Keeping Up with Consumers - Duke-Margolis · 2019-05-23 · through an app on her phone. Ingestible sensors track her medication adherence. Christie authenticates her apps and devices

D k I MARGOLIS CENTER U e for Health Policy

Keeping Up with Consumers:

How can mHealth apps and wearables

help evaluate medical products?

mHealth as a Source of Real-World Data (RWD) – Working Group

Duke-Margolis Center for Health Policy

June 26, 2017, 2 – 4pm EST

Tweet: @dukemargolis #RealWorldDigitalHealth

Partic:ipants

v Participants

CS El > Cl'nisti111a Silcox I me,. internal}


X

V

Notes

Housekeeping • We will be recording this presentation, including any comments or questions from

the participants.

• The slides from the presentation as well as the audio recording will be publically posted on the Duke-Margolis Center for Health Policy website: https://healthpolicy.duke.edu/

• To make a comment or to ask a question during the comment periods of this presentation, please type into the Q&A box or use the “raise hand” feature .

- Longer comments can be emailed to [email protected] by July 12, 2017

You can only use the “raise hand” feature if you are using audio through the computer or put in an individual attendee ID when calling in. When we call on you to make your comment, we will unmute you.

2

https://healthpolicy.duke.edu/mailto:[email protected]


Disclosure

Funding for this working group was made possible by the Food and Drug Administration through grant 7U01FD004969. Views expressed in the written materials and by speakers and moderators do not necessarily reflect the official policies of the Department of Health and Human Services nor does any mention of trade names, commercial practices, or organization imply endorsement by the United States Government.

3


Working Group Members

•Ashish Atreja (Sinai AppLab)

•David Bates (Brigham and Women’s Hospital)

•Seth Clancy (Edwards Lifesciences)

•Gregory Daniel (Duke-Margolis Center for Health Policy)

•Megan Doerr (Sage Bionetworks)

•Patricia Franklin (FORCE-TJR, University of Massachusetts Medical School)

•Adrian Hernandez (Duke University)

•Martin Ho (Food and Drug Administration)

•Mohit Kaushal (Stanford University)

•John Mattison (Kaiser Permanente)

•Mike McConnell (Verily Life Sciences, Stanford University School of Medicine)

•Greg Pappas (Food and Drug Administration)

•Bakul Patel (Food and Drug Administration)

•Ravi Ramachandran (PatientsLikeMe)

•John Reites (THREAD)

•Annie Saha (Food and Drug Administration)

•Kevin Weinfurt (Duke University)

4


Webinar Agenda • Project Scope • Addressing Needs to Build Engagement

What causes patients to regularly and sustainably engage with mHealth apps and wearables?

Why would mHealth companies add functionality necessary for evidence generation?

How can researchers and mHealth companies work together to promote methods for frequent, real-

life, and/or novel measurements?

- Public Feedback

• Proposed mHealth Data Types Person-Reported Data

Task-Based Measures Data

Passive Sensing Data

- Public Feedback

• Next Steps and Closing Comments - Public comment period details

5


mHealth – Challenges and Opportunities

• Inflection point in the availability and usability of mHealth technologies - Empowers patients to participate in their own health

- Proliferation and advancement of consumer wellness and health technology

- Development of analytical tools enabling the transformation of data into knowledge

- NEST can support the use of mHealth as real-world data for the evaluation of medical devices

• Enormous growth in opportunity, but also in the complexity and cost of both healthcare and clinical research.

• Patients are becoming an integral part of how medical products are researched, developed, marketed, and used

• mHealth is a disruptive technology including: - Feeding into the next generation of evidence development platforms

- Regulatory and research policies need to keep up with promising technologies and methods

6


Project Scope

• mHealth apps and wearables can collect data from patients about their experiences and symptoms, as well as objective data about activities

- Medical studies often collect similar data (PROs, PerfOs) - Opportunity for novel outcome measures important to patients

• Real-world evidence generation and NEST - Already being used for consumer and clinical use

• Requirements for research use - Accuracy measurements

Validation studies in patient group of interest

- Sustained longitudinal collection

- Predictability for how the data will be verified, transmitted, and linked

7


Outside of the Scope

• Evaluating mHealth apps and wearables as medical devices

- Does using a mHealth technology in research studies make it a “medical

device”?

App/wearable is a medical device when it is intended for use in the diagnosis of disease

or other conditions, or in the cure, mitigation, treatment or prevention of disease

FTC tool to find out when FDA, Federal Trade Commission (FTC) or Office of Civil

Rights (OCR) laws apply - https://www.ftc.gov/tips-advice/business-

center/guidance/mobile-health-apps-interactive-tool

8

https://www.ftc.gov/tips-advice/business-center/guidance/mobile-health-apps-interactive-tool


“Christie” is a 39-year-old female with a history of high blood pressure. She completes her annualphysical exam each year. Like most people in 2024, Christie monitors her weight, steps, energyexpenditure, sleep, blood pressure, stress, and heart rate through wearable trackers, some of which have been provided through her insurer or employer. She also keeps food- and mood-tracking journals through an app on her phone. Ingestible sensors track her medication adherence.

Christie authenticates her apps and devices so that her PGHD could be fully integrated into centralizedsoftware that also records EHR and pharmacy information (such as flu shots received at her local pharmacy and cold medicine prescribed at the community health clinic) without needing individual action from Christie. Within a centralized app, Christie controls who can see and use her information byauthorizing and revoking access.

With her permission, algorithms linking data from these multiple sources continuously analyze Christie’s health status and alert Christie and/or her doctor’s practice if there is a significant concern.

“Christie” is a 39-year-old female with a history of high blood pressure. She completes her annualphysical exam each year. Like most people in 2024, Christie monitors her weight, steps, energyexpenditure, sleep, blood pressure, stress, and heart rate through wearable trackers, some of which have been provided through her insurer or employer. She also keeps food- and mood-tracking journals through an app on her phone. Ingestible sensors track her medication adherence.

Christie authenticates her apps and devices so that her PGHD could be fully integrated into centralizedsoftware that also records EHR and pharmacy information (such as flu shots received at her local pharmacy and cold medicine prescribed at the community health clinic) without needing individual action from Christie. Within a centralized app, Christie controls who can see and use her information byauthorizing and revoking access.


Connected/Quantified Life in 2024 “Christie” is a 39-year-old female with a history of high blood pressure. She completes her annual physical exam each year. Like most people in 2024, Christie monitors her weight, steps, energy expenditure, sleep, blood pressure, stress, and heart rate through wearable trackers, some of which have been provided through her insurer or employer. She also keeps food- and mood-tracking journals through an app on her phone. Ingestible sensors track her medication adherence.

Christie authenticates her apps and devices so that her PGHD could be fully integrated into centralized software that also records EHR and pharmacy information (such as flu shots received at her local pharmacy and cold medicine prescribed at the community health clinic) without needing individual action from Christie. Within a centralized app, Christie controls who can see and use her information by authorizing and revoking access.


Accenture Draft White Paper for a PGHD Policy Framework https://www.healthit.gov/sites/default/files/Draft_White_Paper_PGHD_Policy_Framework.pdf

9

https://www.healthit.gov/sites/default/files/Draft_White_Paper_PGHD_Policy_Framework.pdf

Olstrt onJry Studies

UDI lncorpor ted

Into EHi

Post•Appro JI St dies


ng

Leveraging Real-World Evidence

• Types of real-world data (RWD) • EHR • Claims • Registries • mHealth

• Efficient linking and analysis of RWD • Standardized informed consent

guidelines, data use agreements, and

data governance practices

• Common data models and data quality standards

• Well-characterized methodologies

10Image: FDA (2013) Strengthening Our National System for Medical Device Postmarket Surveillance: Updates and Next Steps.


How do we get from 2017 to 2024?

• What types of mHealth data are useful now?

• What reasonable changes could technology companies incorporate to make mHealth data more useful for evidence generation in the near-term?

- What challenges are impeding these changes?

• How do we engage people to securely share data for evidence generation? - What would meaningful informed consent look like?

- How can the use of these tools be sustained to ensure the data is useful?

• How can researchers and mHealth companies work together to promote methods for frequent, real-life, and/or novel measurements?

- What information about the validity and reliability of mHealth technology must be characterized for the data to be used for evidence generation?

- What makes an off-the-shelf mHealth solution more attractive than a proprietary solution designed for a specific study?

11


What will the Action Plan address? • Contribution of mHealth data for novel real-world evidence generation

- Role of NEST in encouraging the inclusion and use of mHealth in evidence development for medical devices

• Ways to think about different types of person-facing mHealth technologies, software, and data

• Recommendations for advancing person-facing mHealth adoption and usage - User engagement

- Researcher/sponsor needs

- mHealth company incentives

• Overarching challenges in digital health data: Current work and resources - Best practices for patient-consumer informed consent

- Data linkages and interoperability

- Accommodating diversity of wearable technology and application

- Fit-for-purpose: validation and reliability of data

• Recommended next steps to advance consumer/clinical mHealth technologies as a viable source of reliable data for evidence generation

12







- Public Feedback


Task-Based Data

Passive Sensing

- Public Feedback


13


Balancing Stakeholder Benefits

User Engagement

Why would people sustainably use mHealth apps and wearables for the time frames necessary for evidence generation?

14

Sponsor/Researcher Needs

What needs do mHealth data fill for researchers?

mHealth Company Incentives

Why add the functionality necessary for evidence generation?


User Engagement

User engagement is necessary for individuals, and/or their care givers, to start and

maintain use of mHealth technology

- What characteristics of mHealth apps and wearables encourage sustained usage?

- How can we ease the transition from being a user of mHealth into being a research participant?

15


User Engagement – What should the action plan highlight?

1. Approaches to cultivate sustainable usage Easy to use

Actionable information to support ongoing input in their own health status and care

• Personal wellness management

• Integration with clinical information in ways that assist clinicians with shared decision-making

Person-centred design

Security and privacy

Emphasis of value to society (i.e., altruism)

2. Features of successful mHealth apps and wearables 3. Potential implications for evidence generation from features meant to

increase user value (e.g., biasing)

4. Recommendations to bridge transition from “user” to “research participant”

16


!ID Original Paper

Formative Evaluation of Participant Experience With Mobile eConsent in the App-Mediated Parkinson mPower Study: A Mixed Methods Study Megan Doe1r, MS, CGC • ; Amy Maguire T1Uong, MS • ; Brian M Bot, BS 41!) ; Jolin Wilbanks, BA 41!) ; Christine Suver, PhD • ;

Lara M Mangravite, PhD •

Real World User Engagement

• From patientslikeme.com

From https://mhealth.jmir.org/2017/2/e14/

- I very much like participating. I feel as if I am helping to reach an overall outcome

- I lost interest/motivation and stopped recording for a while...

17

https://mhealth.jmir.org/2017/2/e14/http:patientslikeme.com

STRONG E GAGE ENT

@ : I ·@~ . -I ······~·····1······ ............ . . .... ~ ..... .

' ' • • I ' ' ••••••~•••••••••••••••••••P••••••P•••••~•••••• I ' I I •

·····{·····

. . --. -~- -... ~- -. -..

-----~----··

. I • ······,·····'\······ Q : . .

. . . . , ...... ······r····· .. . . •••••• ~ •••••• ~ ••••• J •••••• . . -..... • .. -.... \. -.. -. .

. . I • . ........... ··,·····'\············

FRICTION ©

POOR ENGAGEMENT

Engagement Predictor Tool

Motivation Factors

Financial incentives

Clinical benefit

Altruistically fulfilling

Enjoyable (education or entertainment)

Passive data collection

Added value to daily living

Friction Factors

Interaction frequency

Duration of interaction

Convenience of interaction

Cognitive load requirements

Active data collection

Reminders / notifications

CONFIDENTI!L ©️THREAD 2017 www.THREADresearch.com

http:www.THREADresearch.com

- - - - -

- - - - - -D k I MARGOLIS CENTER U e for Health Policy

Currently using PGHD

Wi 11 s tart neJ(t year

No plans to use PGHD

Wi 11 start in 2+ years

Clinical Use of Patient Generated mHealth Data Do clinicians use patient generated mHealth

data (PGHD) in clinical decision-making?

NO • Only 6% currently use PGHD in clinical decision-

making.

• Majority use the information as part of patient engagement in health status and activity goals

• Few received data directly from a device

• 81% do not use wearable data from patients for clinical decision-making.

19 Leventhal R. MGM! Poll: Providers Not Yet Using Data from Patients’ Wearables. Healthcare informatics. June 20, 2017. Retrieved from https://www.healthcare informatics.com/news item/patient engagement/mgma poll providers not yet using data patients wearables

- - - -- - - - - -


2011 KUCIC HEALTI-1 CONSUMER SURVEY HEALTHCARE INNOIIAT 0

Help me stay healthy Help me better manage my condition

Help me locate the best treatment options Help me provide better care to a loved one

Help me prevent sickness or disease

Help me communicate with others about my health condition Help me locate the best healthcare professionals

Help my physician or caregiver(s) better empathize with my condition

Help my physician provide earlier diagnosis

Other It won't have a positive impact on my health

24% 15% 10%

3% 14% 4%

3% 2% 14%

1% 10%

Q. Where do you thlnk technology will have the most posltl11e Impact on your health In the future?

26% 12%

10%

3% 12%

3% 4%

3% 13%

1%

12%

lh"'lifaii iiii!iit¥11PAM 21% 29% 23% 19%

17% 10% 15% 18%

10% 8% 9% 13%

2% 4% 3% 1% 15% 14% 15% 13%

5% 6% 4% 2%

3% 4% 3% 3% 1% 2% 2% 3%

15% 11% 17% 14%

2% 0% 1% 3%

9% 12% 9% 11%

And How Are People Using PGHD?

Survey by Klick Health of 1012 US adults

70% say technology will help them personally manage their health

41% use mHealth technology NOW to manage their health

20https://www.klick.com/health/wp content/uploads/2017/06/2017 Klick Health Consumer Survey on Healthcare Innovation Report_LR 1.pdf


How do we Bridge the PGHD Gap?

• 70% of people want to use PGHD to manage their health

• Only 6% of clinicians use PGHD in clinical decision making

• PGHD gap - Solve for the right user needs

- Incentivize mHealth companies to develop the right tools

- Address sponsor/clinician barriers to adopting PGHD

21


mHealth Company Incentives

What incentives are necessary for mHealth companies to

intentionally design their products to have the capability to collect

and share data to support secondary evidence generation?

• The types of mHealth technologies of interest (including mobile apps, wearables, sensors, etc.) record data directly from patients

and can characterize the accuracy and reliability of that data

22


mHealth Company Incentives – What should the action plan highlight?

1. Potential market opportunities for mHealth technologies to support evidence generation

- Emerging markets and payment models (e.g., employer wellness programs, insurance

programs)

- Characteristics of payer and delivery systems that promote effective partnerships

2. Current challenges for mHealth Technology - Lack of interoperability and data standards

- Regulatory concerns (privacy, data security, SaMD)

3. Lessons learned from mHealth technologies developed for different markets/uses (e.g., specifically for clinical trials)

- Identifying approaches that add value to consumers as well as collect appropriate data

23


Does Size Affect Incentives? Large multinational technology corporations

• Trusted known company

• Optics and public relations (e.g., branding)

• Using their size and broader market presence to: • Balancing extended development timelines (e.g.,

access to existing capital)

• Manage risk

• Leverage skills and expertise

Shared opportunities

Start-ups and small companies

• More likely to embrace open source and open innovation in non-competitive areas

• Access to incubator and/or accelerator programs to support startups

• Digital health remains a growing vertical for capital infusion

• Connectivity with other apps and/or platforms enhance usability and offer new value to end-users

• Markets interested in disrupting healthcare with digital and consumer-facing solutions • Fostering altruism contributing to better health and care in society by creating better

services

• Partnering with life sciences/healthcare companies allows innovators to leverage their experience in regulatory, data privacy, etc.

24


Sponsor/Researcher Needs

•Sponsor/Researcher needs must be defined in order to understand how the relative data needs of organizations conducting and/or sponsoring research on medical products actively

incorporating mHealth data can differ depending on purpose, as well as what novel

outcome measures based on mHealth would be useful to them.

- “Sponsors” include any organization that is funding evidence generation (e.g., medical device and drug manufacturers, clinical societies, patient organizations, hospitals, employers and

payers).

•Potential Use Cases

1. Regulatory decision making

2. Quality measurement

3. Value-based payment models

4. Novel Outcome Measures

5. Evidence generation for shared decision-making models

25


Sponsor/Researcher Needs – What should the action plan highlight?

1. Existing mHealth technologies capable of collecting data appropriate for evidence generation

2. Opportunities for improved collection and sharing of data - Convenience and accessibility of mobile phones makes them ideal for data collection in the

field

- Improved and more representative recruitment by reducing time and travel costs associated with the study

3. Lessons learned from other mHealth areas (e.g., clinical support, medical device regulation, clinical trials implementation) - Potential to bias data through user engagement features (e.g., progress tracking could

encourage some people and discourage others)

- Large amounts of data require increased level of expertise and labor needed to analyze data

(e.g., high drop-out rates, missing data)

4. Relative needs on data quality, specificity, validation methods, and longitudinality for different purposes for mHealth data

26


What Affects “Fit-for-Purpose”? The utility of mHealth data for secondary research is highly dependent on the study objective, design, and other data types

• Use of the data • Outcome measure • Primary • Secondary • Ancillary • Exploratory

• Exclusion/Inclusion criteria • Recruitment

• Characterizing subgroups • Analysis • Risk-adjustment

• Type of Study • Regulatory • Premarket clearance/approval of medical

products

• Postmarket studies • Postmarket surveillance • Quality Assurance/Quality Control

• Quality Improvement/Clinical Guidelines • Predictive analytics for high-risk identification • Comparative effectiveness • Clinical practice research studies

27


Demystifying “Validation”

• What information is most helpful?

• Where is accuracy needed?

• Who is responsible for validation? - Accuracy in healthy/target population

- Validation in specific disease population

- Surrogate measure validation (often a medical claim)

• Validation is dependent on use (fit-for-purpose)

28


Validation of a mHealth Surrogate Marker • A secondary analysis of Multiple Sclerosis (MS) patients

- 786 persons with Multiple Sclerosis (MS)

157 healthy controls.

- Participants wore an accelerometer or pedometer over a 7 day period during waking hours

provided demographic information

People with MS also reported:

• MSWS-12 (Multiple Sclerosis Walking Scale-12) • PDDS (Patient Determined Disease Steps) scales • Other clinical and health information

• Conclusion - Change in motion sensor output of 800 steps/day represents a lower-bound

estimate of clinically meaningful change in free-living walking behavior in

interventions of MS

29Clinical Importance of Steps Taken per Day among Persons with Multiple Sclerosis Robert W. Motl, Lara A. Pilutti, Yvonne C. Learmonth, Myla D. Goldman, Ted Brown Retrieved from https://doi.org/10.1371/journal.pone.0073247


Participants

v Participants

CS EJ J Chri_Stiina Silcox , me,. int~rnal)

X

V

Notes

Public Feedback

• Please use the “raise hand” feature next to

your name or the chat feature

• Duke-Margolis also welcomes written comments on these topics. • Please send your thoughts to

[email protected] by July 12, 2017.

30

mailto:[email protected]







- Public Feedback


Task-Based Data

Passive Sensing

- Public Feedback


31

.... -

****


Potential mHealth Data Types

32

Pain? User-Reported Data What people say

Task-Based Measures Measures effort and physiology

Passive Sensing What people actually do day to day


User-Reported Data

• Data reported manually by the person themselves (or their caregiver if the person is unable to enter the data)

• Examples include: - Questionnaires/surveys

- Symptom tracking

- Person-reported outcomes

- Patient diary entries

• The data could be but is not limited to a validated outcome measure (i.e. Patient-Reported Outcome Measure or PROM)

- Historically captured through paper-based approaches, web surveys, phone calls, etc.

• Could be collected through mHealth apps (devices can be given out for the study or users can download apps onto their own device)

33


Task-Based Measures

• Objective measurement of a person’s mental and/or physical ability to perform a test consisting of a defined task or set of tasks

• Examples include: - Physical functioning (e.g., 6-minute walk test)

- Cognitive functioning (e.g., digital symbol substitution)

- Physiological tests performed by the user (e.g., glucose self-measurements)

• Typically collected in a clinical setting with appropriate clinical/task procedure validation

• Could be collected through remote sensors and/or mobile apps which may utilize smartphone hardware, but would require specific instruction and confirmation the task was performed

34


Passive Sensing

• A measurement of a person’s daily activities/mental state, where the measurement does not interrupt the person’s normal activities (i.e. it

measures what the person actually does in daily life)

• Could be collected through wearable and remote sensors, mobile devices/apps and other tools that monitor behavior (e.g., social

media)

35

mPower helps decipher Parkinson's disease. The variability in Parkinson's disea.se symptoms

has llleft many questions u nan swe·re·d. So the

University of Roe hester and Sage, Bionetworlks

created the 1mPower app to precisely measure

data such as de·xterity, ballla nee, me·mory, and

garnt Thils information could help researchers

better understand how various symptoms. are

con ne·cted to Parlki nson's disease .. In turn,

pa rtilcilpants couild start to re,cog nize their own

srngns and symptoms.


Robert Wood Johnson Foundation

~ ParkinsonNet

36


motor initiation

gps -displacement

vectors

tapping activity

MDS-UPDRS PDQ8

gait/balance hypophonia memory

gps -displacement

vectors

walking/ memory

standing voice activity game

activity

MDS-UPDRS MDS-UPDRS MDS-UPDRS PDQ8 PDQ8 PDQ8

37

motor initiation

StepJ or5

Tapping Interval Test Re:st your phone on a t sutfa.:e. il'len

use two linger$ on 1he s.eme hsnd to altem tely tap the buttons th t appear. Keep pplng for2(hecoo.ds and lime

your taps to be as C>CJm&Sten as possit>le.

Tap Ge $tar1eid to begin the lest


gait/balance

_,... Step1 ors

Gaiit and Balance Test

This lest measu~ your gal and balance a:s yoo wa and stand s ' I. To co piete lihis te:st, yoo'I need to put you, phone in your pocke and 0ClffleCt neadphones lo

foll!Qw audio ~iQl'l,:i.,

Ge eel

hypophonia

""' Step 5 or 6 Caricel

Say "Aaaaah" into the microphone for as long as you can.

memory

""' Step or4 Caricel

Spatial Memory Test This te:st mBasl.l'l!I$ )'Ola' ~tial memory by showing you p;Jllems an


..

z

@

.. X

gait/ balance

user Acceleration

gravity

rota ti on Rate

att itude

device motion readings at 100 .hz

39

N

"' 9

0 N

0

0

0

userAcceleration + grav ity, x-axis

2 4 6 8 10 ~me

userAcceleratlon + gra.vlty, y-axis

2 4 6 8 10 Wli

userAcceleratlon + gravity, z-ax ls

2 4 6 8 10 time:


12

12

12

C!

., 0

j l!o

lo ., 0

' 0

j

0 2

q

"' 0

1~ i .. j

0 2

0 -"' 0

C

lo Jo "' 9

0

' 0 2

gravity, x-axis

4

e ~me

6 ,,me

8

8

gravity, z-axls

4 6 8 crme

10 12

10 12

---

10 12

., 0

§"' l!o ,! ~~

"' 9 0 ., I

0 2

"1 -

= -

1: "' -

"' Cj> -q 'i - I ' 0 2

"1 -

q -

" id m ~o mo

-t I

"' 0 -I 0

' -

0 2

userAcceleration, x-axis

4 6 8 10 Um

userAccelerntlon, y-axis

1/

I

I I

6 liiTI

(

I ' 8 10

user.Acceleratlon, z-a.xls

' J J J , I 6 8 10

time

12

[

I

' ~ '

12

[

40

D k I MARGOLIS CENTER U e for Health Policy 41

.,

•

"'

1: C

i

0

.. "'

0

1u,serAcoeleration gravity, x- axis

2 4 6 llme

a 10

user Acceleration+ gravity, z-;axls

2 4 6 bme

a 0


gravity., x-axis

~- ---------------i .. C

.., y-

0 -----'

... C

0 2

0 2

4

..

6 lme

11

gravity, z-axls

e 10

8 10

; --1-:~~~~~-------------l

0 2 4 6 8 10 lime

"'

N

"'

"'

-I

0

0

0

userAcceleration, x- axis

2 4 e me

6

userA.cceleratlon, y-axis

2 e me

!I

user Ac~I e ra.tlo n,, z-nxls

2 4 6 rne

8

10

10

10

42

leration + gravity, x-axis

N

i-I! .!2

8 '° c

-I 0 5 10 15 20 25

time

userAcceleratlo111 + gravity, y -axis

"'

!-.. l'l ., .. 0

T

0 5 10 15 20 time

userAcceleration + gravity,

"'

C .2 -e e

Jo

--;-

0 5 10

D k I MARGOLIS CENTER U e for Health Policy 15 20

time

25

z-axis

25

C!

"" 0 j !o

f' "" 0 I

C! I

30 0 5

rAcce'lerat'ion + graviity, x-axis

...

.a i"' M_

II

';:'

0 5 10 15 20 25 iJme

userAcceleratlon + gravUy, - axis

N

-I·

N

' 1 1-i

0 !i 10 15 20 2 llme

userAccete ratlon + gravity, z•a,c l.s

...

C

t~ 1l ij~

' ...

II

0 5 10 ·15 lll 25 lime


9 ~

IQ d

.a l!!Q jo M ..,

d ,,

Cl

' 30 0

9 -:I

!i 1~ ;

30 0

e .,, 0

C ,, 0.

1l 0 .ij

"1 'i'

s I

30 0

gravity, x- axis userA.cceleration, x- axis

IQ ~

9

Ill

.ad I!!

il"' M9

:'.l '

5 10 15 20 25 30 0 5 10 15 20 2 30 fmll iJme

gravi , y-axis userA.cceleratlon, y-axis

"1 -q

"' r 'ii' .,,

' t-r !i 10 15 20 0 !i 10 15 20 2 30 - llme gravity, z-axls userA.cceleraUon, z·axls

"'1 ~

.,

..:

"' id 1l.., ij9

IQ ;

5 10 15 lll 25 30 0 5 10 15 lll is 30 1me 1Jme

44

0

g

~

1; ;

' I

~ I

-

-

-

-

I

0

0

-

-

-

-

I

0

userAcceleraUon + gravity, x-axis

I I I I I

5 10 15 20 2 IJme

userAcceleraUon + gravity, y-axis

5 10 15 2D 25 um

userAcce leraUon + gravity, z-ax ls

I I

5 10 I

15 bme

I I

20 25


C,

"' -0 ~ 1~-

"' o -' 0 ,.; -' I I

JO 0

0

~ -

lg-"' '? -

0 --' 30 0

-

-"' ? -

;l I

I I

30 0

I

I

5

gravity, x-axis

I I I

10 115 20 IJme

g.ravlty, y-axis

ID 115 Ii

gravity, z-axls

' 10 I

115 hme

I

20

I I

25 JO

25 30

I I

25 30

., userAcceleratlon,, x-axis 8-,----------------, Q

0

8 .§:0 l'l"

1~ 'i'

0

6 0 I

0 10 15 20 30 bme

§ userAcceleratlon,, y .. axls 0

§

1; ?

0 0 y

D 5 10 15 20 2%o 30 ~me

§ userAc,celeratlon, z"'axls 0

8 0

lo li

'i'

0

0 ?

0 5 10 15 20 25 30 bme

45


Participants

v Participants

CS EJ J Chri_Stiina Silcox , me,. int~rnal)

X

V

Notes

Public Feedback

• Please use the “raise hand” feature next to

your name or the chat feature

• Duke-Margolis also welcomes written comments on these topics. • Please send your thoughts to

[email protected] by July 12, 2017.

46

mailto:[email protected]







- Public Feedback


Task-Based Data

Passive Sensing

- Public Feedback


47


What will the Action Plan address? • Contribution of mHealth data for novel real-world evidence generation

- Role of NEST in encouraging the inclusion and use of mHealth in evidence development for medical devices

• Ways to think about different types of person-facing mHealth technologies, software, and data

• Recommendations for advancing person-facing mHealth adoption and usage - User engagement

- Researcher/sponsor needs

- mHealth company incentives

• Overarching challenges in digital health data: Current work and resources - Best practices for patient-consumer informed consent

- Data linkages and interoperability

- Accommodating diversity of wearable technology and application

- Fit-for-purpose: validation and reliability of data

• Recommended next steps to advance consumer/clinical mHealth technologies as a viable source of reliable data for evidence generation

48


Next Steps

• Public Comment period through July 12, 2017 - Please send your thoughts to [email protected]

- Comments will be shared with the working group but will not be made public.

- The webinar slides and recording will be posted to the Duke-Margolis website within 48 hours.

• Action Plan Recommendations - Public Release Meeting - September 15, 2:00-4:00 pm EST In-person at our DC office and webcast

- Details will be posted on the Duke-Margolis website or you can sign up for event notifications at https://healthpolicy.duke.edu/newsletter

49

mailto:[email protected]://healthpolicy.duke.edu/newsletter

Tweet dukemargolis RealWorldDigitalHealth: undefined: undefined_2: undefined_3: 1: 2: Ge ed: undefined_4: Step 5 or 6 Caooel Say Aaaaah into the microphone for as long as you can: 1_2: 2_2: undefined_5: C1: q 1: undefined_6: undefined_7: I Ill: V: Row1: Row1_2: Row1_3: Row1_4: Row1_5: i: 1_3: undefined_8: JRow1: undefined_9: 7J: 0: 2_3: 4 6: 8: 10: undefined_10: N: C: i_2: i_3: 80: undefined_11: Row1_6: j: Row1_7: undefined_12: 9: l: il: N_2: undefined_13: Cl: Row1_8: undefined_14: undefined_15: undefined_16: undefined_17: 8_2: C 0 Row1: undefined_18: lg: Row3: 0 0Row1:

Documents

Keeping Up with Consumers - Duke-Margolis · 2019-05-23 · through an app on her phone. Ingestible sensors track her medication adherence. Christie authenticates her apps and devices