Blog

Digital Medicine/Personal Devices

STC 100

2018: David Mercer

Digital Medicine and Personal Devices as parts of socio-technical systems Not just single devices, apps, or platforms, but parts of socio-technical

systems.

• Hardware/software eg: programs, material objects, interfaces, infrastructure to maintain them.

• Knowledges/practices eg: computing, medical knowledge, patient experiences.

• Different actors with convergent and divergent interests: Authorities/institutions, government and health agencies, medical entrepreneurs/ corporations $, designers, users

• Cultural expectations, Hopes/fears, Utopia’s/Distopia’s, Socio- technical imaginaries, Images of medicine, health and personal identity, Technological Determinist myths.

Visibility of ‘technology’ in new socio-technical systems • Technologies of digital medicine of health apps, personal medical devices are

currently more visible than parts of other more settled socio-technical systems

• In well established socio-technical systems ‘technology’ is often less visible as systems are stable and there is less ‘interpretive flexibility’ regarding how technologies can be used and given meaning.

• Do we think of electric light-bulbs, refrigerators or microwave ovens, as ‘technologies’ in the same way as we think of new technology such as iPhones as ‘technological’?

• When old technologies such as refrigerators were new they symbolized important parts of new technological systems and attracted debates about their uses and implications and alternative technological options. They are now largely invisible.

• Well established socio-technical systems often display inertia/resistance to change ( this will be returned to as a theme in later lectures on cars)

• When socio-technical systems are new there is normally more opportunity to shape their directions and more possibilities for social disruption good and bad.

Multiple often intersecting forms of Personal Digital Medicine: Some examples

• ‘On line’ minimally interactive medical information services/web sites (web 1.0).

• ‘Measurement Based Medicine’ using web 2.0 capacity to do both medical research and provide patient/user support via ‘user-generated data’ eg: Patients Like Me some links with traditions of ‘patient-movements’:

• ‘23 and Me’ Online genetic screening

• Health apps and wearable technologies extending possibilities of gathering real time patient data and continuous self monitoring. Eg: Fitbit etc

• Government initiatives across the world to create individual digital health records for their populations eg: The Australian Digital Health Agency wants everyone to have their own ‘my health record’ by the end of 2018.

Web 1.0 First Generation ‘On Line’ Medicine

Medical Information web-sites (web 1.0)

• Patients relatively passive consumers

• Significant ease of access to greater quantities of medical information

• Issues of quality control /patient competancies

• Many concerns in professional medical community • Debates about cyberchondria • These web sites still being used but are steadily being transformed

and overtaken by health apps and other forms of digital medicine.

Patients in control or cyberchondria (or both)?

“The use of Web search as a diagnostic methodology—where queries describing symptoms are input and the rank and information of results are interpreted as diagnostic conclusions—can lead users to believe that common symptoms are likely the result of serious illnesses. Such escalations from common symptoms to serious concerns may lead to unnecessary anxiety, investment of time, and expensive engagements with healthcare professionals. We use the term cyberchondria to refer to the unfounded escalation of concerns about common symptomatology, based on the review of search results and literature on the Web” White and Horvitz (2008) Cyberchondria: Studies of Medical Concerns in Web Search’ ACM transactions on Information Systems, 24(4).

Escalation

• According to White and Horvitz (ibid) who first popularized the term cyberchondria, the syndrome is set off by the problem of escalation: where a significant number of those conducting web searches of basic symptoms quickly escalate searches from mundane diagnoses to searching for serious ones.

• Eg Go from Headache to Caffeine withdrawl or stress to Brain Tumour.

Explanations for Escalations

Technical issues(possibility of improvement via changes to web design) • Problems of design of search engines and user designer feedback loops.

• Users erroneously treating Web search engines as medical expert (A.I.) systems.

• Rankings of web searches confused with likelihood of symptom being linked to illness.

• Over-reliance on links to pages appearing higher on a result page.

• Possibility of reinforcement of this ‘ranking ‘as search engines respond to more hits for these pages and maintains their preferential placement.

Broader Issues

Unreliability

• Research from within medical community has noted the unreliability of Web content in general. Most have noted less an issue of straightforward errors than incompleteness, lack of clarity and accuracy and poor design and readability.

Limitations of Audiences/Users

• Medical terminology difficult for lay person to grasp. Easy for misunderstandings to arise.

• White and Horvitz noted that whilst as many as 8 in 10 American adults had searched for healthcare information online, 75% hadn’t checked key quality indicators, such as the validity of the source and the creation

date of medical information.

Lack of Expertise in Diagnosis

• Expert diagnosis relies on ‘Tacit Knowledge’ experience based knowledge ie: experience in seeing various symptoms and better appreciations of probabilities of links between symptoms and illness

• Eg: demographic info and some personal info less likely to be accurately factored into web search processes.

• Are we witnessing the beginning of a ‘de-skilling’ of medical experts and a reconfiguration of the relationships between doctors (experts) and patients (users)?

• Are patients becoming more empowered to make inadequate health decisions….?

(more on this later in lecture)

Are the risks of cyberchondria overstated ? • Natural curiosity may mean ‘escalation’ doesn’t really equate with

cyberchondria

• Some surveys suggest that younger internet users in particular are more sceptical of information they obtain

• Cyberchondria critics often don’t factor in the links between the use of the internet alongside other information sources and the capacity for users to enagage in social learning.

• How much better in reality are the decisions made by GP’s under the ‘de-skilling’ pressures of dealing with huge patient loads and a culture of referral to specialists/ and ‘objective tests for symptoms eg: scans and pathology?

• Web 2.0 and beyond offers new more interactive web-sites and ‘authorized’ forms of information increasing patient control and quality of information diminishing risks of cyberchondria (maybe?)

Patients Like Me: Self Learning Health Care Systems: Measurement Based Medicine • ‘Patients Like Me’ Founded in 2004 by MIT trained

engineer Jamie Heywood. • Web 2.0: ‘User generated data’ • ‘Measurement Based Medicine’ • 500k + Members • 2700 + Conditions currently being monitored • 38 + million data points about disease • Patients register on-line and share experiences with

symptoms and treatments • Chart side-effects of medications • Patient and Profit orientated (not a charity or public

service)

12

Positive Impacts: Good for patients, medical innovation and business

• Published 100 papers in academic journals • Improved treatment regime for ALL sufferers • Questions mimic clinical trials but are open to more room for

patient defined experiences/conditions • Much faster/real time data gathering and potential for much

larger cohorts than current clinical trials • Has started selling information to many big Pharma companies • Recently formed deal to supply information to US FDA:

surveillance of drug side-effects • In Jan 2017 formed 100 million US $ partnership with Chinese

based company iCarbonX (more than US Billion $ assets) as part of iCarbonX’s broader vision to link Patients Like Me and other self learning health platforms to Artificial Intelligence systems for life time health monitoring to build an ‘ecosystem of digital life’

Business Wire, Jan 5, 2017.

13

But be aware of promotional hype: Utopian ‘socio- technical imaginary’ of Ecosystems of Digital Life ?

• “ Founded in October 2015, iCarbonX aims to build an ecosystem of digital life based on a combination of an individual’s biological , behavioural and psychological data, the internet and artificial intelligence…Together with the world’s leading partners, we plan to observe, study, guide and take care of one’s health from the beginning of one’s life”

iCarbonX home-page accessed 14/02/17

• Like a lot of people I thought patient Web groups were merely chat boards and self-help news distribution. “That’s the Internet of the 1990’s” …. “Qualitative data gets high-jacked by people with an agenda, or extroverts or people who write well”…I want an electric sensor on your head, in your toilet, on your pancreas, on your steering wheel and in your Google Glasses”

Paul Wicks research director for ‘Patients Like Me’ quoted in Forbes March 1, 2013.

Some issues raised by Measurement Based Medicine (MBM)

• Possibility of generating massively large amount of meaningless data • Experts still needed to oversee algorithm development and provide

interpretation –lack of transparency of ways this is done. • Which conditions will be chosen to focus on for further research possibly

shaped by political economic considerations • How will these systems impact on existing health professionals and

government providers? • A number of insurers already collaborating/impacts on insurance- access

and premiums • Possible limits on user learning, participate often, but is it in a way that

encourages deeper learning • Generational, gender and social class issues re: shapes who opts in to a self

learning health system to start with • Web 2.0 Offshore/global data gathering avoids legal/ethical controls • Privacy and Surveillance/Dual use issues • Links with $ big pharmaceutical companies lack of independent assessment

of risks and efficacy. Who profits?

Health Apps and wearable technology

• Market research firm Markets and Markets estimated by 2020 mHeath market –connected devices, apps and monitoring services worth 60$ Billion US.

• In 2017 estimates of worldwide shipments of wearables at 125.5 million a 20% increase on 2016 with an estimate of 240.1 million to be shipped in 2021 (IDC June 2017)

But beware of Hype!

• Apple watch sales which have some health monitoring functions are strong Fitbit sales are actually declining (Glance The Conversation March 1, 2017)

The Rise and Fall of Fitbit?

• In early 2017 Fitbit reported a financial loss and announced laying off 6% of its staff in a bid to save 200 million $ with a share price 90% down from its peak in 2015. Only sold 6.5 million products in the last quarter of 2016 about 8% less than anticipated

Various problems:

• Lack of expansion of its features still primarily counting steps and measuring heart rate/ features perceived as mundane incorporated into other digital devices ‘old-hat’.

• Security issues

• Lack of verification of health benefits: eg: medical criticism of its default goal of 10,000 steps a day as having limited evidence supporting it as a one size fits all health measure (Glance ibid, 2017)

Motivations influencing whether or not to use a wearable means social context of user important

(D. Lupton ‘The Social Factors that Influence whether you’ll use your wearable device’ The Conversation, Dec 20, 2017)

Survey respondents who used Fitbit did:

• Because they believed the technology allowed them to: take control over health and well-being, provided motivation to achieve personal best, spurred competition with fellow ‘Fitbiters’, gave real time re- inforcement when 10k steps were achieved, facilitated learning healthy behaviours, encouraged sense of community with other online ‘Fitbiters.

Survey Respondents who stopped using Fitbit:

• Goals seemed unachievable (people can feel bad when targets can’t be reached); reminders and constant alerts become annoying, lack of peer support

Susanna Trnka survey of NZ young people between 16-24 and health app use Engaging Science,Technology, and Society 2, 2016 (Tute reading)

Users take control

• Availability portability speed of access and possible anonymity when seeking advice all highly valued

• Privacy re: personal details highly valued but not much concern with the unauthorized collating of data for commercial purposes

• Valued a sense of control via doing their own searching/ interaction

• Respondents regularly reported that health apps enhanced their capacity to reconstitute themselves, re-inforce their identities and help them achieve goals, something particularly important in this period of life

• Possible increases in broader sociality via; engaging in comparing apps Fitbits etc. sharing data and information

• Satisfaction in mastering web-sites app technology

The B side: Planet of the Apps? Apps in control?

• A sense that apps were themselves exerting control/demands over their lives: always there, all places times.

• Strong senses of guilt when goals couldn’t be met and accompanying sense of failure/ awareness of possible obsessiveness

• Perceptions of overwhelming amounts of possible information but compulsion to still check symptoms of health conditions

• Cheating in depiction of ‘on-line’ self to satisfy ‘expectations of the app.

How effective are health apps?

(Oyuka Byambasuren (et.al.) ‘Prescribable mHealth apps identified from an overview of systematic reviews’ N Digital Medicine (2018/1-12)

•Over 250, 000 health apps available researchers attempted to answer the question using standard medical reviews of: Which apps had sufficient evidence of efficacy to satisfy the normal criteria used by doctors to prescribe them as a therapy?

•The answer was very few and practically no clear high quality research, to date only 23 comprehensive reports

•Only 1 clear efficacy Get Happy. •Promillekol app designed to curb youth drinking in Sweden appeared to increase the frequency of drinking of users whilst using it !!

•Popular MyfitnessPal app no significant reduction in weight of users over time

In Victoria Drinking app for uni-students being trialed: Will it work?

• A government-backed app that texts reminders to university students to stop drinking will largely be ineffective, students say.

• The app, designed by the Burnet Institute, is part of a $300,000 campaign from VicHealth to curb binge drinking among university- age Australians, but has been likened to a nagging cyber-parent.

• Users will activate the app before they embark on a night out, and the app will ask questions such as “Do you have work tomorrow?” or “What time do you intend on going home?”

• Students will also be asked to complete periodic surveys about how they are feeling, how much money they have spent, and how inebriated they feel.

• The app will keep sending text messages during the night. The Guardian 12th July 2017 (Naaman Zhou) ‘Healthy drinking app for uni-students likened to

nagging cyber-parent’

How private are APPS: HealthEngine

(Points below drawn from) ‘Medical appointment booking app HealthEngine sharing clients' personal information with lawyers’ Pat McGrath, Clare Blumer and Jeremy Story Carter, ABC Investigations June 2018)

•HealthEngine a Perth-based startup, part-owned by Telstra and SevenWest Media boasts 1.5 million monthly and 15 million annual users and is Australia’s largest on-line medical booking service it is linked to a Data-Sharing arrangement with the new MyPatient Australian government personal medical records system.

•HealthEngine asks users to include details of their symptoms and medical conditions, including whether they have suffered a workplace injury or been in a traffic accident, as part of the process of booking appointments with GPs, dentists, physiotherapists, optometrists and other medical practitioners.

Privacy Breaches !

• HealthEngine has boasted to advertisers it can tailor advertising to patients' symptoms and been sharing the information with legal firms

• The ABC obtained secret documents from plaintiff law giant Slater and Gordon that reveal HealthEngine was passing on a daily list of prospective clients to the firm, based on their personal medical information( an average of 200 a month between March and August 2017)as part of a "referral partnership pilot"

• If a patient wants to use the app, there is limited opportunity to opt-out of the fine print about giving information to third parties

• Health Minister Greg Hunt has ordered an "urgent review“ of HealthEngine.

Overview Health apps

• New technology so very little detailed robust research into how well they work

• Many governments health departments keen to encourage their development. Will they end up in the future linking back in mundane ways to government sponsored digital medical health records?

• Global complex issues regarding regulation or standard setting. Developers have huge freedom in creating medical self diagnostic tools and advice without medical input and lack of accountability in how data is used.

• Immediate accessible/fashionable possibly encouraging their popularity large investments of venture capital corporate opportunities for profits….but does the recent decline in Fitbit represent a new trend? Will mHealth apps become less interesting/profitable as they become less technologically visible?

• Many privacy issues remain unresolved • Possibly much more effective in encouraging positive health outcomes when

coupled with a supportive social context

Overview Continued

• Do patients become more expert and in what sense?

• Is the status health professionals potentially diminished? • Apps can empower users with strong perceptions of being in control

but also increase perceptions of ‘self responsibility’, mismanagement of one’s health letting society down and evidence of moral deficiency if they don’t achieve their health goals.

• Apps can begin to take on a sense of agency in themselves as users feel controlled by them, can’t escape them, apps always there any place-any-time.

Health as Data

• Do users begin to interpret their health and aspects of their identity in terms of data and numbers. Q: “Did you enjoy your walk? ” A: “I’m not sure but my wearable device suggests I burnt off 10% more calories than yesterday and walked 15% more steps, so I must have”

• Do they still encourage Cyberchondria ? • NB Also need to evaluate mHealth apps keeping in mind traditional

health delivery has plenty of problems and the overall aim of empowering patients is generally very positive and apps have room for improvement