) in 4 countries (Peru, South Africa, Thailand, United States) reported product use daily via short message service (SMS). To prevent disclosure of study participation, the SMS system program included privacy and security features. We evaluated participants’ perceptions of privacy while using the system and acceptability of privacy/security features. Materials and Methods To protect privacy, the SMS system: (1) confirmed participant availability before sending the study questions, (2) required a password, and (3) did not reveal product name or study participation. To ensure security, the system reminded participants to lock phone/delete messages. A computer-assisted self-interview (CASI), administered at the final visit, measured burden of privacy and security features and SMS privacy concerns. A subsample of 33 participants underwent an in-depth interview (IDI). Results Based on CASI, 85% had no privacy concerns; only 5% were very concerned. Most were not bothered by the need for a password (73%) or instructions to delete messages (82%). Based on IDI, reasons for low privacy concerns included sending SMS in private or feeling that texting would not draw attention. A few IDI participants found the password unnecessary and more than half did not delete messages. Discussion Most participants were not concerned that the SMS system would compromise their confidentiality. SMS privacy and security features were effective and not burdensome. Conclusion Short ID-related passwords, ambiguous language, and reminders to implement privacy and security-enhancing behaviors are recommended for SMS systems. text messaging, privacy, HIV, LGBT persons BACKGROUND AND SIGNIFICANCE Protecting participants’ privacy and security is key to conducting ethical research. Privacy refers to a participant’s right to control when, how, and to what extent personal information is shared with others. Security refers to protection against unauthorized access to data or devices.1,2 Researchers are obligated to protect participants from potential harm resulting from a privacy or security breach, including disclosure of study participation or personal health information, especially on sensitive topics such as HIV, sexual behavior, sexual orientation, and gender identity. Disclosure of sexual orientation by gay or bisexual men can increase exposure to negative responses and stressors, and men who have same-sex sexual behavior may elect not to disclose it to avoid stigmatizing reactions.3,4 Data collection via short message service (SMS, or text messaging) is convenient but also presents new challenges to protecting participants’ privacy and security. SMS is a useful and practical data collection method that is widely accessible, with 95% of the global population living in areas covered by mobile-cellular networks.5 Messages can be received silently and even when the phone is turned off.6 However, SMS messages leave a record on the phone, which can have relevant implications for the protection of privacy and security,6 particularly in cases where a cell phone is shared by 2 or more people.7,8 Institutional Review Boards (IRBs) at research institutions charged with ensuring the protection of participants’ privacy and security often raise concerns regarding the use of SMS in research. These concerns focus on perceptions that text messages are not secure and could breach participant confidentiality or expose personal health information. Given that only 30% of IRBs have had training regarding new media technology in research,9 board members must take into account a plurality of views from information technology specialists, technology developers, study investigators, ethicists, and legal officials when evaluating the use of SMS in research. In addition, research on participants’ perceptions of privacy and security during SMS-based mHealth interventions should also inform IRB discussions. Participants’ perceptions may differ from researchers’ and health care workers’ expectations. For example, in a study of an SMS reminder intervention for HIV and tuberculosis patients, a majority of health care workers perceived a risk of unintentional disclosure of patients’ health status, while a vast majority of patients did not perceive similar risks.7 Therefore, although participants’ perceptions of privacy may be context-dependent and malleable,10 as end users their perspectives are valuable in informing researchers and ethical boards of the potential for study-related social harms when using SMS in research. Current literature on participants’ views on privacy and security when receiving health information via SMS has demonstrated that concerns exist. In research exploring the use of SMS-based reminders for antiretroviral treatment adherence, some participants reported a fear of unintentional disclosure of health status, with the potential for resulting stigma and discrimination.7,8,11,12 In one such study, participants reported deleting SMS messages immediately after receipt to prevent others from viewing them.12 However, in another study, participants did not consider the reminders intrusive to their privacy and did not report negative consequences to receiving messages, even when others inadvertently received them.13 Finally, for some, social norms around SMS privacy dictate an unspoken etiquette by which people do not read others’ messages.14 To date, most of the studies that have examined privacy in the context of SMS focused on systems that provide reminders or health information but are not interactive, as ours was.7,12,15 We designed an interactive SMS-based adherence reminder and data collection system for MTN-017, an international Phase 2 safety and acceptability study comparing an oral tablet (emtricitabine/tenofovir disoproxil fumarate, or FTC/TDF) and a gel (reduced glycerin tenofovir, or TFV-RG 1% gel) applied rectally for HIV prevention among men who have sex with men and transgender women. To protect participants’ privacy/security and mitigate the potential for related concerns, we implemented several privacy and security design features. We evaluated participants’ perceptions of privacy while using the system and the acceptability of the system’s privacy features. Our report can contribute to the development of future SMS-based data collection systems and inform ethical review of research involving SMS for data collection. MATERIALS AND METHODS Study design The data for this study come from an international Phase 2 safety and acceptability trial comparing 2 HIV-prevention products (TFV-RG 1% gel and FTC/TDF tablets) and 3 study regimens: (1) daily use of the gel applied rectally with an applicator, (2) gel applied before and after receptive anal intercourse (RAI) with an applicator, and (3) daily oral tablet. All participants completed an 8-week study period for each regimen (daily oral tablet, daily rectal gel, RAI rectal gel) in a randomly assigned order, with a 1-week washout period between each study period.16 A secondary aim of the study was to evaluate adherence to product use. Sample Participants included men and transgender women ≥18 years of age, non-HIV infected, in good health, with a history of RAI at least once in the past 3 months. The study took place at 8 sites in 4 countries: Cape Town, South Africa; Bangkok and Chiang Mai, Thailand; Lima, Peru; San Juan, Puerto Rico (US territory); and San Francisco, Pittsburgh, and Boston, continental United States. Participants were recruited from health clinics and community-based organizations. The study was reviewed and approved by the IRBs at all participating institutions, and participants provided written informed consent. Procedures This paper describes a substudy focusing on use of an SMS system for adherence monitoring among participants in the larger study comparing safety and acceptability of 3 HIV prevention regimens. At enrollment, participants were randomized to 1 of 6 study-product sequences and given study product, either a bottle of oral FTC/TDF tablets to take daily or 30 white plastic applicators prefilled with RG-TFV 1% gel to insert rectally either daily or before and after RAI. Participants were given a study-owned mobile phone if they did not have their own and were trained to use the SMS system to report product use daily during the trial. We created alphanumeric passwords, made up of 2-letter abbreviations for each study site plus 1- or 2-digit numbers that corresponded to the participant identification numbers. For example, participant 1 in Cape Town was assigned the password CT1. During the SMS training, participants chose their preferred reminder time, programmed the SMS system mobile number into their mobile phones, and joined the system. Joining the system entailed sending the word “Join,” the participant’s password, and the preferred reminder time (using the 24-h clock for clarity and brevity) to the system (eg, “Join CT1 2200”). After choosing a reminder time and joining the system, participants completed their first SMS session at the clinic under staff supervision. They responded to a reminder by sending their passwords, reported the number of times they used the study product, and then deleted the messages from their phones. They were given an SMS instruction card, which included the SMS system mobile number; their password; and a reminder to lock their phone, protect it with a password, and delete the study text messages to safeguard their privacy. During each study period, participants were asked to respond to daily text messages from the system. They were compensated for 1 completed text message session per day (defined as responding to the reminder with a password and sending the number of times the product was used). Each study site determined a locally appropriate compensation structure, ranging from US$0.15 to US$2.00 per session. Participants were informed that they would receive a bonus at the end of the month, ranging from US$1.00 to US$10.00, for completing at least 6 sessions per week. SMS system description The SMS script consisted of a generic daily reminder, programmed to arrive at the chosen reminder time, requesting that participants send their password to initiate the session. Upon sending the password, participants received a text message asking them to report the number of times they had used the product since their prior report. After sending a number, participants received a final daily text reminding them to delete the study messages and lock their phone (see Figure 1). Messages were programmed in the regional language of the study site, either English, Spanish, Thai, Xhosa, or Afrikaans. Additional system specifications, challenges to and solutions for implementation, and evaluations of system acceptability and feasibility are detailed elsewhere.17,18 Figure 1. View largeDownload slide MTN-017 SMS question tree. Figure 1. View largeDownload slide MTN-017 SMS question tree. CASI questionnaires At enrollment, participants underwent a baseline assessment via computer-assisted self-interview (CASI), which captured demographic data, among other items. At the final clinic visit, after completing all study product use, they completed a final CASI assessment, which asked about experiences using the SMS system. Participants were asked to indicate how much they were bothered by having to use a password and by having to erase the SMS session from their phone (answer choices: not at all, a little, a lot); they were also asked, “How concerned were you about privacy while using the SMS system?” (not concerned at all, a little concerned, very concerned). Questionnaire items were developed by our team, based on questions tested in previous studies. Based on prior research in this area,19 we anticipated that there would not be adequate dispersion if the scale asked about positive feelings associated with these features, hence items measured how much participants were bothered by the SMS system protocol for privacy and security. In-depth interviews Upon returning to the clinic after the first 8-week study period, a subsample of 33 participants, evenly distributed by country and study regimen, underwent in-depth interviews (IDIs) to address the secondary study aim of evaluating adherence to the study product. The IDIs were conducted by phone by a trained interviewer based at Columbia University in New York. Stratified purposeful sampling was used to select participants for the interviews based on having excellent (100%) or poor (<80%) adherence to study product use.20 During the IDIs, participants discussed their experiences using the SMS system during the study, including about privacy issues with the question “How did you feel about privacy while texting?” and the probes “Were you worried other people would see?” and “Did you erase the questions after answering them?,” as well as “How did you feel about having a password for the system?” To better understand participants’ perceptions of privacy and disclosure of study participation and product use, and also to add context to their responses about privacy while texting, participants were also asked, “Who knew you were using the product?,” “What was their reaction?,” “Who didn't you tell that you were using the product?,” and “How did you feel about the possibility of them finding out about it?” The interview guide was developed by our team, based on items tested in previous studies. Interviews were audio-recorded, transcribed, and translated into English when necessary (Spanish interviews were not translated, since many of the research staff are bilingual). Data analysis We used a convergent parallel design for our mixed-methods analyses.21 For our quantitative analyses, we generated descriptive statistics for demographic variables, how much people were bothered by having to send a password and erase the text messages from their phones, and how concerned they were about privacy while using the SMS system. In addition, we tested for differences between countries with respect to reaction to the SMS system using analysis of variance. SPSS v.23 software was used for all quantitative analyses. We used the qualitative data to elucidate the findings noted in the survey. To analyze data from qualitative interviews, a codebook was developed jointly by the interviewers, the project manager, and the project’s principal investigator22 on the basis of prior research and inductively. It incorporated categories and themes from the interview guide and included definitions, inclusion and exclusion criteria, and examples of passages for inclusion. To validate and finalize the codebook, 3 researchers coded an initial set of 3 transcripts independently and then compared the codes to assess concordance. Any discrepancies were discussed until consensus was reached, and the codebook was modified where necessary, refining inclusion and exclusion criteria and adding new codes. Researchers coded the remaining transcripts independently using NVivo 10.0 software. Every transcript was coded in pairs to ensure intercoder agreement. Spot checks for intercoder agreement consistently yielded >90% agreement. Coding reports were generated for the codes titled “Privacy while texting” and “Privacy of product use.” The coding reports were analyzed using content analysis,20 then summarized and discussed by team members. We further analyzed coding reports by country to identify any country-level differences in attitudes toward the privacy and security features or perceptions of privacy while using SMS. Finally, we selected quotes that contributed to understanding the participants’ experiences and, when necessary, translated them from Spanish into English. RESULTS Demographics Participants included 187 men and transgender women, of whom 180 responded to the CASI section evaluating the SMS system. Demographic information for this subsample is included in Table 1. Participants had a mean age of 31 years, and ranged in age from 18 to 64 years. Eleven percent considered themselves transgender women and 89% identified as gay or homosexual. Almost half of the participants (43%) were from the United States (including Puerto Rico), close to one-third (29%) from Thailand, one-fifth from Peru, and the rest (8%) from South Africa. Table 1. Demographics of MTN-017 participants (N = 180) Demographic characteristics  Mean (SD)  Range  Age  31.5 (9.3)  18–64  Education (years)  12.3 (1.9)  0–13    Na (%)    Currently working full or part time  140 (80)    Currently in school full or part time  51 (29)    Country of residence       Peru  36 (20)     South Africa  15 (8)     Thailand  52 (29)     United States (including Puerto Rico)  77 (43)    Gender       Male  159 (89)     Transgender female  20 (11)    Sexual identity       Gay/homosexual  157 (89)     Bisexual  13 (7)     Straight/heterosexual  3 (2)    Demographic characteristics  Mean (SD)  Range  Age  31.5 (9.3)  18–64  Education (years)  12.3 (1.9)  0–13    Na (%)    Currently working full or part time  140 (80)    Currently in school full or part time  51 (29)    Country of residence       Peru  36 (20)     South Africa  15 (8)     Thailand  52 (29)     United States (including Puerto Rico)  77 (43)    Gender       Male  159 (89)     Transgender female  20 (11)    Sexual identity       Gay/homosexual  157 (89)     Bisexual  13 (7)     Straight/heterosexual  3 (2)    aNs may not sum to 180 due to missing data; percentages are of those with non-missing data. SMS system privacy and security In the final follow-up questionnaire, we assessed 3 components of privacy and security while using the SMS system: use of a password to access the system, instructions to erase all SMS messages from the system after each report, and overall privacy concerns while using the system (see Table 2). Table 2. MTN-017 participant reactions to privacy while texting and privacy-enhancing features of SMS system (N = 180) Participant reactions  Na (%)  Bothered by having to enter password     Not at all  128 (73)   A little  34 (19)   A lot  14 (8)  Bothered by having to erase SMS session     Not at all  143 (82)   A little  24 (14)   A lot  7 (4)  Concerned about privacy while using system     Not at all  151 (85)   A little  18 (10)   A lot  9 (5)  Participant reactions  Na (%)  Bothered by having to enter password     Not at all  128 (73)   A little  34 (19)   A lot  14 (8)  Bothered by having to erase SMS session     Not at all  143 (82)   A little  24 (14)   A lot  7 (4)  Concerned about privacy while using system     Not at all  151 (85)   A little  18 (10)   A lot  9 (5)  aNs do not sum to 180 due to missing data; percentages are of those with non-missing data. Reactions to using a password to access the SMS system The majority of participants (73%) reported not being bothered at all by having to enter a password, while 19% were bothered a little and 8% were bothered a lot. There were no significant differences between countries. In the IDIs, participants elaborated on this topic (see Table 3). Table 3. Qualitative responses on SMS privacy and security Age, study site  Reactions to using a password to access the SMS system  The password was useful  a  42, Bangkok  I think it’s something that is needed to identify that I am the person who is responding to the message.  b  36, Cape Town  I tried to send an SMS using my friend’s phone because mine, I had no airtime. I tried to send an SMS using my friend’s phone and when I am done I texted “one.” So after that I phoned [the study coordinator] and asked her what is going to happen if I reply using a friend’s phone. Because I did enter my password, and after that I texted “one.”  c  44, Lima  Passwords are part of our lives, right, because you are always authorizing them or changing them, but this one, as I said, since it is so short – it’s four digits – it’s easier to remember, so …  The password was unnecessary  d  32, Boston  I mean, I can see the need for confidentiality in the system, but I didn’t feel personally a need for that, so in some ways the, like, doing your password and then doing your answer seemed a little bit more than I needed.  d  26, San Francisco  [Entering the password] was a little bit precarious. You know, especially since it’s both letters and numbers, you know? Or, I would think that since it’s so easy to track where a message is coming from by phone number, that it’s a little obsolete to have the passcode.  Age, study site  Reactions to using a password to access the SMS system  The password was useful  a  42, Bangkok  I think it’s something that is needed to identify that I am the person who is responding to the message.  b  36, Cape Town  I tried to send an SMS using my friend’s phone because mine, I had no airtime. I tried to send an SMS using my friend’s phone and when I am done I texted “one.” So after that I phoned [the study coordinator] and asked her what is going to happen if I reply using a friend’s phone. Because I did enter my password, and after that I texted “one.”  c  44, Lima  Passwords are part of our lives, right, because you are always authorizing them or changing them, but this one, as I said, since it is so short – it’s four digits – it’s easier to remember, so …  The password was unnecessary  d  32, Boston  I mean, I can see the need for confidentiality in the system, but I didn’t feel personally a need for that, so in some ways the, like, doing your password and then doing your answer seemed a little bit more than I needed.  d  26, San Francisco  [Entering the password] was a little bit precarious. You know, especially since it’s both letters and numbers, you know? Or, I would think that since it’s so easy to track where a message is coming from by phone number, that it’s a little obsolete to have the passcode.  Age, study site  Reactions to instructions to erase messages from cell phones  Participants were not concerned about deleting study-related messages  e  30, Boston  It's funny, I delete – delete messages, things you don’t want people to see or whatever. In this case, it’s one of those that I have not even touched, I just let it – you know, it’s there, there’s no reason to erase it … there’s nothing there that’s incriminating or weird, or stands out, so I just haven’t even touched it.  Participants deleted messages after showing study staff  f  21, Cape Town  Sometimes just for proof I was keeping [the messages on my phone] to show them when I go to the clinic. I was showing them so they won’t criticize me for not texting, to show them it says no response sometimes.  Age, study site  Reactions to instructions to erase messages from cell phones  Participants were not concerned about deleting study-related messages  e  30, Boston  It's funny, I delete – delete messages, things you don’t want people to see or whatever. In this case, it’s one of those that I have not even touched, I just let it – you know, it’s there, there’s no reason to erase it … there’s nothing there that’s incriminating or weird, or stands out, so I just haven’t even touched it.  Participants deleted messages after showing study staff  f  21, Cape Town  Sometimes just for proof I was keeping [the messages on my phone] to show them when I go to the clinic. I was showing them so they won’t criticize me for not texting, to show them it says no response sometimes.  Age, study site  Privacy concerns  Participants felt others would not be suspicious of study text messages  Upon arrival of reminder message  g  38, Lima  [The daily reminder] is a quick message, and it’s anonymous because it just asks for a password, so, there’s no problem. Whether it be your partner or your mother, there is no reason for them to feel, I don’t know, to mistrust you because it is just asking for a password.  While texting  h  28, Chiang Mai  Other people might look at me and say that I was just using the Internet, Facebook, or sending messages online with my friends. … I am not concerned that other people might know what I am sending. I am not concerned that other people will sneak in and read the SMS.  h  36, Cape Town  I was not worried about [other people seeing what I was texting]. I was doing it even in front of them. They just didn't know what I was doing.  After texting  i  22, Cape Town  I don’t worry about that because nobody noses around my phone because I lock it.  Participants felt message content would not reveal study participation  j  22, Cape Town  It’s pretty much private itself because you only text your password, it doesn’t have any name or anything, and doesn’t say what it is all about. And it reminds you to delete everything when you are done.  j  51, Lima  I wouldn’t want my family to find out right now [that I am gay]. … So, I am assuming that all of this is private and that my name doesn’t appear. The text messages, well, it is something that is much more private because there are no names, no last names, you don't enter any of that information … and no one else sees my phone.  Participants had suggestions for how to make the system more private  k  43, San Francisco  Have something really, really vague where only the study participant and you guys know what the real question is, so that if they got the text across, nobody would be really suspicious, they’d be like, oh, what’s the weather like today.    Age, study site  Privacy concerns  Participants felt others would not be suspicious of study text messages  Upon arrival of reminder message  g  38, Lima  [The daily reminder] is a quick message, and it’s anonymous because it just asks for a password, so, there’s no problem. Whether it be your partner or your mother, there is no reason for them to feel, I don’t know, to mistrust you because it is just asking for a password.  While texting  h  28, Chiang Mai  Other people might look at me and say that I was just using the Internet, Facebook, or sending messages online with my friends. … I am not concerned that other people might know what I am sending. I am not concerned that other people will sneak in and read the SMS.  h  36, Cape Town  I was not worried about [other people seeing what I was texting]. I was doing it even in front of them. They just didn't know what I was doing.  After texting  i  22, Cape Town  I don’t worry about that because nobody noses around my phone because I lock it.  Participants felt message content would not reveal study participation  j  22, Cape Town  It’s pretty much private itself because you only text your password, it doesn’t have any name or anything, and doesn’t say what it is all about. And it reminds you to delete everything when you are done.  j  51, Lima  I wouldn’t want my family to find out right now [that I am gay]. … So, I am assuming that all of this is private and that my name doesn’t appear. The text messages, well, it is something that is much more private because there are no names, no last names, you don't enter any of that information … and no one else sees my phone.  Participants had suggestions for how to make the system more private  k  43, San Francisco  Have something really, really vague where only the study participant and you guys know what the real question is, so that if they got the text across, nobody would be really suspicious, they’d be like, oh, what’s the weather like today.    Note: a–k quotations correspond to in-text descriptions. Five people felt the password was useful.a This was especially salient in cases where participants had to use a phone that was not registered with the system; the password allowed them to be identified.b Five others had positive comments: 2 mentioned the password was good protection, while another said it was a nice thing to have. Two compared it to passwords for other systems. One said it was simple and convenient, and another said it was short and easy to remember.c Ten people had a neutral stance, stating it was fine and did not cause them any problems. Four participants had trouble remembering their password at first, but most were able to memorize it quickly, except for one who said he could not remember it. Finally, 3 people felt that the password was unnecessary, whether due to lack of privacy concerns or perceived inconvenience.d Reactions to instructions to erase messages from cell phones In the quantitative assessment, participants were also asked how much they were bothered by having to erase the SMS sessions from their cell phones. Again, the majority (82%) said they were not bothered at all, a few (14%) were bothered a little, and 4% were bothered a lot. There were no significant differences by country. We explored with participants in the IDIs whether they deleted the messages from their phones following each report, as instructed. A little over half of them (18 participants) said they did not delete the messages, and some specifically mentioned that they were not concerned about others seeing them.e Given that participants received a financial incentive when they returned for their clinic visit based on the number of SMS reports sent, many mentioned keeping the messages on their phone to show them to clinic staff as proof of sent messages. Some participants deleted the messages after receiving that month’s incentive. This was the case particularly at the Cape Town site, where issues with cellular service caused some messages not to be received by the system.f Those who deleted the messages often had reasons other than privacy, like keeping their message inbox from getting too full. Others deleted only the incoming messages from the SMS system, but kept the messages they had sent on their phones. Privacy concerns Finally, participants were asked via CASI how concerned they were about privacy while using the SMS system during the study. Most (85%) were not concerned at all about privacy, 10% were a little concerned, and only 5% reported being very concerned. Participants in South Africa showed a significantly higher level of concern about privacy while using the SMS system than participants in other countries, with a mean rating of 2.00 on a 3-point scale (1 = not concerned at all, 2 = a little concerned, 3 = very concerned), while mean scores for other countries ranged from 1.08 to 1.21 (P < .001). During the IDIs, participants did not report concerns about the privacy of study text messages at any point in the cycle of interaction with the system (upon receiving the first message, while texting the system, or after completing the text report), and no differences between countries were identified. One participant stated that the initial reminder only asked for a password and could not arouse the suspicion of others who might see it.g Participants were not worried that others might wonder with whom they were texting while responding to the system, stating that no one could know what they were doing by simply observing their texting behavior.h Finally, participants were unconcerned about others viewing their messages after they completed their responses.i Others made reference to the harmless content of the messages, which were worded carefully to prevent revealing study participation. This gave them assurance that their privacy was protected within the SMS system.j Finally, one had suggestions for making the first question more private so that no one would be suspicious if it flashed on the screen.k Overall, the great majority of participants did not have concerns about privacy while using the SMS system. For those who had concerns about revealing study participation to others, the privacy elements worked well to protect them. Privacy of HIV prevention product use, study participation, and sexual identity In addition to participants’ perceptions of SMS system privacy, we analyzed IDI data on the privacy of product use (see Table 4). Themes regarding disclosure of HIV prevention product use, study participation, and sexual identity emerged, allowing us to contextualize responses on SMS privacy within overall study-related privacy concerns. Table 4. Qualitative responses on privacy of sexual identity, HIV prevention product use, and study participation Age, study site  Overall privacy concerns  Participants were concerned about others’ assumptions upon disclosing study product use  l  30, Boston  I’ve told a couple of people, and they’d look at me like, like is he telling me that he’s HIV positive?  Participants did not want family members to learn of study participation for fear of disclosure of sexual identity  m  51, Lima  Well, I would also have to tell them, “So, this is how I am, I am gay, and so, my health is fine, I am taking [the pills] for a study” and all that, right, and my family, whom I live with currently, is a little – I could say – gossipy, right … and I have family members on my father’s side who are very religious. I think that, well it would be better to avoid all of this, by not telling them.  m  31, Cape Town  Ehh … it’s my mom. I never told her. Because she still hasn’t accepted the fact that I am gay. So things like this study I did not want to talk to her about it because of that.  Age, study site  Overall privacy concerns  Participants were concerned about others’ assumptions upon disclosing study product use  l  30, Boston  I’ve told a couple of people, and they’d look at me like, like is he telling me that he’s HIV positive?  Participants did not want family members to learn of study participation for fear of disclosure of sexual identity  m  51, Lima  Well, I would also have to tell them, “So, this is how I am, I am gay, and so, my health is fine, I am taking [the pills] for a study” and all that, right, and my family, whom I live with currently, is a little – I could say – gossipy, right … and I have family members on my father’s side who are very religious. I think that, well it would be better to avoid all of this, by not telling them.  m  31, Cape Town  Ehh … it’s my mom. I never told her. Because she still hasn’t accepted the fact that I am gay. So things like this study I did not want to talk to her about it because of that.  Note: l–m quotations correspond to in-text descriptions. Out of 33 participants interviewed, less than half discussed product use with friends (n = 16) and/or their partners or boyfriends (n = 14). Only 9 (27%) discussed it with family members, and 5 did not tell anyone they were using the study gel or tablets. Some felt comfortable talking about the study and their involvement in it. However, others expressed concern that people could assume they were HIV-infected if they disclosed product use, and some even had such an experience during the study.l In countries other than the United States, many participants lived with their parents and did not tell them about the study to avoid having to disclose their sexual identity. Four people specifically mentioned that their families either did not know or did not accept that they were gay, and they hid their study participation and study product use from family members for this reason.m Overall, many people in our sample did not feel comfortable discussing study participation or product use with others, indicating the relevance of exploring privacy issues around the SMS system. DISCUSSION Our study of participants’ perceptions of privacy while using SMS to report adherence to HIV-prevention product use during a Phase 2 study showed that most participants were not highly concerned about privacy while texting. Although some participants reported concerns about disclosure of study participation, sexual identity, or HIV-prevention product use, they were not concerned that the SMS system would compromise their confidentiality. Whether they considered texting behavior to be private or the text message language sufficiently ambiguous to protect their privacy, most did not have privacy concerns when responding to the SMS system. The use of SMS, a ubiquitous technology, did not exacerbate privacy concerns for participants, given that texting behavior is very common, and therefore did not arouse suspicion. Nevertheless, a subsample of 9 participants, 6 of whom were from South Africa and most of whom were unemployed, reported being very concerned about privacy while using the SMS system. Unfortunately, the subsample participants interviewed were not selected based on responses to CASI questions on SMS privacy, and none who reported SMS privacy concerns were interviewed. Given that phone sharing in South Africa is common23,24 and that these participants may have had fewer economic resources and an increased likelihood of sharing phones, this could be the reason for higher privacy concerns among this group. Prior research in African settings has found that participants with the lowest incomes who may be more likely to share mobile phones reported the greatest concerns with SMS privacy.7 Design features were effective in protecting participants’ privacy and mitigating any concerns. First, participants were able to select the time of day they preferred to receive the study messages. Participants in other studies also emphasized self-selection of time and/or day of SMS messages to protect privacy.12 Second, the first message, asking if they were “available to text,” was instrumental in preparing the participants for answering study questions. It also allowed them to physically position themselves in a place that could give them greater privacy while texting, if necessary. Third, the password had 2 functions. It was effective in preventing access to sensitive study questions without authorization, and it was linked to the participant ID, which allowed for participant tracking if a different phone number was used to report. This was critical in the data analysis phase to quickly link reports sent from multiple phones by a single participant. While a previous study found that having to enter a personal identification number to access the study SMS resulted in poorer access to the messages,15 in our study most participants were not bothered by having to enter a password, and some felt it provided an increased sense of privacy. Fourth, the security feature reminding participants to lock their cell phone and delete the study-related messages was not burdensome. While in other studies participants preferred to delete messages upon receipt to prevent others from viewing them,11,12 in our study many participants found it unnecessary and deleted the messages only after showing them to study coordinators and receiving compensation or when their message inbox became too full. Finally, the carefully constructed language of our SMS system prevented disclosure of study participation or the type of product being used. In prior research, participants suggested that language used in messages and how much it revealed about the nature of the study could either bolster or hinder confidentiality,25 and in this study the ambiguous language was key. Research has shown that participants may prefer coded messages, in which only they can understand the meaning,11 which can be as effective as direct messages about health information.15 Nevertheless, there is the potential for confusion with coded language, and participants prefer messages that clearly convey meaning to minimize uncertainty.11 We achieved this by designing a clear message that was carefully worded to preclude revealing study participation or the nature of the study product. Limitations While the qualitative data provide a rich description of SMS privacy and security concerns in this population, a larger sample size could allow for greater generalizability and comparative analyses. Furthermore, the IDI subsample was selected to explore adherence to the study products; therefore, we could not purposefully select those who reported fewer or greater SMS privacy concerns. In addition, our sample included different countries and cultural contexts, so we cannot make generalizations, given that small numbers of participants from each context were interviewed and others may have different perspectives. Variations in participants’ technological literacy or proficiency may have influenced their privacy concerns or lack thereof; however, these data were not available to supplement the analysis. CONCLUSION Privacy and security are concerns for researchers charged with the protection of research participants. Studies like ours can inform both investigators and IRBs on effective SMS practices to safeguard privacy and provide data security. Based on our evaluation of privacy and security features for an interactive SMS data collection system, the most useful features for health informatics practitioners who are developing SMS-based interventions include: (1) reminders that reinforce privacy and security-enhancing behaviors (ie, lock your phone, delete study-related messages), (2) ambiguous language that can be understood by study participants without revealing confidential information to others, and (3) the use of short, participant ID–related passwords that allow for participant tracking. These last 2 features are innovative and have yet to be highlighted in in the published SMS literature in combined use. COMPETING INTERESTS The authors have no competing interests to declare. CONTRIBUTORS RG, WB, CD, TH, AS, and MI collected and cleaned the data. RG, WB, and CD analyzed and interpreted the data. IB, WB, CD, JL, IM, RC, and ACD contributed to the study design. RG, WB, CD, IB, AS, and ACD drafted the report. TH, MI, JL, IM, and RC provided critical revision of the article. All authors gave input to the final version and provided final approval of the version to be published. FUNDING MTN-017 was sponsored by the US National Institutes of Health (NIH) and CONRAD (Contraception Research and Development). The study was designed and implemented by the Microbicide Trials Network (MTN). The MTN is funded by the National Institute of Allergy and Infectious Diseases (UM1AI068633, UM1AI068615, UM1AI106707), with co-funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Institute of Mental Health (NIMH), all components of the US National Institutes of Health. The study products used in this study were supplied by the Investigational New Drug sponsor, CONRAD, and the pharmaceutical collaborator, Gilead. This work was also supported by a center grant from the NIMH to the HIV Center for Clinical and Behavioral Studies at New York State Psychiatric Institute and Columbia University (P30-MH43520). WB was supported by the National Library of Medicine (R01-LM012355-01A1 and T15-LM007079) and the NIMH/National Institute on Drug Abuse (R03-MH103957). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. ACKNOWLEDGMENTS We are grateful to the study participants for their participation and dedication. We also thank the study team members at the research sites, who worked tirelessly to assist in implementing the SMS system and to train study participants to use it. We thank the protocol management team and the Microbicide Trials Network leadership operations center for their contributions to the study. References 1 Andriole KP. Security of electronic medical information and patient privacy: what you need to know. J Am Coll Radiol.  2014; 11: 1212– 16. Google Scholar CrossRef Search ADS PubMed  2 Tavani H. Privacy and security. Internet Ethics  .   London :  Macmillan; 2000: 65– 95. 3 White Y, Sandfort T, Morgan Ket al.  , Family relationships and sexual orientation disclosure to family by gay and bisexual men in Jamaica. Int J Sex Heal.  2016; 28: 306– 17. Google Scholar CrossRef Search ADS   4 Schrimshaw EW, Downing MJ, Cohn DJ. Reasons for non-disclosure of sexual orientation among behaviorally bisexual men: non-disclosure as stigma management. Arch Sex Behav.  2016; 1– 15. [Epub ahead of print]. 5 Sanou B. ICT Facts and Figures 2016 . Published online first: 2016. http://www.itu.int/en/ITU-D/Statistics/Documents/facts/ICTFactsFigures2016.pdf. Accessed July 29, 2016. 6 Kaplan WA. Can the ubiquitous power of mobile phones be used to improve health outcomes in developing countries? Global Health.  2006; 2: 9. Google Scholar CrossRef Search ADS PubMed  7 Nhavoto JA, Grönlund Å, Klein GO. Mobile health treatment support intervention for HIV and tuberculosis in Mozambique: Perspectives of patients and healthcare workers. PLoS One.  2017; 12: e0176051. Google Scholar CrossRef Search ADS PubMed  8 Crankshaw T, Corless IB, Giddy Jet al.  , Exploring the patterns of use and the feasibility of using cellular phones for clinic appointment reminders and adherence messages in an antiretroviral treatment clinic, Durban, South Africa. AIDS Patient Care STDs.  2010; 24: 729– 34. Google Scholar CrossRef Search ADS PubMed  9 Allison S, Bauermeister JA, Bull Set al.  , The intersection of youth, technology, and new media with sexual health: moving the research agenda forward. J Adolesc Heal.  2017; 51: 207– 12. Google Scholar CrossRef Search ADS   10 Acquisti A, Brandimarte L, Loewenstein G. Privacy and human behavior in the age of information. Science (80–).  2015; 347: 509 LP– 514. Google Scholar CrossRef Search ADS   11 Curioso WH, Alex Quistberg D, Cabello Ret al.  , “It’s time for your life”: How should we remind patients to take medicines using short text messages? AMIA Annu Symp Proc.  2009; 2009: 129– 33. Google Scholar PubMed  12 Rodrigues R, Poongulali S, Balaji Ket al.  , “The phone reminder is important, but will others get to know about my illness?” Patient perceptions of an mHealth antiretroviral treatment support intervention in the HIVIND trial in South India. BMJ Open.  2015; 5:e007574. 13 Sidney K, Antony J, Rodrigues Ret al.  , Supporting patient adherence to antiretrovirals using mobile phone reminders: patient responses from South India. AIDS Care.  2012; 24: 612– 17. Google Scholar CrossRef Search ADS PubMed  14 Hakkila J, Chatfield C. It’s like if you opened someone else’s letter: user perceived privacy and social practices with SMS communication. In: Proceedings of the 7th International Conference on Human Computer Interaction with Mobile Devices and Services.  2005; 219– 22. 15 Siedner JM, Santorino D, Haberer EJet al.  , Know your audience: predictors of success for a patient-centered texting app to augment linkage to HIV care in rural Uganda. J Med Internet Res.  2015; 17: e78. CrossRef Search ADS PubMed  16 Cranston RD, Lama JR, Richardson BAet al.  , MTN-017: a rectal phase 2 extended safety and acceptability study of tenofovir reduced-glycerin 1% gel. Clin Infect Dis.  2017; 64: 614– 20. Google Scholar PubMed  17 Brown III W, Giguere R, Ibitoye Met al.  , Successfully addressing challenges to implemeting a multinational SMS-based reminder and data collection system in a biomedical HIV prevention trial. AIDS Res Hum Retroviruses.  2014; 30: A87. Google Scholar CrossRef Search ADS   18 Brown III W, Giguere R, Sheinfil Aet al.  , Feasibility and acceptability of an international SMS text message–based adherence and survey system in a biomedical HIV prevention study (MTN-017). AIDS Res Hum Retroviruses.  2016; 32: 386. Google Scholar CrossRef Search ADS PubMed  19 Brown III W. Instrumentation and mobile technology for monitoring sexual behavior and product use in clinical trials. In: Microbicide Trials Network Annual Meeting . 2013. http://www.mtnstopshiv.org/sites/default/files/attachments/BROWN-MTN 017 SMS meeting presentation.pdf. Accessed July 24, 2017. 20 Patton MQ, Patton MQ. Qualitative Research and Evaluation Methods . 3rd ed. Thousand Oaks, CA: SAGE Publications; 2002. 21 John W. Creswell, Vicki L. Plano Clark. Chapter 3. Choosing a mixed methods research design. In: Designing and Conducting Mixed Methods Research . Thousand Oaks, CA: SAGE Publications; 2011: 53– 106. 22 MacQueen KM, McLellan E, Kay Ket al.  , Codebook development for team-based qualitative analysis. Cult Anthropol Methods.  1998; 10: 31– 36. 23 James J, Versteeg M. Mobile phones in Africa: how much do we really know? Soc Indic Res.  2007; 84: 117– 26. Google Scholar CrossRef Search ADS PubMed  24 Kreutzer T. Generation mobile: online and digital media usage on mobile phones among low-income urban youths in South Africa. Centre for Film and Media Studies, University of Cape Town, Cape Town; 2009. www.tinokreutzer.org/mobile. Accessed July 24, 2017. 25 Baranoski AS, Meuser E, Hardy Het al.  , Patient and provider perspectives on cellular phone-based technology to improve HIV treatment adherence. AIDS Care.  2014; 26: 26– 32. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the American Medical Informatics Association. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of the American Medical Informatics AssociationOxford University Press

Published: Apr 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off