yuri quintana of bidmc - november 11th health innovators presentation
TRANSCRIPT
Global Trends in e-Health and Medication Adherence
Yuri Quintana, Ph.D. Division of Clinical Informa8cs
BIDMC and Harvard Medical School
An Academic Division of the Dept of Medicineat Harvard Medical Faculty Physicians at BIDMC, Inc.
TM
11/11/14
AGENDA
11/11/14 2
o Global health trends
o Global m-health trends
o Evaluation of medication adherence systems
o Design and evaluation considerations
o Road ahead
Global E-Health and Medication Adherence
Global Health Trend 1 – Aging Population
• The global popula8on age 60 or above is expected to more than triple by 2050
11/11/14 Global E-Health and Medication Adherence 3
Source: Deloitte 2014 Global health care outlook https://www2.deloitte.com/content/dam/Deloitte/global/Documents/Life-Sciences-Health-Care/dttl-lshc-2014-global-health-care-sector-report.pdf
Global Health Trend 2 – Cost and Quality
• 5.3% annual spending increase in health costs expected globally over next five years
11/11/14 4 Global E-Health and Medication Adherence
Sector: Deloitte 2014 Global health care sector outlook www.deloitte.com/2014healthcareoutlook
Global Health Trend 3 – Access to Care
• There will be a shortage of 230,000 physicians across Europe in the near future.
• The number of caregivers in 36 countries in Africa is inadequate to deliver even the most basic immuniza8on and maternal health services.
11/11/14 5 Global E-Health and Medication Adherence
Source: Deloitte 2014 Global health care outlook https://www2.deloitte.com/content/dam/Deloitte/global/Documents/Life-Sciences-Health-Care/dttl-lshc-2014-global-health-care-sector-report.pdf
Global Health Trend 4 – Technology
• New technologies (m-‐health, cloud, wearable) are driving change in the way physicians, payers, pa8ents and stakeholders interact
11/11/14 6 Global E-Health and Medication Adherence
Source: mHealth in an mWorld - Delloitte Center for Health Solutions http://www.deloitte.com/view/en_US/us/Industries/life-sciences/2545b66b8dc4b310VgnVCM2000003356f70aRCRD.htm
Global Medication Adherence Trends
11/11/14 7
Source: Economic aspect of medication adherence using mobile medication reminder in French Health System http://www.medetel.lu/download/2014/parallel_sessions/presentation/day2/Economic_aspect_of_medication.pdf
Global m-Health Trends
11/11/14 8 Global E-Health and Medication Adherence
• 247 million Americans have downloaded a health app
• In 2013, 95 million Americans are using mobile phones as health tools
• 77% of U.S. seniors own a cell phone and their smartphone ownership has
increased 55% in the past year
• 42% of U.S. hospitals are using digital health technology to treat pa8ents
• Mobile remote pa8ent monitoring expected to save the U.S. $36 billion in
health care costs by 2018
• Wireless pill bo^les helped increase medica8on compliance
• Mobile health is a $1.3 billion industry and by 2018 is expected to reach $20 billion
Source: Mobile Health Tools Make Big Impact by Jonathan Spalter http://mobilefuture.org/mobile-health-tools-make-big-impact/
Evaluation of Mobile Medication Systems
11/11/14 9 Global E-Health and Medication Adherence
Source: Phansalkar S, Zachariah M, Seidling HM, Mendes C, Volk L, Bates DW. Evalua8on of medica8on alerts in electronic health records for compliance with human factors principles. J Am Med Inform Assoc. 2014 Oct;21(e2):e332-‐40. doi: 10.1136/amiajnl-‐2013-‐002279. Epub 2014 Apr 29. PubMed PMID: 24780721; PubMed Central PMCID: PMC4173170. http://www.ncbi.nlm.nih.gov/pubmed/?term=24780721
each highlighting a different feature, such as appearance,reminders, drug information, drug interactions, and connectivity(Table 1, Figure 1).
We chose MyMedRec (Version 1.0.4) for its simple featuresand linear data entry. MyMedRec was developed as acollaboration between the Institute of Safe Medication PracticesCanada, Canada’s Research Based Pharmaceutical Companies(Rx&D), and several health professional association acrossCanada. We chose Pillboxie (Version 2.6) for its graphicalinterface. A registered nurse in the United States developedPillboxie to be a virtual medicine cabinet. We chose DrugHub(Version 1.3) for its drug information feature. The Great-West
Life Assurance Company, a large provider of health insurancein Canada, developed DrugHub as a service to the generalpublic. We chose PocketPharmacist (Version 3.1.8, Danike,Inc.) for its drug interaction feature. A pharmacist in the UnitedStates designed PocketPharmacist to provide users withmedication information and the ability to check multiple drugsfor any interaction. Finally, we chose MediSafe (Version 2.3.2,MediSafe Project) for its cloud-synced, family-centered profilesharing features. MediSafe was designed in Israel. At the timeof the study, MyMedRec, Pillboxie, DrugHub, and PocketPharmacist were available for the iOS system, and PocketPharmacist and MediSafe were available for the Android OS.
Table 1. Features of the mobile medication management applications selected for review.
MediSafePocketPharmacist
DrugHubPillboxieMyMedRecFull name
✓✓✓✓✓Medication list
✓✓✓✓✓Reminder alarms
✓✓Drug information
✓Drug interactions
✓✓✓✓Multiple user profiles
✓✓✓✓✓Profile sharing viaemail
✓Sharing across multipledevices
Figure 1. Screenshots of the mobile medication management applications included in the assessment.
JMIR Mhealth Uhealth 2014 | vol. 2 | iss. 1 | e11 | p.3http://mhealth.jmir.org/2014/1/e11/(page number not for citation purposes)
Grindrod et alJMIR MHEALTH AND UHEALTH
XSL•FORenderX
User Perceptions of Mobile Medication Systems
11/11/14 10 Global E-Health and Medication Adherence
purse. Participants would only speculate future use under theassumption of declining health, declining memory, or the needto manage medications for a relative.
I’m looking at it from the point of view of my motherwhen she was elderly. She was confined to awheelchair and okay, she wasn’t computer literate.But had she been, you know, if she’d had it with her,in her chair, she could’ve looked at it and said ‘yeahhey I need to take this pill’ or there’s a reminder, or‘no I can’t’, somebody’s making lunch for her, ‘no Ican’t have grapefruit because I’ve taken Lipitor’ orsomething like that. I’m sure that kind of information
would have been good. If you’re not terribly mobile,I think something like that, and in this day and age,as time goes on, people are much more computerliterate and can handle these things much more easilyand how do you say, you know, that would be muchmore useful, if you’re concerned, if you’re taking alot of medications. Because I know some people whoare taking seven or eight a day. [Female, Group 1]
Most also felt it was only appropriate for health careprofessionals to recommend an mHealth application if they hadused it themselves and if the patient was willing, needed it, andwas technologically literate.
Table 5. Participants summarize the experience of trying mobile medication management applications in one word.
Positive Words (Count)Neutral Words (Count)Negative Words (Count)
Fascinating (2)Different (1)Frustrating (5)
Fun (2)Perplexing (1)Challenging (3)
Enlightening (1)Overwhelming (2)
Doable (1)Stressful/nerve-wracking (2)
Interesting (1)Confusing (1)
Useful (1)Exhausting (1)
Informative (1)Complicated (1)
FunctionalityMobile medication management applications can beconceptually separated into two categories: adherence(MyMedRec, Pillboxie) and information (DrugHub,PocketPharmacist). An “ideal" application was described asincluding both features:
I found MyMedRec covers everything, it’s orderedproperly. But it did miss the other little features, thelittle pillbox [in Pillboxie] and then the [druginteraction] check [in Pocket Pharmacist] and then[DrugHub]. I guess it’s the access to the informationand whether you could check interactions and thingslike that. If somehow you could incorporate that intothe [MyMedRec] then it would be perfect. [Female,Group 3]
In their daily lives, all participants sought information abouttheir medications in order to stay aware and avoid adverseevents. The drug information features were seen as providingbackground information on a new prescription, supplementingthe information given by a health care professional, andsatisfying curiosity.
For me, personally, I take a lot of pills everyday…I’vegot it so down and whenever I take a prescription,well the pharmacist is very good to go over things,but I always, always make a point of reading theliterature when I get it. [Female, Group 2]
While the most popular source of drug information was thepharmacist, some participants worried that too much informationwas dangerous and that the applications were replacing theexpertise of the pharmacist. Given the choice, participantspreferred an in-person conversation for important information:
Something like drug interactions? I don’t want to bebothered by anything like that. I mean I know I should,but I want my pharmacist to say to me when I go in,don’t take this or do take that. You know what I mean?I didn’t go to school, I don’t want to have theresponsibility of worrying about that... [Female,Group 1]
When asked to estimate the cost of applications, mostparticipants valued drug information applications over adherenceapplications. Participants who had purchased applicationsexpected to pay less than Can $5 (or often nothing at all), butthose who had never purchased an application expected to payup to Can $100 or a monthly fee. Most did not take intoconsideration the cost of the device itself.
SimplicityThere was a competing relationship between functionality andcomplexity. The “ideal” application may actually be twoapplications, one for adherence and another for information.Separating the features into two applications would maximizethe functionality of both features rather than trying to do bothincompletely.
I think there’s two parts of it. There’s remindingpeople to take the medication but then there’s thewhole information side with what’s working withwhat. So it almost seems like you should have twoapps. [Female, Group 7]
For many participants, linear navigation was preferred.Participants commonly struggled with going “back and forth”,essentially, moving forward to enter a medication into theirprofile and once completed, going back to enter a newmedication (Table 6). Moving backwards also referred to fixing
JMIR Mhealth Uhealth 2014 | vol. 2 | iss. 1 | e11 | p.10http://mhealth.jmir.org/2014/1/e11/(page number not for citation purposes)
Grindrod et alJMIR MHEALTH AND UHEALTH
XSL•FORenderX
Source: Phansalkar S, Zachariah M, Seidling HM, Mendes C, Volk L, Bates DW. Evalua8on of medica8on alerts in electronic health records for compliance with human factors principles. J Am Med Inform Assoc. 2014 Oct;21(e2):e332-‐40. doi: 10.1136/amiajnl-‐2013-‐002279. Epub 2014 Apr 29. PubMed PMID: 24780721; PubMed Central PMCID: PMC4173170. http://www.ncbi.nlm.nih.gov/pubmed/?term=24780721
Evaluation of Mobile Medication Systems
11/11/14 11 Global E-Health and Medication Adherence
a mistake. Participants moved from one main menu and followedsingle pathways to perform or correct a task. As such, mostfound MyMedRec and DrugHub to be easier and more logical.In contrast, Pillboxie and PocketPharmacist did not flowbecause, rather than advancing through different screens,navigation was broken into submenus or different windows ona single screen. Inconsistencies also caused confusion. To returnto a previous screen or menu, MyMedRec and DrugHub usedthe standard iOS arrow pointing to the left corner. To back outof a task in Pillboxie, users had to tap outside the task window.
I’m impatient as hell. So when it comes to an app, it’sgot to be simple. See…it wasn’t easy for me to findthe prompts, you know, partly from [my] glasses, butalso I’m impatient and I quit looking. And I said ‘OhI’ve spent all this time entering the stuff but if I put
CANCEL, does that mean it’s gone? [Male, Group5]
Similarly, the first screen a user saw with PocketPharmacistcontained both a menu and submenu. Participants expressedfeeling overwhelmed by the amount of information beingpresented all at once. Many first time users struggled with basictouchscreen features, such as accessing and using the keyboardand employing application-specific gestures. One participantfelt this should be standardized,
So ultimately, you’d want a universal language, anda universal kind of, you know, this is where the backbutton is, this is where the forward button is. But ifthat doesn’t happen, then every app has its ownunique way. [Female, Group 2]
Table 6. Application actions and features identified by participants as being nonintuitive or difficult to interpret.
Description of challengeAction/feature
Though typically used to add a new item, the symbol had little meaning for first time users. Also, because it is often foundin the top corners it is easily missed.
A “+” to add a new item
The back arrow is used to return the user to the previous screen but instead of testing the button, the research team wasoften asked, “How do I go back without losing my information?”
Go back
The word “cancel” typically means “undo” but many participants felt it implied finality and described how they “cancel”social or service contracts such as memberships, subscriptions, and appointments.
Cancel
Without a scrollbar, participants rarely looked for additional information.Scrolling
The audio alarms were inaudible to many participants, especially males.Audio reminders
When typing, many participants focused on the keyboard and missed the autocorrect feature that would change drug namesor dosage units (eg, “mcg” to “mg”).
Autocorrect
Inconsistent terminology led participants miss features. For example, reminder features were called “schedule,” “dose re-minder,” or “first dose” in each application.
Inconsistent terminology
Greyed text was used to provide examples of data that could be entered into a field, but participants typically misinterpretedthe grey text to be the information of another user.
Sample text
Participants associated a black frame as being outside of the application and noninteractive, thus overlooking peripheralbuttons completely.
Peripheral buttons
AccessibilityOne of the challenges faced by participants was that theadherence features we examined (dose reminders, refillreminders) made assumptions about the end user. For example,the reminder strategies (alarms, notification boxes) assumedusers were “attached” to mobile devices. Participants said, forexample,
Like, [young people] live with their cell, live withtheir Blackberry, and that becomes, you know whatI mean. Like, I could see, even obviously, when thosekids get to be 50, they will still be attached to the hipwith those Blackberries. [Female, Group 1]
Comparing the use of the applications on the tablet to thesmartphone, one participant noted,
Reminders would probably be the best [feature] butit would be inconvenient unless I had one of the otherdevices that you could carry in your pocket or yourshirt pocket or a woman could carry in her purse.(Male, Group 4).
This is an important distinction because though the tablets areless portable, they are more accessible to individuals withage-related vision loss. In one case, a participant with severelow vision noted that touchscreen devices were surprisinglyaccessible,
I was always afraid to even look at them or try them,because I just thought that I wouldn’t be able to see,so why even bother. But I was surprised…yeah.[Female, Group 4, low vision]
The participants, as older adults, also described how they poweroff devices between use to conserve battery power or save itfor emergencies.
These tablet things, they’re not plugged in, so, mostof the time…you tend to turn them off to conserve thebattery and maybe they could be designed so that theyautomatically turn themselves on, give a signal, andthen go back to rest. [Male, Group 8]
The reminder strategies also assumed users were physicallyable to hear alerts. In every session, we observed at least 1
JMIR Mhealth Uhealth 2014 | vol. 2 | iss. 1 | e11 | p.11http://mhealth.jmir.org/2014/1/e11/(page number not for citation purposes)
Grindrod et alJMIR MHEALTH AND UHEALTH
XSL•FORenderX
Source: Phansalkar S, Zachariah M, Seidling HM, Mendes C, Volk L, Bates DW. Evalua8on of medica8on alerts in electronic health records for compliance with human factors principles. J Am Med Inform Assoc. 2014 Oct;21(e2):e332-‐40. doi: 10.1136/amiajnl-‐2013-‐002279. Epub 2014 Apr 29. PubMed PMID: 24780721; PubMed Central PMCID: PMC4173170. http://www.ncbi.nlm.nih.gov/pubmed/?term=24780721
User Interface Design Problems
11/11/14 12 Global E-Health and Medication Adherence
• Simple Screen Design -‐ Linear Naviga8on
• Most used features must be simple to use
• Avoiding feature creep and clu^ered design
• Design for Seniors – font size, color, naviga8on • Need to follow User Center Design Process
• Need more coordina8on and itera8on between user requirements, wire framing and usability tes8ng
m-health Evaluations
• Global case studies demonstrate that mHealth solu8ons for NCD management are feasible ..but.. there is li^le evidence about the costs and savings of these technologies Source: dx.doi.org/10.1145/2093698.2093868
• A review found 18 of 29 mobile systems using text messaging improved medica8on adherence. Nega8ve studies tended to have more basic and repe88ous content with a simple medica8on reminder, while posi8ve studies delivered a variety of educa8onal and mo8va8onal content with ‘tailored’ or ‘personalized’ SMS
Source: dx.doi.org/10.1111/jan.12400
• Diversity and mul8modal reminder methods should be
considered to increase usability for seniors Source: www.ncbi.nlm.nih.gov/pubmed/?term=25099993
11/11/14 Global E-Health Conference 13
Global Design Considerations
• Care coordina8on needs to be considered in design of systems – engagement with family and care providers
• Cultural and language customiza8ons needed
• Health literacy needs to be considered in each popula8on
• Deep integra8on with local healthcare IT systems is key
• Training of staff and pa8ents needs to be planned • Evalua8on and ROI needs to define the metrics for adherence,
u8liza8on, and costs that relate to local business models
11/11/14 14 Global E-Health and Medication Adherence
Road Ahead
• Need to define evalua8on metrics and ROI models for compara8ve analysis
• Need to have more outcome evalua8ons of implemented systems
• Need to understand the modifica8ons need to adapt
systems to other healthcare networks and countries for global deployments
• Need to have best prac8ces on design for user interfaces, implementa8on, cultural and language localiza8ons
11/11/14 15 Global E-Health and Medication Adherence
References
Health literacy http://www.nlm.nih.gov/medlineplus/healthliteracy.html mHealth in an mWorld - Delloitte Center for Health Solutions http://www.deloitte.com/view/en_US/us/Industries/life-sciences/2545b66b8dc4b310VgnVCM2000003356f70aRCRD.htm mhealth Competence Center - Barcelona http://www.mobilehealthglobal.com/catalogue/ Mobile Health Tools Make Big Impact by Jonathan Spalter http://mobilefuture.org/mobile-health-tools-make-big-impact/ Deloitte 2014 Global health care outlook https://www2.deloitte.com/content/dam/Deloitte/global/Documents/Life-Sciences-Health-Care/dttl-lshc-2014-global-health-care-sector-report.pdf Emerging mHealth: Paths for growth http://www.pwc.co.nz/healthcare-industry-sector/publications/emerging-mhealth-mobile-health-from-patients-payers-and-providers/ The effectiveness of interventions using electronic reminders to improve adherence to chronic medication: a systematic review of the literature http://www.ncbi.nlm.nih.gov/pubmed/22534082 Economic aspect of medication adherence using mobile medication reminder in French Health System http://www.medetel.lu/download/2014/parallel_sessions/presentation/day2/Economic_aspect_of_medication.pdf Adherence to medication among chronic patients in Middle Eastern countries: review of studies http://apps.who.int/iris/bitstream/10665/118131/1/17_4_2011_0356_0363.pdf?ua=1 Experiences in mHealth for Chronic Disease Management in 4 Countries http://www.ghdonline.org/uploads/Piette2011-ExperienceMHealth4Countries_1.pdf A quantitative systematic review of the efficacy of mobile phone interventions to improve medication adherence. http://www.ncbi.nlm.nih.gov/pubmed/24689978 Evaluating User Perceptions of Mobile Medication Management Applications With Older Adults: A Usability Study http://mhealth.jmir.org/article/viewFile/mhealth_v2i1e11/2
Thank you!
11/11/14 17
An Academic Division of the Dept of Medicineat Harvard Medical Faculty Physicians at BIDMC, Inc.
TM
Yuri Quintana, Ph.D. Director, Global Health Informa8cs Division of Clinical Informa8cs
BIDMC and Harvard Medical School [email protected]
Global E-Health and Medication Adherence
http://www.hmfpinformatics.org