1. 2025: 20% doctor included? an exercise in technology speculation & musings vinod khosla vk@khoslaventures.com twitter: @vkhosla

2. 10% to 20% of cases: delayed, missed, and incorrect diagnosis graber, et al., jama, 2005 2

3. 40,000+ patients in u.s. icus may die with a misdiagnosis annually winters, et al., bmj quality & safety, 2012 3

4. 50% of MDs are below-average math 4

5. human doctors cognitive limitations cognitive biases 5

6. a study of one hundred cases of diagnostic error involving internists found… Article: http://archinte.jamanetwork.com/article.aspx?articleid=486642 6

7. …system-related factors contributed to the diagnostic error in 65% of the cases and cognitive factors in 74%... Article: http://archinte.jamanetwork.com/article.aspx?articleid=486642 7

8. …premature closure was the single most common cause Article: http://archinte.jamanetwork.com/article.aspx?articleid=486642 8

9. the value of second opinions Source: http://www.businessinsurance.com/article/20111204/NEWS05/312049987?tags=|74|305|339|342 9 cleveland clinic doctors’ review of initial diagnosis

10. the American College of Cardiology and the American Heart Association made 7,196 recommendations leading to 53 practice guidelines on 22 topics… …48% have level C evidence (the worst kind)… …11% have level A evidence (the best kind)… …and only 19% of recommendations in class I guidelines had level A evidence 10 Source: www.ncbi.nlm.nih.gov/pubmed/19244190www.ncbi.nlm.nih.gov/pubmed/19244190

11. surgeons were given detailed diagnoses & asked if patients should get surgery … half said yes … the other half said no … when asked again two years later, 40% of the docs gave a different answer Source: http://jama.jamanetwork.com/article.aspx?articleid=380215 11

12. four cardiologists were asked to diagnose stenosis in patients using high-quality angiograms … … they disagreed 60% of the time …and disagreed with themselves (on re- reads) 8-37% of the time 12

13. fifty-eight experts’ estimates of the chance of an outcome of an important procedure 0% 0.2% 0.5% 1% 1% 1% 1.5% 1.5% 2% 3% 3% 4% 5% 5% 5% 5% 5% 5% 5% 6% 6% 6% 8% 10% 10% 10% 10% 13% 13% 15% 15% 18% 20% 20% 20% 25% 25% 25% 30% 30% 40% 50% 50% 50% 62% 70% 73% 75 75% 75% 75% 80% 80% 80% 80% 80% 80% 100% what does a consensus of a group whose perceptions might vary from 0% to 100% even mean? 13

14. wide ranges of uncertainty Source: http://jama.jamanetwork.com/article.aspx?articleid=380215 14 seventeen experts’ estimates of the effect of screening on colon cancer deaths 0%25%50%75%100% proportion of colon cancer deaths prevented = one expert’s response

15. conventional wisdom and the “tradition of medicine” Source: http://www.newscientist.com/article/mg20727711.400-fever-friend-or-foe.html 15 should fever be reduced in critically ill patients? “there were seven deaths in people getting standard treatment and only one in those allowed to have fever” “the team felt compelled to call a halt, feeling it would be unethical to allow any more patients to get standard treatment”

16. nearly half of all american adults have difficulty understanding and acting upon health information institute of medicine of the national academies, 2004 http://www.nap.edu/openbook.php?record_id=10883&page=1 16

17. there is good reason to challenge the assumption that every individual practitioner's decision is necessarily correct eddy, et al., jama, 1990 17

18. for most study designs and settings, it is more likely for a research claim to be false than true ioannidis, plos med, 2005 18

19. inescapable conflicts of interest… …most physicians are motivated to deliver quality care… …but the typical physician response also includes a desire to protect high salaries… http://online.wsj.com/article/SB10001424052702303983904579093252573814132.html19

20. entrepreneurs will ask the naïve questions that uncover hidden assumptions… …and move us to the grey zone of “speculations” 20

21. in the future, patients will have the data & analysis to become the CEO of your own health peter diamandis 21

22. 80% of what MDs do can be replaced (with better care than the average MD)… …but not every MD function will be replaced 22

23. the “human” element of care can be provided by the most “humane” humans (and MDs can be humane) 23

24. machines are better at integrative medicine… …across “all symptoms”, demeanor, patient history, phone activity, 1000s of data points, genomics, population management guidelines, … …and machines won’t have to win every time… …they’ll just be better overall 24

25. Lifecom CHAMP in acute care I …distributed care with medical assistants were 91% accurate without labs, imaging, or exams II …“safe triage” with 75% physician bypass rate for acute care encounters 25

26. isabel II matched expert diagnoses 91-95% of the time 26

27. dr. algorithm v0 27

28. the transition will start with “toddler MDs” and digital first-aid kits 28

29. 29 Cellscope: ENT+ derm images… Adamant: breath analysis Eyenetra: auto-optometrist Ginger.io: mental health Alivecor: frequent EKG+ analysis Quanttus: physiological metrics (HR, BP, SV, CO, RR, T, …) Medgle: graph of medicine Healthtap, Crowdmed: crowdsourced answers Kyron: practice based evidence Jawbone, Misfit: wellness wearables

30. don’t wait days to take your daughter to the hospital… …check her ear infection as soon as it hurts 30 *a khosla ventures investment CellScope

31. don’t go to the hospital and get connected to a bunch of electrodes… …take your own ecg for less than a buck… …and know you have heart disease before you have an attack! 31 *a khosla ventures investment AliveCor

32. don’t go to the optometrist… …get measured for glasses at home 32 *a khosla ventures investment EyeNetra

33. don’t guess what’s going on inside your body… …get vital intelligence 33 *a khosla ventures investment Quanttus

34. don’t wait for an asthma attack… …know when it’s coming 34 *a khosla ventures investment Adamant

35. forget kappas of 0.2 in the DSM-5… …get reliable, consistent diagnoses 35 *a khosla ventures investment Ginger.io

36. graph the world of medicine… …and see where you fit 36 *a khosla ventures investment MEDgle

37. evidence-based medicine isn’t enough… …think practice-based evidence 37 *a khosla ventures investment Kyron … use data-mining to learn ethnicity-specific drug interactions (e.g. statins work differently in Indians)

38. thousands of physicians… …no waiting room 38 *a khosla ventures investment HealthTap

39. ...and your innovation here (call us) 39

40. healthcare service stations & digital first aid kits 40

41. keep people out of the doctor’s office… …with point innovations in cardiology, dermatology, optometry, psychiatry, internal medicine, … 41

42. innocuous point innovations… 42 …will evolve into a wave and explode into a tsunami

43. dr. algorithm v0 v1 – 2015 v2 – 2017 v3 – 2019 v4 – 2021 v5 – 2023 v6 – 2025 … 43

44. we’ll start with clumsy point innovations like alivecor, cellscope, adamant, ginger.io, neurotrek, consumer physics, jawbone, misfit, … …“insighted” by machine learning… …leading us to discover things we never knew were right in front of us 44

45. the best MDs will train systems over 10 years… …systems will symbiotically provide “bionic assist” and “AMPLIFY” MDs 45

46. dr. house+++ will be the trainer for dr. algorithm …no manners required! …but manners learned! 46

47. findings thanks to data using statins for in-hospital stroke patients reduced the death rate by 40%! kaiser permanente47

48. 48 the practice of medicine the science of medicine

49. I will be wrong on the specifics but directionally right 49

50. the shift to “computerization” has already happened in other areas… …airline pilots, stock trading, car driving 50

51. there aren’t enough rural doctors in india and few have access to jama journals, mris, … …the world of medicine is under-resourced globally 51

52. 20% doctor included? vinod khosla vk@khoslaventures.com twitter: @vkhosla 52

53. 53

54. in many records of patients with a high-risk diagnosis, high-information clinical findings were present before the diagnosis was established feldman, et al., jamia 54

55. 90M adults with limited health literacy cannot fully benefit from much that the healthcare system has to offer institute of medicine of the national academies 55

56. better data for post-study studies overcome ioannidis’ problem of regressions to the mean 56

57. smart computers can be objective cost minimizers… …while being care optimizers 57

58. the problem is not MDs… …they have served us well in the “practice of medicine”… …but humans are not good at the integrative “science of medicine”… …or the misalignment of incentives 58

59. would a hospital cure you in half the time… …if it meant cutting their business in half? 59

60. would big pharma give you a better drug… …if it meant losing subscription revenue? 60

61. would most psychiatrists cure you… …if it meant cutting off the recurring revenue visits? 61

62. would a medical device manufacturer sell you a less- expensive sensor… …if it meant cannibalizing sales of the really expensive equipment? 62

63. findings thanks to data depression is a risk factor for diabetes 63

64. devices analytics data improved patient outcomes at lower cost architecting healthcare 64

65. social and technological forces are conspiring to make the traditional role of the doctor irrelevant bryan vartabedian (attending, texas children's hospital) 65

66. baseline systems models of patients 66

67. computerized dialog managers 67

68. innovation starts at the fringes 68

69. hundreds of “experts” … 28,000+ forecasts over 20+ years results: “experts” are poorer forecasters than dart-throwing monkeys the folly of experts: tetlock study 69

70. don’t prescribe pills … … prescribe apps 70

71. adverse drug events increase hospital costs by >$3000 per stay… …and billions (estimates range from 110 - 180 B) in annual spending nationwide 71

72. in a study of 3 hospitals,  40% of adverse events were medication-related 72

73. the cost of drug-related morbidity and mortality exceeded $177.4 billion in 2000. 73

74. nearly half of all american adults have difficulty understanding and acting upon health information institute of medicine of the national academies 74

75. most [patients] preferred receiving their discharge information from the [computer] agent compared to their doctors or nurses in the hospital bickmore, et al., interacting with computers 75

76. virtually all of the current quality assurance and cost-containment mechanisms assume that there is "accuracy in numbers." why should we assume that the physician offering the second opinion knows the correct answer? what does a consensus of a group whose perceptions might vary from 0% to 100% even mean? eddy, et al., jama, 1990 76

77. physicians are in the impossible position of not knowing the outcomes of different actions, but having to act anyway eddy, et al., jama, 1990 77

78. while there are ineffective and unnecessary practices, the real questions pertain to which treatments work best and whether the costs and risks of more risky and expensive practices are matched by proportionate benefits eddy, et al., jama, 1990 78

79. don’t just smile and nod… …hear what your “doc” is telling you 79 *a khosla ventures investment Gamgee

80. get the answers to questions… …you never knew to ask 80 *a khosla ventures investment Ayasdi

81. get second opinions… …from everyone 81 *a khosla ventures investment CrowdMed

82. track your life… …all from your wrist 82 *a khosla ventures investment Jawbone

83. track your life… …with a fashion statement 83 *a khosla ventures investment Misfit