0:00:00.809,0:00:02.270
I'm going to talk to you today

0:00:02.294,0:00:05.777
about the design of medical technology[br]for low-resource settings.

0:00:05.801,0:00:08.154
I study health systems in these countries.

0:00:08.178,0:00:10.093
And one of the major gaps in care,

0:00:10.117,0:00:11.776
almost across the board,

0:00:11.800,0:00:13.967
is access to safe surgery.

0:00:13.991,0:00:16.650
Now one of the major[br]bottlenecks that we've found

0:00:16.674,0:00:19.909
that's sort of preventing[br]both the access in the first place,

0:00:19.933,0:00:23.369
and the safety of those surgeries[br]that do happen, is anesthesia.

0:00:23.846,0:00:26.238
And actually, it's the model[br]that we expect to work

0:00:26.262,0:00:28.961
for delivering anesthesia[br]in these environments.

0:00:29.620,0:00:33.473
Here, we have a scene that you would find[br]in any operating room across the US,

0:00:33.497,0:00:34.974
or any other developed country.

0:00:34.998,0:00:36.185
In the background there

0:00:36.209,0:00:38.780
is a very sophisticated[br]anesthesia machine.

0:00:38.804,0:00:42.653
And this machine is able[br]to enable surgery and save lives

0:00:42.677,0:00:45.976
because it was designed[br]with this environment in mind.

0:00:46.329,0:00:49.261
In order to operate,[br]this machine needs a number of things

0:00:49.285,0:00:51.185
that this hospital has to offer.

0:00:51.209,0:00:54.577
It needs an extremely[br]well-trained anesthesiologist

0:00:54.601,0:00:57.052
with years of training[br]with complex machines

0:00:57.076,0:00:59.497
to help her monitor the flows of the gas

0:00:59.521,0:01:02.084
and keep her patients[br]safe and anesthetized

0:01:02.108,0:01:03.559
throughout the surgery.

0:01:03.583,0:01:06.350
It's a delicate machine[br]running on computer algorithms,

0:01:06.374,0:01:09.592
and it needs special care, TLC,[br]to keep it up and running,

0:01:09.616,0:01:11.474
and it's going to break pretty easily.

0:01:11.498,0:01:14.410
And when it does, it needs[br]a team of biomedical engineers

0:01:14.434,0:01:17.743
who understand its complexities,[br]can fix it, can source the parts

0:01:17.767,0:01:19.999
and keep it saving lives.

0:01:20.688,0:01:22.222
It's a pretty expensive machine.

0:01:22.246,0:01:24.887
It needs a hospital[br]whose budget can allow it

0:01:24.911,0:01:29.805
to support one machine[br]costing upwards of 50 or $100,000.

0:01:30.296,0:01:32.195
And perhaps most obviously,

0:01:32.219,0:01:33.733
but also most importantly --

0:01:33.757,0:01:36.593
and the path to concepts[br]that we've heard about

0:01:36.617,0:01:37.903
kind of illustrates this --

0:01:37.927,0:01:43.073
it needs infrastructure that can supply[br]an uninterrupted source of electricity,

0:01:43.097,0:01:46.109
of compressed oxygen,[br]and other medical supplies

0:01:46.133,0:01:50.411
that are so critical[br]to the functioning of this machine.

0:01:50.435,0:01:53.734
In other words, this machine[br]requires a lot of stuff

0:01:53.758,0:01:55.731
that this hospital cannot offer.

0:01:56.332,0:01:59.352
This is the electrical supply[br]for a hospital in rural Malawi.

0:02:00.153,0:02:01.314
In this hospital,

0:02:01.338,0:02:04.009
there is one person qualified[br]to deliver anesthesia,

0:02:04.033,0:02:05.185
and she's qualified

0:02:05.209,0:02:09.785
because she has 12, maybe 18 months[br]of training in anesthesia.

0:02:09.809,0:02:12.056
In the hospital and in the entire region

0:02:12.080,0:02:14.087
there's not a single biomedical engineer.

0:02:14.111,0:02:15.680
So when this machine breaks,

0:02:15.704,0:02:17.913
the machines that they have[br]to work with break,

0:02:17.937,0:02:19.706
they've got to try and figure it out,

0:02:19.730,0:02:22.022
but most of the time,[br]that's the end of the road.

0:02:22.046,0:02:24.263
Those machines go the proverbial junkyard.

0:02:24.287,0:02:26.616
And the price tag[br]of the machine that I mentioned

0:02:26.640,0:02:28.701
could represent maybe a quarter or a third

0:02:28.725,0:02:31.537
of the annual operating budget[br]for this hospital.

0:02:32.580,0:02:35.964
And finally, I think you can see[br]that infrastructure is not very strong.

0:02:35.988,0:02:38.662
This hospital is connected[br]to a very weak power grid,

0:02:38.686,0:02:40.468
one that goes down frequently.

0:02:40.492,0:02:43.038
So it runs frequently,[br]the entire hospital,

0:02:43.062,0:02:44.485
just on a generator.

0:02:44.509,0:02:46.665
And you can imagine,[br]the generator breaks down

0:02:46.689,0:02:48.001
or runs out of fuel.

0:02:48.361,0:02:50.388
And the World Bank sees this

0:02:50.412,0:02:53.943
and estimates that a hospital[br]in this setting in a low-income country

0:02:53.967,0:02:57.043
can expect up to[br]18 power outages per month.

0:02:58.658,0:03:01.851
Similarly, compressed oxygen[br]and other medical supplies

0:03:01.875,0:03:03.057
are really a luxury,

0:03:03.081,0:03:06.363
and can often be out of stock[br]for months or even a year.

0:03:06.387,0:03:09.551
So it seems crazy, but the model[br]that we have right now

0:03:09.575,0:03:11.780
is taking those machines[br]that were designed

0:03:11.804,0:03:14.087
for that first environment[br]that I showed you

0:03:14.111,0:03:17.975
and donating or selling them[br]to hospitals in this environment.

0:03:18.861,0:03:20.437
It's not just inappropriate,

0:03:20.461,0:03:22.999
it becomes really unsafe.

0:03:23.516,0:03:25.438
One of our partners at Johns Hopkins

0:03:25.462,0:03:29.641
was observing surgeries in Sierra Leone[br]about a year ago.

0:03:30.225,0:03:33.558
And the first surgery of the day[br]happened to be an obstetrical case.

0:03:33.582,0:03:36.565
A woman came in,[br]she needed an emergency C-section

0:03:36.589,0:03:39.318
to save her life and the life of her baby.

0:03:39.805,0:03:41.759
And everything began pretty auspiciously.

0:03:41.783,0:03:44.036
The surgeon was on call and scrubbed in.

0:03:44.060,0:03:45.348
The nurse was there.

0:03:45.372,0:03:48.233
She was able to anesthetize her quickly,[br]and it was important

0:03:48.257,0:03:50.593
because of the emergency[br]nature of the situation.

0:03:50.617,0:03:52.334
And everything began well

0:03:52.358,0:03:54.430
until the power went out.

0:03:55.773,0:03:57.597
And now in the middle of this surgery,

0:03:57.621,0:04:00.690
the surgeon is racing[br]against the clock to finish his case,

0:04:00.714,0:04:03.066
which he can do -- he's got a headlamp.

0:04:03.090,0:04:07.151
But the nurse is literally running[br]around a darkened operating theater

0:04:07.175,0:04:10.130
trying to find anything[br]she can use to anesthetize her patient,

0:04:10.154,0:04:11.843
to keep her patient asleep.

0:04:11.867,0:04:14.750
Because her machine doesn't work[br]when there's no power.

0:04:15.437,0:04:18.491
This routine surgery that many of you[br]have probably experienced,

0:04:18.515,0:04:22.989
and others are probably the product of,[br]has now become a tragedy.

0:04:23.655,0:04:26.391
And what's so frustrating[br]is this is not a singular event;

0:04:26.415,0:04:28.914
this happens across the developing world.

0:04:28.938,0:04:31.891
35 million surgeries[br]are attempted every year

0:04:31.915,0:04:33.511
without safe anesthesia.

0:04:34.210,0:04:37.122
My colleague, Dr. Paul Fenton,[br]was living this reality.

0:04:37.146,0:04:38.976
He was the chief of anesthesiology

0:04:39.000,0:04:41.500
in a hospital in Malawi,[br]a teaching hospital.

0:04:42.028,0:04:43.522
He went to work every day

0:04:43.546,0:04:45.514
in an operating theater like this one,

0:04:45.538,0:04:48.677
trying to deliver anesthesia[br]and teach others how to do so

0:04:48.701,0:04:49.976
using that same equipment

0:04:50.000,0:04:54.311
that became so unreliable,[br]and frankly unsafe, in his hospital.

0:04:55.039,0:04:56.819
And after umpteen surgeries

0:04:56.843,0:04:59.411
and, you can imagine,[br]really unspeakable tragedy,

0:04:59.435,0:05:01.777
he just said, "That's it.[br]I'm done. That's enough.

0:05:01.801,0:05:03.871
There has to be something better."

0:05:04.229,0:05:05.776
He took a walk down the hall

0:05:05.800,0:05:09.231
to where they threw all those machines[br]that had just crapped out on them,

0:05:09.255,0:05:10.932
I think that's the scientific term,

0:05:10.956,0:05:12.298
and he started tinkering.

0:05:12.322,0:05:14.675
He took one part from here[br]and another from there,

0:05:14.699,0:05:17.281
and he tried to come up[br]with a machine that would work

0:05:17.305,0:05:18.933
in the reality that he was facing.

0:05:18.957,0:05:20.171
And what he came up with:

0:05:20.195,0:05:21.518
was this guy.

0:05:21.542,0:05:25.079
The prototype for the Universal[br]Anesthesia Machine --

0:05:25.103,0:05:28.400
a machine that would work[br]and anesthetize his patients

0:05:28.424,0:05:31.999
no matter the circumstances[br]that his hospital had to offer.

0:05:32.519,0:05:34.000
Here it is, back at home

0:05:34.024,0:05:37.453
at that same hospital, developed[br]a little further, 12 years later,

0:05:37.477,0:05:40.657
working on patients[br]from pediatrics to geriatrics.

0:05:41.137,0:05:43.907
Let me show you a little bit[br]about how this machine works.

0:05:43.931,0:05:45.809
Voila!

0:05:46.452,0:05:47.659
Here she is.

0:05:47.683,0:05:48.927
When you have electricity,

0:05:48.951,0:05:51.969
everything in this machine[br]begins in the base.

0:05:51.993,0:05:54.590
There's a built-in[br]oxygen concentrator down there.

0:05:54.614,0:05:57.743
Now you've heard me mention[br]oxygen a few times at this point.

0:05:57.767,0:06:01.555
Essentially, to deliver anesthesia,[br]you want as pure oxygen as possible,

0:06:01.579,0:06:05.002
because eventually you're going[br]to dilute it, essentially, with the gas.

0:06:05.026,0:06:07.462
And the mixture that the patient inhales

0:06:07.486,0:06:09.797
needs to be at least[br]a certain percentage oxygen

0:06:09.821,0:06:11.507
or else it can become dangerous.

0:06:11.531,0:06:13.449
But so in here when there's electricity,

0:06:13.473,0:06:16.348
the oxygen concentrator takes in room air.

0:06:16.372,0:06:19.651
Now we know room air is gloriously free,

0:06:19.675,0:06:20.976
it is abundant,

0:06:21.000,0:06:22.999
and it's already 21 percent oxygen.

0:06:23.436,0:06:27.113
So all this concentrator does[br]is take that room air in, filter it

0:06:27.137,0:06:30.957
and send 95 percent pure oxygen[br]up and across here,

0:06:30.981,0:06:33.187
where it mixes with the anesthetic agent.

0:06:33.743,0:06:37.367
Now before that mixture[br]hits the patient's lungs,

0:06:37.391,0:06:39.605
it's going to pass by here --[br]you can't see it,

0:06:39.629,0:06:41.411
but there's an oxygen sensor here --

0:06:41.435,0:06:45.640
that's going to read out on this screen[br]the percentage of oxygen being delivered.

0:06:46.284,0:06:48.258
Now if you don't have power,

0:06:48.282,0:06:51.888
or, God forbid, the power cuts out[br]in the middle of a surgery,

0:06:51.912,0:06:54.256
this machine transitions automatically,

0:06:54.280,0:06:56.029
without even having to touch it,

0:06:56.053,0:06:58.384
to drawing in room air from this inlet.

0:06:58.805,0:07:00.141
Everything else is the same.

0:07:00.165,0:07:01.761
The only difference is that now

0:07:01.785,0:07:04.999
you're only working[br]with 21 percent oxygen.

0:07:05.406,0:07:08.227
Now that used to be[br]a dangerous guessing game,

0:07:08.251,0:07:10.637
because you only knew[br]if you gave too little oxygen

0:07:10.661,0:07:12.037
once something bad happened.

0:07:12.061,0:07:14.774
But we've put a long-life[br]battery backup on here.

0:07:14.798,0:07:16.992
This is the only part[br]that's battery backed up.

0:07:17.016,0:07:20.421
But this gives control to the provider,[br]whether there's power or not,

0:07:20.445,0:07:22.038
because they can adjust the flows

0:07:22.062,0:07:26.030
based on the percentage of oxygen[br]they see that they're giving the patient.

0:07:26.054,0:07:29.399
In both cases,[br]whether you have power or not,

0:07:29.423,0:07:31.463
sometimes the patient[br]needs help breathing.

0:07:31.487,0:07:34.688
It's just a reality of anesthesia,[br]the lungs can be paralyzed.

0:07:34.712,0:07:36.796
And so we've just added[br]this manual bellows.

0:07:36.820,0:07:39.811
We've seen surgeries[br]for three or four hours

0:07:39.835,0:07:41.693
to ventilate the patient on this.

0:07:42.424,0:07:45.412
So it's a straightforward machine.

0:07:45.436,0:07:48.341
I shudder to say simple;[br]it's straightforward.

0:07:48.936,0:07:50.837
And it's by design.

0:07:50.861,0:07:56.061
You do not need to be a highly trained,[br]specialized anesthesiologist

0:07:56.085,0:07:57.256
to use this machine,

0:07:57.280,0:08:00.032
which is good because,[br]in these rural district hospitals,

0:08:00.056,0:08:02.651
you're not going to get[br]that level of training.

0:08:02.675,0:08:05.803
It's also designed for the environment[br]that it will be used in.

0:08:05.827,0:08:07.604
This is an incredibly rugged machine.

0:08:07.628,0:08:11.083
It has to stand up to the heat[br]and the wear and tear

0:08:11.107,0:08:14.225
that happens in hospitals[br]in these rural districts.

0:08:14.249,0:08:16.644
And so it's not going[br]to break very easily,

0:08:16.668,0:08:19.942
but if it does, virtually[br]every piece in this machine

0:08:19.966,0:08:21.976
can be swapped out and replaced

0:08:22.000,0:08:23.994
with a hex wrench and a screwdriver.

0:08:25.750,0:08:28.010
And finally, it's affordable.

0:08:28.034,0:08:31.687
This machine comes in[br]at an eighth of the cost

0:08:31.711,0:08:34.789
of the conventional machine[br]that I showed you earlier.

0:08:34.813,0:08:39.301
So in other words, what we have here[br]is a machine that can enable surgery

0:08:39.325,0:08:40.497
and save lives,

0:08:40.521,0:08:43.511
because it was designed[br]for its environment,

0:08:43.535,0:08:45.953
just like the first machine I showed you.

0:08:45.977,0:08:47.981
But we're not content to stop there.

0:08:48.005,0:08:49.180
Is it working?

0:08:49.204,0:08:51.786
Is this the design[br]that's going to work in place?

0:08:51.810,0:08:53.577
Well, we've seen good results so far.

0:08:53.601,0:08:56.986
This is in 13 hospitals in four countries,

0:08:57.010,0:09:00.943
and since 2010, we've done[br]well over 2,000 surgeries

0:09:00.967,0:09:03.200
with no clinically adverse events.

0:09:03.540,0:09:04.707
So we're thrilled.

0:09:04.731,0:09:09.153
This really seems like[br]a cost-effective, scalable solution

0:09:09.177,0:09:11.210
to a problem that's really pervasive.

0:09:11.698,0:09:13.298
But we still want to be sure

0:09:13.322,0:09:15.739
that this is the most effective[br]and safe device

0:09:15.763,0:09:17.820
that we can be putting into hospitals.

0:09:17.844,0:09:20.368
So to do that, we've launched[br]a number of partnerships

0:09:20.392,0:09:21.878
with NGOs and universities,

0:09:21.902,0:09:24.551
to gather data on the user interface,

0:09:24.575,0:09:26.767
on the types of surgeries[br]it's appropriate for,

0:09:26.791,0:09:29.178
and ways we can enhance the device itself.

0:09:29.202,0:09:31.856
One of those partnerships[br]is with Johns Hopkins

0:09:31.880,0:09:33.253
just here in Baltimore.

0:09:33.277,0:09:37.555
They have a really cool anesthesia[br]simulation lab out in Baltimore.

0:09:37.579,0:09:39.460
So we're taking this machine

0:09:39.484,0:09:43.058
and recreating some[br]of the operating theater crises

0:09:43.082,0:09:44.589
that this machine might face

0:09:44.613,0:09:46.976
in one of the hospitals[br]that it's intended for,

0:09:47.000,0:09:49.124
and in a contained, safe environment,

0:09:49.148,0:09:50.865
evaluating its effectiveness.

0:09:51.347,0:09:54.979
We're then able to compare[br]the results from that study

0:09:55.003,0:09:56.423
with real-world experience,

0:09:56.447,0:09:58.644
because we're putting[br]two of these in hospitals

0:09:58.668,0:10:00.819
that Johns Hopkins[br]works with in Sierra Leone,

0:10:00.843,0:10:03.957
including the hospital[br]where that emergency C-section happened.

0:10:05.385,0:10:08.513
So I've talked a lot about anesthesia,[br]and I tend to do that.

0:10:08.537,0:10:12.629
I think it is incredibly fascinating[br]and an important component of health.

0:10:12.653,0:10:15.284
And it really seems peripheral,[br]we never think about it,

0:10:15.308,0:10:17.512
until we don't have access to it,

0:10:17.536,0:10:19.498
and then it becomes a gatekeeper.

0:10:19.522,0:10:21.491
Who gets surgery and who doesn't?

0:10:21.824,0:10:24.554
Who gets safe surgery and who doesn't?

0:10:25.216,0:10:27.679
But you know,[br]it's just one of so many ways

0:10:27.703,0:10:30.810
that design, appropriate design,

0:10:30.834,0:10:32.810
can have an impact on health outcomes.

0:10:33.318,0:10:35.437
If more people[br]in the health-delivery space

0:10:35.461,0:10:39.030
really working on some of these[br]challenges in low-income countries

0:10:39.054,0:10:42.707
could start their design process,[br]their solution search,

0:10:42.731,0:10:44.669
from outside of that proverbial box

0:10:44.693,0:10:46.631
and inside of the hospital --

0:10:46.655,0:10:48.486
In other words, if we could design

0:10:48.510,0:10:51.701
for the environment that exists[br]in so many parts of the world,

0:10:51.725,0:10:54.134
rather than the one[br]that we wished existed --

0:10:54.158,0:10:56.324
we might just save a lot of lives.

0:10:56.787,0:10:58.255
Thank you very much.

0:10:58.279,0:11:02.714
(Applause)