HIV, Medication Adherence
Conditions
Keywords
biological markers, monitoring device, wireless technology
Brief summary
Currently available methods to monitor antiretroviral (ARV) adherence to HIV treatment and prevention, such as self-report, pill counts, medication electronic monitoring system (MEMS), and devices which wirelessly monitor adherence in real-time, have multiple limitations, including over-reporting, inability to assess pill ingestion, and size/expense. Our multidisciplinary research team at the University of California, San Francisco (UCSF) and in India has designed a new adherence-monitoring device called Tel-Me-Box which is small, low-cost, rechargeable, inconspicuous, and could be programmed to deliver tailored real-time adherence reminders following additional hardware modifications. The aim of this study is to modify and validate this discreet Tel-Me-Box (TMB) adherence monitoring and reminder device against hair ARV concentrations as a pharmacologic measure of drug ingestion/adherence, a measure pioneered and validated by our team, with the expectation that this device and intermittent hair monitoring could have widespread utility for HIV and non-HIV adherence science, both in India and globally.
Detailed description
Novel, validated methods to monitor adherence to HIV treatment in real time are urgently needed given the well-known limitations of self-reported adherence, pill counts, and MEMS caps. In order to be scalable and sustainable in both resource-rich and resource limited settings, such devices need to be low-cost and designed in a way that does not increase the risk of accidental HIV disclosure and subsequent stigma. Devices which both monitor adherence in real time and provide tailored reminders to patients for medication-taking have the potential to greatly improve adherence to HIV treatment as well as to treatment for other chronic diseases. Our Indo-US collaborative team has been conducting research on HIV stigma, ARV adherence patterns and barriers, treatment outcome, and drug resistance for over a decade. The study investigators have also pioneered the use of ARV concentrations in small hair samples to measure long-term ARV adherence in resource-rich and limited settings (RLS). In response to the need for novel adherence measures, our team has recently developed Tel-Me-Box, a small, low-cost adherence device that monitors adherence in real time, by transmitting a wireless signal to a server when opened. The investigators now propose to add hardware that will enable the server software to wirelessly program the device to activate tailored medication-taking reminders (via a beep, vibration, or LED light) after a period of inactivity. The device fits in a pocket, can hold 1-2 weeks' worth of medications, has a long battery life, simple charging capability, and has been found acceptable to Indian patients with adherence challenges in pilot studies. Since HIV stigma serves as a profound barrier to ARV adherence in many settings, including India, the small, inconspicuous nature of Tel-Me-Box, along with its ability to deliver tailored reminder features honors participants' privacy concerns. Since this electronic device monitoring cannot assess actual drug ingestion, validating it against a biological measure of adherence, such as ARV hair concentrations, is crucial. This study includes a pilot randomized control trial (RCT) to examine acceptability and feasibility and to estimate the effect size of automated tailored real-time adherence reminders on hair ARV concentrations, device- monitored adherence, and viral load suppression in adherence-challenged patients. The aim of this study is to validate TMB as an innovative tool to assess adherence and predict treatment outcomes, and demonstrate the feasibility of hair analyses in RLS. If successful, pilot data will be available on the efficacy of the Tel-Me-Box automated, tailored reminders for a future large scale RCT.
Interventions
BEHAVIORALTel-me-box with reminder features
Participants will meet with a study staff member to go over their daily routines, including their pill taking schedule, and to work out a plan for tailored reminders. The participant will choose from the reminders available and convey their preferences regarding when reminders should be sent for the TMB. Preferences will be stored in an online secure database. If requested by the participant, the reminder schedule and type of reminder can be reprogrammed by study staff via an online portal. When the reminder is reprogrammed the TMB receives a new configuration SMS to update the device reminder settings.
BEHAVIORALTel-me-box
Participants receive tel-me-box device with no reminder features added.
Sponsors
National Institute of Mental Health (NIMH)
University of California, San Francisco
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
OTHER
Masking
SINGLE (Outcomes Assessor)
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* HIV-positive * Currently on antiretroviral therapy (ART) * Speaks one of the local languages * Able and willing to participate in the study, provide informed consent, contact information, and express a willingness to return for follow-up visits every three months for 12 months after baseline. * Adherence-challenged per self-report, i.e. report of \<90% ART adherence or \>2 ART treatment interruptions of at least 2 days in the past 3 months
Exclusion criteria
* Unwilling or unable to participate in the study and/or return for follow-up appointments * younger than 18 years old * Not HIV-infected * Not currently on ART * Does not speak one of the local languages
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Tel-me-box Adherence | 30 days prior to each follow-up assessment (3, 6, 9 and 12 months post baseline). | Number of days over past 30 days with at least one device-recorded pill box opening. |
| HIV Viral Load (VL) | Baseline, and 6 and 12 month follow-up | HIV viral load (number of copies/ml of blood), dichotomized as undetectable VL (= suppressed, meaning \<50 copies/ml detected) vs. detectable (or unsuppressed, meaning \>= 50 copies/ml detected) |
| Hair Concentrations of Anti-retroviral Medications (ARV), in ng/mg | 12 month follow-up | Hair concentrations of Efavirenz (EFV) or Dolutegravir (DTG) in ng/mg, and log transformed to improve normality of the distribution if necessary. The vast majority of participants were on an ARV regimen that included either EFV or DTG at the time of the study. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Self-reported Adherence in the Past Month | 30 days prior to baseline and 3, 6, 9, and 12 month follow-up | Participant self-report measured using the visual analogue scale to assess percent of pills taken in the past month, further dichotomized into optimal (100%) vs. suboptimal adherence (\<100%) |
Countries
India
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Pillbox ('Tel-me-box') With Reminder Features Participants randomly assigned to the intervention reminder condition will choose from the reminders available and convey their preferences regarding when reminders should be sent for the tel-me-box device.
Tel-me-box with reminder features: Participants will meet with a study staff member to go over their daily routines, including their pill taking schedule, and to work out a plan for tailored reminders. The participant will choose from the reminders available and convey their preferences regarding when reminders should be sent for the TMB. Preferences will be stored in an online secure database. If requested by the participant, the reminder schedule and type of reminder can be reprogrammed by study staff via an online portal. When the reminder is reprogrammed the TMB receives a new configuration SMS to update the device reminder settings. | 69 |
| Pillbox ('Tel-me-box') Without Reminder Features The control arm will include tel-me-box monitoring only. No reminder features will be included with the device.
Tel-me-box: Participants receive tel-me-box device with no reminder features added. | 62 |
| Total | 131 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| 12 Month Follow-up = Final Assessment | Death | 0 | 1 |
| 12 Month Follow-up = Final Assessment | Lost to Follow-up | 7 | 3 |
| 3 Month Follow-up Assessment | Lost to Follow-up | 5 | 4 |
| 3 Month Follow-up Assessment | Withdrawal by Subject | 2 | 0 |
| 6 Month Follow-up Assessment | Lost to Follow-up | 7 | 6 |
| 6 Month Follow-up Assessment | Withdrawal by Subject | 2 | 1 |
| 9 Month Follow-up Assessment | Lost to Follow-up | 9 | 4 |
Baseline characteristics
| Characteristic | Pillbox ('Tel-me-box') With Reminder Features | Total | Pillbox ('Tel-me-box') Without Reminder Features |
|---|---|---|---|
| Age, Continuous | 37.3 years STANDARD_DEVIATION 10.1 | 39.4 years STANDARD_DEVIATION 10.1 | 41.8 years STANDARD_DEVIATION 9.6 |
| Race (NIH/OMB) American Indian or Alaska Native | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Asian | 69 Participants | 131 Participants | 62 Participants |
| Race (NIH/OMB) Black or African American | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) More than one race | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) White | 0 Participants | 0 Participants | 0 Participants |
| Region of Enrollment India | 69 participants | 131 participants | 62 participants |
| Sex: Female, Male Female | 35 Participants | 61 Participants | 26 Participants |
| Sex: Female, Male Male | 34 Participants | 70 Participants | 36 Participants |
| Type of reminder feature (in intervention group only) Buzzer | 34 Participants | 34 Participants | 0 Participants |
| Type of reminder feature (in intervention group only) Light + vibration | 35 Participants | 35 Participants | 0 Participants |
| Type of reminder feature (in intervention group only) None (only applicable to arm without reminder features) | 0 Participants | 62 Participants | 62 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 69 | 1 / 62 |
| other Total, other adverse events | 0 / 69 | 0 / 62 |
| serious Total, serious adverse events | 0 / 69 | 0 / 62 |
Outcome results
Primary
Hair Concentrations of Anti-retroviral Medications (ARV), in ng/mg
Hair concentrations of Efavirenz (EFV) or Dolutegravir (DTG) in ng/mg, and log transformed to improve normality of the distribution if necessary. The vast majority of participants were on an ARV regimen that included either EFV or DTG at the time of the study.
Time frame: 12 month follow-up
Population: Of the 131 participants who started the pilot trial, 111 were interviewed at 12 mo. follow-up. Ten of these participants were on a regimen that did not include EFV or DTG, resulting in n=101 for analysis.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Pillbox ('Tel-me-box') With Reminder Features | Hair Concentrations of Anti-retroviral Medications (ARV), in ng/mg | 0.12 log10 transformed concentration (ng/mg) | Standard Deviation 0.71 |
| Pillbox ('Tel-me-box') Without Reminder Features | Hair Concentrations of Anti-retroviral Medications (ARV), in ng/mg | 0.28 log10 transformed concentration (ng/mg) | Standard Deviation 0.64 |
p-value: <0.0595% CI: [-0.11, 0.43]t-test, 2 sided
Primary
HIV Viral Load (VL)
HIV viral load (number of copies/ml of blood), dichotomized as undetectable VL (= suppressed, meaning \<50 copies/ml detected) vs. detectable (or unsuppressed, meaning \>= 50 copies/ml detected)
Time frame: Baseline, and 6 and 12 month follow-up
Population: Overall number analyzed = number of participants enrolled at baseline. Numbers analyzed reported below reflect missing data for VL due to failed lab tests or loss to follow-up.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Pillbox ('Tel-me-box') With Reminder Features | HIV Viral Load (VL) | Undetectable VL at Baseline | 37 participants |
| Pillbox ('Tel-me-box') With Reminder Features | HIV Viral Load (VL) | Undetectable VL at 6mo FU | 34 participants |
| Pillbox ('Tel-me-box') With Reminder Features | HIV Viral Load (VL) | Undetectable VL at 12 mo FU | 34 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | HIV Viral Load (VL) | Undetectable VL at Baseline | 34 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | HIV Viral Load (VL) | Undetectable VL at 6mo FU | 45 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | HIV Viral Load (VL) | Undetectable VL at 12 mo FU | 44 participants |
Comparison: statistical analyses is a GEE model with logit link function and binomial distribution with odds of undetectable VL as outcome and as predictors: intervention, time and time\*intervention. We hypothesize that over time, the odds of undetectable VL will be higher for the intervention than the control group, which would be reflected in a significant interaction effect time\*intervention.~H0: no significant interaction effect time\*interventionp-value: 0.08Chi-squared
Primary
Tel-me-box Adherence
Number of days over past 30 days with at least one device-recorded pill box opening.
Time frame: 30 days prior to each follow-up assessment (3, 6, 9 and 12 months post baseline).
Population: Overall number analyzed = number of participants enrolled at baseline. Numbers analyzed reported below reflect missing data due to loss to follow-up.
| Arm | Measure | Group | Value (MEDIAN) |
|---|---|---|---|
| Pillbox ('Tel-me-box') With Reminder Features | Tel-me-box Adherence | No. days box opened (0-30) at 9 mo. follow-up | 0 days (out of 30) |
| Pillbox ('Tel-me-box') With Reminder Features | Tel-me-box Adherence | No. days box opened (0-30) at 3 mo. follow-up | 10 days (out of 30) |
| Pillbox ('Tel-me-box') With Reminder Features | Tel-me-box Adherence | No. days box opened (0-30) at 12 mo. follow-up | 0 days (out of 30) |
| Pillbox ('Tel-me-box') With Reminder Features | Tel-me-box Adherence | No. days box opened (0-30) at 6 mo. follow-up | 1.5 days (out of 30) |
| Pillbox ('Tel-me-box') Without Reminder Features | Tel-me-box Adherence | No. days box opened (0-30) at 12 mo. follow-up | 0 days (out of 30) |
| Pillbox ('Tel-me-box') Without Reminder Features | Tel-me-box Adherence | No. days box opened (0-30) at 3 mo. follow-up | 16.5 days (out of 30) |
| Pillbox ('Tel-me-box') Without Reminder Features | Tel-me-box Adherence | No. days box opened (0-30) at 9 mo. follow-up | 1 days (out of 30) |
| Pillbox ('Tel-me-box') Without Reminder Features | Tel-me-box Adherence | No. days box opened (0-30) at 6 mo. follow-up | 7 days (out of 30) |
Comparison: We used zero inflated negative binomial regression due to the count variable being overdispersed and containing a large number of 0. The analysis aggregated over multiple time points per participant. It simultaneously tests the odds of having 0 openings and the difference in the rate of openings between the 2 intervention groups. These 2 results are reported separately.~Null hypothesis 1: the odds of having 0 openings is lower in intervention than control group.95% CI: [0.9, 1.93]
Comparison: We used zero inflated negative binomial regression due to the count variable being overdispersed and containing many 0s. The analysis aggregated over multiple time points per participant. It simultaneously tests the odds of having 0 openings and the difference in the rate of openings between the 2 intervention groups. These 2 results are reported separately.~Null hypothesis 2: the incidence rate ratio for # days with box opening is \>1, i.e. more openings in intervention than control group.95% CI: [0.61, 0.98]
Secondary
Self-reported Adherence in the Past Month
Participant self-report measured using the visual analogue scale to assess percent of pills taken in the past month, further dichotomized into optimal (100%) vs. suboptimal adherence (\<100%)
Time frame: 30 days prior to baseline and 3, 6, 9, and 12 month follow-up
Population: Overall number of participants = number enrolled at baseline. At follow-up waves the number analyzed is lower due to loss to follow-up.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Pillbox ('Tel-me-box') With Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at 3 month follow up | 49 participants |
| Pillbox ('Tel-me-box') With Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at 9 month follow up | 45 participants |
| Pillbox ('Tel-me-box') With Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at 6 month follow up | 44 participants |
| Pillbox ('Tel-me-box') With Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at 12 month follow up | 45 participants |
| Pillbox ('Tel-me-box') With Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at baseline | 4 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at 12 month follow up | 50 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at baseline | 5 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at 3 month follow up | 37 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at 6 month follow up | 35 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | Self-reported Adherence in the Past Month | Optimally adherent at 9 month follow up | 40 participants |
Comparison: statistical analyses is a GEE model with logit link function and binomial distribution with odds of optimal adherence as outcome and as predictors: intervention, time and time\*intervention. We hypothesize that over time, the odds of optimal adherence will be higher for the intervention than the control group, which would be reflected in a significant interaction effect time\*intervention.~H0: no significant interaction effect time\*interventionp-value: 0.08Chi-squared
Post Hoc
Presence of Tel-me-box Adherence Data (>0) in Past 30 Days
The number of participants for whom at least 1 pill box opening was registered in the 30 days before an assessment
Time frame: 30 days prior to assessment at 3, 6, 9, and 12 mo follow-up
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Pillbox ('Tel-me-box') With Reminder Features | Presence of Tel-me-box Adherence Data (>0) in Past 30 Days | any box use data at 3 mo FU | 47 participants |
| Pillbox ('Tel-me-box') With Reminder Features | Presence of Tel-me-box Adherence Data (>0) in Past 30 Days | any box use data at 6 mo FU | 38 participants |
| Pillbox ('Tel-me-box') With Reminder Features | Presence of Tel-me-box Adherence Data (>0) in Past 30 Days | any box use data at 9 mo FU | 26 participants |
| Pillbox ('Tel-me-box') With Reminder Features | Presence of Tel-me-box Adherence Data (>0) in Past 30 Days | any box use data at 12 mo FU | 16 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | Presence of Tel-me-box Adherence Data (>0) in Past 30 Days | any box use data at 12 mo FU | 22 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | Presence of Tel-me-box Adherence Data (>0) in Past 30 Days | any box use data at 3 mo FU | 48 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | Presence of Tel-me-box Adherence Data (>0) in Past 30 Days | any box use data at 9 mo FU | 31 participants |
| Pillbox ('Tel-me-box') Without Reminder Features | Presence of Tel-me-box Adherence Data (>0) in Past 30 Days | any box use data at 6 mo FU | 41 participants |