1
00:00:01,579 --> 00:00:04,500
We live in a medication nation.

2
00:00:04,524 --> 00:00:09,476
Four point five billion drug prescriptions
will be prescribed by doctors like me

3
00:00:09,500 --> 00:00:11,347
this year, in the United States alone.

4
00:00:11,371 --> 00:00:13,839
That's 15 for every man, woman and child.

5
00:00:13,863 --> 00:00:15,022
And for most of us,

6
00:00:15,046 --> 00:00:18,585
our experience with this medication
is often confusing number of pills,

7
00:00:18,609 --> 00:00:22,204
instructions, side effects,
one-size-fits-all dosing,

8
00:00:22,228 --> 00:00:24,704
which all too often
we aren't taking as prescribed.

9
00:00:24,728 --> 00:00:26,751
And this comes at tremendous expense,

10
00:00:26,775 --> 00:00:29,387
costing us our time,
our money and our health.

11
00:00:29,696 --> 00:00:32,577
And in our now exponential,
connected, data-driven age,

12
00:00:32,601 --> 00:00:35,068
I think we can and we must do better.

13
00:00:35,919 --> 00:00:38,996
So let's take a dive at some
of the challenges we have,

14
00:00:39,020 --> 00:00:41,783
and some potential solutions.

15
00:00:42,133 --> 00:00:43,371
Let's start with the fact

16
00:00:43,395 --> 00:00:46,672
that many drugs don't work
for those who are prescribed to them.

17
00:00:46,696 --> 00:00:51,296
The top 10 grossing drugs
in the United States this year,

18
00:00:51,320 --> 00:00:55,001
they only benefit one in four
to one in 23 of who take them.

19
00:00:55,025 --> 00:00:58,518
That's great if you're number one,
but what about everybody else?

20
00:00:58,997 --> 00:01:02,164
And what's worse, drugs,
when they sometimes don't work,

21
00:01:02,188 --> 00:01:03,625
can still cause side effects.

22
00:01:03,649 --> 00:01:05,525
Take aspirin -- about one in four of us

23
00:01:05,550 --> 00:01:08,463
who take aspirin to reduce
our risk of cardiovascular disease,

24
00:01:08,487 --> 00:01:10,400
are unknowingly aspirin-resistant.

25
00:01:10,424 --> 00:01:13,433
And still have the same risks
of gastrointestinal bleeds

26
00:01:13,457 --> 00:01:15,524
that kill thousands every year.

27
00:01:16,069 --> 00:01:19,029
It's adverse drug reactions like these,

28
00:01:19,053 --> 00:01:22,203
that are, by some estimates,
the number four leading cause of death

29
00:01:22,227 --> 00:01:23,387
in the United States.

30
00:01:23,411 --> 00:01:24,847
My own grandfather passed away

31
00:01:24,871 --> 00:01:27,966
after a single dose of antibiotic
caused his kidneys to fail.

32
00:01:27,990 --> 00:01:31,014
Now, adverse drug reactions
and side effects

33
00:01:31,038 --> 00:01:33,998
are often tied to challenges in dosing.

34
00:01:34,022 --> 00:01:37,760
I trained in pediatrics -- little people,
internal medicine -- big people,

35
00:01:37,784 --> 00:01:40,220
so one night I might have been
on call in the NICU,

36
00:01:40,244 --> 00:01:42,506
carefully dosing
to the fraction of a milligram,

37
00:01:42,530 --> 00:01:43,847
a medication for NICU baby.

38
00:01:43,871 --> 00:01:46,132
The next night -- on call
in the emergency room,

39
00:01:46,156 --> 00:01:49,776
treating a 400-pound lineman
or a frail nursing-home patient,

40
00:01:49,800 --> 00:01:53,744
who, by most accounts, usually
would get the same dose of medications

41
00:01:53,768 --> 00:01:55,014
from the formulary.

42
00:01:55,038 --> 00:01:58,371
Which would mean, most of the time
I would be under-dosing the lineman,

43
00:01:58,395 --> 00:02:01,062
and overdosing the nursing-home patient.

44
00:02:01,498 --> 00:02:02,832
And beyond age and weight,

45
00:02:02,856 --> 00:02:06,149
we tend to ignore differences
in sex and race, in dosing.

46
00:02:06,704 --> 00:02:10,854
Now, beyond this, we know
we have a massive challenge

47
00:02:10,878 --> 00:02:13,077
with noncompliance or low adherence.

48
00:02:13,101 --> 00:02:15,156
Many of us who need
to take our medications,

49
00:02:15,180 --> 00:02:18,069
aren't taking them
or are taking them incorrectly.

50
00:02:18,093 --> 00:02:20,791
You know, 40 percent
of adults in the US over 65

51
00:02:20,815 --> 00:02:23,268
are on five or more
prescription medications.

52
00:02:23,292 --> 00:02:24,617
Sometimes 15 or more.

53
00:02:24,641 --> 00:02:30,130
And even small improvements in adherence
can dramatically save dollars and lives.

54
00:02:31,452 --> 00:02:34,500
So, as we think into the future,

55
00:02:34,524 --> 00:02:36,283
you think that where we are today,

56
00:02:36,307 --> 00:02:40,292
as we often hear about smart,
personalized, targeted drugs,

57
00:02:40,316 --> 00:02:42,196
internet of things, gene therapy, AI,

58
00:02:42,220 --> 00:02:45,053
that we'd already arrived
in this era of precision medicine.

59
00:02:45,077 --> 00:02:46,760
In reality, we still live in an age

60
00:02:46,784 --> 00:02:49,783
of empiric trial and error
imprecision medicine.

61
00:02:49,807 --> 00:02:51,102
I think we can do better.

62
00:02:51,126 --> 00:02:55,503
What if we could reimagine ways
to help make your medicine-taking easier?

63
00:02:55,527 --> 00:02:58,743
To get the right doses
and combinations to match you?

64
00:02:59,161 --> 00:03:01,829
What if we could move beyond
today's literal cutting edge

65
00:03:01,853 --> 00:03:04,640
of pill cutters and fax machines,

66
00:03:04,664 --> 00:03:07,458
to an era where we could have
better outcomes, lower costs,

67
00:03:07,482 --> 00:03:10,629
saving lives and space
in your medicine cabinet?

68
00:03:11,466 --> 00:03:14,244
Well, I think part of the solution
is all the emerging ways

69
00:03:14,268 --> 00:03:17,413
that we can measure and connect
our health care information.

70
00:03:17,815 --> 00:03:20,825
Today, we pretty much live
in a reactive, sick-care world,

71
00:03:20,849 --> 00:03:23,141
siloed information that doesn't flow.

72
00:03:23,165 --> 00:03:26,142
We have the potential to move
into more continuous, real-time

73
00:03:26,166 --> 00:03:28,420
proactive world of true health care.

74
00:03:28,444 --> 00:03:31,609
And part of that starts with
the emerging world of quantified self,

75
00:03:31,633 --> 00:03:34,958
we can measure so much of our
physiology and behaviors today,

76
00:03:34,982 --> 00:03:37,133
and often is siloed
on our phones and scales,

77
00:03:37,157 --> 00:03:40,145
but it's starting to connect
to our clinicians, our caregivers,

78
00:03:40,169 --> 00:03:43,534
so they can better optimize prevention,
diagnostics and therapy.

79
00:03:44,851 --> 00:03:47,755
And when we can do that,
we can do some interesting things.

80
00:03:47,779 --> 00:03:49,847
Take, for example, hypertension.

81
00:03:49,871 --> 00:03:53,656
It's the number one risk factor
for early death and morbidity worldwide.

82
00:03:53,680 --> 00:03:56,974
Half of adult Americans,
on approximation, have hypertension,

83
00:03:56,998 --> 00:03:58,950
less than half have it well-controlled.

84
00:03:58,974 --> 00:04:02,491
It's often because it takes two
or three different classes of medications,

85
00:04:02,515 --> 00:04:05,819
it's tough to do adherence and adjust
your blood pressure medications.

86
00:04:05,843 --> 00:04:08,845
We have 500 preventable deaths
from non-controlled hypertension

87
00:04:08,869 --> 00:04:10,626
in the US, every day.

88
00:04:10,929 --> 00:04:13,794
But now we're in the era
of connected blood pressure cuffs --

89
00:04:13,818 --> 00:04:15,827
the FDA just approved
a blood pressure cuff

90
00:04:15,851 --> 00:04:17,545
that can go into your watch.

91
00:04:17,569 --> 00:04:21,060
There are now prototypes of cuffs,
radar-based blood pressure devices

92
00:04:21,084 --> 00:04:23,371
that can continuously stream
your blood pressure.

93
00:04:23,774 --> 00:04:25,198
So, in the future, I could

94
00:04:25,222 --> 00:04:27,895
instead of spot-checking
my blood pressure in the clinic,

95
00:04:27,919 --> 00:04:30,506
my doctor could see my real-time
numbers and my trends,

96
00:04:30,530 --> 00:04:32,699
and adjust them as necessary,

97
00:04:32,723 --> 00:04:35,173
with the help of a blood pressure
dosing algorithm,

98
00:04:35,197 --> 00:04:37,499
or using the internet of things.

99
00:04:37,523 --> 00:04:40,189
Now, technology today can do even more.

100
00:04:40,904 --> 00:04:43,291
My smart watch, already today,
has an EKG built in,

101
00:04:43,315 --> 00:04:46,656
that can be read
by artificial intelligence,

102
00:04:46,680 --> 00:04:49,601
I'm wearing a small, band-aid sized patch,

103
00:04:49,625 --> 00:04:51,871
that is live-streaming
my vital signs right now,

104
00:04:51,895 --> 00:04:53,085
let's take a look.

105
00:04:53,109 --> 00:04:55,514
They're actually a little
concerning at the moment.

106
00:04:55,538 --> 00:04:57,887
(Laughter)

107
00:04:57,911 --> 00:05:00,349
Now, it's not just my real-time vitals

108
00:05:00,373 --> 00:05:02,569
that can be seen
by my medical team or myself,

109
00:05:02,593 --> 00:05:04,268
it could be my retrospective data,

110
00:05:04,292 --> 00:05:07,878
and again, that'd be used to modify
dosing and medication going forward.

111
00:05:08,298 --> 00:05:10,704
Even my weight can be super quantified.

112
00:05:10,728 --> 00:05:15,604
My weight, now my shape,
how much body mass, fat,

113
00:05:15,628 --> 00:05:16,886
muscle mass I might have,

114
00:05:16,910 --> 00:05:19,830
and use that to optimize
my prevention or therapy.

115
00:05:20,823 --> 00:05:22,633
And it's not just for the tech savvy.

116
00:05:22,657 --> 00:05:25,832
Now, MIT engineers have modified wi-fi

117
00:05:25,856 --> 00:05:29,260
so we can seamlessly
connect and collect our vital signs,

118
00:05:29,284 --> 00:05:31,529
from our connected rings
and smart mattresses,

119
00:05:31,553 --> 00:05:34,268
we can start to share
this digital exhaust, our digitome,

120
00:05:34,292 --> 00:05:37,498
and even potentially crowdsource it,
sharing our health information,

121
00:05:37,522 --> 00:05:39,982
just like we share
with our Google maps and driving,

122
00:05:40,006 --> 00:05:43,773
to improve our -- not our driving,
but our health experience globally.

123
00:05:44,911 --> 00:05:46,308
So, that's great.

124
00:05:46,332 --> 00:05:48,604
We can potentially now
collect this information.

125
00:05:48,628 --> 00:05:52,434
What if your labs can go
from the central lab

126
00:05:52,458 --> 00:05:54,990
to your home, to your phone,
to even inside our bodies

127
00:05:55,014 --> 00:05:57,616
to measure drug levels or other varieties?

128
00:05:58,085 --> 00:06:00,149
And of course, we're in
the age of genomics,

129
00:06:00,173 --> 00:06:03,046
I've been sequenced,
it's just less than 1,000 dollars today.

130
00:06:03,070 --> 00:06:05,569
And I can start to understand
my pharmacogenomics --

131
00:06:05,593 --> 00:06:08,252
how my genes impact,
whether I need high dose, low dose,

132
00:06:08,276 --> 00:06:10,355
or maybe a different
medication altogether.

133
00:06:10,379 --> 00:06:13,054
Let's imagine if your physician
or your pharmacist,

134
00:06:13,078 --> 00:06:15,738
had this information integrated
into their workflow,

135
00:06:15,762 --> 00:06:19,610
augmented with artificial intelligence,
AI, or as I like to refer to it, IA --

136
00:06:19,634 --> 00:06:21,229
intelligence augmentation,

137
00:06:21,253 --> 00:06:22,660
to leverage that information,

138
00:06:22,684 --> 00:06:25,256
to understand of the
18,000 or more approved drugs,

139
00:06:25,280 --> 00:06:28,601
which would be the right dose
and combination for you.

140
00:06:28,625 --> 00:06:31,649
So great, now maybe we can
optimize your drugs and your doses,

141
00:06:31,673 --> 00:06:34,823
but the problem today is,
we're still using this amazing technology

142
00:06:34,847 --> 00:06:36,744
to keep track of our drugs.

143
00:06:36,768 --> 00:06:39,335
And of course, these technologies evolve,

144
00:06:39,359 --> 00:06:42,513
there's connected dispensers,
reminder apps,

145
00:06:42,537 --> 00:06:45,553
smart pill bottle caps that can text
or tweet you or your mother

146
00:06:45,577 --> 00:06:47,410
if you haven't taken your medications.

147
00:06:47,434 --> 00:06:49,252
PillPack was just acquired by Amazon,

148
00:06:49,276 --> 00:06:53,208
so assuming we may have same-day delivery
of our drugs, delivered by drone.

149
00:06:53,807 --> 00:06:55,958
So, all these things are possible today,

150
00:06:55,982 --> 00:06:57,895
but we're still taking multiple pills.

151
00:06:57,919 --> 00:06:59,411
What if we can make it simpler?

152
00:06:59,435 --> 00:07:02,641
I think one of the solutions
is to make better use of the polypill.

153
00:07:02,665 --> 00:07:06,196
A polypill is the integration of multiple
medications into a single pill.

154
00:07:06,220 --> 00:07:09,736
And we have these today in common,
over the counter cold and flu remedies.

155
00:07:09,760 --> 00:07:12,331
And there have been prevention
polypill studies done,

156
00:07:12,355 --> 00:07:15,045
giving combinations of statins,
blood pressure, aspirin,

157
00:07:15,069 --> 00:07:17,110
which in randomized studies
have been shown

158
00:07:17,134 --> 00:07:19,450
to dramatically reduce risk,
compared to placebo.

159
00:07:19,474 --> 00:07:23,506
But these polypills weren't personalized,
they weren't optimized to the individual.

160
00:07:23,530 --> 00:07:26,117
What if we could optimize
your personalized polypill?

161
00:07:26,141 --> 00:07:29,263
So it would be built for you,
based on you, it could adapt to you,

162
00:07:29,287 --> 00:07:30,906
even every single day.

163
00:07:30,930 --> 00:07:34,850
Well, we're now in the era of 3D printing,
you can print personalized braces,

164
00:07:34,874 --> 00:07:36,509
hearing aids, orthopedic devices,

165
00:07:36,533 --> 00:07:39,975
even I've been scanned
and had my genes tailored to fit to me.

166
00:07:39,999 --> 00:07:41,254
So, this got me thinking,

167
00:07:41,278 --> 00:07:44,800
what if we could 3D print
your personalized polypill?

168
00:07:44,824 --> 00:07:47,213
So instead of taking six
medications, for example,

169
00:07:47,237 --> 00:07:48,935
I could integrate them into one.

170
00:07:48,959 --> 00:07:51,251
So it would be easier to take,
improve adherence,

171
00:07:51,275 --> 00:07:53,926
and potentially, it could even
integrate in supplements,

172
00:07:53,950 --> 00:07:56,823
like vitamin D or CoQ10.

173
00:07:57,173 --> 00:08:00,117
So with some help --
I call these intelli-meds --

174
00:08:00,141 --> 00:08:02,887
and with the help of my
IntelliMedicine engineering team,

175
00:08:02,911 --> 00:08:05,879
we built the first IntelliMedicine
prototype printer.

176
00:08:05,903 --> 00:08:07,069
And here's how it works:

177
00:08:07,093 --> 00:08:09,553
instead of full tablets,
we have small micromeds,

178
00:08:09,577 --> 00:08:10,998
one or two milligrams each,

179
00:08:11,022 --> 00:08:14,061
which are sorted and selected
based on the dose and combination

180
00:08:14,085 --> 00:08:15,696
needed for an individual.

181
00:08:15,720 --> 00:08:18,180
And of course, these would be
doses and combinations

182
00:08:18,204 --> 00:08:20,656
you could already take together,
FDA-approved drugs.

183
00:08:20,680 --> 00:08:22,545
We could change the pharmacokinetics,

184
00:08:22,569 --> 00:08:26,699
by professionally layering on different
elements to the individual micromeds.

185
00:08:27,569 --> 00:08:29,196
And when we hit print,

186
00:08:29,220 --> 00:08:32,482
you print your combination of medications
that might be needed by you

187
00:08:32,506 --> 00:08:34,244
on any individual day.

188
00:08:34,268 --> 00:08:38,188
And we'd start with, again, generic drugs
for the most common problems.

189
00:08:38,212 --> 00:08:41,863
About 90 percent of prescribed drugs today
are low-cost generics.

190
00:08:41,886 --> 00:08:45,225
And once we've printed the pill,
we can do some fun bells and whistles.

191
00:08:45,249 --> 00:08:47,276
We could print the name of the patient,

192
00:08:47,300 --> 00:08:50,069
the date, the day of the week, a QR code.

193
00:08:50,093 --> 00:08:54,086
We could print different meds for tapering
for patient on a steroid taper,

194
00:08:54,110 --> 00:08:56,377
or tapering from pain medications.

195
00:08:56,958 --> 00:09:01,926
So, this is actually a look at our
prototype in telemedicine printer.

196
00:09:01,950 --> 00:09:03,284
See, unveil it here.

197
00:09:03,974 --> 00:09:06,895
It has about 16 different silos,

198
00:09:06,919 --> 00:09:09,643
each containing individual micromeds.

199
00:09:10,704 --> 00:09:14,450
And I can now adjust on the software,
individual dosings.

200
00:09:14,926 --> 00:09:18,363
And when I do that,
the robotic arm will adjust

201
00:09:18,387 --> 00:09:21,498
the height of these [unclear]
and the micromeds will release.

202
00:09:21,522 --> 00:09:22,672
I can now --

203
00:09:23,093 --> 00:09:25,514
Automated process would rotate
and cycle through,

204
00:09:25,538 --> 00:09:27,442
to make sure the micromeds are loaded.

205
00:09:27,466 --> 00:09:29,268
And when I hit print,

206
00:09:29,292 --> 00:09:31,704
these will all fall through the device,

207
00:09:31,728 --> 00:09:35,328
I now pull out my
personalized printed polypill,

208
00:09:35,352 --> 00:09:38,873
with the doses and
medications meant for me.

209
00:09:39,379 --> 00:09:42,315
And we can take a look,
if you look back to the slides,

210
00:09:42,339 --> 00:09:45,958
sort of the whole process,
we can see the drug silos being selected,

211
00:09:45,982 --> 00:09:47,990
the pills doing down the different silos,

212
00:09:48,014 --> 00:09:50,796
and being collected
in the individual capsule.

213
00:09:50,820 --> 00:09:52,547
Now, this is great,

214
00:09:52,571 --> 00:09:54,635
I can potentially print
my meds based on me,

215
00:09:54,659 --> 00:09:56,000
instead of taking six pills.

216
00:09:56,024 --> 00:09:59,192
I can now be looking
at my individual dosing,

217
00:09:59,216 --> 00:10:01,410
my smart watch is looking
at my blood pressure,

218
00:10:01,434 --> 00:10:03,982
I needed an adjustment
in my blood pressure medicines,

219
00:10:04,006 --> 00:10:06,339
my coumadin level, my blood is too thin,

220
00:10:06,363 --> 00:10:10,291
so I lower my micromed
dose of coumadin, a blood thinner.

221
00:10:10,315 --> 00:10:12,482
So, this could be
smartly adapted, day to day,

222
00:10:12,506 --> 00:10:15,188
programmed by my physician
or cardiologist.

223
00:10:15,212 --> 00:10:18,764
And you can imagine that larger printers,
fast printers like this,

224
00:10:18,788 --> 00:10:21,553
could be in your corner pharmacy,

225
00:10:21,577 --> 00:10:23,756
in your doctor's office,
in a rural clinic.

226
00:10:23,780 --> 00:10:27,740
But it could eventually merge and shrink
to small ones, that could be in your home,

227
00:10:27,764 --> 00:10:30,859
with integrated cartridges like this,
that are delivered by drone.

228
00:10:31,176 --> 00:10:33,205
Could print your personalized polypill,

229
00:10:33,229 --> 00:10:36,545
each morning on your kitchen
or your bathroom cabinet.

230
00:10:36,569 --> 00:10:38,379
And this could evolve, I think

231
00:10:38,403 --> 00:10:42,212
into an incredible way to improve
adherence in medications across the globe.

232
00:10:42,236 --> 00:10:46,275
So, I hope we can reimagine
the future of medicine in new ways,

233
00:10:46,299 --> 00:10:47,828
moving from poly pharmacy,

234
00:10:47,852 --> 00:10:51,307
one-size-fits-all,
low adherence complications,

235
00:10:51,331 --> 00:10:55,358
to an era of personalized, precise,
on-demand medications,

236
00:10:55,382 --> 00:10:58,759
that can take us and individualize
our own health,

237
00:10:58,783 --> 00:11:00,969
and health and medicine around the planet.

238
00:11:01,525 --> 00:11:02,859
Thank you very much.

239
00:11:02,883 --> 00:11:08,318
(Applause)

240
00:11:09,882 --> 00:11:11,953
Host: Daniel, that's kind of awesome.

241
00:11:11,977 --> 00:11:13,143
Really cool.

242
00:11:13,167 --> 00:11:14,561
Question for you, though.

243
00:11:14,585 --> 00:11:18,141
How long is it until, say that
nursing-home patient that you mentioned

244
00:11:18,165 --> 00:11:20,442
is able to print
their pills in their home?

245
00:11:20,752 --> 00:11:23,553
Daniel Kraft: Well, again,
this is just a prototype.

246
00:11:23,577 --> 00:11:26,704
We think that the regular [unclear]
be automated compounding,

247
00:11:26,728 --> 00:11:30,069
and especially in nursing homes,
folks are taking multiple medications,

248
00:11:30,093 --> 00:11:31,395
and they're often mixed up,

249
00:11:31,419 --> 00:11:34,451
so it would be a perfect place
to start with these technologies.

250
00:11:34,475 --> 00:11:38,150
These aren't going to evolve and start
with printers on your bathroom counter.

251
00:11:38,174 --> 00:11:41,532
We need to be intelligent and smart
about how we roll these things out,

252
00:11:41,556 --> 00:11:45,287
but realizing there's so many challenges
with dosing, adherence and precision,

253
00:11:45,311 --> 00:11:47,915
and now that we have all these
amazing new technologies

254
00:11:47,939 --> 00:11:49,686
that can integrate and be leveraged,

255
00:11:49,710 --> 00:11:52,838
I think we need approaches like this
to really catalyze and foster

256
00:11:52,862 --> 00:11:54,643
a true future of health and medicine.

257
00:11:54,667 --> 00:11:56,326
Host: Great, thank you.
DK: Thanks.

258
00:11:56,350 --> 00:11:58,636
(Applause)