Video Directly Observed Therapy for Treatment Adherence Monitoring

Video Directly Observed Therapy for Treatment Adherence Monitoring
Richard S. Garfein
11/09/2018
UCSD HIV & Global Health Rounds

2. Video Directly Observed Therapy (VDOT) for
Tuberculosis Treatment Monitoring
Presented by:
Richard S. Garfein, PhD, MPH
UC San Diego, School of Medicine
for
HIV & Global Health Rounds
November 9, 2018
1

3. Disclosures
•Dr. Garfein is a co-founder of SureAdhere Mobile Technology
(a VDOT service provider).
•No services or funding were provided by SureAdhere for any data
presented in this talk
2

4. Collaborators
•UC San Diego, School of Medicine: Kevin Patrick, Jazmine Cuevas-Mota , Kelly Collins, Fatima Munoz, Valerie
Mercer, Donald Catanzaro, Maria Luisa Zuniga, Jose Luis Burgos, Timothy Rodwell, Lin Liu, Michelle Bulterys, Erin
South, Diana Do
•UC San Diego, Qualcomm Institute: Fredric Raab, Phillip Rios, Allison Flick, Mark Sullivan, Ganz Chockalingam,
David McCarter
•San Diego County Health and Human Services Agency
•San Francisco Department of Public Health
•Santa Clara, San Joaquin and Imperial County Health Departments
•California Department of Public Health
•New York City Department of Health: Christine Chuck, Nikolas Mitropoulos, John Soma, Virginia Vasquez-Stewart
•ISESALUD, Tijuana, BC, Mexico: Dra. Liliana Andrade, Dra. Lidia Perez, Dr. Luis Garcia, Dr. HéctorZepeda Cisneros,
Dr. Gustavo López, CristhianColin
•Mexico-US Border Health Commission, BC, Mexico: Dra. Gudelia Rangel Gómez, Carlos Cota, Gabriela Escalante,
•Secretary of Health of the State of Baja California: Dr. Jonathan Figueroa
•Tijuana General Hospital -Integrated Care Services (SAI):Dr. Samuel Navarro
•CAPASITS, Tijuana, BC, Mexico: Dr. Mario Lam
Funded by the National Institutes of Health (R21-AI088326; U01-AI116392), Verizon Foundation, Alliance Healthcare
Foundation and California HealthCare Foundation, UC San Diego CFAR.3

5. 4
Does adherence really matter in TB?
Source: TB ReFLECTConsortium, unpublished data3
TB patients taking RIPE with <90%
adherence had 5.6 times increased risk of TB
recurrence in a meta-analysis of the
OFLOTUB, REMox, and Rifaquintrials.
Study of 534 smear-positive patients in India
found a strong relationship between adherence
and post-treatment TB recurrence.
Severity of non-
adherence
TBrecurrence
rate,18 months
after completing
treatment
“Regular”
adherence9%
“Irregular”
adherence15%
“Very irregular”
adherence25%
Source: Thomas et al. IntJ TB Lung Dis 2005; 9(5): 556-614
Slide adapted from Bruce Thomas, The Arcady Group

6. •YES!Poor medication adherence may
have the following impact on health
outcomes:
§Increased disease relapse
§Increased acquired drug resistance
ØModeling study found adherence to be the
strongest predictor for the emergence of MDR TB
in retreatment patients8
5
Does adherence really matter in TB?
“[P]robablythe cheapest and most effective
way to ensure a positive treatment outcome
while minimizing the risk for the emergence
of MDR-TB is to maintain proper patient
compliance with the treatment.”8
Source: Cadoschet al. PlosComp Bio 2016;12:e10047498
Slide adapted from Bruce Thomas, The Arcady Group

7. Monitoring TB Medication Adherence
•Purpose:
•Document that all doses were taken
•Promote treatment completion
•Detect drug adverse effects
•Goals:
•Reduce TB morbidity and mortality
•Prevent TB transmission
•Prevent acquired drug resistance

8. Directly Observed Therapy (DOT)
•Patient observed swallowing each dose
•Recommended by CDC & WHO
•Rationale:
•Improves adherence
•Reduces risk of acquired drug resistance,
treatment failure, and relapse
•Establishes rapport with patients
•Allows intermittent dosing
•May reduce pill burden
•Saved 6.8 million lives in 1995-2010
Community-Based DOT
Clinic-Based DOT

9. First-Line TB Treatment
Initial phase (8 weeks):
•4 drugs daily (~500 pills)
Continuation phase (18 weeks):
•2 drugs daily (~500 pills)
~1000 pills over 6 months
CDC, http://www.cdc.gov/tb/topic/treatment/tbdisease.htm

10. “VDOT”
9

11. Types of Video Directly Observed
Therapy (VDOT)
•SynchronousVDOT
•Videoconferencing
•“Real-time” or “Live”
•AsynchronousVDOT
•Recorded videos
•“Store-and-forward”
(Wireless)(Wired)
10

12. SynchronousVDOT Process
Scheduling
Connecting
Observing
Documenting
11

13. SynchronousVDOT
Evidence
•High adherence/completion1-3,5-7
•High patient & provider
satisfaction1,3-6
•Resource saving1-5,7
Limitations
•Limited to business hours
•Scheduling requires staff time
•Dosing times not always optimal
for patients
•Requires consistent and reliable
connectivity
•Not all video-conferencing apps
meet security standards for PHI
(i.e., HIPAA or GDPR)
1DeMaio, et al., ClinInfect Dis, 2001; 2Krueger, et al., IntJ TubercLung Dis, 2010; 3Gassanov, et al., Can J Public Health, 2013; 4Bethel, et al., ATS
Conf., 2006; ; 5Mirsaeidi, et al., EurRespJ, 2015; 6Wade, et al., PLoSOne, 2012; 7Chuck, et al., IntJ TubercLung Dis, 201612

14. AsynchronousVDOT Process
Review
ObserveDocument
Store
Secure cloud or local server
Record
Date/ Time stampEncryptUpload
13

15. Patient Dashboard helps providers
visualize how each dose was taken,
and immediately identify missed
doses.
Provider Dashboard shows adherence at the
patient and program level. Missed and taken
doses are color-coded for easy identification

16. VDOT Pilot Study -2010
•Objective:
•To assess the feasibility, acceptability and potential
efficacy of VDOT for monitoring TB treatment
•Setting:
•TB control programs in San Diego and Tijuana
•Conducted by program staff
•Design:
•Phase I = Focus groups
•Phase II = Single-arm pilot trial
•Eligibility: Age>18, pansensitivepulmonary TB, DOT-experienced
•Patients interviewed pre & post VDOT
Funding: NIH R21-AI088326; PI: R. Garfein

17. Patient Education
Materials (English/Spanish)

18. VDOT Pilot Trial: Demographics
San DiegoTijuana
Number enrolled in VDOT439
Number of participants who spent time in both cities60
Age: mean(range)38.9 (18-86)28.1 (18-65)
Hispanic or Latino n(%)18 (41.9)9 (100)
Race: n(%)
Asian
African American/Black
Pacific Islander/Native Hawaiian
Caucasian/White
Other/Mixed Race
13 (30.2)
3 (6.9)
2 (4.7)
9 (21.0)
16 (37.2)
0 (0)
0 (0)
0 (0)
3 (33.4)
6 (66.6)
Gender: n(%)
Male
Female
23 (53.5)
20 (46.5)
5 (55.5)
4 (45.5)
Number of participants who switched back to in-person DOT*6 (13.9)1 (11.1)
Telephones stolen/replaced1/22/1
*3 patients with drug resistant TB returned to in-person DOT per protocol.

19. 93%
7%
VDOT Pilot Study: Results I
VDOT adherence
(observed / expected
videos)
96%
4%
San Diego (n=43)Tijuana (n=9)
How often did you have
problems recording a video?
Is VDOT more or less
confidential than in-person
DOT?
93%
7%
89%
11%
80%
15%
5%
76%
22%
Videos
received
Videos not
received
Never/rarely
≥1/2 the time
More
No difference
Less

20. VDOT Pilot Study Results II
Would you rather use VDOT
or in-person DOT ifyou had
to repeat your TB treatment?
Would you recommend
VDOT to other TB patients?
Are you more or less
comfortable using a
smartphone since using
VDOT?
93%
5%
…
89%
11%
VDOT
Either
In-Person
100%100%Yes
No
68%
32%
89%
11%
More
comfortable
No
difference
San Diego (n=43)Tijuana (n=9)

21. $4,167
$458
$1,293
$174
$0
$500
$1,000
$1,500
$2,000
$2,500
$3,000
$3,500
$4,000
$4,500
SAN DIEGOTIJUANA
COST PER CASE FOR 6 MONTHS
In-Person DOT
VDOT
VDOT Pilot Study: Cost Analysis
In-PersonDOT Costs
•Staff salaries and transportation
VDOT Costs
•Staff salaries, transportation, and phoneservice
•Excludesthecostof theVDOT application

22. Results Published in International Journal
of TB and Lung Disease

23. 1. Focus Groups 2010 (N=52)
-San Diego, CA & Tijuana, Mexico
-Included patients, providers, health officials
2. Pilot Study 2010-2012 (N=52)
-San Diego, CA & Tijuana, Mexico
-Included only pansusceptiblepulmonary TB
3. Expansion Study 2012-2015 (N=149)
-San Diego, San Francisco, New York City
-Permitted DR-TB and suspected TB
4. P3Study 2015-2016 (N=274)
-(Urban) San Diego, San Francisco, Santa Clara; (Rural) San Joaquin, Imperial Counties
-Permitted any type of TB
-Compared in-person DOT with VDOT
Common Study Features
-Multisite, single arm, prospective study
-Pre & post VDOT interviews (demographics, perceptions)
-Eligibility:
qAge >18 years
q≥30 days of treatment left
qPatient initially on DOT (min. 2 weeks)
qNo plans to move
qNo physical/cognitive conditions that prevent VDOT

24. Baseline Characteristics by Study
Variable
San Diego Pilot
(n=43)
Expansion
(n=149)
P3Study
(n=275)
VDOT Observation Rate–Mean % (SD)
Range %
93.0%
(51-100)
84.0% (14.5)
(33-100)
87.9% (13.5)
(28-100)
Months on VDOT –Mean (SD)
Range
5.5
(1-11)
5.4 ( 3.2)
(0.3-18.1)
5.5 (2.9)
(0.3-18.1)
Age –Mean (SD)
Range
39.0 (17.6)
(18-86)
40.9 (16.0)
(18-87)
43.8 (16.5)
(18-87)
Education (n)
<High school
>High school
38% (16)
62% (26)
41% (60)
59% (85)
49% (131)
51% (139)
Gender (n)
Male
Female
54% (23)
46% (20)
59% (88)
41% (61)
61% (167)
39% (106)
Race (n)
Asian
Black/African-American
Caucasian/White
Hispanic/Latino
Other/Mixed Race
31% (13)
7% ( 3)
14% ( 6)
40% (17)
7% ( 3)
45% (67)
13% (20)
7% (10)
30% (44)
6% ( 8)
57% (154)
1% ( 3)
7% ( 19)
30% ( 82)
6% ( 17)
Country Born (n)
U.S.
Mexico
Other
23% (34)
13% (19)
64% (96)
17% ( 47)
16% ( 44)
67% (181)
Owned Smartphone at Baseline (n) 69% (103)72% (196)23

25. Multivariable Linear RegressionAnalysis of Factors Associated with VDOT Observation Rate
Expansion Study
(n=149)
P3Study
(n=274)
VariableBeta (SE)P-valueBeta (SE)P-value
Intercept0.878 (0.02)0.886 (0.02)
Country of Birth: (ref: Other*)
Mexico
U.S.
-0.131 (0.04)
-0.071 (0.03)
0.001
0.018
-0.096 (0.02)
-0.052 (0.02)
<0.001
0.023
Months on VDOT (per month)--------0.007 (0.01)0.012
Found VDOT process to be: (ref: Very Easy)
‘Somewhat Easy’
‘Somewhat/Very Difficult’
0.055 (0.03)
-0.225 (0.08)
0.078
0.004
----
-0.171 (0.04)
----
<0.001
Problems using VDOT >1/2 the time (ref: Never)---------0.090 (0.02)<0.001
*Most countries in “Other” category were in Asia.
Garfein, et al., NTCA Meeting, Palm Springs, CA, May 21-25, 201824

26. VDOT Observation Rates (2010-2016)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
San Diego (n=43)*San Diego (n=100)
Tijuana (n=9)
Tijuana HIV+ (n=18)*San Francisco (n=99)New York City (n=48)*Santa Clara (n=49)*San Joaquin (n=15)*Imperial (n=11)
% of Expected Doses Observed
*Rates increased when unviewable videos and partial doses were counted as “observed”
(SD=88%; SF=94%; SC=92%; SJ=98%; Imp=93%)
25

27. VDOTFEDODOT FEDODOT Adherence
0.0
0.2
0.4
0.6
0.8
1.0
Fraction of Expected Doses Observed
Median=93%
Median=66%
Median=100%
P<.001
P<.001
FEDO= number of doses observed divided by the number of doses expected
Adherence= number of doses observed via DOT divided by the number of prescribed doses
VDOT vs. DOT in 5 California TB Programs
26Garfein, et al., EmergingInfectiousDiseases 2018;24:1806-1815

28. •DOT personnel costs included time for patient contact, administrative tasks, and travel.
•VDOT personnel costs included time for DOT visits prior to initiating VDOT, patient VDOT training, administrative tasks,
video observation, and follow-up when expected videos were not received.
•Corporate pricing used for smartphones ($100) and service plans ($54/mth). VDOT app priced at $35/patient/mth.
•Excludes costs for antibiotics, laboratory tests, chest radiographs and clinical exams.
$4609$4549
$4888
$3212
$5788
$3141$3179
$3911
$3031$3137
0
1000
2000
3000
4000
5000
6000
7000
Cost (USD)
In-person DOT
VDOT
-30%-20%-6%-46%
TotalSan DiegoSan FranciscoSan JoaquinImperial
-32%
Cost to Monitor Standard 6-Months of
TB Treatment, California (2015-2016)
27Garfein, et al., EmergingInfectiousDiseases 2018;24:1806-1815

29. Patient Perspectives
Split Dosing (31 y.o., male, U.S.)
“The video was really helpful because of dual doses at different times and it would be really hard to do it in person”
Transportation (21 y.o., female, Mexico)
“I think the phone is better because I don’t have to call in to the hospital and go there to take my medication. Also I sometimes don’t have enough money to take the bus to the clinic”
Autonomy (70 y.o., female, U.S.)
“When I started I was depressed because I got sick from someone I was caring for. At first it was difficult for me to accept the treatment b/c they have to see me in person. I felt like I lost my freedom and felt like a prisoner. But this alternative was great”
(37 y.o., male, Mexico)
“VDOT is really good. It gives you confidence and makes you responsible for yourself and make a conscious decision to take your drugs without having someone else tell you to take them.” 28

30. VDOT in Belarus
Projectmilestones
•2015 WHO/ERS digital health strategy published
•Jan 2015 : Feasibility assessment of VDOT (WHO)
•Feb 2015: VDOT included in GF project proposal;
MoHVDOT working group
•May 2015 : technical specifications for VOT app; software finished
by Jan 2016
•Jan 2016 : pilot single center VOT project
•Oct 2016: programmatic expansion of VOT with GF support
Slide courtesy of Alena Skrahina, Belarus NTP

31. Programmatic expansion of VOT in Belarus
Patients regional distribution
-52
35
Minsk City71
Minskregion40
Gomel region52
Mogilev region21
Vitebskregion
Brest region
Grodno region
24
5
18
9
16
425
15
Slide courtesy of Alena Skrahina, Belarus NTP

32. Programmatic Expansion of VDOT in Belarus, 2017
n=231
Gender m/f 139/92 (60%/40%)
Age
18-25
26-35
36-45
46-55
56-65
>65
37 (16%)
75 (32%)
60 (26%)
31 (13%)
26 (11%)
2 (1%)
Social data
employed
student
unemployed
maternity leave
military service
121 (52%)
12 (5%)
86 (37%)
10 (4%)
2 (1%)
DR profile
Drug susceptible
Monoresistance(R)
MDR
Pre-XDR
XDR
110 (47%)
6 (2%)
45 (19%)
36 (16%)
34 (15%)
n=231
Total MDR-TB -50%
Slide courtesy of Alena Skrahina, Belarus NTP

33. Belarus VDOT Treatment Outcomes
Final treatment outcomes in 61 TB patients on VDOT:
•Treatment success (DS-TB) –59 (96%)
•Death (XDR-TB) –1 (2%)
•LTFU (MDR-TB) –1 (2%)
•170 patients are still on treatment
Nationwide scale-up ongoing
Slide courtesy of Alena Skrahina, Belarus NTP

34. VDOT forTB TreatmentamongHIV+
Patients, Tijuana, Mexico
UCSD CFAR DevelopmentalPilotGrant (PI: F. Munoz)
•Objectives
•ToevaluateVDOT formonitoringTB treatmentadherence
amongHIV/TB co-infectedpersonsin Tijuana, Mexico
•Methods
•HIV-positive patients (N=19) prescribed oral TB medications recruited from CAPASITS and Tijuana General Hospital
•Eligibility
§>18yearsold
§AbletospeakSpanish
§NoplanstomovefromTijuanawithinthestudyperiod
§Willingandabletoprovideinformedconsent
§DataCollection
§15-minute baseline and follow-up interviews assessed sociodemographic, clinical, risk factor and VDOT perception variables

35. VDOT
(n=19)
n (%)
In-Person DOT*
(n=50)
n (%)
Mean Age (SD)
Range
34.1(8.2)
25-56
34.3 (16.0)
21-55
Gender:
Male
Female
13 (68.4)
6 (31.6)
38 (76.0)
12 (24.0)
Treatment Outcome
Continue in treatment
Completion of treatment (Cured)
Lost to follow-up (Abandonment)
Died during TB treatment
Treatment suspended
12 (57.1)
4 (19.0)
2 ( 9.5)
0
1 ( 5.2)
n/a
12 (28.5)
12 (28.5)
6 (14.2)
1 ( 2.3)
Fraction of Expected Doses Observed (mean)**94.7% 95.5%
TB Treatment Using VDOT among HIV+ Patients, Tijuana, Mexico
*Adherence for historical controls obtained through medical record review for a
random sample of patients treated for TB in the 1 year prior to the study.

36. n (%)
Overall,I’m “somewhat/very satisfied”with TB treatment using VDOT17 (100)
I wouldchoose VDOT over DOT if treatment had to be repeated17 (100)
I would you recommend VDOT to other patients17 (100)
Confidentiality of VDOT compared to DOT
VDOT > DOT11 (64.7)
VDOT = DOT6 (35.3)
Overall, how difficult was the VDOT process?
Somewhat/Very easy14 (82.4)
Somewhat Difficult2 (11.7)
Very difficult1 ( 5.9)
How often did you have problems with the VDOT application?
Never 12 (70.6)
Rarely 5 (29.4)
HIV+ Patient Satisfaction with VDOT for TB Treatment, Tijuana, Mexico (n=17)

37. WHO Endorsement
*Results not available at the time of WHO Guidelines publication showed that VDOT had twice the efficacy and was more acceptable
to patients and providers compared to DOT. Story A. VDOT in the UK: Results of a Randomized Control Trial. 2017 Union NAR Meeting,
Vancouver, Canada. https://bc.lung.ca/lung-disease/union-north-america/2017-meeting-archives.
*
36

38. United Kingdom TB Reach Study
•Design –Randomized controlled trial
•Location –London and Birmingham, England
•Study Period –2013-2017
•Eligibility
•Include-TB patient age >16 years eligible for DOT at participating
clinics
•Exclude-can’t charge phone / <2 months treatment remaining
•Primary outcome
•Proportion of patients with >80% of doses observed
•Data collection –Baseline and follow-up interviews
Story A, Aldridge RW, Smith CM, et al., Smartphone-enabled video observed versus directly observed treatment for
tuberculosis: a randomisedcontrolled trial. Lancet(in press)

39. TB Reach Flow
Diagram

40. TB Reach Baseline Characteristics
0
10
20
30
40
50
60
70
80
90
16-2425-3435-4445-5455-64>=65MaleFemaleNo YesNo Yes
Age group Sex Born in UKPrevious TB
VDOT DOT

41. Complex cases
* Any current social risk factor = Homeless, problem drug use, alcohol, prison
0
10
20
30
40
50
60
70
Social risk
factor (any) *
Homeless PrisonDrug misuseAlcohol
misuse
Mental
illness
Immigration
concerns
DOT
VOT

42. Proportion of Subjects who had >80% of
Scheduled Doses Observed by Month
Numbers above bars are numbers of patients who had scheduled treatment
observations in each month following randomisation and numbers who completed
>80% of scheduled observations. Error bars are 95% confidence intervals.
VDOT DOT

44. VDOT Cost Analysis: United Kingdom
•Cost to provide DOT for six months
•£5,700 per patient for five times per week
•£3,420 per patient for three times per week
•Cost to provide daily VDOT for six months
•£1,645 per patient for seven times per week

45. VDOT for Monitoring Adherence to LTBI
Treatment (V-MALT)
NIH U01-AI116392; PI: R. Garfein
Study Aims
ØTo determine whether latent TB infection (LTBI) treatment adherence
and completion are greater for patients treated with 12 doses of
Isoniazid and Rifapentine (3HP) administered through VDOT
compared to DOT
ØTo compare treatment perceptions between VDOT and DOT
ØTo measure the cost-effectiveness between VDOT and DOT

46. VMALT Study Population & Eligibility
•Sample
•310 patients with LTBI (n=155 per arm)
•Recruited from San Diego County TB Control Program LTBI clinics and UCSD
Student Health Services
•Eligibility Criteria
•Candidate for 3HP (based on CDC and HHSA guidelines)
•Age ≥13 years
•Plan to stay in San Diego area for next 4 months
•Nophysical or cognitive disabilities that preclude VDOT
•Unless household member can assist for the duration of the study
•Willing to follow study procedures and provide informed consent
•Timeline
•Feb. 2016 –ongoing = Enrollment period
•Aug. 2019 –Feb. 2020 = Data analysis and reporting

47. VDOT in High Burdon Low Healthcare
Access Regions
Ongoing Projects:
•Haitian Prisons
•Philippines MDR-TB Patients
•Kazakhstan TB Patients
•Vietnam Private Sector
•Uganda
UNITAID and StopTBPartnership
have awarded several million
dollars for pilot and implementation
projects in over 8 countries to
scale-up VDOT globally

48. Are Digital Adherence
Technology Apps Enough?
•In-person DOT
•Synchronous VOT
•Asynchronous VOT
•99DOTS
•WisePill
•SMS
•Ingestible sensors
•AI observation
•Others
•Physicians
•DOT workers
•Case managers
•Health educators
•Family supporters
•Surveillance
?
?
Patient InputsPatient Supporters

49. Opportunities in HIV for VDOT?
•LTBI treatment for HIV+
•Establishing treatment routines for ARV initiates
•Supporting patients with low ARV adherence
•Monitoring adherence to PrEP
•Monitoring adherence in drug trials
•Other uses?

50. Questions?
49