0:00:03.675,0:00:07.166
Hello, I'm Chris Anderson.[br]Welcome to The TED Interview.

0:00:07.190,0:00:10.825
We're gearing up for season four[br]with some extraordinary guests,

0:00:10.849,0:00:14.319
but I don't want to wait for that[br]for today's episode,

0:00:14.343,0:00:17.263
because we're in the middle of a pandemic,

0:00:17.287,0:00:20.642
and there's a guest[br]I really wanted to talk to now.

0:00:21.514,0:00:23.593
He is Adam Kucharski,

0:00:23.617,0:00:26.053
an infectious diseases scientist

0:00:26.077,0:00:29.466
who focuses on the mathematical[br]modeling of pandemics.

0:00:29.990,0:00:31.316
He's an associate professor

0:00:31.340,0:00:33.901
at the London School of Hygiene[br]and Tropical Medicine

0:00:33.925,0:00:35.266
and a TED Fellow.

0:00:35.290,0:00:37.391
(Music)

0:00:37.415,0:00:39.813
(TED Talk) Adam Kucharski:[br]So what kind of behavior

0:00:39.837,0:00:41.598
is actually important for epidemics?

0:00:41.622,0:00:45.260
Conversations, close physical contacts?

0:00:45.284,0:00:47.815
What sort of data should we be collecting

0:00:47.839,0:00:49.022
before an outbreak

0:00:49.046,0:00:51.580
if we want to predict[br]how infection might spread?

0:00:52.268,0:00:56.426
To find out, our team[br]built a mathematical model ...

0:00:56.450,0:00:59.331
Chris Anderson: When it comes[br]to figuring out what to make of

0:00:59.355,0:01:01.785
this pandemic, [br]known technically as COVID-19,

0:01:01.809,0:01:04.760
and informally as just the coronavirus,

0:01:04.784,0:01:07.681
I find his thinking unbelievably helpful.

0:01:07.705,0:01:10.054
And I'm excited to dive into it with you.

0:01:10.467,0:01:12.483
A special callout to my friends on Twitter

0:01:12.507,0:01:14.824
who offered up many[br]suggestions for questions.

0:01:14.848,0:01:18.022
I know this topic[br]is on everyone's mind right now.

0:01:18.046,0:01:19.998
And what I hope this episode does

0:01:20.022,0:01:22.346
is give us all a more nuanced way

0:01:22.370,0:01:26.085
of thinking about how this pandemic[br]has unfolded so far,

0:01:26.109,0:01:27.847
what might be to come

0:01:27.871,0:01:30.499
and what we all collectively[br]can do about it.

0:01:31.212,0:01:32.405
Let's dive in.

0:01:32.429,0:01:33.579
(Music)

0:01:37.777,0:01:39.792
Adam, welcome to the TED Interview.

0:01:39.816,0:01:41.499
Adam Kucharski: Thank you.

0:01:41.523,0:01:44.840
CA: So let's just start[br]with a couple of basics.

0:01:44.864,0:01:48.341
A lot of skeptical people's response --

0:01:48.365,0:01:51.309
certainly over the last few weeks,[br]maybe less so now --

0:01:51.333,0:01:53.929
has been, "Oh, come on,[br]this isn't such a big deal,

0:01:53.953,0:01:56.487
there's a relatively tiny number of cases.

0:01:56.511,0:01:58.903
Compare it to the flu,[br]compare it to anything else.

0:01:58.927,0:02:01.014
There are much bigger[br]problems in the world.

0:02:01.038,0:02:04.378
Why are we making such a fuss about this?"

0:02:04.847,0:02:08.475
And I guess the answer to that fuss[br]is that it comes down to the mathematics.

0:02:08.499,0:02:12.481
We're talking about the mathematics[br]of exponential growth,

0:02:12.505,0:02:13.871
fundamentally, right?

0:02:13.895,0:02:15.045
AK: Exactly.

0:02:15.069,0:02:19.903
And there's a number that we use[br]to get a sense of how easy things spread

0:02:19.927,0:02:22.364
and the level of transmission[br]we're dealing with.

0:02:22.864,0:02:24.661
We call that the reproduction number,

0:02:24.685,0:02:26.122
and conceptually, it's just,

0:02:26.146,0:02:27.873
for each case you have, on average,

0:02:27.897,0:02:29.589
how many others are they infecting?

0:02:29.613,0:02:32.625
And that gives you a sense[br]of how much is this scaling,

0:02:32.649,0:02:34.649
how much this growth[br]is going to look like.

0:02:34.673,0:02:37.875
For coronavirus, we're now seeing,[br]across multiple countries,

0:02:37.899,0:02:41.354
we're seeing each person on average[br]giving it to two or three more.

0:02:42.518,0:02:44.129
CA: So that reproduction number,

0:02:44.153,0:02:48.377
the first thing to understand[br]is that any number above one

0:02:48.401,0:02:50.768
means that this thing is going to grow.

0:02:50.792,0:02:54.880
Any number below one[br]means it's going to diminish.

0:02:55.507,0:02:57.501
AK: Exactly -- if you have it above one,

0:02:57.525,0:02:59.593
then each group of people infected

0:02:59.617,0:03:02.626
are going to be generating more infection[br]than there was before.

0:03:02.650,0:03:04.585
And you will see the exponential effect,

0:03:04.609,0:03:07.920
so if it's two, you will be doubling[br]every round of infection,

0:03:07.944,0:03:09.109
and if it's below one,

0:03:09.133,0:03:12.229
you're going to get something[br]that's going to decline, on average.

0:03:12.602,0:03:14.201
CA: So that number two or higher,

0:03:14.225,0:03:17.514
I think everyone here is maybe familiar[br]with the famous story

0:03:17.538,0:03:19.779
of the chessboard and the grains of rice,

0:03:19.803,0:03:24.316
and if you double the number of grains[br]for every square of the chessboard,

0:03:24.340,0:03:27.712
for the first 10 or 15 squares[br]nothing much happens,

0:03:27.736,0:03:30.403
but by the time you've got[br]to the 64th square,

0:03:30.427,0:03:33.828
you suddenly have tons of rice[br]for every individual on the planet.

0:03:33.852,0:03:34.905
(Laughs)

0:03:34.929,0:03:38.010
Exponential growth is an incredible thing.

0:03:38.034,0:03:39.375
And the small numbers now

0:03:39.399,0:03:42.032
are really not what you[br]should be paying attention to --

0:03:42.056,0:03:45.397
you should be paying attention[br]to the models of what could be to come.

0:03:45.982,0:03:47.133
AK: Exactly.

0:03:47.157,0:03:49.514
Obviously, if you continue[br]the exponential growth,

0:03:49.538,0:03:51.618
you do sometimes get[br]these incredibly large,

0:03:51.642,0:03:53.205
maybe implausibly large numbers.

0:03:53.229,0:03:55.592
But even looking at a timescale[br]of say, a month,

0:03:55.616,0:03:57.378
if the reproduction number is three,

0:03:57.402,0:03:59.442
each person is infecting three on average.

0:03:59.466,0:04:02.903
The gap between these rounds[br]of infection is about five days.

0:04:02.927,0:04:05.188
So if you imagine[br]that you've got one case now,

0:04:05.212,0:04:09.386
that's, kind of, six of these[br]five-day steps in a month.

0:04:09.410,0:04:10.776
So by the end of that month,

0:04:10.800,0:04:12.958
that one person could have generated,

0:04:12.982,0:04:15.712
I think it works out at about 729 cases.

0:04:15.736,0:04:17.291
So even in a month,

0:04:17.315,0:04:19.578
just the scale of this thing[br]can really shoot up

0:04:19.602,0:04:21.135
if it's not controlled.

0:04:22.236,0:04:23.400
CA: And so certainly,

0:04:23.424,0:04:26.495
that seems to be happening[br]on most numbers that you look at now,

0:04:26.519,0:04:29.399
certainly where the virus[br]is in the early stages

0:04:29.423,0:04:31.447
of entering a country.

0:04:31.471,0:04:32.649
You've given a model

0:04:32.673,0:04:38.029
whereby we can much more clearly[br]understand this reproduction number,

0:04:38.053,0:04:41.820
because it seems to me this is almost[br]like the core to how we think of the virus

0:04:41.844,0:04:46.434
and how we respond to it[br]and how much we should fear it, almost.

0:04:46.458,0:04:48.006
And in your thinking,

0:04:48.030,0:04:50.895
you sort of break it down[br]into four components,

0:04:50.919,0:04:54.355
which you call DOTS:

0:04:54.379,0:04:57.029
Duration, Opportunities,

0:04:57.053,0:04:59.069
Transmission probability

0:04:59.093,0:05:00.292
and Susceptibility.

0:05:00.316,0:05:02.425
And I think it would be[br]really helpful, Adam,

0:05:02.449,0:05:04.299
for you to just explain each of these,

0:05:04.323,0:05:07.260
because it's quite a simple equation

0:05:07.284,0:05:11.520
that links those four things[br]to the actual reproduction number.

0:05:11.544,0:05:13.154
So talk about them in turn.

0:05:13.178,0:05:14.662
Duration, what does that mean?

0:05:14.686,0:05:18.364
AK: Duration measures[br]how long someone is infectious for.

0:05:18.742,0:05:19.917
If, for example,

0:05:19.941,0:05:23.780
intuitively, if someone is infectious[br]for a longer period of time,

0:05:23.804,0:05:25.966
say, twice as long as someone else,

0:05:25.990,0:05:27.723
then that's twice the length

0:05:27.747,0:05:30.127
that they've got[br]to be spreading infection.

0:05:30.981,0:05:35.821
CA: And what is the duration[br]number for this virus,

0:05:35.845,0:05:39.901
compared with, say, flu[br]or with other pathogens?

0:05:40.439,0:05:41.808
AK: It depends a little bit

0:05:41.832,0:05:43.855
on what happens[br]when people are infectious,

0:05:43.879,0:05:47.459
if they're being isolated very quickly,[br]that shortens that period of time,

0:05:47.483,0:05:49.749
but potentially, we're looking[br]at around a week

0:05:49.773,0:05:54.090
people are effectively infectious[br]before they might be isolated in hospital.

0:05:54.844,0:05:58.741
CA: And during that week,[br]they may not even be showing symptoms

0:05:58.765,0:06:00.965
for that full week either, right?

0:06:01.395,0:06:05.220
So someone gets infected,[br]there's an incubation period.

0:06:05.680,0:06:08.569
There's a period some way[br]into that incubation period

0:06:08.593,0:06:11.141
where they start being infectious,

0:06:11.165,0:06:14.645
and there may be a period after that,[br]where they start to show symptoms,

0:06:14.669,0:06:17.425
and it's not clear, quite,[br]how those dates align.

0:06:17.449,0:06:18.600
Is that right?

0:06:18.624,0:06:20.561
AK: No, we're getting more information.

0:06:20.585,0:06:24.799
One of the signals we see in data

0:06:24.823,0:06:27.851
that suggest that you may have[br]that early transmission going on

0:06:27.875,0:06:31.893
is when you have this delay[br]from one infection to the next.

0:06:32.349,0:06:34.621
So that seems to be around five days.

0:06:35.191,0:06:37.835
That incubation period,[br]the time for symptoms to appear,

0:06:37.859,0:06:39.483
is also about five days.

0:06:39.507,0:06:41.311
So if you imagine that most people

0:06:41.335,0:06:44.811
are only infecting others[br]when they're symptomatic,

0:06:44.835,0:06:46.422
you'd have that incubation period

0:06:46.446,0:06:49.506
and then you'd have some more time[br]when they're infecting others.

0:06:49.530,0:06:51.848
So the fact that those values[br]seem to be similar,

0:06:51.872,0:06:53.948
suggesting that some people[br]are transmitting

0:06:53.972,0:06:57.808
either very early on or potentially[br]before they're showing clear symptoms.

0:06:58.397,0:07:01.881
CA: So almost implies that on average,

0:07:01.905,0:07:03.649
people are infecting others

0:07:03.673,0:07:06.910
as much before[br]they show symptoms as after.

0:07:07.403,0:07:08.569
AK: Potentially.

0:07:08.593,0:07:10.348
Obviously these are early data sets,

0:07:10.372,0:07:13.348
but I think there's good evidence[br]that a fair number of people,

0:07:13.372,0:07:15.966
either before they're[br]showing clear symptoms

0:07:15.990,0:07:19.421
or maybe they're not showing the kind of[br]very distinctive fever and cough

0:07:19.429,0:07:21.924
but they're feeling unwell[br]and they're shedding virus

0:07:21.938,0:07:23.205
during that period.

0:07:23.625,0:07:28.341
CA: And does that make it[br]quite different from the flu, for example?

0:07:29.421,0:07:32.016
AK: It makes it actually[br]similar to flu in that regard.

0:07:32.040,0:07:34.541
One of the reasons pandemic flu[br]is so hard to control

0:07:34.565,0:07:36.389
and so feared as a threat

0:07:36.413,0:07:40.738
is because so much transmission happens[br]before people are severely ill.

0:07:40.762,0:07:44.149
And that means that by the time[br]you identify these cases,

0:07:44.173,0:07:47.251
they've probably actually spread it[br]to a number of other people.

0:07:47.275,0:07:49.522
CA: Yeah, so this is[br]the trickery of the thing,

0:07:49.546,0:07:54.497
and why it's so hard[br]to do anything about it.

0:07:54.521,0:07:56.387
It is ahead of us all the time,

0:07:56.411,0:08:00.166
and you can't just pay attention[br]to how someone feels

0:08:00.190,0:08:01.410
or what they're doing.

0:08:01.434,0:08:03.426
I mean, how does that happen, by the way?

0:08:03.450,0:08:05.323
How does someone infect someone else

0:08:05.347,0:08:08.109
before they're even showing[br]symptoms themselves,

0:08:08.133,0:08:11.752
because classically, we think of,[br]you know, the person sneezing

0:08:11.776,0:08:14.895
and droplets go through the air[br]and someone else breathes them in

0:08:14.919,0:08:16.546
and that's how infection happens.

0:08:16.570,0:08:20.760
What is actually going on[br]for infection pre-symptoms?

0:08:21.950,0:08:24.529
AK: So the level of transmission[br]we see with this virus

0:08:24.553,0:08:26.673
isn't what we see,[br]for example, with measles,

0:08:26.697,0:08:29.029
where someone sneezes[br]and a lot of virus gets out

0:08:29.053,0:08:32.005
and potentially lots of susceptible[br]people can get exposed.

0:08:32.029,0:08:34.159
So potentially, it could be quite early on

0:08:34.183,0:08:36.292
that even if someone[br]has quite mild symptoms,

0:08:36.316,0:08:37.530
maybe a bit of a cough,

0:08:37.554,0:08:39.720
that's enough for some virus[br]to be getting out

0:08:39.744,0:08:40.917
and particularly,

0:08:40.941,0:08:42.481
some of the work that we've done

0:08:42.505,0:08:44.520
trying to look at sort of[br]close gatherings,

0:08:44.544,0:08:45.871
so very tight-knit meals,

0:08:45.895,0:08:47.784
there was an example in a ski chalet --

0:08:47.808,0:08:50.868
and even in those situations,[br]you might have someone mildly ill,

0:08:50.892,0:08:53.799
but enough virus is getting out[br]and somehow exposing others,

0:08:53.823,0:08:55.823
we're still trying[br]to work out exactly how,

0:08:55.847,0:08:58.315
but there's enough there[br]to cause some infection.

0:08:58.958,0:09:03.125
CA: But if someone's mildly ill,[br]don't they still have symptoms?

0:09:03.149,0:09:07.886
Isn't there evidence that even before[br]they know that they're ill,

0:09:07.910,0:09:11.942
something is going on?

0:09:11.966,0:09:14.062
There was a German paper[br]published this week

0:09:14.086,0:09:18.073
that seemed to suggest[br]that even really early on,

0:09:18.097,0:09:20.968
you take a swab from the back[br]of someone's throat

0:09:20.992,0:09:24.268
and they have hundreds[br]of thousands of these viruses

0:09:24.292,0:09:26.301
already reproducing there.

0:09:26.325,0:09:30.172
Like, can someone just[br]literally just be breathing normally

0:09:30.196,0:09:33.466
and there is some transmission[br]of virus out into the air

0:09:33.490,0:09:35.081
that they don't even know about

0:09:35.105,0:09:36.969
and is either infecting people directly

0:09:36.993,0:09:39.002
or settling on surfaces, is that possible?

0:09:39.026,0:09:41.335
AK: I think that's what[br]we're trying to pin down,

0:09:41.359,0:09:42.546
how much that [unclear].

0:09:42.570,0:09:43.743
As you said,

0:09:43.767,0:09:46.482
there's evidence that you can have[br]people without symptoms

0:09:46.506,0:09:48.450
and you can get virus out their throats.

0:09:48.474,0:09:51.434
And so certainly there's potential[br]that it can be breathed out,

0:09:51.458,0:09:54.804
but is that a fairly rare event[br]for that actual transmission to happen,

0:09:54.828,0:09:58.384
or are we potentially seeing more[br]infections occur through that route?

0:09:58.408,0:10:00.779
So it's really early data,

0:10:00.803,0:10:02.676
and I think it's a piece of the puzzle,

0:10:02.700,0:10:04.910
but we're trying to work out[br]where that fits in

0:10:04.934,0:10:08.388
with what we know about the kind of[br]other transmission events we've seen.

0:10:08.412,0:10:13.649
CA: Alright, so, duration is the duration[br]of the period of infectiousness.

0:10:13.673,0:10:18.291
We think is five to six days,[br]is that what I heard you say?

0:10:18.315,0:10:19.772
AK: Potentially around a week,

0:10:19.796,0:10:23.026
depending on exactly what happens[br]to people when they're infectious.

0:10:23.050,0:10:25.971
CA: And there are cases[br]of people testing positive

0:10:25.995,0:10:28.726
way, way later,[br]after they've got infected.

0:10:29.384,0:10:32.447
That may be true, but they are probably[br]not as infectious then.

0:10:32.471,0:10:35.161
Is that basically right,[br]that's the way to think of this?

0:10:35.185,0:10:37.003
AK: I think that's our working theory,

0:10:37.027,0:10:39.408
that a lot of that infection[br]is happening early on.

0:10:39.432,0:10:42.050
And we see that for a number[br]of respiratory infections,

0:10:42.074,0:10:44.349
that when people obviously[br]become severely ill,

0:10:44.373,0:10:45.912
their behavior is very different

0:10:45.936,0:10:49.188
to when they may be walking around[br]and going about their normal day.

0:10:49.908,0:10:52.827
CA: And so again, comparing[br]that D number to other cases,

0:10:52.851,0:10:54.042
like the flu,

0:10:54.066,0:10:55.534
is flu similar?

0:10:55.558,0:10:57.608
What's the D number for flu?

0:10:58.414,0:11:01.463
AK: So for flu,[br]it's probably slightly shorter

0:11:01.487,0:11:04.776
in terms of the period[br]that people are actively infectious.

0:11:04.800,0:11:07.249
I mean, for flu,[br]it's a very quick turnover

0:11:07.273,0:11:09.054
from one case to the next, actually.

0:11:09.078,0:11:11.514
Even a matter of[br]about three days, potentially,

0:11:11.538,0:11:13.995
from one infection[br]to the person that they infect.

0:11:14.852,0:11:17.958
And then at the other end of the scale,[br]you get things like STDs,

0:11:17.982,0:11:20.678
where that duration could be[br]several months, potentially.

0:11:21.228,0:11:22.379
CA: Right.

0:11:22.403,0:11:26.958
OK, really nothing that unusual so far,[br]in terms of this particular virus.

0:11:26.982,0:11:29.791
Let's look at the O, opportunity.

0:11:29.815,0:11:30.973
What is that?

0:11:30.997,0:11:34.437
AK: So opportunity is a measure[br]of how many chances

0:11:34.461,0:11:37.217
the virus has to spread[br]through interactions

0:11:37.241,0:11:38.709
while someone is infectious.

0:11:38.733,0:11:41.058
So typically, it's a measure[br]of social behavior.

0:11:41.646,0:11:45.014
On average, how many[br]social contacts do people make

0:11:45.038,0:11:48.529
that create opportunities for transmission[br]while they're infectious.

0:11:48.553,0:11:53.696
CA: So it's how many people[br]have you got close enough to

0:11:53.720,0:11:56.095
during a day, during a given day,

0:11:56.119,0:11:58.063
to have a chance of infecting them.

0:11:58.087,0:12:00.990
And that number could be,

0:12:01.014,0:12:04.611
if people aren't taking precautions[br]in a normal, sort of, urban setting,

0:12:04.635,0:12:07.155
I mean, that could run[br]into the hundreds, presumably?

0:12:07.179,0:12:08.751
AK: Potentially, for some people.

0:12:08.775,0:12:11.734
We've done a number of studies[br]looking at that in recent years,

0:12:11.758,0:12:13.992
and the average,[br]in terms of physical contacts,

0:12:14.016,0:12:15.452
is about five people per day.

0:12:15.476,0:12:17.633
Most people will have[br]conversation or contacts

0:12:17.657,0:12:19.007
generally with about 10, 15,

0:12:19.031,0:12:20.195
but obviously,

0:12:20.219,0:12:22.518
between cultures,[br]we see quite a lot of variation

0:12:22.542,0:12:25.101
in the level of physical greetings[br]that might happen.

0:12:25.125,0:12:29.710
CA: And presumably, that number[br]again is no different for this virus

0:12:29.734,0:12:30.900
than for any other.

0:12:30.924,0:12:34.367
I mean, that's just a feature[br]of the lives that we live.

0:12:35.280,0:12:37.049
AK: I think for this one,

0:12:37.073,0:12:39.438
if it's driven through[br]these kind of interactions,

0:12:39.462,0:12:42.367
and we've seen for flu,[br]for other respiratory infections,

0:12:42.391,0:12:46.316
those kinds of fairly close contacts[br]and everyday physical interactions

0:12:46.340,0:12:49.435
seem to be the ones[br]that are important for transmission.

0:12:49.459,0:12:51.650
CA: Perhaps there is one difference.

0:12:51.674,0:12:56.400
The fact that if you're[br]infectious pre-symptoms,

0:12:56.424,0:12:59.790
perhaps that means that actually,[br]there are more opportunities here.

0:12:59.814,0:13:02.784
This is part of the virus's[br]genius, as it were,

0:13:02.808,0:13:06.609
that by not letting on[br]that it's inside someone,

0:13:06.633,0:13:09.976
people continue to interact and go to work

0:13:10.000,0:13:11.777
and take the subway and so forth,

0:13:11.801,0:13:13.729
not even knowing that they're sick.

0:13:13.753,0:13:14.916
AK: Exactly.

0:13:14.940,0:13:16.293
And for something like flu,

0:13:16.317,0:13:20.276
you see when people get ill, clearly,[br]their social contacts drop off.

0:13:20.300,0:13:23.283
So to have a virus that can be infectious

0:13:23.307,0:13:26.149
while people are going around[br]their everyday lives,

0:13:26.173,0:13:28.712
really gives it an advantage[br]in terms of transmission.

0:13:28.736,0:13:29.903
CA: In your modeling,

0:13:29.927,0:13:34.755
do you actually have this[br]opportunities number higher than for flu?

0:13:35.406,0:13:39.940
AK: So for the moment,[br]we're kind of using similar values,

0:13:39.964,0:13:43.023
so we're trying to look at, for example,

0:13:43.047,0:13:45.450
physical contacts[br]within different populations.

0:13:45.474,0:13:47.879
But what we are doing is scaling the risk.

0:13:47.903,0:13:50.212
So that's coming on to the T term.

0:13:50.236,0:13:52.474
So that between each contact,

0:13:52.498,0:13:55.347
what's the risk that a transmission[br]event will occur.

0:13:55.371,0:13:57.620
CA: Alright, so let's go on[br]to this next number,

0:13:57.644,0:14:00.299
the T, transmission probability.

0:14:00.323,0:14:01.982
How do you define that?

0:14:02.006,0:14:04.609
AK: So this measures the chance

0:14:04.633,0:14:07.212
that essentially,[br]the virus will get across

0:14:07.236,0:14:10.110
during a particular opportunity[br]or a particular interaction.

0:14:10.134,0:14:13.120
So you may well have[br]a conversation with somebody,

0:14:13.144,0:14:16.573
but actually, you don't cough[br]or you don't sneeze

0:14:16.597,0:14:19.295
or for some reason,[br]the virus doesn't actually get across

0:14:19.319,0:14:20.969
and expose the other person.

0:14:20.993,0:14:23.794
And so, for this virus, as I mentioned,

0:14:23.818,0:14:25.929
say people are having[br]10 conversations a day,

0:14:25.953,0:14:28.810
but we're not seeing infected people[br]infect 10 others a day.

0:14:28.834,0:14:31.254
So it suggests that not all[br]of those opportunities

0:14:31.278,0:14:33.710
are actually resulting[br]in the virus getting across.

0:14:34.540,0:14:38.921
CA: But people say[br]that this is an infectious virus.

0:14:38.945,0:14:42.452
Like, what is that transmission[br]probability number,

0:14:42.476,0:14:44.407
again, compared with, say, the flu?

0:14:44.858,0:14:48.617
AK: So, we did some analyses[br]looking at these very close gatherings.

0:14:48.641,0:14:51.061
We looked at about 10[br]different case studies

0:14:51.085,0:14:54.889
and we found that about a third[br]of the contacts in those settings

0:14:54.913,0:14:56.764
subsequently got infected

0:14:56.788,0:14:59.145
in these early stages,[br]when people weren't aware.

0:14:59.169,0:15:01.440
So if you had these kind of,[br]big group meals,

0:15:01.464,0:15:05.177
potentially, each contact[br]had about kind of, one in three chance

0:15:05.201,0:15:06.558
of getting exposed.

0:15:06.582,0:15:10.189
For seasonal flu,[br]that tends to be slightly lower,

0:15:10.213,0:15:12.323
but even within households[br]in close settings,

0:15:12.347,0:15:14.482
you don't necessarily[br]get values that high.

0:15:14.506,0:15:18.366
And even for something like SARS,[br]those values have, kind of --

0:15:18.390,0:15:21.700
the risk per interaction you had

0:15:21.724,0:15:24.466
was lower than what we seem[br]to be getting for coronavirus.

0:15:24.490,0:15:25.934
Which intuitively makes sense,

0:15:25.958,0:15:28.120
there must be a higher risk[br]per interaction

0:15:28.144,0:15:30.096
if this thing is spreading so easily.

0:15:30.120,0:15:31.270
CA: Hm.

0:15:32.350,0:15:35.904
OK, and then the fourth letter of DOTS

0:15:35.928,0:15:38.482
is S for susceptibility.

0:15:39.873,0:15:41.045
What's that?

0:15:41.069,0:15:45.591
AK: So that is a measure of the proportion[br]of the population who are susceptible.

0:15:45.615,0:15:48.170
If you imagine you have[br]this interaction with someone,

0:15:48.194,0:15:50.381
the virus gets across, it exposes them,

0:15:50.405,0:15:52.350
but some people may have been vaccinated

0:15:52.374,0:15:54.175
or otherwise have some immunity

0:15:54.199,0:15:55.992
and not develop infection themselves

0:15:56.016,0:15:57.776
and not be infectious to others.

0:15:57.800,0:16:01.291
So we've got to account for this[br]potential proportion of people

0:16:01.315,0:16:04.382
who are not actually[br]going to turn into cases themselves.

0:16:06.371,0:16:11.701
CA: And obviously, there's no vaccine yet[br]for this coronavirus,

0:16:11.725,0:16:16.312
nor is anyone, at least initially,[br]immune, as far as we know.

0:16:16.336,0:16:20.360
So are you modeling that[br]susceptibility number pretty high,

0:16:20.384,0:16:22.097
is that part of the issue here?

0:16:22.121,0:16:23.728
AK: Yeah, I think the evidence

0:16:23.752,0:16:26.577
is that this is going to fully[br]susceptible populations,

0:16:26.601,0:16:29.141
and even in areas,[br]for example, like China,

0:16:29.165,0:16:31.093
where there's been a lot of transmission

0:16:31.117,0:16:33.268
but there's been very strong[br]control measures,

0:16:33.292,0:16:35.339
we estimated that up to[br]the end of January,

0:16:35.363,0:16:38.115
probably about 95 percent of Wuhan[br]are still susceptible.

0:16:38.139,0:16:39.816
So there's been a lot of infection

0:16:39.840,0:16:43.056
but it hasn't really taken much[br]of that component

0:16:43.080,0:16:46.096
of the DOTS, of those four things[br]that drive transmission.

0:16:47.359,0:16:49.371
CA: And so the way the mathematics works,

0:16:49.395,0:16:55.219
I have to confess, amidst the stress[br]of this whole situation,

0:16:55.243,0:16:58.387
the nerd in me kind of loves[br]the elegance of the mathematics here,

0:16:58.411,0:17:00.822
because I'd never really[br]thought about it this way,

0:17:00.846,0:17:03.546
but you basically just multiply[br]those numbers together

0:17:03.570,0:17:05.509
to get the reproduction number.

0:17:05.533,0:17:06.684
Is that right?

0:17:06.708,0:17:07.890
AK: Exactly, yeah,

0:17:07.914,0:17:10.843
you almost take the path[br]of the infection during transmission

0:17:10.867,0:17:12.374
as you multiply those together,

0:17:12.398,0:17:14.541
and that gives you[br]the number for that virus.

0:17:14.565,0:17:17.781
CA: And so there's just[br]a total logic to that.

0:17:17.805,0:17:20.567
It's the number of days,[br]duration that you're infectious,

0:17:20.591,0:17:23.432
it's the number of people[br]you're seeing on average

0:17:23.456,0:17:26.058
during those days[br]that you have a chance to infect.

0:17:27.188,0:17:32.180
Then you multiply that[br]by the transmission probability,

0:17:32.204,0:17:34.880
is virus getting into them, essentially,

0:17:34.904,0:17:36.752
that's what you mean by crossing over.

0:17:36.776,0:17:38.863
And then by the susceptibility number.

0:17:38.887,0:17:41.948
By the way, what do you think[br]the susceptibility probability is

0:17:41.972,0:17:43.312
for this case?

0:17:43.796,0:17:47.050
AK: I think we have to assume[br]that it's near 100 percent

0:17:47.074,0:17:49.110
in terms of spread, yeah.

0:17:50.133,0:17:52.434
CA: Alright, you multiply[br]those numbers together,

0:17:52.458,0:17:57.454
and right now, it looks like,[br]for this coronavirus,

0:17:57.478,0:18:02.228
that you say two to three[br]is the most plausible current number.

0:18:02.530,0:18:04.663
Which implies very rapid growth.

0:18:05.141,0:18:06.308
AK: Exactly.

0:18:06.332,0:18:07.878
In these uncontrolled outbreaks,

0:18:07.902,0:18:10.338
we're seeing now a number[br]of countries in the stage

0:18:10.362,0:18:13.443
that you are going to get[br]this really rapid growth occurring.

0:18:13.804,0:18:18.638
CA: And so how does[br]that two to three compare with flu?

0:18:18.662,0:18:22.217
And I guess, there's seasonal flu,

0:18:22.241,0:18:23.942
in the winter, when it's spreading,

0:18:23.966,0:18:28.244
and at other times during the year[br]drops well below one

0:18:28.268,0:18:30.403
as a reproduction number, right?

0:18:30.427,0:18:33.093
But what is it during seasonal flu time?

0:18:33.831,0:18:36.045
AK: During the early stage[br]when it's taking off

0:18:36.069,0:18:37.596
at the start of the flu season,

0:18:37.620,0:18:41.736
it's probably, we reckon,[br]somewhere between about maybe 1.2-1.4.

0:18:41.760,0:18:43.706
So it's not incredibly transmissible,

0:18:43.730,0:18:47.735
if you imagine you do have some immunity[br]in your population from vaccination

0:18:47.759,0:18:48.949
and from other things.

0:18:48.973,0:18:50.560
So it can spread, it's above one,

0:18:50.584,0:18:54.139
but it's not taking off, necessarily,[br]as quickly as the coronavirus is.

0:18:55.020,0:18:58.268
CA: So I want to come back[br]to two of those elements,

0:18:58.292,0:19:00.856
specifically opportunity[br]and transmission probability,

0:19:00.880,0:19:05.553
because those seem to have the most[br]chance to actually do something

0:19:05.577,0:19:07.387
about this infection rate.

0:19:07.411,0:19:08.561
Before we go there,

0:19:08.585,0:19:10.927
let's talk about another[br]key number on this,

0:19:10.951,0:19:13.910
which is the fatality rate.

0:19:13.934,0:19:15.609
First of all, could you define --

0:19:15.633,0:19:18.410
I think there's two different versions[br]of the fatality rate

0:19:18.434,0:19:20.442
that maybe confuse people.

0:19:20.466,0:19:22.466
Could you define them?

0:19:22.490,0:19:26.584
AK: So the one that we often talk about[br]is what's known as the case fatality rate,

0:19:26.608,0:19:30.482
and that's of the proportion who show up[br]with symptoms as cases,

0:19:30.506,0:19:33.559
what proportion of those[br]will subsequently be fatal.

0:19:34.177,0:19:36.313
And we also sometimes talk[br]about what's known

0:19:36.337,0:19:37.829
as the infection fatality rate,

0:19:37.853,0:19:39.789
which is, of everyone who gets infected,

0:19:39.813,0:19:41.027
regardless of symptoms,

0:19:41.051,0:19:43.736
how many of those infections[br]will subsequently be fatal.

0:19:43.760,0:19:45.866
But most of the values[br]we see kicking around

0:19:45.890,0:19:49.032
are the case fatality rate, or the CFR,[br]as it's sometimes known.

0:19:49.963,0:19:54.032
CA: And so what is[br]that fatality rate for this virus,

0:19:54.056,0:19:57.264
and again, how does that compare[br]with other pathogens?

0:19:57.668,0:20:00.482
AK: So there's a few numbers[br]that have been bouncing around.

0:20:00.506,0:20:04.189
One of the challenges in real time[br]is you often don't see all of your cases,

0:20:04.213,0:20:06.688
you have people symptomatic[br]not being reported.

0:20:06.712,0:20:07.871
You also have a delay.

0:20:07.895,0:20:09.236
If you imagine, for example,

0:20:09.260,0:20:11.815
100 people turn up to a hospital[br]with coronavirus

0:20:11.839,0:20:13.307
and none have died yet,

0:20:13.331,0:20:15.680
that doesn't imply[br]that the fatality rate is zero,

0:20:15.704,0:20:18.569
because you've got to wait to see[br]what might happen to them.

0:20:18.593,0:20:21.482
So when adjust for that[br]underreporting and delays,

0:20:21.506,0:20:24.601
best estimate for the case fatality[br]is about one percent.

0:20:24.625,0:20:26.729
So about one percent[br]of people with symptoms,

0:20:26.753,0:20:28.133
on average,

0:20:28.157,0:20:29.403
those outcomes are fatal.

0:20:29.427,0:20:32.282
And that's probably about 10 times[br]worse than seasonal flu.

0:20:33.514,0:20:37.174
CA: Yeah, so that's a scary[br]comparison right there,

0:20:37.198,0:20:40.351
given how many people die of flu.

0:20:40.375,0:20:44.844
So when the World Health Organization[br]mentioned a higher number

0:20:44.868,0:20:47.608
a little while back, of 3.4 percent,

0:20:47.632,0:20:50.331
they were criticized a bit for that.

0:20:50.355,0:20:54.102
Explain why that[br]might have been misleading

0:20:54.126,0:20:57.173
and how to think about it[br]and adjust for that.

0:20:57.197,0:21:00.300
AK: It's incredibly common that people[br]look at these raw numbers,

0:21:00.324,0:21:03.238
they say, "How many deaths[br]are there so far, how many cases,"

0:21:03.262,0:21:04.633
and they look at that ratio,

0:21:04.657,0:21:08.141
and even a couple of weeks ago,[br]that number produced a two percent value.

0:21:08.165,0:21:10.791
But if you imagine[br]you have this delay effect,

0:21:10.815,0:21:13.061
then even if you stop all your cases,

0:21:13.085,0:21:15.918
you will still have these kind of[br]fatal outcomes over time,

0:21:15.942,0:21:17.688
so that number will creep up.

0:21:17.712,0:21:21.617
This has occurred in every epidemic[br]from pandemic flu to Ebola,

0:21:21.641,0:21:23.228
we see this again and again.

0:21:23.252,0:21:26.982
And I made the point to a number of people[br]that this number is going to go up,

0:21:27.006,0:21:28.536
because as China's cases slow,

0:21:28.560,0:21:30.220
it will look like it's increasing,

0:21:30.244,0:21:32.694
and that's just kind of[br]a statistical quirk.

0:21:32.718,0:21:35.131
There's nothing really[br]kind of, behind the change,

0:21:35.155,0:21:37.638
there's no mutations or anything going on.

0:21:38.972,0:21:41.940
CA: If I have this right,[br]there are two effects going on.

0:21:41.964,0:21:45.056
One is that the number of fatalities

0:21:45.080,0:21:47.618
from the existing caseload will rise,

0:21:47.642,0:21:51.634
which actually would boost[br]that 3.4 even higher.

0:21:51.658,0:21:55.482
But then you have to offset that[br]against the fact that apparently,

0:21:55.506,0:21:58.093
huge numbers of cases[br]have just gone undetected

0:21:58.117,0:21:59.579
and that we haven't,

0:21:59.603,0:22:02.025
due to bad testing,

0:22:02.049,0:22:04.589
that the number of fatalities don't --

0:22:04.613,0:22:07.956
They probably reflect[br]a much larger number of early cases.

0:22:07.980,0:22:09.151
Is that it?

0:22:09.175,0:22:10.349
AK: Exactly.

0:22:10.373,0:22:12.445
So you have one thing[br]pulling the number up,

0:22:12.469,0:22:13.926
and one thing pulling it down.

0:22:13.950,0:22:16.243
And it means that on these[br]kind of early values,

0:22:16.267,0:22:18.284
if you actually just adjust for the delay,

0:22:18.308,0:22:20.411
and don't think[br]about these unreported cases,

0:22:20.435,0:22:22.873
you start getting really[br]very scary numbers indeed.

0:22:22.897,0:22:24.882
You get up to 20, 30 percent potentially,

0:22:24.906,0:22:26.206
which really doesn't align

0:22:26.230,0:22:29.191
with what we know[br]about this virus in general.

0:22:29.542,0:22:30.874
CA: Alright.

0:22:31.621,0:22:33.137
There's a lot more data in now.

0:22:33.161,0:22:36.781
From your point of view,[br]you think the likely fatality rate,

0:22:36.805,0:22:41.503
at least in the earlier stage[br]of an infection,

0:22:41.527,0:22:43.687
is about two percent?

0:22:44.330,0:22:45.481
AK: I think overall,

0:22:45.505,0:22:48.806
I think we can put something probably[br]in the 0.5 to two percent range,

0:22:48.830,0:22:51.644
and that's on a number[br]of different data sets.

0:22:51.668,0:22:53.668
And that's for people who are symptomatic.

0:22:53.692,0:22:56.736
I think on average, one percent[br]is a good number to work with.

0:22:56.760,0:22:58.301
CA: OK, one percent,

0:22:58.325,0:23:01.370
So flu is often quoted[br]as a tenth of a percent,

0:23:01.394,0:23:06.513
so it's five to ten times or more[br]more dangerous than flu.

0:23:06.537,0:23:10.006
And that danger is not symmetric[br]across age groups,

0:23:10.030,0:23:11.395
as is well known.

0:23:11.419,0:23:14.466
It primarily affects the elderly.

0:23:14.490,0:23:16.815
AK: Yeah, we've seen[br]that one percent on average,

0:23:16.839,0:23:19.458
but once you start getting[br]into the over 60s, over 70s,

0:23:19.482,0:23:21.323
that number really starts to shoot up.

0:23:21.347,0:23:24.109
I mean, we're estimating[br]potentially in these older groups,

0:23:24.133,0:23:28.879
you're looking at maybe five,[br]ten percent fatality.

0:23:29.243,0:23:31.013
And then of course, on top of that,

0:23:31.037,0:23:33.656
you've got to add what[br]are going to be the severe cases

0:23:33.680,0:23:35.950
and people are going to require[br]hospitalization.

0:23:35.974,0:23:38.853
And those risks get very large[br]in the older groups indeed.

0:23:41.193,0:23:43.296
CA: Adam, put these numbers[br]together for us.

0:23:43.320,0:23:44.470
In your models,

0:23:44.494,0:23:49.407
if you put together[br]a reproduction rate of two to three

0:23:49.431,0:23:53.548
and a fatality rate[br]of 0.5 percent to one percent,

0:23:53.572,0:23:56.062
and you run the simulation,

0:23:56.086,0:23:57.620
what does it look like?

0:23:58.699,0:24:01.517
AK: So if you have[br]this uncontrolled transmission,

0:24:01.541,0:24:04.030
and you have this reproduction[br]number of two or three

0:24:04.054,0:24:05.783
and you don't do anything about it,

0:24:05.807,0:24:07.480
the only way the outbreak ends

0:24:07.504,0:24:11.289
is enough people get it[br]and immunity builds up

0:24:11.313,0:24:15.400
and the outbreak kind of ends on its own.

0:24:15.424,0:24:16.587
And in that case,

0:24:16.611,0:24:19.861
you would expect very large numbers[br]of the population to be infected.

0:24:19.885,0:24:21.449
It's what we see, for example,

0:24:21.473,0:24:24.019
with many other uncontained outbreaks,

0:24:24.043,0:24:26.355
that it essentially burns[br]through the population,

0:24:26.379,0:24:27.856
you get large numbers infected

0:24:27.880,0:24:30.869
and with this kind of fatality rate[br]and hospitalization rate,

0:24:30.893,0:24:35.022
that would really be hugely damaging[br]if that were to occur.

0:24:35.046,0:24:37.252
Certainly at the country level,[br]we're seeing --

0:24:37.276,0:24:39.133
Italy is a good example at the moment,

0:24:39.157,0:24:41.768
that if you have that early[br]transmission is undetected,

0:24:41.792,0:24:43.045
that rapid growth,

0:24:43.069,0:24:47.260
you very quickly get to a situation[br]where your health systems are overwhelmed.

0:24:47.284,0:24:50.895
I think one of the nastiest[br]aspects of this virus

0:24:50.919,0:24:55.002
is that because you have the delay[br]between infection and symptoms

0:24:55.026,0:24:57.279
and people showing up in health care,

0:24:57.303,0:24:59.145
if your health system is overwhelmed,

0:24:59.169,0:25:00.382
even on that day,

0:25:00.406,0:25:02.430
if you completely stop transmission,

0:25:02.454,0:25:05.328
you've got all of these people[br]who have already been exposed,

0:25:05.352,0:25:08.260
so you're still going to have cases[br]and severe cases appearing

0:25:08.284,0:25:10.180
for maybe another couple of weeks.

0:25:10.204,0:25:13.149
So it's really this huge[br]accumulation of infection and burden

0:25:13.173,0:25:16.144
that's coming through the system[br]on your population.

0:25:17.497,0:25:19.561
CA: So there's another[br]key number, actually,

0:25:19.585,0:25:24.098
is how does the total case number

0:25:24.122,0:25:27.791
compare to the capacity[br]of a country's health system

0:25:27.815,0:25:30.219
to process that number of cases.

0:25:30.846,0:25:33.014
Presumably that issue[br]makes a huge difference

0:25:33.038,0:25:34.235
to the fatality rate,

0:25:34.259,0:25:37.037
the difference between people[br]coming in with severe illness

0:25:37.061,0:25:40.410
and a health system that's able[br]to respond and one that's overwhelmed.

0:25:40.434,0:25:43.363
The fatality rate is going to be[br]very different at that point.

0:25:43.387,0:25:45.141
AK: If someone requires an ICU bed,

0:25:45.165,0:25:47.816
that's a couple of weeks[br]they're going to require it for

0:25:47.840,0:25:50.331
and you've got more cases[br]coming through the system,

0:25:50.355,0:25:52.045
so it very quickly gets very tough.

0:25:52.069,0:25:54.942
CA: So talk about the difference[br]between containment

0:25:54.966,0:25:56.529
and mitigation.

0:25:56.553,0:25:59.895
These are different terms[br]that we're hearing a lot about.

0:25:59.919,0:26:06.110
In the early stages of the virus,[br]governments are focused on containment.

0:26:06.134,0:26:07.737
What does that mean?

0:26:07.761,0:26:11.467
AK: Containment is this idea[br]that you can focus your effort on control

0:26:11.491,0:26:13.737
very much on the cases and their contacts.

0:26:13.761,0:26:16.467
So you're not causing disruption[br]to the wider population,

0:26:16.491,0:26:19.167
you have a case that comes in,[br]you isolate them,

0:26:19.191,0:26:21.459
you work out who they've come[br]into contact with,

0:26:21.483,0:26:24.550
who's potentially these[br]opportunities for exposure,

0:26:24.574,0:26:26.537
and then you can follow up those people,

0:26:26.561,0:26:30.133
maybe quarantine them to make sure[br]that no further transmission happens.

0:26:30.157,0:26:32.680
So it's a very focused, targeted method,

0:26:32.704,0:26:35.304
and for SARS, it worked remarkably well.

0:26:35.982,0:26:37.823
But I think for this infection,

0:26:37.847,0:26:41.397
because some cases are going to be missed,[br]they're going to be undetected,

0:26:41.421,0:26:44.627
you've really got to be capturing[br]a large chunk of people at risk.

0:26:44.651,0:26:46.153
If a few slip through the net,

0:26:46.177,0:26:48.356
potentially, you're going[br]to get an outbreak.

0:26:48.380,0:26:49.749
CA: Are there any countries

0:26:49.773,0:26:51.855
that have been able[br]to employ this strategy

0:26:51.879,0:26:55.188
and effectively contain the virus?

0:26:55.212,0:26:58.760
AK: So Singapore have been doing[br]a really remarkable job of this

0:26:58.784,0:27:00.324
for the last six weeks or so.

0:27:00.744,0:27:03.157
So as well as some wider measures,

0:27:03.181,0:27:04.903
they've been working incredibly hard

0:27:04.927,0:27:07.482
to trace people[br]who have come into contact.

0:27:08.117,0:27:09.585
Looking at CCTV,

0:27:09.609,0:27:12.569
going through to find out[br]which taxi someone might have gotten,

0:27:12.593,0:27:13.767
who might be at risk --

0:27:13.791,0:27:15.450
really, really thorough follow up.

0:27:15.474,0:27:18.656
And for about six weeks,[br]that has kept a lid on transmission.

0:27:19.355,0:27:20.529
CA: So that's amazing.

0:27:20.553,0:27:22.974
So someone comes into the country,

0:27:22.998,0:27:24.553
they test positive --

0:27:24.577,0:27:27.133
they go to work, and with a massive team,

0:27:27.157,0:27:29.268
and trace everything,

0:27:29.292,0:27:31.133
to the level of actually saying,

0:27:31.157,0:27:33.157
"Oh, you don't know what taxi you went in?

0:27:33.181,0:27:34.664
Let us find that out for you."

0:27:34.688,0:27:36.926
And presumably,[br]when they find the taxi driver,

0:27:36.950,0:27:40.344
they then have to try and figure out[br]everyone else who was in that taxi?

0:27:40.368,0:27:43.463
AK: So they will focus on[br]close contacts of people most at risk,

0:27:43.487,0:27:47.327
but they're really minimizing the chance[br]that anyone slips through the net.

0:27:47.925,0:27:51.680
CA: But even in Singapore,[br]if I'm not mistaken,

0:27:51.704,0:27:54.054
numbers started to trend[br]back down to zero,

0:27:54.078,0:27:56.522
but I think recently,[br]they've picked up again a bit.

0:27:56.546,0:27:58.435
It's still unclear

0:27:58.459,0:28:01.617
whether they will actually[br]be able to sustain containment.

0:28:01.641,0:28:02.801
AK: Exactly.

0:28:02.825,0:28:05.029
If we talk in terms[br]of the reproduction number,

0:28:05.053,0:28:07.443
we saw it dipped to maybe 0.8, 0.9,

0:28:07.467,0:28:09.402
so under that crucial value of one.

0:28:10.201,0:28:11.550
But in the last week or two,

0:28:11.574,0:28:14.876
it does seem to be taking up[br]and they're getting more cases appearing.

0:28:14.900,0:28:16.138
I think a lot of it is,

0:28:16.162,0:28:18.130
even if they are containing it,

0:28:18.154,0:28:20.065
the world is experiencing outbreaks

0:28:20.089,0:28:22.157
and just keeps throwing[br]sparks of infection,

0:28:22.181,0:28:23.760
and it becomes harder and harder

0:28:23.784,0:28:26.522
with that level of intensive effort[br]to stamp them all out.

0:28:26.546,0:28:32.349
(Music)

0:28:47.712,0:28:49.578
CA: In the case of this virus,

0:28:49.602,0:28:52.688
you know, there was warning[br]to most countries in the world

0:28:52.712,0:28:54.172
that this thing was happening.

0:28:54.196,0:28:58.380
The news out of China[br]very quickly became very bleak

0:28:58.404,0:29:01.118
and people had time to prepare.

0:29:01.142,0:29:05.630
I mean, what would an ideal[br]preparation look like

0:29:05.654,0:29:07.805
if you know that something[br]like this is coming

0:29:07.829,0:29:09.832
and you know that there's[br]a lot on the line

0:29:09.856,0:29:12.807
if you can successfully contain it[br]before it really escapes?

0:29:13.291,0:29:16.022
AK: I think two things[br]would make a really big difference.

0:29:16.046,0:29:20.767
One is having as thorough follow up[br]and detection as possible.

0:29:20.791,0:29:22.416
We've done some modeling analyses,

0:29:22.440,0:29:25.529
looking at how effective[br]that kind of early containment is.

0:29:25.553,0:29:29.671
And it can be, if you're identifying[br]maybe 70 or 80 percent

0:29:29.695,0:29:32.799
of the people who might have[br]come into contact.

0:29:32.823,0:29:36.348
But if you're not detecting[br]those cases coming in,

0:29:36.372,0:29:38.368
if you're not detecting their contacts --

0:29:38.392,0:29:41.948
and a lot of the early focus, for example,[br]was on travel history to China,

0:29:41.972,0:29:44.669
and then it became clear[br]that the situation was changing,

0:29:44.693,0:29:47.852
but because you were relying on that[br]as your definition of a case,

0:29:47.876,0:29:50.860
it meant a lot of maybe other cases[br]that matched the definition

0:29:50.884,0:29:52.051
weren't being tested

0:29:52.075,0:29:54.617
because they didn't seem[br]to be potentially at risk.

0:29:54.641,0:29:59.227
CA: So I mean, if you know[br]that early detection is key to this,

0:29:59.251,0:30:01.131
an essential early measure, I guess,

0:30:01.155,0:30:06.268
would be to rapidly ensure[br]that you had enough tests available,

0:30:06.292,0:30:07.990
and where needed,

0:30:08.014,0:30:09.783
so that you could respond,

0:30:09.807,0:30:13.720
be ready to swing into action[br]as soon as someone was detected,

0:30:13.744,0:30:19.029
you then have to very quickly,[br]I guess, test their contacts and so forth,

0:30:19.053,0:30:21.910
to have a chance[br]of keeping this under control.

0:30:21.934,0:30:23.220
AK: Exactly.

0:30:23.244,0:30:26.410
In my line of work, we say[br]there's value in a negative test,

0:30:26.434,0:30:29.760
because it shows that you're looking[br]for something and it's not there.

0:30:29.784,0:30:33.283
And so I think having[br]small numbers of people tested

0:30:33.307,0:30:36.315
doesn't give you confidence[br]that you're not missing infections,

0:30:36.339,0:30:39.308
whereas if you are doing[br]really thorough follow up on contacts,

0:30:39.332,0:30:41.316
we've seen countries even like Korea now,

0:30:41.340,0:30:42.901
huge numbers of people tested.

0:30:42.925,0:30:45.013
So although there are still[br]cases appearing,

0:30:45.037,0:30:46.426
it gives them more confidence

0:30:46.450,0:30:49.241
that they have some sense[br]of where those infections are.

0:30:49.265,0:30:51.963
CA: I mean, you're in the UK right now,

0:30:51.987,0:30:54.767
I'm in the US.

0:30:54.791,0:30:58.368
How likely is it that the UK[br]is going to be able to contain,

0:30:58.392,0:31:02.194
how likely is it that the US[br]is going to be able to contain this?

0:31:03.162,0:31:06.625
AK: I think it's pretty[br]unlikely in both cases.

0:31:06.649,0:31:10.051
I think the UK is going to have[br]to introduce some additional measures.

0:31:10.075,0:31:12.410
I think when that happens[br]obviously depends a bit

0:31:12.434,0:31:13.656
on the current situation,

0:31:13.680,0:31:16.013
but we've tested almost 30,000 people now.

0:31:17.403,0:31:21.735
Frankly, I think the US[br]may well be moving beyond that point,

0:31:21.759,0:31:24.688
given how much evidence[br]of extensive transmission that has,

0:31:24.712,0:31:27.808
and I think without clear ideas[br]of how much infection there is

0:31:27.832,0:31:30.103
and that level of testing,

0:31:30.127,0:31:33.516
it's quite hard to actually see[br]what the picture currently is in the US.

0:31:35.308,0:31:39.028
CA: I mean, I definitely don't want to get[br]too political about this,

0:31:39.052,0:31:40.894
but I mean, does this strike you as --

0:31:40.918,0:31:43.441
you just said that the UK[br]has tested 30,000 people --

0:31:43.465,0:31:45.964
the US is five or six times bigger

0:31:45.988,0:31:49.181
and I think the total number[br]of tests here is five or six thousand,

0:31:49.205,0:31:50.458
or it was a few days ago.

0:31:50.482,0:31:52.673
Does that strike you as bizarre?

0:31:52.697,0:31:56.644
I don't understand, honestly,[br]how that happened in an educated country

0:31:56.668,0:31:59.460
that has so much knowledge[br]about infectious diseases.

0:32:00.173,0:32:01.340
AK: It does,

0:32:01.364,0:32:04.542
and I think there's obviously[br]a number of factors playing in there,

0:32:04.566,0:32:05.891
logistics and so on,

0:32:05.915,0:32:08.109
but there has been that period of warning

0:32:08.133,0:32:10.212
that this is a threat[br]and this is coming in.

0:32:10.236,0:32:13.561
And I think countries need to make sure[br]that they've got the capacity

0:32:13.585,0:32:16.697
to really do as much detection as they can[br]in those early stages,

0:32:16.721,0:32:18.871
because that's where[br]you're going to catch it

0:32:18.895,0:32:22.450
and that's where you're going to have[br]a better chance of containing it.

0:32:22.474,0:32:24.752
CA: OK, so if you fail to contain,

0:32:24.776,0:32:28.090
then you have to move[br]to some kind of mitigation strategy.

0:32:28.114,0:32:31.495
So what comes into play there?

0:32:31.519,0:32:34.574
And I think I almost want[br]to bring that back

0:32:34.598,0:32:38.419
to two of your DOTS factors,

0:32:38.443,0:32:41.443
opportunity and transmission probability,

0:32:41.467,0:32:43.720
because it seems like[br]the virus is what it is,

0:32:43.744,0:32:46.506
the actual duration when someone[br]is potentially infectious,

0:32:46.530,0:32:47.764
we can't do much about.

0:32:47.788,0:32:49.879
The susceptibility side,

0:32:49.903,0:32:52.505
we can't do much about[br]until there's a vaccine.

0:32:52.529,0:32:54.760
We could maybe talk about that in a bit.

0:32:54.784,0:32:58.451
But the middle two of opportunity[br]and transmission probability,

0:32:58.475,0:32:59.871
we can do something about.

0:32:59.895,0:33:03.418
Do you want to maybe talk[br]about those in turn,

0:33:03.442,0:33:04.665
of what that looks like,

0:33:04.689,0:33:09.273
how would you build a mitigation strategy?

0:33:09.297,0:33:11.623
I mean, first of all,[br]thinking about opportunity,

0:33:11.647,0:33:13.781
how do you reduce[br]the number of opportunities

0:33:13.805,0:33:15.072
to pass on the bug?

0:33:15.885,0:33:17.643
AK: And so I think in that respect,

0:33:17.667,0:33:21.328
it would be about massive changes[br]in our social interactions.

0:33:21.352,0:33:23.877
And if you think in terms[br]of the reproduction number

0:33:23.901,0:33:26.068
of being about two or three,

0:33:26.092,0:33:27.473
to get that number below one,

0:33:27.497,0:33:30.763
you've really got to cut[br]some aspect of that transmission

0:33:30.787,0:33:32.192
in half or in two thirds,

0:33:32.216,0:33:33.620
to get that below one.

0:33:34.017,0:33:35.747
And so that would require,

0:33:35.771,0:33:38.128
of the opportunities[br]that could spread the virus

0:33:38.152,0:33:40.101
are these kind of close contacts.

0:33:40.125,0:33:42.379
Everybody in the population, on average,

0:33:42.403,0:33:45.323
will be needing to reduce[br]those interactions

0:33:45.347,0:33:47.966
potentially by two thirds[br]to bring it under control.

0:33:47.990,0:33:50.704
That might be through working from home,

0:33:50.728,0:33:52.735
from changing lifestyle

0:33:52.759,0:33:56.102
and kind of where you go[br]in crowded places and dinners.

0:33:56.653,0:33:59.490
And of course, these measures,[br]things like school closures,

0:33:59.514,0:34:01.577
and other things[br]that just attempt to reduce

0:34:01.601,0:34:03.386
the social mixing of a population.

0:34:03.712,0:34:06.513
CA: Well, actually, talk to me more[br]about school closures,

0:34:06.537,0:34:08.934
because that, if I remember,

0:34:08.958,0:34:15.307
often in past pandemics has been cited[br]as an absolutely key measure,

0:34:15.331,0:34:18.597
that schools represent this sort of[br]coming together of people,

0:34:18.621,0:34:21.454
children are often --

0:34:21.478,0:34:23.451
certainly when it comes to flu and colds,

0:34:23.475,0:34:25.545
they're carriers.

0:34:25.569,0:34:27.188
But on this case,

0:34:27.212,0:34:31.093
children don't seem to be getting sick[br]from this particular virus,

0:34:31.117,0:34:33.958
or at least very few of them are.

0:34:34.371,0:34:39.301
Do we know whether they[br]can still be infectious?

0:34:39.325,0:34:41.896
They can be the unintended carriers of it.

0:34:41.920,0:34:45.342
Or actually, is there evidence[br]that school closures

0:34:45.366,0:34:48.659
may not be as important[br]in this instance as it is in others?

0:34:49.093,0:34:51.156
AK: So that point[br]on what role children play

0:34:51.180,0:34:52.339
is a crucial one,

0:34:52.363,0:34:54.680
and there's still not[br]a good evidence base there.

0:34:54.704,0:34:56.942
From following up of contacts of cases,

0:34:56.966,0:34:59.617
there's now evidence[br]that children are getting infected,

0:34:59.641,0:35:01.974
so when you're testing,[br]they are getting exposed,

0:35:01.998,0:35:05.629
it's not that somehow they're just[br]not getting the infection at all,

0:35:05.653,0:35:09.117
but as you said, they're not showing[br]symptoms in the same way,

0:35:09.141,0:35:11.445
and particularly for flu,

0:35:11.469,0:35:14.201
when we see the implications[br]of school closures,

0:35:14.225,0:35:17.120
even in the UK in 2009 during swine flu,

0:35:17.144,0:35:19.954
there was a dip in the outbreak[br]during the school holidays,

0:35:19.978,0:35:21.883
you could see it on the epidemic curve,

0:35:21.907,0:35:25.319
it kind of comes back down in the summer[br]and goes back up in the autumn.

0:35:25.343,0:35:28.403
But of course, in 2009,[br]there was some immunity in older groups.

0:35:28.427,0:35:31.537
That kind of shifted more the transmission[br]onto the younger ones.

0:35:31.561,0:35:34.832
So I think it's really something[br]we're trying to work to understand.

0:35:34.856,0:35:37.733
Obviously, it will reduce interactions,[br]with school closures,

0:35:37.757,0:35:39.749
but then there's knock-on social effects,

0:35:39.773,0:35:41.876
there's potential[br]knock-on changes in mixing,

0:35:41.900,0:35:45.743
maybe grandparents and their role[br]in terms of alternative carers

0:35:45.767,0:35:47.109
if parents have to work.

0:35:47.133,0:35:50.720
So I think there's a lot of pieces[br]that need to be considered.

0:35:52.101,0:35:56.865
CA: I mean, based on all of the different[br]pieces of evidence you've seen,

0:35:56.889,0:35:58.166
if it were down to you,

0:35:58.190,0:36:01.579
would you be recommending[br]that most countries at this point

0:36:01.603,0:36:05.878
do look hard at extensive school closures[br]as a precautionary measure,

0:36:05.902,0:36:09.458
that it's just worth it to do that

0:36:09.482,0:36:14.504
as a sort of painful two,[br]three, four, five-month strategy?

0:36:14.528,0:36:16.163
What would you recommend?

0:36:16.187,0:36:17.505
AK: I think the key thing,

0:36:17.529,0:36:20.684
given the age distribution of risk[br]and the severity in older groups

0:36:20.708,0:36:25.489
is reduce interactions that bring[br]the infection into those groups.

0:36:25.513,0:36:28.975
And then amongst everyone else,[br]reduce interactions as much as possible.

0:36:28.999,0:36:30.602
I think the key thing is

0:36:30.626,0:36:34.473
we've got so much of the disease burden[br]in the kind of 60-plus group

0:36:34.497,0:36:37.990
that it's not just about[br]everyone trying to avoid

0:36:38.014,0:36:39.180
everyone's interactions,

0:36:39.204,0:36:40.680
but it's the kind of behaviors

0:36:40.704,0:36:43.147
that would drive infections[br]into those groups.

0:36:44.448,0:36:47.264
CA: Does that mean[br]that people should think twice

0:36:47.288,0:36:50.464
before, I don't know, visiting a loved one

0:36:50.488,0:36:55.454
in an old people's home[br]or in a residential facility?

0:36:55.478,0:36:59.281
Like that, we should just pay[br]super special attention to that,

0:36:59.305,0:37:02.155
should all these facilities[br]be taking great care

0:37:02.179,0:37:03.630
about who they admit,

0:37:03.654,0:37:06.910
taking temperature and checking[br]for symptoms or something like that?

0:37:06.934,0:37:09.832
AK: I think those measures[br]definitely need to be considered.

0:37:09.856,0:37:11.387
In the UK, we're getting plans

0:37:11.411,0:37:13.863
for potentially what's known[br]as a cocooning strategy

0:37:13.887,0:37:15.172
for these older groups

0:37:15.196,0:37:17.810
that we can really try[br]and seal off interactions

0:37:17.834,0:37:19.212
as much as possible

0:37:19.236,0:37:21.847
from people who might[br]be bringing infection in.

0:37:22.202,0:37:24.879
And ultimately, because as you said,

0:37:24.903,0:37:27.371
we can't target these other[br]aspects of transmission,

0:37:27.395,0:37:30.331
it is just reducing the risk[br]of exposure in these groups,

0:37:30.355,0:37:34.099
and so I think anything[br]at the individual level you can do

0:37:34.123,0:37:36.321
to get people reducing their risk,

0:37:36.345,0:37:39.385
if either they're elderly[br]or in other risk groups,

0:37:39.409,0:37:40.561
I think is crucial.

0:37:40.585,0:37:43.020
And I think more at the general level

0:37:43.044,0:37:47.077
those kind of more large-scale measures[br]can help reduce interactions overall,

0:37:47.101,0:37:49.696
but I think if those[br]reductions are happening,

0:37:49.720,0:37:50.955
and not reducing the risk

0:37:50.979,0:37:53.244
for people who are going[br]to get severe disease,

0:37:53.268,0:37:56.912
then you're still going to get[br]this really remarkably severe burden.

0:37:58.107,0:38:03.361
CA: I mean, do people have to almost[br]apply this double lens

0:38:03.385,0:38:04.901
as they think about this stuff?

0:38:04.925,0:38:07.292
There's the risk to you[br]as you go about your life,

0:38:07.316,0:38:09.348
of you catching this bug.

0:38:09.657,0:38:12.839
But there's also the risk[br]of you being, unintentionally, a carrier

0:38:12.863,0:38:15.897
to someone who would suffer[br]much more than you might.

0:38:15.921,0:38:19.569
And that both those things[br]have to be top of mind right now.

0:38:19.593,0:38:22.204
AK: Yeah, and it's not just[br]whose hand you shake,

0:38:22.228,0:38:24.434
it's whose hand that person[br]goes on to shake.

0:38:24.458,0:38:27.323
And I think we need to think[br]about these second-degree steps,

0:38:27.347,0:38:29.529
that you might think you have low risk,

0:38:29.553,0:38:31.331
and you're in a younger group,

0:38:31.355,0:38:34.450
but you're often going to be[br]a very short step away

0:38:34.474,0:38:37.077
from someone who is going to get hit[br]very hard by this.

0:38:37.101,0:38:39.744
And I think we really need[br]to be socially minded

0:38:39.768,0:38:43.082
and think this could be quite dramatic[br]in terms of change of behavior,

0:38:43.106,0:38:44.765
but it needs to be

0:38:44.789,0:38:47.238
to reduce the impact[br]that we're potentially facing.

0:38:48.556,0:38:51.135
CA: So the opportunity[br]number, we bring down

0:38:51.159,0:38:53.805
by just reducing the number[br]of physical contacts we have

0:38:53.829,0:38:55.532
with other people.

0:38:55.556,0:38:58.810
And I guess the transmission[br]probability number,

0:38:58.834,0:39:01.334
how do we bring that down?

0:39:01.358,0:39:03.188
That impacts how we interact.

0:39:03.212,0:39:04.558
You mentioned hand-shaking,

0:39:04.582,0:39:06.876
I'm guessing you're going to say[br]no handshaking.

0:39:06.900,0:39:08.910
AK: Yeah, so changes like that.

0:39:08.934,0:39:10.341
I mean, another one, I think,

0:39:10.365,0:39:12.026
handwashing operates in a way

0:39:12.050,0:39:16.363
that we might be still be doing[br]activities that we've previously done,

0:39:16.387,0:39:20.601
but handwashing reduces the chance[br]that from one interaction to another,

0:39:20.625,0:39:22.172
we might be spreading infection,

0:39:22.196,0:39:23.633
so it's all of these measures

0:39:23.657,0:39:26.338
that mean that even[br]if we're having these exposures,

0:39:26.362,0:39:29.655
we're taking additional steps[br]to avoid any transmission happening.

0:39:30.346,0:39:32.822
CA: I still think most people[br]don't fully understand

0:39:32.846,0:39:34.886
or don't have a clear model of the pathway

0:39:34.910,0:39:38.886
by which thins thing spreads.

0:39:38.910,0:39:40.879
So you think definitely people understand

0:39:40.903,0:39:42.664
that you don't breathe in

0:39:42.688,0:39:46.941
the water droplets of someone[br]who has just coughed or sneezed.

0:39:46.965,0:39:49.117
So how does it spread?

0:39:49.141,0:39:51.403
It gets onto surfaces. How?

0:39:51.427,0:39:54.586
Do people just breathe out[br]and it goes on from people who are sick,

0:39:54.610,0:39:56.784
they touch their mouth[br]or something like that,

0:39:56.808,0:39:59.110
and then touch a surface[br]and it gets on that way?

0:39:59.134,0:40:01.022
How does it actually get onto surfaces?

0:40:01.046,0:40:03.864
AK: I think a lot of it would be[br]that you cough in your hand

0:40:03.888,0:40:05.497
and it ends up on a surface.

0:40:05.871,0:40:09.565
But I think the challenge, obviously,[br]is untangling these questions

0:40:09.589,0:40:10.947
of how transmission happens.

0:40:10.971,0:40:12.789
You have transmission in a household,

0:40:12.813,0:40:15.403
and is it that someone coughed[br]and it got on a surface,

0:40:15.427,0:40:17.332
is it direct contact, is it a handshake,

0:40:17.356,0:40:18.779
and even for things like flu,

0:40:18.803,0:40:22.014
that's something that we work[br]quite hard to try and unpick,

0:40:22.038,0:40:25.250
how does social behavior correspond[br]to infection risk.

0:40:25.274,0:40:28.537
Because it's clearly important,[br]but pinning it down is really tough.

0:40:29.228,0:40:32.442
CA: It's almost like embracing the fact

0:40:32.466,0:40:34.990
that for a lot of these things,[br]we actually don't know

0:40:35.014,0:40:38.855
and that we're all[br]in this game of probabilities.

0:40:39.156,0:40:42.117
Which, in a way, is why I think[br]the math is so important here.

0:40:42.141,0:40:48.148
That you have to think of this[br]as these multiple numbers

0:40:48.172,0:40:49.887
working together on each other,

0:40:49.911,0:40:51.545
they all have their part to play.

0:40:51.569,0:40:56.562
And any of them that you can[br]take down by a percentage

0:40:56.586,0:40:58.062
is likely contributing,

0:40:58.086,0:41:00.807
not just to you, but to everyone.

0:41:00.831,0:41:04.609
And people don't actually know[br]in detail how the numbers go together,

0:41:04.633,0:41:06.728
but they know that they[br]probably all matter.

0:41:06.752,0:41:11.664
We almost need people to, somehow,[br]you know, embrace that uncertainty

0:41:11.688,0:41:16.799
and then try to get some satisfaction[br]by acting on every single part of it.

0:41:16.823,0:41:18.085
AK: I think it's this idea

0:41:18.109,0:41:21.797
that if on average,[br]you're infecting, say, three people,

0:41:21.821,0:41:24.695
what's driving that and how can you[br]chip away at that value?

0:41:24.719,0:41:26.123
If you're washing your hands,

0:41:26.147,0:41:29.403
how much might that chip away[br]in terms of the handshakes,

0:41:29.427,0:41:32.006
you know, you may have had virus[br]and you no longer do,

0:41:32.030,0:41:35.744
or if you are changing[br]your social behavior in a certain way,

0:41:35.768,0:41:37.903
is that taking away[br]a couple of interactions,

0:41:37.927,0:41:39.131
is that taking away half,

0:41:39.155,0:41:42.584
how can you really chip into that number[br]as much as you possibly can?

0:41:43.698,0:41:47.203
CA: Is there anything else to say[br]about how we could reduce

0:41:47.227,0:41:51.693
that transmission probability[br]in our interactions?

0:41:51.717,0:41:54.525
Like, what is the physical distance

0:41:54.549,0:42:00.140
that it's wise to stay away[br]from other people if we can?

0:42:00.665,0:42:02.680
AK: I think it's hard to pin down exactly,

0:42:02.704,0:42:06.216
but I think one thing to bear in mind[br]is that there's not so much evidence

0:42:06.240,0:42:08.871
that this is a kind of aerosol[br]and it goes really far --

0:42:08.895,0:42:10.428
it's reasonably short distances.

0:42:10.452,0:42:11.759
I don't think it's the case

0:42:11.783,0:42:14.846
that you're sitting a few meters[br]away from someone

0:42:14.870,0:42:17.028
and the virus is somehow[br]going to get across.

0:42:17.735,0:42:19.085
It's in closer interactions,

0:42:19.109,0:42:22.096
and it's why we're seeing[br]so many transmission events

0:42:22.120,0:42:25.906
occur in things like meals[br]and really tight-knit groups.

0:42:25.930,0:42:27.271
Because if you imagine

0:42:27.295,0:42:30.310
that's where you can get[br]a virus out and onto surfaces

0:42:30.334,0:42:31.882
and onto hands and onto faces,

0:42:31.906,0:42:36.031
and it's really situations like that[br]we've got to think more about.

0:42:37.818,0:42:39.024
CA: So in a way,

0:42:39.048,0:42:41.993
some of the fears that people have[br]may actually be overstated,

0:42:42.017,0:42:44.720
like, if you're in the middle[br]of an airplane

0:42:44.744,0:42:47.268
and someone at the front sneezes,

0:42:47.292,0:42:49.077
I mean, that's annoying,

0:42:49.101,0:42:53.458
but it's actually not the thing[br]you should be most freaked out about.

0:42:53.482,0:42:57.220
There are much smarter ways[br]to pay attention to your well-being.

0:42:57.244,0:43:00.697
AK: Yeah, if it was measles[br]and the plane is susceptible people,

0:43:00.721,0:43:02.847
you would see a lot[br]of infections after that.

0:43:02.871,0:43:05.395
I think it is, bear in mind,[br]that this is, on average,

0:43:05.419,0:43:07.184
people infecting two or three others,

0:43:07.208,0:43:11.029
so it's not the case of your[br]maybe 50 interactions over a week,

0:43:11.053,0:43:12.982
all of those people are at risk.

0:43:13.006,0:43:14.633
But it's going to be some of them,

0:43:14.657,0:43:16.625
particularly those close contacts,

0:43:16.649,0:43:19.331
that are going to be[br]where transmission's occurring.

0:43:19.355,0:43:21.498
CA: So talk about,

0:43:21.522,0:43:25.632
from a sort of national[br]strategy point of view.

0:43:25.656,0:43:29.863
There's a lot of talk about the need[br]to "flatten the curve."

0:43:29.887,0:43:31.212
What does that mean?

0:43:31.236,0:43:35.946
AK: I think it refers to this idea[br]that for your health systems,

0:43:35.970,0:43:38.641
you don't want all your cases[br]to appear at the same time.

0:43:38.665,0:43:40.712
So if we sat back and did nothing,

0:43:40.736,0:43:42.260
and just let the epidemic grow,

0:43:42.284,0:43:44.625
and you had this growth rate[br]that, at the moment,

0:43:44.649,0:43:46.380
in some places is looking like maybe

0:43:46.404,0:43:48.550
three to four days,[br]you're getting doubling.

0:43:48.574,0:43:51.098
So every three or four days,[br]the epidemic is doubling.

0:43:51.122,0:43:52.804
It will skyrocket and you'll end up

0:43:52.828,0:43:55.982
with a whole bunch[br]of really severely ill people

0:43:56.006,0:43:58.042
needing hospital care[br]all at the same time,

0:43:58.066,0:44:00.029
and you just won't have capacity for it.

0:44:00.053,0:44:03.204
So the idea of flattening the curve[br]is if we can slow transmission,

0:44:03.228,0:44:05.276
if we can get that[br]reproduction number down,

0:44:05.300,0:44:07.164
then there may still be an outbreak,

0:44:07.188,0:44:08.545
but it will be much flatter,

0:44:08.569,0:44:09.736
it will be longer,

0:44:09.760,0:44:12.054
and there will be fewer[br]severe cases showing up,

0:44:12.078,0:44:14.759
which means that they can get[br]the health care they need.

0:44:16.220,0:44:23.220
CA: Does it imply that there will be[br]fewer cases overall, or ---

0:44:24.323,0:44:26.776
When you look at the actual images[br]of people showing

0:44:26.800,0:44:28.577
what flattening the curve looks like,

0:44:28.601,0:44:31.879
it almost looks as if you've got[br]the same area still under the graph,

0:44:31.903,0:44:34.855
i.e. that the same number of people,[br]ultimately, are infected,

0:44:34.879,0:44:37.672
but over a longer period.

0:44:37.696,0:44:40.264
Is that typically what happens,

0:44:40.288,0:44:45.247
and even if you adopt[br]all these strategies of social distancing

0:44:45.271,0:44:49.006
and washing hands and etc.

0:44:49.030,0:44:52.149
that the best you can hope for[br]is that you slow the thing down,

0:44:52.173,0:44:54.910
you actually will get[br]as many people infected in the end?

0:44:54.934,0:44:57.792
AK: Not necessarily --[br]it depends on the measures that go in.

0:44:57.816,0:45:00.212
There are some measures like,[br]shutting down travel,

0:45:00.236,0:45:03.418
which typically delay the spread[br]rather than reduce it.

0:45:03.442,0:45:05.719
So you're still going to get[br]the same outbreaks,

0:45:05.743,0:45:07.845
but you're stretching out the outbreaks.

0:45:08.338,0:45:09.783
But there are other measures.

0:45:09.807,0:45:11.675
If we talk about reducing interactions,

0:45:11.699,0:45:13.568
if your reproduction number's lower,

0:45:13.592,0:45:15.849
you would expect fewer cases overall.

0:45:16.384,0:45:18.183
And eventually, in your population,

0:45:18.207,0:45:20.111
you will get some buildup of immunity,

0:45:20.135,0:45:22.964
which would help you out[br]if you think about the components,

0:45:22.988,0:45:24.243
reducing susceptibility,

0:45:24.267,0:45:26.553
alongside what's going on elsewhere.

0:45:26.577,0:45:29.164
So the hope is that the two things[br]will work together.

0:45:29.848,0:45:33.727
CA: So help me understand[br]what the endgame is here.

0:45:34.904,0:45:37.117
So, take China, for example.

0:45:38.967,0:45:42.570
Whatever you make[br]of the early suppression of data

0:45:42.594,0:45:43.759
and so forth,

0:45:43.783,0:45:47.713
that seems pretty troubling there.

0:45:47.737,0:45:53.426
The intensity of the response[br]come January time or whatever,

0:45:53.450,0:45:57.449
with the shutdown[br]of this huge area of the country,

0:45:57.473,0:45:59.356
seems to have actually been effective.

0:45:59.380,0:46:04.879
The number of cases there are falling[br]at a shockingly high rate in some ways.

0:46:04.903,0:46:06.791
Falling to almost nothing.

0:46:06.815,0:46:09.966
And I can't understand that.

0:46:09.990,0:46:14.115
You are talking about a country[br]of, whatever, 1.4 billion people.

0:46:14.139,0:46:16.242
There have been a huge[br]number of cases there,

0:46:16.266,0:46:19.718
but it was a tiny fraction[br]of the population have actually got sick.

0:46:19.742,0:46:24.613
And yet, they've got the number way down.

0:46:24.637,0:46:29.458
It's not like every other person in China[br]has somehow developed immunity.

0:46:29.482,0:46:33.396
Is it that they have been[br]absolutely disciplined

0:46:33.420,0:46:37.507
about shutting down travel[br]from the infected regions

0:46:37.531,0:46:43.335
and somehow really dialed up,[br]massively dialed up

0:46:43.359,0:46:46.363
testing at any sign of any problem,

0:46:46.387,0:46:50.065
so that literally, they are back[br]in containment mode

0:46:50.089,0:46:52.272
in most parts of China?

0:46:52.296,0:46:55.623
I can't get my head around it,[br]help me understand it.

0:46:55.647,0:46:58.197
AK: So we estimated,[br]in the last two weeks of January,

0:46:58.221,0:46:59.585
when these measures went in,

0:46:59.609,0:47:02.014
the reproduction number[br]went from about 2.4 to 1.1.

0:47:02.038,0:47:04.403
So about 60 percent decline[br]in transmission

0:47:04.427,0:47:06.418
in the space of a week or two.

0:47:06.442,0:47:08.919
Which is remarkable and really,

0:47:08.943,0:47:13.490
a lot of it is likely to be driven[br]by just fundamental change

0:47:13.514,0:47:14.801
in social behavior,

0:47:14.825,0:47:16.262
huge social distancing,

0:47:16.286,0:47:19.055
really intensive follow up,[br]intensive testing.

0:47:20.166,0:47:21.540
And it got to the point

0:47:21.564,0:47:23.825
where it took enough[br]off the reproduction number

0:47:23.849,0:47:25.053
to cause the decline,

0:47:25.077,0:47:28.166
and of course,[br]we're seeing, in many areas,

0:47:28.190,0:47:30.863
a transition back to more[br]of this kind of containment,

0:47:30.887,0:47:33.257
because there's few cases,[br]it's more manageable.

0:47:34.320,0:47:36.493
But we're also seeing them[br]face a challenge,

0:47:36.517,0:47:40.312
because a lot of these cities[br]have basically been locked down

0:47:40.336,0:47:41.507
for six weeks

0:47:41.531,0:47:43.974
and there's a limit to how long[br]you can do that for.

0:47:43.998,0:47:47.164
And so some of these measures[br]are gradually starting to be lifted,

0:47:47.188,0:47:48.806
which of course creates the risk

0:47:48.830,0:47:51.570
that cases that are appearing[br]from other countries

0:47:51.594,0:47:54.708
may subsequently go in[br]and reintroduce transmission.

0:47:57.610,0:48:00.847
CA: But given how infectious the bug is,

0:48:00.871,0:48:05.371
and how many theoretical pathways[br]and connection points there are

0:48:05.395,0:48:09.053
between people in Wuhan, even in shutdown,

0:48:09.077,0:48:10.371
or relatively shut down,

0:48:10.395,0:48:13.323
or the other places[br]where there's been some infection

0:48:13.347,0:48:14.736
and the rest of the country,

0:48:14.760,0:48:21.009
does it surprise you how quickly[br]that curve has gone down to nearly zero?

0:48:21.865,0:48:23.016
AK: Yes.

0:48:23.040,0:48:26.863
Early on when we saw[br]that flattening off in cases

0:48:26.887,0:48:28.696
in those first few days,

0:48:28.720,0:48:32.268
we did wonder whether it was just[br]they hit a limit in testing capacity

0:48:32.292,0:48:34.053
and they were reporting 1,000 a day,

0:48:34.077,0:48:36.109
because that's all the kits they had.

0:48:36.133,0:48:38.371
But it continued, thankfully,

0:48:38.395,0:48:41.490
and it shows that it is possible[br]to turn this over

0:48:41.514,0:48:43.276
with that level of intervention.

0:48:43.300,0:48:46.529
I think the key thing now[br]is seeing how it works in other settings.

0:48:46.553,0:48:51.221
Italy now are putting in[br]really dramatic interventions.

0:48:51.245,0:48:53.293
But of course,[br]because of this delay effect,

0:48:53.317,0:48:54.896
if you put them in today,

0:48:54.920,0:48:57.094
you won't necessarily see[br]the effects on cases

0:48:57.118,0:48:58.277
for another week or two.

0:48:58.301,0:49:00.418
So I think working out[br]what impact that's had

0:49:00.442,0:49:02.649
is going to be key for helping[br]other countries

0:49:02.673,0:49:04.140
work on how to contain this.

0:49:04.711,0:49:06.045
CA: To have a picture, Adam,

0:49:06.069,0:49:09.990
of how this is likely to play out[br]over the next month or two,

0:49:10.014,0:49:13.585
give us a couple of scenarios[br]that are in your head.

0:49:14.560,0:49:17.315
AK: I think the optimistic scenario

0:49:17.339,0:49:20.465
is that we're going to learn a lot[br]from places like Italy

0:49:20.489,0:49:22.532
that have unfortunately[br]been hit very hard,

0:49:22.556,0:49:25.176
and that countries are going to take[br]this very seriously

0:49:25.200,0:49:27.688
and that we're not going to get[br]this continued growth

0:49:27.712,0:49:29.337
that's going to overwhelm totally,

0:49:29.361,0:49:33.138
that we're going to be able[br]to sufficiently slow it down,

0:49:33.162,0:49:35.409
that we are going to get[br]large numbers of cases,

0:49:35.433,0:49:38.210
we're probably going to get[br]a lot of severe cases,

0:49:38.234,0:49:40.035
but that will be more manageable,

0:49:40.059,0:49:41.951
that's the kind of optimistic scenario.

0:49:41.975,0:49:43.260
I think if we have a point

0:49:43.284,0:49:45.674
where countries either[br]don't take this seriously

0:49:45.698,0:49:49.721
or populations don't respond well[br]to control measures

0:49:49.745,0:49:51.389
or it's not detected,

0:49:51.413,0:49:52.696
we could get situations --

0:49:52.720,0:49:55.387
I think Iran is probably[br]the closest one at the moment --

0:49:55.411,0:49:58.495
where there's been extensive[br]widespread transmission,

0:49:58.519,0:50:01.480
and by the time it's being responded to,

0:50:01.504,0:50:03.536
those infections are already in the system

0:50:03.560,0:50:06.313
and they are going to turn up[br]as cases and severe illness.

0:50:06.337,0:50:08.202
So I'm hoping we're not at that point,

0:50:08.226,0:50:10.209
but we've certainly got, at the moment,

0:50:10.233,0:50:13.547
potentially about 10 countries[br]on that trajectory

0:50:13.571,0:50:15.799
to have the same outlook as Italy.

0:50:15.823,0:50:18.918
So it's really crucial what happens[br]in the next couple of weeks.

0:50:19.811,0:50:22.302
CA: Is there a real chance[br]that quite a few countries

0:50:22.326,0:50:25.087
end up having, this year,

0:50:25.111,0:50:30.774
substantially more deaths from this virus[br]than from seasonal flu?

0:50:31.942,0:50:34.609
AK: I think for some countries[br]that is likely, yeah.

0:50:34.633,0:50:36.817
I think if control is not possible,

0:50:36.841,0:50:38.759
and we've seen it happen in China,

0:50:38.783,0:50:43.116
but that was really just an unprecedented[br]level of intervention.

0:50:43.140,0:50:45.632
It was really just changing[br]the social fabric.

0:50:45.656,0:50:51.767
I think people, many of us,[br]don't really appreciate, at a glance,

0:50:51.791,0:50:53.022
just what that means,

0:50:53.046,0:50:55.780
to reduce your interactions[br]to that extent.

0:50:55.804,0:50:58.968
I think many countries just simply[br]won't be able to manage that.

0:51:00.769,0:51:03.410
CA: It's almost a challenge[br]to democracies, isn't it --

0:51:03.434,0:51:07.936
"OK, show us what you can do[br]without that kind of draconian control."

0:51:07.960,0:51:10.065
If you don't like the thought of that,

0:51:10.089,0:51:12.862
come on, citizens, step up,[br]show us what you're capable of,

0:51:12.886,0:51:14.859
show that you can be wise about this

0:51:14.883,0:51:17.450
and smart and self-disciplined,

0:51:17.474,0:51:19.863
and get ahead of the damn bug.

0:51:19.887,0:51:21.291
AK: Yeah.

0:51:21.315,0:51:25.117
CA: I mean, I'm not personally[br]superoptimistic about that,

0:51:25.141,0:51:29.805
because there's such conflicting messaging[br]coming out in so many different places,

0:51:29.829,0:51:35.753
and people don't like[br]to short-term sacrifice.

0:51:35.777,0:51:38.177
I mean, is there almost a case that --

0:51:38.673,0:51:40.580
I mean, what's your view

0:51:40.604,0:51:43.568
on whether the media has played[br]a helpful role here

0:51:43.592,0:51:44.763
or an unhelpful role?

0:51:44.787,0:51:46.575
Is it actually, in some ways, helpful

0:51:46.599,0:51:50.529
to, if anything, overstate[br]the concern, the fear,

0:51:50.553,0:51:53.212
and actually make people[br]panic a little bit?

0:51:53.236,0:51:55.617
AK: I think it's a really[br]tough balance to strike,

0:51:55.641,0:51:58.133
because of course, early on,[br]if you don't have cases,

0:51:58.157,0:52:01.464
if you don't have any evidence[br]of potential pressure,

0:52:01.488,0:52:05.029
it's very hard to get that message[br]and convince people to take it seriously

0:52:05.053,0:52:06.260
if you're over hyping it.

0:52:06.284,0:52:08.728
But equally, if you're waiting too long,

0:52:08.752,0:52:11.783
and saying it's not a concern yet,[br]we're OK for the moment,

0:52:11.807,0:52:14.207
a lot of people think it's just flu.

0:52:14.553,0:52:17.711
By the time it hits hard, as I've said,

0:52:17.735,0:52:20.855
you're going to have weeks[br]of an overburdened health system,

0:52:20.879,0:52:23.671
because even if you take interventions,

0:52:23.695,0:52:26.531
it's too late to control the infections[br]that have happened.

0:52:26.555,0:52:28.087
So I think it's a fine line

0:52:28.111,0:52:30.516
and my hope is there is[br]this ramp-up in messaging,

0:52:30.540,0:52:33.015
now people have these[br]tangible examples like Italy,

0:52:33.039,0:52:36.555
where they can see what's going to happen[br]if they don't take it seriously.

0:52:37.038,0:52:39.799
But certainly, of all[br]the diseases I've seen,

0:52:39.823,0:52:42.465
I think many of my colleagues[br]who are much older than me

0:52:42.489,0:52:44.276
and have memories of other outbreaks,

0:52:44.300,0:52:47.752
it's the scariest thing we've seen[br]in terms of the impact it could have,

0:52:47.776,0:52:49.656
and I think we need to respond to that.

0:52:49.680,0:52:52.311
CA: It's the scariest disease you've seen.

0:52:53.042,0:52:54.192
Wow.

0:52:54.216,0:52:58.732
I've got some questions for you[br]from my friends on Twitter.

0:52:58.756,0:53:04.973
Everyone is obviously[br]superexercised about this topic.

0:53:04.997,0:53:06.536
Hypothetically,

0:53:06.560,0:53:09.052
if everyone stayed home for three weeks,

0:53:09.076,0:53:11.791
would that effectively wipe this out?

0:53:11.815,0:53:14.661
Is there a way to socially[br]distance ourselves out of this?

0:53:15.352,0:53:20.296
AK: Yeah, I think in certain countries[br]with reasonably small household sizes,

0:53:20.320,0:53:22.863
I think average in the UK, US[br]is about two and a half,

0:53:22.887,0:53:26.331
so even if you had a round of infection[br]within the household,

0:53:26.355,0:53:28.019
that would probably stamp it out.

0:53:28.043,0:53:29.270
As a secondary benefit,

0:53:29.294,0:53:31.643
you may well stamp out[br]a few other infections too.

0:53:31.667,0:53:33.348
Measles only circulates in humans,

0:53:33.372,0:53:35.149
so you may have some knock-on effect,

0:53:35.173,0:53:37.529
if, of course, that were[br]ever to be possible.

0:53:37.553,0:53:41.743
CA: I mean, obviously that would be[br]a huge dent to the economy,

0:53:41.767,0:53:46.043
and this is in a way almost, like,[br]one of the underlying challenges here

0:53:46.067,0:53:49.561
is that you can't optimize public policy

0:53:49.585,0:53:54.946
for both economic health[br]and fighting a virus.

0:53:54.970,0:53:57.628
Like, those two things are,[br]to some extent, in conflict,

0:53:57.652,0:54:01.565
or at least, short-term[br]economic health and fighting a virus.

0:54:01.589,0:54:03.565
Those two things are in conflict, right?

0:54:03.589,0:54:06.503
And societies need to pick one.

0:54:06.988,0:54:10.686
AK: It is tough to convince[br]people of that balance,

0:54:10.710,0:54:12.910
the thing we always say[br]of pandemic planning

0:54:12.934,0:54:15.193
is it's cheap to put[br]this stuff in place now --

0:54:15.217,0:54:17.218
otherwise, you've got to pay for it later.

0:54:18.207,0:54:20.287
But unfortunately,[br]as we've seen with this,

0:54:20.311,0:54:22.849
that a lot of early money[br]for response wasn't there.

0:54:23.341,0:54:26.746
And it's only when it has an impact[br]and when it's going to get expensive

0:54:26.770,0:54:30.725
that people are happy to take[br]that cost on board, it seems.

0:54:31.612,0:54:33.652
CA: OK, some more Twitter questions.

0:54:33.676,0:54:36.438
Will the rising temperature[br]in coming weeks and months

0:54:36.462,0:54:39.215
slow down the COVID-19 spread?

0:54:39.784,0:54:42.109
AK: I haven't seen any convincing evidence

0:54:42.133,0:54:44.490
that there's that strong pattern[br]with temperature,

0:54:44.514,0:54:48.768
and we've seen it for other infections[br]that there is this seasonal pattern,

0:54:48.792,0:54:51.403
but I think the fact[br]we're getting widespread outbreaks

0:54:51.427,0:54:53.895
makes it hard to identify, and of course,

0:54:53.919,0:54:55.355
there's other things going on.

0:54:55.379,0:54:58.506
So even if one country doesn't have[br]as big an outbreak as another,

0:54:58.530,0:55:00.887
that's going to be influenced[br]by control measures,

0:55:00.911,0:55:04.100
by social behavior, by opportunities[br]and these things as well.

0:55:04.124,0:55:07.045
So it would be really reassuring[br]if this was the case,

0:55:07.069,0:55:09.323
but I don't think[br]we can say that just yet.

0:55:10.038,0:55:11.760
CA: Continuing from Twitter,

0:55:11.784,0:55:15.366
I mean, is there a standardized[br]global recommendation

0:55:15.390,0:55:16.571
for all countries

0:55:16.595,0:55:18.365
on how to do this?

0:55:18.389,0:55:19.723
And if not, why not?

0:55:20.628,0:55:23.454
AK: I think that's what[br]people are trying to piece together,

0:55:23.478,0:55:25.039
first in terms of what works.

0:55:25.345,0:55:28.490
It's only really in the last[br]sort of few weeks

0:55:28.514,0:55:31.330
we've got a sense that this thing[br]can be controllable

0:55:31.354,0:55:32.981
with this extent of interventions,

0:55:33.005,0:55:35.864
but of course, not all countries[br]can do what China have done,

0:55:35.888,0:55:37.314
some of these measures

0:55:37.338,0:55:41.077
incur a huge social, economic,[br]psychological burden

0:55:41.101,0:55:42.547
on populations.

0:55:42.925,0:55:44.784
And of course, there's the time limit.

0:55:44.808,0:55:46.490
In China, they've had six weeks,

0:55:46.514,0:55:48.078
it's tough to maintain that,

0:55:48.102,0:55:49.942
so we need to think of these tradeoffs

0:55:49.966,0:55:52.799
of all the things we can ask people to do,

0:55:52.823,0:55:56.696
what's going to have the most impact[br]on actually reducing the burden.

0:55:57.807,0:55:59.006
CA: Another question:

0:55:59.030,0:56:02.008
How did this happen[br]and is it likely to happen again?

0:56:03.295,0:56:07.544
AK: So it's likely that this originated[br]with the virus that was circling in bats

0:56:07.568,0:56:10.807
and then probably made its way[br]through another species

0:56:10.831,0:56:12.045
into humans somehow,

0:56:12.069,0:56:14.934
there's a lot of bits[br]of evidence around this,

0:56:14.958,0:56:16.910
there's not kind of single, clear story,

0:56:16.934,0:56:18.925
but even for SARS, it took several years

0:56:18.949,0:56:22.253
for genomics to piece together[br]the exact route that it happened.

0:56:22.277,0:56:25.331
But certainly, I think it's plausible[br]that it could happen again.

0:56:25.355,0:56:27.876
Nature is throwing out[br]these viruses constantly.

0:56:28.419,0:56:31.005
Many of them aren't[br]well-adapted to humans,

0:56:31.029,0:56:32.207
don't pick up,

0:56:32.231,0:56:35.437
you know, there may well have been[br]a virus like this a few years ago

0:56:35.461,0:56:37.207
that just happened to infect someone

0:56:37.231,0:56:40.054
who just didn't have any contacts[br]and didn't go any further.

0:56:40.078,0:56:42.046
I think we are going to face these things

0:56:42.070,0:56:44.474
and we need to think[br]about how can we get in early,

0:56:44.498,0:56:47.291
at the stage where we're talking[br]small numbers of cases,

0:56:47.315,0:56:49.363
and even something like this[br]is containable,

0:56:49.387,0:56:51.339
rather than the situation we've got now.

0:56:51.363,0:56:53.379
CA: It seems like[br]this isn't the first time

0:56:53.403,0:56:56.709
that a virus seems to have emerged[br]from, like, a wild meat market.

0:56:57.664,0:57:00.346
That's certainly how[br]it happens in the movies. (Laughs)

0:57:00.370,0:57:03.029
And I think China has already taken[br]some steps this time

0:57:03.053,0:57:05.696
to try to crack down on that.

0:57:05.720,0:57:08.505
I think that's potentially[br]quite a big deal for the future

0:57:08.529,0:57:11.656
if that can be properly maintained.

0:57:11.680,0:57:13.569
AK: It is, and we saw, for example,

0:57:13.593,0:57:15.633
the H7N9 avian flu,

0:57:15.657,0:57:19.759
over the last few years, in 2013,[br]it was a big emerging concern

0:57:19.783,0:57:21.803
and China made a very extensive response

0:57:21.827,0:57:24.489
in terms of changing[br]how they operate their markets

0:57:24.513,0:57:26.217
and vaccination of birds

0:57:26.241,0:57:29.791
and that seems[br]to have removed that threat.

0:57:29.815,0:57:33.585
So I think these measures can be effective[br]if they're identified early on.

0:57:34.323,0:57:35.847
CA: So talk about vaccinations.

0:57:35.871,0:57:37.744
That's the key measure, I guess,

0:57:37.768,0:57:41.170
to change that susceptibility[br]factor in your equation.

0:57:44.720,0:57:49.053
There's obviously a race on[br]to get these vaccinations out there,

0:57:49.077,0:57:51.687
there are some candidate[br]vaccinations there.

0:57:51.711,0:57:54.047
How do you see that playing out?

0:57:55.099,0:57:58.647
AK: I think there's certainly some[br]promising development happening,

0:57:58.671,0:58:00.686
but I think the timescales of these things

0:58:00.710,0:58:03.797
are really on the order[br]of maybe a year, 18 months

0:58:03.821,0:58:05.728
before these things be widely available.

0:58:05.752,0:58:08.666
Obviously, a vaccine has to go[br]through these stages of trials,

0:58:08.690,0:58:11.125
that takes time,[br]so even if by the end of the year,

0:58:11.149,0:58:13.491
we have something[br]which is viable and works,

0:58:13.515,0:58:16.831
we're still going to see a delay[br]before everyone can get ahold of it.

0:58:16.855,0:58:18.784
CA: So this really puzzles me, actually,

0:58:18.808,0:58:21.760
and I'd love to ask you[br]as a mathematician about this as well.

0:58:21.784,0:58:23.717
There are already several companies

0:58:23.741,0:58:27.745
believing that they have[br]plausible candidate vaccines.

0:58:28.061,0:58:31.839
As you say, the process[br]of testing takes forever.

0:58:32.909,0:58:37.552
Is there a case that we're not[br]thinking about this right

0:58:37.576,0:58:42.553
when we're looking at the way[br]that testing is done

0:58:42.577,0:58:45.279
and that the safety calculations are made?

0:58:45.303,0:58:47.879
Because it's one thing[br]if you're going to introduce

0:58:47.903,0:58:49.471
a brand new drug or something --

0:58:49.495,0:58:53.681
yes, you want to test to make sure[br]that there are no side effects,

0:58:53.705,0:58:55.125
and that can take a long time

0:58:55.149,0:58:58.482
by the time you've done all[br]the control trials and all the rest of it.

0:58:58.506,0:59:00.482
If there's a global emergency,

0:59:00.506,0:59:02.994
isn't there a case,

0:59:03.018,0:59:04.903
both mathematically and ethically,

0:59:04.927,0:59:07.276
that there should just be[br]a different calculation,

0:59:07.300,0:59:08.744
the question shouldn't be

0:59:08.768,0:59:13.807
"Is there any possible case[br]where this vaccine can do harm,"

0:59:13.831,0:59:15.855
the question surely should be,

0:59:15.879,0:59:18.291
"On the net probabilities,

0:59:18.315,0:59:22.188
isn't there a case[br]to roll this out at scale

0:59:22.212,0:59:27.250
to have a shot at nipping[br]this thing in the bud?"

0:59:27.274,0:59:30.170
I mean, what am I missing[br]in thinking that way?

0:59:30.654,0:59:32.885
AK: I mean, we do see that[br]in other situations,

0:59:32.909,0:59:36.751
for example, the Ebola vaccine in 2015

0:59:36.775,0:59:39.720
showed, within a few months,[br]very promising evidence

0:59:39.744,0:59:44.601
and interim results of the trial in humans

0:59:44.625,0:59:47.180
showed what seemed very high efficacy.

0:59:47.204,0:59:50.434
And even though[br]it hadn't been licensed fully,

0:59:50.458,0:59:53.141
it was employed for what is known[br]as compassionate use

0:59:53.165,0:59:54.680
in subsequent other outbreaks.

0:59:54.704,0:59:56.605
So there are these mechanisms

0:59:56.629,0:59:59.469
where vaccines can be[br]fast-tracked in this way.

0:59:59.934,1:00:03.221
But of course, we're currently[br]in a situation where we have no idea

1:00:03.245,1:00:05.172
if these things will do anything at all.

1:00:05.196,1:00:08.021
So I think we need[br]to accrue enough evidence

1:00:08.045,1:00:10.141
that it could have an impact,

1:00:10.165,1:00:12.566
but obviously, fast-track that[br]as much as possible.

1:00:13.780,1:00:16.946
CA: But the skeptic in me[br]still doesn't fully get this.

1:00:17.439,1:00:19.439
I don't understand

1:00:19.463,1:00:24.786
why there isn't more energy[br]behind bolder thinking on this.

1:00:24.810,1:00:28.252
Everyone seems, despite the overall risk,

1:00:28.276,1:00:31.212
incredibly risk-averse[br]about how to build the response to it.

1:00:31.759,1:00:33.164
AK: So with the caveat that,

1:00:33.188,1:00:35.371
yeah, there's a lot[br]of good questions on this,

1:00:35.395,1:00:37.981
and some of them are slightly[br]outside my warehouse,

1:00:38.005,1:00:40.822
but I agree that we need to do more[br]to get timescales out.

1:00:40.846,1:00:42.276
The example I always quote

1:00:42.300,1:00:45.047
is it takes us six months[br]to choose a seasonal flu strain

1:00:45.071,1:00:47.039
and get the vaccines out there to people.

1:00:47.063,1:00:51.085
We always have to try and predict ahead[br]which strains are going to be circulating.

1:00:51.109,1:00:53.228
And that's for something[br]we know how to make

1:00:53.252,1:00:55.367
and has been manufactured for a long time.

1:00:56.212,1:00:58.537
So there is definitely more[br]that needs to be done

1:00:58.561,1:01:00.514
to get these timescales shorter.

1:01:00.538,1:01:02.620
But I think we do have to balance that,

1:01:02.644,1:01:05.788
especially if we're exposing[br]large numbers of people to something

1:01:05.812,1:01:08.081
to make sure that we're confident and safe

1:01:08.105,1:01:10.930
and that it's going to have[br]some benefit, potentially.

1:01:12.546,1:01:14.903
CA: And so, finally,

1:01:14.927,1:01:18.019
Adam, I guess going into this --

1:01:18.942,1:01:23.172
There's another set of infectious things[br]happening around the world

1:01:23.196,1:01:24.362
at the same time,

1:01:24.386,1:01:28.124
which is ideas and the communication[br]around this thing.

1:01:28.148,1:01:33.800
They really are two very dynamic[br]interactive systems of infectiousness --

1:01:33.824,1:01:36.998
there's some very damaging[br]information out there.

1:01:37.022,1:01:41.829
Is it fair to think of this as battle[br]of credible knowledge and measures

1:01:41.853,1:01:44.045
against the bug,

1:01:44.069,1:01:47.735
and just bad information --

1:01:47.759,1:01:50.149
You know, part of what[br]we have to think about here

1:01:50.173,1:01:55.022
is how to suppress one set of things[br]and boost the other, actually,

1:01:55.046,1:01:56.705
turbocharge the other.

1:01:56.729,1:01:58.109
How should we think of this?

1:01:58.133,1:02:02.074
AK: I think we can definitely think of it[br]almost as competition for our attention,

1:02:02.098,1:02:03.813
and we see similarly, with diseases,

1:02:03.837,1:02:06.416
you have viruses competing[br]to infect susceptible hosts.

1:02:06.440,1:02:08.227
And I think we're now seeing,

1:02:08.251,1:02:11.331
I guess over the last few years[br]with fake news and misinformation

1:02:11.355,1:02:12.904
and the emergence of awareness,

1:02:12.928,1:02:14.101
more of a transition

1:02:14.125,1:02:16.733
to thinking about how do we[br]reduce that susceptibility

1:02:16.757,1:02:19.391
if we have people that can be[br]in these different states,

1:02:19.415,1:02:21.876
how can we try and preempt[br]better with information.

1:02:21.900,1:02:24.292
I think the challenge[br]for an outbreak is obviously,

1:02:24.316,1:02:26.569
early on, we have[br]very little good information,

1:02:26.593,1:02:30.844
and it's very easy for certainty[br]and confidence to fill that vacuum.

1:02:30.868,1:02:32.971
And so I think that is something --

1:02:33.320,1:02:36.583
I know platforms are working[br]on how can we get people exposed

1:02:36.607,1:02:38.106
to good information earlier,

1:02:38.130,1:02:40.867
so hopefully protect them[br]against other stuff.

1:02:41.830,1:02:44.410
CA: One of the big unknowns to me[br]in the year ahead --

1:02:44.434,1:02:47.925
let's say that the year ahead includes[br]many, many more weeks,

1:02:47.949,1:02:49.291
for many people,

1:02:49.315,1:02:52.657
of actually self-isolating.

1:02:52.681,1:02:57.728
Those of us who are lucky enough[br]to have jobs where you can do that.

1:02:57.752,1:02:59.434
You know, staying home.

1:02:59.458,1:03:01.807
By the way, the whole injustice[br]of this situation,

1:03:01.831,1:03:06.291
where so many people can't do that[br]and continue to make a living,

1:03:06.315,1:03:10.331
is, I'm sure, going to be[br]a huge deal in the year ahead

1:03:10.355,1:03:16.122
and if it turns out that death rates[br]are much higher in the latter group

1:03:16.146,1:03:17.502
than in the former group,

1:03:17.526,1:03:19.493
and especially in a country like the US,

1:03:19.517,1:03:22.577
where the latter group[br]doesn't even have proper health insurance

1:03:22.601,1:03:24.326
and so forth.

1:03:25.342,1:03:30.610
That feels like right there,[br]that could just become a huge debate,

1:03:30.634,1:03:33.754
hopefully a huge source[br]of change at some level.

1:03:33.778,1:03:36.084
AK: I think that's an incredibly[br]important point,

1:03:36.108,1:03:37.730
because it's very easy --

1:03:37.754,1:03:41.164
I similarly have a job[br]where remote working is fairly easy,

1:03:41.188,1:03:44.987
and it's very easy to say[br]we should just stop social interactions,

1:03:45.011,1:03:48.276
but of course, that could have[br]an enormous impact on people

1:03:48.300,1:03:51.100
and the choices and the routine[br]that they can have.

1:03:51.124,1:03:53.299
And I think those do need[br]to be accounted for,

1:03:53.323,1:03:55.705
both now and what the effect[br]is going to look like

1:03:55.729,1:03:57.259
a few months down the line.

1:03:57.283,1:03:58.701
CA: When all's said and done,

1:03:58.725,1:04:03.783
is it fair to say that the world has[br]faced, actually, much graver problems

1:04:03.807,1:04:04.958
in the past,

1:04:04.982,1:04:08.069
that on any scenario,

1:04:08.093,1:04:11.762
it's highly likely that at some point[br]in the next 18 months, let's say,

1:04:11.786,1:04:16.095
a vaccine is there and starts[br]to get wide distribution,

1:04:16.119,1:04:22.101
that we will have learned[br]lots of other ways to manage this problem?

1:04:22.125,1:04:24.257
But at some point, next year probably,

1:04:24.281,1:04:30.412
the world will feel like[br]it's got on top of this

1:04:30.436,1:04:31.839
and can move on.

1:04:31.863,1:04:33.602
Is that likely to be it,

1:04:33.626,1:04:36.990
or is this more likely to be,[br]you know, it escapes,

1:04:37.014,1:04:42.061
it's now an endemic nightmare[br]that every year picks off far more people

1:04:42.085,1:04:44.634
than are picked off by the flu currently.

1:04:44.658,1:04:47.481
What are the likely ways forward,

1:04:47.505,1:04:49.550
just taking a slightly longer-term view?

1:04:49.574,1:04:52.589
AK: I think there's plausible ways[br]you could see all of those

1:04:52.613,1:04:54.204
potentially playing out.

1:04:54.228,1:04:58.838
I think the most plausible is probably[br]that we'll see very rapid growth this year

1:04:58.862,1:05:03.464
and lots of large outbreaks[br]that don't recur, necessarily.

1:05:03.488,1:05:05.638
But there is a potential[br]sequence of events

1:05:05.662,1:05:09.743
that could end up with these kind of[br]multiyear outbreaks in different places

1:05:09.767,1:05:10.925
and reemerging.

1:05:10.949,1:05:12.826
But I think it's likely we'll see

1:05:12.850,1:05:15.701
most transmission concentrated[br]in the next year or so.

1:05:15.725,1:05:18.761
And then, obviously,[br]if there's a vaccine available,

1:05:18.785,1:05:21.338
we can move past this,[br]and hopefully, learn from this.

1:05:21.362,1:05:24.554
I think a lot of the countries[br]that responded very strongly to this

1:05:24.578,1:05:26.006
were hit very hard by SARS.

1:05:26.030,1:05:29.156
Singapore, Hong Kong,[br]that really did leave an impact,

1:05:29.180,1:05:31.910
and I think that's something[br]they've drawn on very heavily

1:05:31.934,1:05:33.276
in their response to this.

1:05:33.300,1:05:34.458
CA: Alright.

1:05:34.482,1:05:37.157
So let's wrap up maybe by just[br]encouraging people

1:05:37.181,1:05:39.101
to channel their inner mathematician

1:05:39.125,1:05:44.414
and especially think[br]about the opportunities,

1:05:44.438,1:05:48.358
and the transmission probabilities[br]that they can help shift.

1:05:48.382,1:05:52.760
Just remind us of the top[br]three or four or five or six things

1:05:52.784,1:05:54.719
that you would love to see people doing.

1:05:54.743,1:05:57.615
AK: I think at the individual level,[br]just thinking a lot more

1:05:57.639,1:06:00.020
about your interactions[br]and your risk of infection

1:06:00.044,1:06:02.012
and obviously, what gets onto your hands

1:06:02.036,1:06:03.731
and once that gets onto your face,

1:06:03.755,1:06:06.340
and how do you potentially[br]create that risk for others.

1:06:06.364,1:06:09.037
I think also, in terms of interactions,

1:06:09.061,1:06:13.896
with things like handshakes[br]and maybe contacts you don't need to have.

1:06:13.920,1:06:16.697
You know, how can we get those down[br]as much as possible.

1:06:16.721,1:06:19.134
If each person's giving it[br]to two or three others,

1:06:19.158,1:06:22.371
how do we get that number[br]down to one, through our behavior.

1:06:22.395,1:06:26.204
And then it's likely that we'll need[br]some larger-scale interventions

1:06:26.228,1:06:29.169
in terms of gatherings, conferences,

1:06:29.193,1:06:31.827
other things where[br]there's a lot of opportunities

1:06:31.851,1:06:33.220
for transmission.

1:06:33.244,1:06:36.197
And really, I think that combination[br]of that individual level,

1:06:36.221,1:06:39.227
you know, if you're ill[br]or potentially you're going to get ill,

1:06:39.251,1:06:40.519
reducing that risk,

1:06:40.543,1:06:42.146
but then also us working together

1:06:42.170,1:06:44.274
to prevent it getting[br]into those groups who,

1:06:44.298,1:06:46.130
if it continues to be uncontrolled,

1:06:46.154,1:06:48.392
could really hit some people[br]very, very hard.

1:06:49.434,1:06:51.060
CA: Yeah, there's a lot of things

1:06:51.084,1:06:54.299
that we may need[br]to gently let go of for a bit.

1:06:54.323,1:06:58.553
And maybe try to reinvent[br]the best aspects of them.

1:06:59.077,1:07:00.418
Thank you so much.

1:07:00.442,1:07:03.235
If people want to keep up with you,

1:07:03.259,1:07:05.974
first of all, they can follow you[br]on Twitter, for example.

1:07:05.998,1:07:07.347
What's your Twitter handle?

1:07:07.371,1:07:09.990
AK: So AdamJKucharski, all one word.

1:07:10.014,1:07:12.524
CA: Adam, thank you so much[br]for your time, stay well.

1:07:12.548,1:07:13.928
AK: Thank you.

1:07:13.952,1:07:20.952
(Music)

1:07:29.263,1:07:32.533
CA: Associate professor[br]and TED Fellow Adam Kucharski.

1:07:33.339,1:07:35.934
We'd love to hear what you think[br]of this bonus episode.

1:07:35.958,1:07:38.799
Please tell us by rating[br]and reviewing us in Apple Podcasts

1:07:38.823,1:07:40.757
or your favorite podcast app.

1:07:41.244,1:07:43.217
Those reviews are influential, actually.

1:07:43.241,1:07:44.617
We certainly read every one,

1:07:44.641,1:07:46.791
and truly appreciate your feedback.

1:07:46.815,1:07:48.768
(Music)

1:07:48.792,1:07:52.592
This week's show was produced[br]by Dan O'Donnell at Transmitter Media.

1:07:52.616,1:07:54.647
Our production manager[br]is Roxanne Hai Lash,

1:07:54.671,1:07:56.988
our fact-checker, Nicole Bode.

1:07:57.012,1:07:59.290
This episode was mixed by Sam Baer.

1:07:59.314,1:08:01.355
Our theme music[br]is by Allison Layton-Brown.

1:08:01.379,1:08:03.926
Special thanks to my colleague[br]Michelle Quint.

1:08:04.252,1:08:06.473
Thanks for listening to the TED Interview.

1:08:06.497,1:08:07.998
We'll be back later this spring

1:08:08.022,1:08:11.370
with a whole new season's worth[br]of deep dives with great minds.

1:08:11.759,1:08:14.934
I hope you'll enjoy them[br]whether or not life is back to normal.

1:08:15.634,1:08:16.850
I'm Chris Anderson,

1:08:16.874,1:08:18.879
thanks for listening and stay well.