Mobile Health for Reducing Disparities: Does it Work and How Will we Know?

2,848 views

Published on

Seminar given at the Medical Effectiveness Research Center, UCSF, June 2011.

Published in: Health & Medicine

Mobile Health for Reducing Disparities: Does it Work and How Will we Know?

  1. 1. Mobile  Health  for  Reducing  HealthDisparities:  Does  it  Work  and  How Will  We  Know? Ida  Sim,  MD,  PhD Director,  Center  for  Clinical  and  Translational  Informatics University  of  California  San  Francisco June  7,  2011
  2. 2. A  Phone  in  73%  of  Pockets 147% 130% 90% 60% 75% 50% 95% 93%
  3. 3. A  Computer  in  73%  of  Pockets 147% 130% 90% 60% 75% 50% 95% 93%
  4. 4. mHealth• using  mobile technologies  in conjunction  with Internet  and  social media  for preventive  and medical  care Corventis Piix EKG Monitor Haiku app, for Epic EHR AsthmaMD app No conflicts with any product mentioned
  5. 5. mHealth  at  Peak  of  Hype Hype Cycle, Gartner Group
  6. 6. Outline• Trends  in  mHealth  Today• The  Digital  Divide,  Restated• Open  Questions• Does  it  Work?• Discussion
  7. 7. Aging-in-placehome monitors Text4Health Devices Enterprise/Doctor Centric AT&T For Health WellDoc FitBit Participatory Health 1Society for Participatory Medicine
  8. 8. Aging-in-placehome monitors Text4Health Devices Enterprise/Doctor Centricself-monitoring and self-care using mobiledevices as “…networked patients AT&T from shift For Healthbeing mere passengers to responsible drivers WellDocof their health, and in which providers FitBitencourage and value them as full partners.”1 Participatory Health 1Society for Participatory Medicine
  9. 9. • “We  can’t  look  at  health  in  isolation.  It’s  not just  in  the  doctor’s  office.  It’s  got  to  be where  we  live,  we  work,  we  play,  we  pray.” U.S.  Surgeon  General  Regina  Benjamin,  LA  Times March  13,  2011
  10. 10. Global  Impact  of  Chronic  Disease WHO | Facts related to Chronic Disease http://www.who.int/dietphysicalactivity/publications/facts/chronic/en/
  11. 11. Aging-in-placehome monitors Text4Health Devices Enterprise/Doctor Centric AT&T For Health WellDoc FitBit Participatory Health LogFrog
  12. 12. mHealth  Assumptions• mHealth  addresses  “last  mile”  of  health  care – objective  is  behavior  change• Technology  +  User  Experience  -­‐-­‐>  Change – “multi-­‐touch”  technology  =  sensors,  phones,  programs – user  experience  =  emotional  experience,  leading  to motivation,  ability,  and  triggers  to  change• Behavior  change  will  lead  to  improved  health outcomes,  reduced  costs,  etc.
  13. 13. Trends  in  Participatory  mHealth• Make  it  simple,  fun,  engaging,  multi-­‐touch – gaming  and  incentives  (e.g.,  rewards  at  Home  Depot) – package  it  like  a  consumer  product• Make  it  hyperlocal – location  doesn’t  matter:  e.g.,  log  your  meals  anytime anywhere – location  is  everything:  e.g.,  text  reminder  NOT  to  walk into  McDonalds• Make  it  social – tie  into  Twitter,  Facebook,  etc.
  14. 14. Open  Questions• Technology  reach  (aka  the  Digital  Divide)• mHealth  usage – going  online/mobile  for  health – social  media  for  health – participatory  health/self-­‐monitoring• Sustainability  of  interventions
  15. 15. Outline• Trends  in  mHealth  Today• The  Digital  Divide,  Restated• Open  Questions• Does  it  Work?• DiscussionData  from  Pew  Internet  and  American  Life  Project,  http://www.pewinternet.org/,  unless  otherwise  stated.
  16. 16. Internet  Access Gap between non-whites (black/Latino) & whites • 66%  of  Americans have  broadband at  home1 – growth  is  flat • Internet  access divide  is  shrinking but  remains  after adjustment  for income  and education21 Home Broadband Survey, Pew Internet, August 20102 http://www.esa.doc.gov/Reports/exploring-digital-nation-home-broadband-internet-adoption-united-states Technology and People of Color 1/25/2011 18
  17. 17. Cell  ownership,  2004-­‐2011Mobile Phone Trends 4/28/2011 19
  18. 18. Asian American: 90% (English-speaking only) • 80%  among  whites; 87%  among  Blacks and  Latinos1 • Smartphone ownership  19% among  Latinos;  23% in  whites2 1Latinos  Online,  Pew,  Sept  2010 2Scarborough  Research,  Dec  2010Mobile Phone Trends 4/28/2011 20
  19. 19. Mobile-­‐only  Households High  Wireless Substitution: • Young  adults (esp.  those ages  24-­‐29) • Renters • Low  income (poverty  line  or below) • Latino/HispanicMobile Phone Trends 4/28/2011 21
  20. 20. “Reverse”  Technology  Divide• Cell  phone  ownership  as  high  as  if  not  higher  in Blacks  and  Latinos•  More  low-­‐income  households  are  cellular  only (no  land  line,  no  broadband) – where  cellphone  is  main  or  only  way  to  get  on  the  web• Overall  trend  is  away  from  broadband/desktop computers  so  overall  technology  divide  will  likely narrow
  21. 21. Digital  Divide  Still  Exists• But  is  in  how  technology  is  used,  not  whether  it  is available• Language  is  strong  predicator – foreign-­‐born  Latino  much  lower  use  of  Internet,  English-­‐ speaking  Latino  equal  to  whites• Also  health  literacy – low  health  literacy  predicts  lower  e-­‐health  use  (Sakar,  J Health  Commun,  2010)• Don’t  automatically  apply  old  assumptions/data from  the  “real”  world  to  the  virtual  world
  22. 22. Outline• Trends  in  mHealth  Today• The  Digital  Divide,  Restated• Open  Questions• Does  it  Work?• Discussion
  23. 23. Open  Questions• mHealth  usage – going  online/mobile  for  health – social  media  for  health – participatory  health/self-­‐monitoring• Sustainability  of  interventions
  24. 24. Internet  Health  Usage %  Internet %  of  US  Adults Users Looked  for  health  info 80% 59%Looked  for  other  people  with 18% 13% similar  health  concerns 1  Social  Life  of  Health  Information,  Pew,  May  2011
  25. 25. Associated withWhites (82% vs. low70s%)Associated withmiddle ages (mid-80%vs. low 70s%)Associated withhigher income
  26. 26. What  Info/Actitivities  Online? %  Internet %  of  US Users Adults Consulted  online  reviews 24% 18% of  drugs/treatmentsConsulted  online  rankings 15% 11%or  reviews  of  hospitals  and other  facilities
  27. 27. Associated with caregiver status and recent health crisis Those with chronic disease and disabilities less likely to look for health info • due to lower Internet access (62% vs. 81%)11  Chronic  Disease  and  the  Internet,  Pew,  Mar  2010
  28. 28. Effect  of  Online  Health  Info?• 60%  say  info  affected  a  real-­‐life  medical  decision• 56%  say  info  changed  their  overall  approach  to maintaining  their  health  or  the  health  of someone  they  help  take  care  of• 38%  say  info  affected  decision  whether  to  see  a doctor• Internet  is  first  source  of  info,  but  doctors  still more  trusted  (increasingly  so) Hesse, et al. NEJM, Mar 4, 2010
  29. 29. Cellphone  Features  Usage • Minorities  use cellphone features  at higher  rates than  WhitesTechnology and People of Color 1/25/2011 31
  30. 30. mHealth  Usage %  Cellphone %  of  US  Adults Users Looked  for  health  info 17% 14% Used  health  apps  for 9% 7.5%tracking/managing  their  health 1  Social  Life  of  Health  Information,  Pew,  May  2011
  31. 31. Mobile  in  action  –  health  apps and  information Technology and People of Color 1/25/2011 33
  32. 32. Internet  and  mHealth  Usage• Increasingly  a  mainstream  Internet  activity• Somewhat  minimal  use  on  mobile  devices – trends  would  suggest  increase  as  Internet  use migrates  to  “mobile  web” – early  indications  of  greater  uptake  among  minorities• Digital  divide  exists,  but  is  non-­‐traditional – less  broadband  use  among  minorities – more  cellphone  owernship  and  use  among  minorities –  greater  interest  in  mHealth  among  those  with  chronic diseases  and  disability,  but  have  lower  Internet  access
  33. 33. Open  Questions• mHealth  usage – going  online/mobile  for  health – social  media  for  health – participatory  health/self-­‐monitoring• Sustainability  of  interventions
  34. 34. Social  Media  Usage  in  General• 62%  of  adult  internet  users  use  social  network sites – 46%  of  all  US  adults• 13%  of  online  Americans  use  Twitter  (Pew,  June  2011) – up  from  8%  in  Nov  2010 – 18-­‐29,  urban,  female,  more  likely  to  Twitter
  35. 35. Technology and People of Color 1/25/2011 37
  36. 36. Daily  Social  Media  Use• Almost  50%  of blacks,  1/3  of whites Daily  Twitter  Use (Tech  Trends  in  People  of  Color,  Pew  Jan.  2011)
  37. 37. Social  Networks  for  Health %  Social %  of  US  Adults Network  Users Followed  friend’s  personal 23% 11%health  or  updates  on  a  social  site Gotten  health  information  from 15% 7% social  networks Memorialized  someone  with  a 17% 8% health  condition 1  Social  Life  of  Health  Information,  Pew,  May  2011
  38. 38. Social  Computing  for  Health• Growing  social  media  use  by  all  Americans – especially  among  minorities – intensity  of  use  higher  in  minorities• Early  use  of  social  media  for  health, uncharted  territory
  39. 39. Open  Questions• mHealth  usage – going  online/mobile  for  health – social  media  for  health – participatory  health/self-­‐monitoring• Sustainability  of  interventions
  40. 40. Self  at  the  Center• Participatory  health,  in  league  with  clinical care  team  and  other  patients – http://www.c3nproject.org/• Self-­‐tracking,  “data-­‐driven  lifestyle”  for  all areas  of  life,  not  just  health – http://quantifiedself.com/
  41. 41. Participatory  Health• Started  strongly  for  patients  with  rare  diseases – e.g.,  http://www.patientslikeme.com/• Now  18%  of  internet  users  find  other  patients – 25%  of  those  with  chronic  health  conditions – transitions  in  health:  new  diagnosis,  pregnancy,  wt. gain/loss,  quitting  smoking – 29%  (?!)  have  contributed  health  content• Professionals  still  the  go-­‐to  for  technical information Peer-to-Peer Health, Pew Internet, Feb 2011
  42. 42. Self-­‐Tracking• 27%  of  internet  users,  or  20%  of  adults,  have tracked  their  weight,  diet,  exercise  routine  or some  other  health  indicators  or  symptoms  online – http://www.medhelp.org/health_tools• Women  more  than  men,  more  if  recent  life change  (gain/lost  wg,  smoking,  pregnancy) 1  Social  Life  of  Health  Information,  Pew,  May  2011
  43. 43. Open  Questions• mHealth  usage – going  online/mobile  for  health – social  media  for  health – participatory  health/self-­‐monitoring• Sustainability  of  interventions
  44. 44. mHealth  Today• Widespread  use  of  Internet  for  health  info• Early  use  of  mobile  tech  for  health  info• Digital  divide  is  with  chronic  health/disabled,  low health  literacy – “reverse  divide”  with  minorities  on  cellphone ownership,  usage  and  social  media  usage• Mostly  people  doing  their  own  thing  with  their own  social  network – mostly  not  integrated  with  clinical  care  team,  other health  professionals,  community,  public  health,
  45. 45. “Full  of  sound  and  fury, signifying  nothing”? Hype Cycle, Gartner Group
  46. 46. App  Usage• 26%  of  downloaded  apps  are  used  only once• Most  (48%)  used  fewer  than  10  times• Little  data  on  sustained  use,  sustained benefit http://www.localytics.com/blog/2011/first-­‐impressions-­‐matter-­‐26-­‐percent-­‐of-­‐apps-­‐ downloaded-­‐used-­‐just-­‐once/
  47. 47. Case  Study:  Text4Baby• Text4Baby  sends  new  (mostly  Medicaid)  mothers brief,  free,  evidence-­‐based  text  messages  for prenatal  and  postpartum  care• A  multi-­‐million  $  public-­‐private  partnership  of 500  partners  (HHS,  wireless  carriers,  Voxiva,  etc.) – launched  Feb  2010,  now  over  157,000  enrollees – spinning  off  into  Text4Baby  Russia,  Text4Health,…• 6  ongoing  evaluations – “96%  would  recommend  Text4Baby” – no  outcomes  data  so  far…
  48. 48. Outline• Trends  in  mHealth  Today• The  Digital  Divide,  Restated• Open  Questions• Does  it  Work?  How  and  when  will  we know??• Discussion
  49. 49. Rephrasing  “Does  it  Work?”(Complexes of) Outcome Exposures strength of association? Increased Text4Baby individual breastfeeding population1With  thanks  to  Rich  Kravitz  MD,  UC  Davis  and  Naihua  Duan,  Columbia
  50. 50. Current  Approaches:  RCT Asthma App ER visits at 1 year 50 people 100 people Usual Care ER visits at 1 year 50 people population• Tests  prespecified  interventions  and  outcomes• To  confirm  a  hypothesis  at  the  population  level• Strong  internal  validity• Problems:  slow  to  set-­‐up,  expensive,  short-­‐term,  lack relevance  to  the  real  world
  51. 51. Current  Approaches:  Data  Mining EHR Exposures Outcomes ? Apps population• Exposures  and  outcomes  from  care  process  systems• To  generate  hypotheses  at  the  population  level• Problems:  limited  to  data  collected,  weak  internal validity  (data  not  complete  or  systematic)
  52. 52. Current  Approaches: N-­‐of-­‐1  Studies Asthma app Usual Care Asthma app peak flow peak flow Usual Care Asthma app Usual Care individual• Within-­‐subject  multiple  crossover• Only  formal  method  for  determining  individual treatment  effectiveness• Problems:  complicated  to  set  up,  analysis  is difficult,  little  known,  not  widely  used
  53. 53. Evidence  Extraction  Attitude• Evidence  is  something  to  be  extracted from  the  care  process – mining  it  from  the  data – directly  manipulating  the  care  process  with rigid  and  pre-­‐defined  protocols
  54. 54. Evidence  Strip  Mining
  55. 55. Evidence  Farming Hay, et al. J Eval Clin Prac 14(2008):707-713.
  56. 56. Rooting  for  Evidence
  57. 57. Industrial  Evidence  Farming Asthma App ER visits at 1 year 50 people100 people Usual Care ER visits at 1 year 50 people population
  58. 58. Personal  Evidence  Gardens Asthma app Usual Care Asthma app peak flow peak flow Usual Care Asthma app Usual Care individual
  59. 59. Personal  Evidence  Gardens Flovent Flovent PRN Flovent dancing dancing Flovent PRN Flovent Flovent PRN individual
  60. 60. Crowdsourcing  What  Matters• (Complexes  of)  Exposures – does  chocolate  trigger  (my)  asthma? – testing  common  regimens  (ACEI,  statin,  b-­‐blocker), complementary  medicines• (Complexes  of)  Outcomes – what  outcomes  do  patients  care  about?
  61. 61. Evidence  MacrosystemRooting for Industrial Evidence Personal Evidence Evidence Farming Gardens
  62. 62. How  can  we  scale  evaluation?
  63. 63. Stovepiped mHealth• Health  apps  built independently – little  data  sharing  and interoperability• Limits  efficiency  and impact  of  quality mHealth
  64. 64. Internet  Hourglass  Model• Standardize  and make  open  the “narrow  waist”• Reduces  duplication, spurs  community innovation,  supports commercial  and  non-­‐ profit  uses
  65. 65. OpenmHealth.org Estrin DE, Sim I. Science; 330: 759-60. 2010.
  66. 66. OpenmHealth.org• The  waist  should  support the  evidence  macrosystem
  67. 67. Open  Architecture  for  an Evidence  Macrosystem• Modules  for  usage  analytics – #  of  text  messages,  #  of  sessions,  etc.• Rooting  for  (glocal)  evidence – data  sharing  with  shared  syntax  and  semantics• Industrial  farming,  e.g.,  with  RCTs – modules  for  informed  consent,  randomization,  adaptive treatment  strategy,  mixed  methods,  etc.• Personal  evidence  gardening,  e.g.,  N-­‐of-­‐1 – modules  for  scripting  and  analyzing  individualized  N-­‐of-­‐ 1  protocols,  etc.
  68. 68. Open  Architecture  for  an Evidence  Macrosystem• Social  media  for  discovery  of  exposures  and outcomes  that  matter• Shared  libraries  of  validated  measures  and instruments  (e.g.,  PROMIS) – measures  that  get  at  finer-­‐grained  mechanisms  based on  theoretical  models  of  change,  etc.
  69. 69. Goal  for  mHealth  Ecosystem• Becomes  a  learning  community  enabled  by  an  open architecture,  to  more  effectively  innovate,  share, and  deploy  best  technology  and  best  practices  for improving  individual  and  population  health
  70. 70. Outline• Trends  in  mHealth  Today• The  Digital  Divide,  Restated• Challenges/Open  Questions• Does  it  Work?• Discussion
  71. 71. • Will  people  really  use  mobile  tech  to  manage  their  health?  Is behavior  change  the  target?• Is  self-­‐tracking  only  for  uber-­‐geeks?• How  much  integration  with  traditional  care  system  is needed?  public  health?  consumer  world?• What  will  be  the  role  of  social  media?• Are  there  fundamentally  different  approaches  needed  for different  population  segments?• How  can  we  learn  as  much  and  as  fast  as  possible  about what  works?• Any  interest  in  establishing  a  trusted  tester  community  in  SF minority  populations?• etc.  etc.

×