HIV/AIDS, Medication Adherence
Conditions
Brief summary
Introduction of antiretroviral therapy (ART) has transformed HIV-infection from a fatal to manageable disease but adherence to ART remains critical to optimize outcomes. Existing measures of ART adherence provide only inferred measures of actual drug intake and most offer no real-time notification capability. Directly observed therapy measures actual drug intake but is not practical. These limitations constrain research into medication adherence and more importantly, limit our ability to develop real-time interventions based on feasible, in vivo monitoring of adherence among HIV-infected people to facilitate medication-taking. The Proteus digital health feedback (PDHF) system, a pill ingestible sensor based adherence measuring and monitoring system developed by Proteus Digital Health, addresses these limitations. It involves use of an ingestible sensor, a tiny edible material that is over-encapsulated along with prescribed medication. The sensor is activated by ingestion and is sensed by a patch worn by the patient with an embedded monitor and sensor. The monitor sends a Bluetooth signal to a mobile device, which in turn sends an encrypted message to a central server, thus effecting real-time monitoring that a dose has been taken. The investigators propose to develop a data receiving hub and add to these components an automated text message that is sent to the patient when a dose is missed. The investigators will evaluate the feasibility, acceptability and sustainability of using the PDHF system; assess the accuracy of the PDHF system in measuring adherence to ART; and evaluate the efficacy of the PDHF system for monitoring and leveraging adherence to ART.
Detailed description
Introduction of antiretroviral therapy (ART) has transformed HIV-infection from a fatal to manageable disease but adherence to ART remains critical to optimize outcomes. Existing measures of ART adherence such as self-report, pill counts, electronic pill-bottle caps, and prescription refills, provide only inferred measures of actual drug intake and most offer no real-time notification capability. Directly observed therapy measures actual drug intake but is not practical. These limitations constrain research into medication adherence and more importantly, limit our ability to develop real-time interventions based on feasible, in vivo monitoring of adherence among HIV-infected people to facilitate medication-taking. The Proteus digital health feedback (PDHF) system, a pill ingestible sensor based adherence measuring and monitoring system developed by Proteus Digital Health, addresses these limitations. It involves use of an ingestible sensor, a tiny edible material that is over-encapsulated along with prescribed medication. The sensor is activated by ingestion and is sensed by a patch worn by the patient with an embedded monitor and sensor. The monitor sends a Bluetooth signal to a mobile device, which in turn sends an encrypted message to a central server, thus effecting real-time monitoring that a dose has been taken. The investigators propose to develop a data receiving hub and add to these components an automated text message that is sent to the patient when a dose is missed. The ingestible sensor and patch monitor system is already FDA-approved as safe, but has yet to be tested in HIV-infected patients in clinical setting. The first goals of this study are to confirm the bioavailability of over-encapsulated antiretrovirals (ARVs) and to pilot-test the use of the PDHF system in 15 participants prescribed ARVs to test and identify approaches that optimize the use of this measuring and monitoring system. The next goals are to determine the system's feasibility, acceptability, sustainability, accuracy and efficacy in fostering ART adherence. Feasibility, acceptability and sustainability will be assessed by patients' rating of the system and the rate of dropping off from using the system. Accuracy will be evaluated by the associations between adherence to ART measured by the PDHF system and other adherence measures such as plasma drug level concentrations of ARVs and self-report. Efficacy will be assessed by comparing adherence of participants assigned to the PDHF system and participants assigned to usual care (UC) over time, with exploratory outcomes of viral load and cluster of differentiation 4 (CD4). The investigators will recruit 120 of HIV-infected patients 18 years or older with sub-optimal adherence. Participants will be randomized to receive the PDHF system or UC for 16 weeks with monthly assessments. The durability of effects of the PDHF system after stopping the use of the system will be determined during a 12-week follow-up stage. In summary, The investigators will evaluate the feasibility, acceptability and sustainability of using the PDHF system; assess the accuracy of the PDHF system in measuring adherence to ART; and evaluate the efficacy of the PDHF system for monitoring and leveraging adherence to ART.
Interventions
DEVICEProteus digital health feedback (PDHF) system
Building on the available Proteus devices, the investigators will design and create a PDHF system to transmit the adherence data using mobile technology to allow treatment monitoring that is, direct confirmation of the type, dose, date and time of oral pharmaceutical ingestion using wirelessly observed therapy (WOT).
Sponsors
Yale University
University of Nebraska
University of California, Los Angeles
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
OTHER
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
17 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* HIV-infected individuals in HIV care * greater than 17 years of age * demonstrated ability to take over-encapsulated ARVs at time of screening; able to provide informed consent * On ART with sub-optimal adherence estimated by either patient (self-reports \< 90% adherence over last 28 days) or treating clinician (e.g., based on gaps in treatment (e.g., missed appointments) or viral load elevations within last 6 months)
Exclusion criteria
* Inability to follow the study procedures manifested during the intake, as evidenced by mental confusion, disorganization, intoxication, withdrawal, risky or threatening behavior
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Adherence Measured by Sensor by Percentage of Prescribed Medications Taken | Adherence to ART measured by PDHF system for 16 weeks | Adherence to antiretroviral therapy (ART) measured by Proteus digital health feedback (PDHF) system for 16 weeks, including percent of prescribed medication taken. |
| Pharmacokinetic Adherence by Integrated Pharmacokinetic Adherence Score | Blood samples will be obtained in all participants before and 2 and 6 hours following an observed dose (baseline) and then at weeks 4, 8, 12, 16, 20, 24 and 28. | The plasma concentration-time data of tenofovir (TFV) from participants who have a tenofovir alafenamide (TAF) in their regimen are used to quantify intra-patient pharmacokinetic (PK) variability as a measure of adherence. Plasma concentrations of TFV are measured by liquid chromatography/tandem mass spectrometry. A population PK model will be developed using a nonlinear mixed-effects approach with data from all the time points. The integrated PK adherence score (IPAM) will be calculated. The IPAM score ranges from 0 to 1. A high score indicates high concentration predictability and relatively higher adherence, while a low score indicates low predictability and lower adherence. |
| Self-Reported Medication Adherence and Change Over Time | Self-report adherence will be measured at baseline, and weeks 4, 8, 12, 16, 20, 24, and 28. | The investigators will use a widely-used measure of self-reported adherence for percent of prescribed dose taken during the preceding seven days. This tool is easy to use and has been significantly associated with virological and immunological outcomes. Due to its potential bias, self-reported adherence will be calibrated by drug level concentration to leverage its accuracy and used in analysis when calibrated self-report adherence is appropriate to be used. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Viral Load | Baseline, weeks 4, 8, 12, 16, and 28. | Viral Load will be measured at baseline, weeks 4, 8, 12, 16, and 28. |
| Cluster of Differentiation 4 (CD4) | on: CD4 will be measured at baseline, weeks 4, 8, 12, 16, and 28. | CD4 cell count is a test that measures the number of CD4 cells (a type of the human T-lymphocyte cells) in a HIV patient's blood. The absolute CD4 cell count is measured by a simple blood test, the results of which are reported as the number of CD4 cells per cubic millimeter of blood. HIV-negative people typically have absolute CD4 cell counts between 600 and 1200 cells per cubic millimeter. HIV is a fatal infection, characterized by the targeting and destruction of CD4 cells. People with advanced HIV can have 200 or fewer CD4 cells per cubic millimeter. |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Intervention Building on the available Proteus devices, the investigators will design and create a Proteus digital health feedback (PDHF) system to transmit the adherence data using mobile technology to allow treatment monitoring that is, direct confirmation of the type, dose, date and time of oral pharmaceutical ingestion using wirelessly observed therapy (WOT).
The investigators will test overall utility (including feasibility, acceptability and sustainability) of the PDHF system, its accuracy for measuring adherence and its impact on enhancing patients' level of adherence and the effect on virologic and clinical outcomes (exploratory), the retention of its impact on keeping up with adherence and improvement of plasma HIV RNA and CD4 cell count after the 16-week usage of the PDHF system.
Proteus digital health feedback (PDHF) system: Building on the available Proteus devices, the investigators will design and create a PDHF system to transmit the adherence data using mobile technology to allow treatment monitoring that is, direct confirmation of the type, dose, date and time of oral pharmaceutical ingestion using wirelessly observed therapy (WOT). | 54 |
| Control UC is chosen as the control condition because it meets ethical and moral requirements to attempt treatment. Eligible patients will be randomized to one of the two conditions using a stratified urn randomization procedure to increase the likelihood of balanced allocation of prognostic variables at baseline. | 58 |
| Total | 112 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| Overall Study | Adverse Event | 1 | 0 |
| Overall Study | COVID-19 | 2 | 1 |
| Overall Study | Death | 1 | 2 |
| Overall Study | Lost to Follow-up | 10 | 7 |
| Overall Study | Out-of-country | 1 | 0 |
| Overall Study | Withdrawal by Subject | 5 | 1 |
Baseline characteristics
| Characteristic | Intervention | Control | Total |
|---|---|---|---|
| Age, Categorical <=18 years | 0 Participants | 0 Participants | 0 Participants |
| Age, Categorical >=65 years | 1 Participants | 4 Participants | 5 Participants |
| Age, Categorical Between 18 and 65 years | 53 Participants | 54 Participants | 107 Participants |
| Age, Continuous | 46.7 years STANDARD_DEVIATION 11.1 | 45.7 years STANDARD_DEVIATION 12.4 | 46.2 years STANDARD_DEVIATION 11.8 |
| Detectable Viral Load at Baseline Greater than 50 copies/mL | 19 Participants | 12 Participants | 31 Participants |
| Detectable Viral Load at Baseline Less than 50 copies/mL | 35 Participants | 46 Participants | 81 Participants |
| Education 8th grade or less/ Some high school but did not graduate | 8 Participants | 10 Participants | 18 Participants |
| Education Completed college/ More than four-year college degree | 8 Participants | 9 Participants | 17 Participants |
| Education High school graduate/ Some college but no degree | 38 Participants | 39 Participants | 77 Participants |
| Employment None/full-time student/retired/disabled | 45 Participants | 39 Participants | 84 Participants |
| Employment Part-time/full-time | 9 Participants | 19 Participants | 28 Participants |
| History of AIDS Diagnosis No | 42 Participants | 46 Participants | 88 Participants |
| History of AIDS Diagnosis Yes | 12 Participants | 12 Participants | 24 Participants |
| HIV+ Years | 14.5 years STANDARD_DEVIATION 8 | 14.0 years STANDARD_DEVIATION 9.3 | 14.2 years STANDARD_DEVIATION 8.7 |
| Race/Ethnicity, Customized Race and Ethnicity Asian | 1 Participants | 1 Participants | 2 Participants |
| Race/Ethnicity, Customized Race and Ethnicity Black | 24 Participants | 30 Participants | 54 Participants |
| Race/Ethnicity, Customized Race and Ethnicity Latino | 19 Participants | 16 Participants | 35 Participants |
| Race/Ethnicity, Customized Race and Ethnicity Other | 4 Participants | 3 Participants | 7 Participants |
| Race/Ethnicity, Customized Race and Ethnicity White | 6 Participants | 8 Participants | 14 Participants |
| Sex/Gender, Customized Gender Female | 5 Participants | 5 Participants | 10 Participants |
| Sex/Gender, Customized Gender Male | 44 Participants | 43 Participants | 87 Participants |
| Sex/Gender, Customized Gender Transgender: Male to Female | 5 Participants | 10 Participants | 15 Participants |
| Years under Antiretroviral Treatment | 12.6 years STANDARD_DEVIATION 7.5 | 11.6 years STANDARD_DEVIATION 7.7 | 12.0 years STANDARD_DEVIATION 7.6 |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 1 / 57 | 1 / 58 |
| other Total, other adverse events | 1 / 57 | 0 / 58 |
| serious Total, serious adverse events | 0 / 57 | 0 / 58 |
Outcome results
Primary
Adherence Measured by Sensor by Percentage of Prescribed Medications Taken
Adherence to antiretroviral therapy (ART) measured by Proteus digital health feedback (PDHF) system for 16 weeks, including percent of prescribed medication taken.
Time frame: Adherence to ART measured by PDHF system for 16 weeks
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Intervention | Adherence Measured by Sensor by Percentage of Prescribed Medications Taken | 87.5 percentage of prescribed medication | Standard Deviation 33.1 |
Primary
Pharmacokinetic Adherence by Integrated Pharmacokinetic Adherence Score
The plasma concentration-time data of tenofovir (TFV) from participants who have a tenofovir alafenamide (TAF) in their regimen are used to quantify intra-patient pharmacokinetic (PK) variability as a measure of adherence. Plasma concentrations of TFV are measured by liquid chromatography/tandem mass spectrometry. A population PK model will be developed using a nonlinear mixed-effects approach with data from all the time points. The integrated PK adherence score (IPAM) will be calculated. The IPAM score ranges from 0 to 1. A high score indicates high concentration predictability and relatively higher adherence, while a low score indicates low predictability and lower adherence.
Time frame: Blood samples will be obtained in all participants before and 2 and 6 hours following an observed dose (baseline) and then at weeks 4, 8, 12, 16, 20, 24 and 28.
Population: For the trial phase, we pick the most common drug with the longest plasma half-life to use for the PK-based method of adherence. That ended up being tenofovir alafenamide (TAF). Only 86 of the participants in the final analysis had available data for the PK analysis.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Intervention | Pharmacokinetic Adherence by Integrated Pharmacokinetic Adherence Score | 0.817 score on a scale | Standard Deviation 0.276 |
| Control | Pharmacokinetic Adherence by Integrated Pharmacokinetic Adherence Score | 0.798 score on a scale | Standard Deviation 0.267 |
Comparison: Compared the IPAM score between two groups.p-value: 0.57Wilcoxon (Mann-Whitney)
Primary
Self-Reported Medication Adherence and Change Over Time
The investigators will use a widely-used measure of self-reported adherence for percent of prescribed dose taken during the preceding seven days. This tool is easy to use and has been significantly associated with virological and immunological outcomes. Due to its potential bias, self-reported adherence will be calibrated by drug level concentration to leverage its accuracy and used in analysis when calibrated self-report adherence is appropriate to be used.
Time frame: Self-report adherence will be measured at baseline, and weeks 4, 8, 12, 16, 20, 24, and 28.
Population: Lost to follow-up, missed visits, or no self-reported adherence.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Intervention | Self-Reported Medication Adherence and Change Over Time | Week 0 | 90.7 percentage of prescribed medication | Standard Deviation 14.5 |
| Intervention | Self-Reported Medication Adherence and Change Over Time | Week 16 | 88.7 percentage of prescribed medication | Standard Deviation 17.5 |
| Intervention | Self-Reported Medication Adherence and Change Over Time | Week 8 | 90.6 percentage of prescribed medication | Standard Deviation 17 |
| Intervention | Self-Reported Medication Adherence and Change Over Time | Week 20 | 90.7 percentage of prescribed medication | Standard Deviation 20.9 |
| Intervention | Self-Reported Medication Adherence and Change Over Time | Week 4 | 91.1 percentage of prescribed medication | Standard Deviation 15.7 |
| Intervention | Self-Reported Medication Adherence and Change Over Time | Week 24 | 94.8 percentage of prescribed medication | Standard Deviation 8.8 |
| Intervention | Self-Reported Medication Adherence and Change Over Time | Week 28 | 93.0 percentage of prescribed medication | Standard Deviation 11.8 |
| Intervention | Self-Reported Medication Adherence and Change Over Time | Week 12 | 88.3 percentage of prescribed medication | Standard Deviation 17 |
| Control | Self-Reported Medication Adherence and Change Over Time | Week 28 | 90.1 percentage of prescribed medication | Standard Deviation 19.8 |
| Control | Self-Reported Medication Adherence and Change Over Time | Week 0 | 87.4 percentage of prescribed medication | Standard Deviation 13.2 |
| Control | Self-Reported Medication Adherence and Change Over Time | Week 4 | 87.4 percentage of prescribed medication | Standard Deviation 15.2 |
| Control | Self-Reported Medication Adherence and Change Over Time | Week 8 | 91.6 percentage of prescribed medication | Standard Deviation 13.5 |
| Control | Self-Reported Medication Adherence and Change Over Time | Week 12 | 82.1 percentage of prescribed medication | Standard Deviation 24.9 |
| Control | Self-Reported Medication Adherence and Change Over Time | Week 16 | 91.7 percentage of prescribed medication | Standard Deviation 16.3 |
| Control | Self-Reported Medication Adherence and Change Over Time | Week 20 | 84.5 percentage of prescribed medication | Standard Deviation 23.2 |
| Control | Self-Reported Medication Adherence and Change Over Time | Week 24 | 87.3 percentage of prescribed medication | Standard Deviation 22.4 |
Secondary
Cluster of Differentiation 4 (CD4)
CD4 cell count is a test that measures the number of CD4 cells (a type of the human T-lymphocyte cells) in a HIV patient's blood. The absolute CD4 cell count is measured by a simple blood test, the results of which are reported as the number of CD4 cells per cubic millimeter of blood. HIV-negative people typically have absolute CD4 cell counts between 600 and 1200 cells per cubic millimeter. HIV is a fatal infection, characterized by the targeting and destruction of CD4 cells. People with advanced HIV can have 200 or fewer CD4 cells per cubic millimeter.
Time frame: on: CD4 will be measured at baseline, weeks 4, 8, 12, 16, and 28.
Population: Lost to follow-up, missed visits, or failed to draw blood CD4.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Intervention | Cluster of Differentiation 4 (CD4) | Week 12 | 501.20 cells/cubic millimeters | Standard Deviation 251.46 |
| Intervention | Cluster of Differentiation 4 (CD4) | Week 0 | 526.17 cells/cubic millimeters | Standard Deviation 279.1 |
| Intervention | Cluster of Differentiation 4 (CD4) | Week 16 | 522.91 cells/cubic millimeters | Standard Deviation 273.61 |
| Intervention | Cluster of Differentiation 4 (CD4) | Week 8 | 522.00 cells/cubic millimeters | Standard Deviation 280.32 |
| Intervention | Cluster of Differentiation 4 (CD4) | Week 28 | 538.80 cells/cubic millimeters | Standard Deviation 301.42 |
| Intervention | Cluster of Differentiation 4 (CD4) | Week 4 | 531.02 cells/cubic millimeters | Standard Deviation 273.85 |
| Control | Cluster of Differentiation 4 (CD4) | Week 28 | 587.69 cells/cubic millimeters | Standard Deviation 318.24 |
| Control | Cluster of Differentiation 4 (CD4) | Week 0 | 542.78 cells/cubic millimeters | Standard Deviation 279.31 |
| Control | Cluster of Differentiation 4 (CD4) | Week 4 | 553.08 cells/cubic millimeters | Standard Deviation 271.42 |
| Control | Cluster of Differentiation 4 (CD4) | Week 8 | 582.25 cells/cubic millimeters | Standard Deviation 293.29 |
| Control | Cluster of Differentiation 4 (CD4) | Week 12 | 577.20 cells/cubic millimeters | Standard Deviation 295.79 |
| Control | Cluster of Differentiation 4 (CD4) | Week 16 | 562.96 cells/cubic millimeters | Standard Deviation 314.09 |
Secondary
Viral Load
Viral Load will be measured at baseline, weeks 4, 8, 12, 16, and 28.
Time frame: Baseline, weeks 4, 8, 12, 16, and 28.
Population: Lost to follow-up, missed visits, or failed to draw blood viral load.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Intervention | Viral Load | Week 28 | 1.957 log10 copies/mL | Standard Deviation 0.816 |
| Intervention | Viral Load | Week 4 | 1.916 log10 copies/mL | Standard Deviation 0.793 |
| Intervention | Viral Load | Week 16 | 1.895 log10 copies/mL | Standard Deviation 0.691 |
| Intervention | Viral Load | Week 8 | 1.851 log10 copies/mL | Standard Deviation 0.68 |
| Intervention | Viral Load | Week 0 | 2.273 log10 copies/mL | Standard Deviation 1.158 |
| Intervention | Viral Load | Week 12 | 1.873 log10 copies/mL | Standard Deviation 0.653 |
| Control | Viral Load | Week 0 | 2.061 log10 copies/mL | Standard Deviation 0.954 |
| Control | Viral Load | Week 16 | 1.988 log10 copies/mL | Standard Deviation 0.895 |
| Control | Viral Load | Week 28 | 2.230 log10 copies/mL | Standard Deviation 1.117 |
| Control | Viral Load | Week 12 | 1.937 log10 copies/mL | Standard Deviation 0.671 |
| Control | Viral Load | Week 4 | 1.906 log10 copies/mL | Standard Deviation 0.516 |
| Control | Viral Load | Week 8 | 1.916 log10 copies/mL | Standard Deviation 0.711 |
Comparison: Compared the change of the plasma HIV RNA levels during the intervention period (week 0-16) between two groups.p-value: 0.08Mixed Models Analysis
Comparison: Compared the change of the plasma HIV RNA levels during the post-intervention period (week 16-28) between two groups.p-value: 0.23Mixed Models Analysis
Comparison: Compared the plasma HIV RNA levels during week 4-28 between two groups.p-value: 0.03Mixed Models Analysis