
we are at a very interesting point where
data science can actually play an
enormous role in healthcare and more
broadly I think digital technologies and
healthcare has a very good meeting point
that probably India has an opportunity
to leapfrog and also you know greatly
improve the delivery of healthcare with
with the use of some of these
technologies I think using the word
epidemiology and for most of you who are
connected with the medical profession
but I I teach at the unit’s Geoscience
and and one of the courses that I’ve
just introduced there is a course on
digital epidemiology and and the of
course the word epidemiology comes from
epi and demos and which means about
populations or upon populations now you
know the New England Journal of Medicine
holds a spring meeting it’s actually
called the Massachusetts Medical Council
it it holds a meeting in the spring
every year and this year’s meeting was
dedicated to epidemics and and the key
lecture was given by Bill Gates and and
it was it’s called the Shattuck lecture
and in his and if you get the New
England Journal of Medicine they made
the first issue the the lecture has
reproduced there so I wrote you to read
it
and in that he talks about how in many
ways you know we’ve learned how to deal
with
the big challenges of HIV we’re now your
people you can manage a patient with HIV
with even malaria is now slowly coming
under control but this whole idea that
they could be a pandemic because all it
takes is for one person to sit in a
plane and move an infection from one
country to another
in fact the other passengers in the
plane or whatever there is now an
increasing sense of threat that the
entire world is sensing and and you know
and of course what you see on the slide
there is a computer simulation of what
would happen if there was a global
epidemic of influenza for which you had
no ability to control and how within six
months or so you would have about 32
million people seriously affected and
there have been many many such studies
carried out and so on
so there’s actually a coalition now for
looking at innovations to address
epidemics
it’s called sepi C epi we will talk more
about it later and can sepi has been
funded by a group of agencies and they
put about 630 million dollars or the
last hour and they’re actually going
after various strategies of how to
address epidemics and pandemics right
now interestingly their first focus on
vaccine development has been on not what
you normally think would think of it
well so anybody wants to guess what
vaccines are being developed for for
to control pandemics yeah it sorry I
couldn’t hear that Zika you know could
be one but right it it turns out that
the first three that they’re going after
our Lhasa MERS and NEPA okay so that
really caught my attention because you
know we faced the challenge with NEPA
right not so long ago and in fact my
friend dr. Arun Kumar who at Manipal
runs the virus research group there was
instrumental in detecting it early right
and and if you look at that timeline
there I’m not sure how visible it is but
the samples were sent to him and by I
think May 18th
he had a diagnosis right now the reason
he was prepared with the diagnosis was
that much earlier on when we had I think
one of the flu epidemics that was
starting to threaten I think it was SARS
and then you know after that swine flu
avian flu and so on the ICMR decided to
set up Sentinel sites around the country
and the money Powell Center for virus
research is sort of overseeing all of
these centers and so arune’s group was
prepared to diagnose nipa because they
were looking at
I think Tripura and various states in
Northeast where they were expecting that
there could be an outbreak because
there’d been outbreaks in West Bengal in
in Bangladesh and they were so they were
prepared for that so they were able to
run the tests and quickly diagnose it
but then it took another couple of days
for for the National Institute of
biology to actually confirm the
diagnosis so they didn’t announce it
till the May 20th I think but the
Kettler public health authorities moved
into action and actually Arun and I did
a webinar a couple of nights back to a
group of people who are taking a course
on epidemiology through the Takshashila
Institute of course epidemiology and
public health there’s is there’s a lot
of challenges and and we know how to how
to get to those challenges but you know
this is all theoretical how do you
actually get to those contagion models
how do you actually know how you’re
going to contain and what are the
strategies and how can you make all this
evidence driven and not a minute driven
right as often happens so you know there
are there are approaches now that are
starting to get very very sophisticated
and this is where data science comes in
and essentially you have to look at
behavioral networks social networks so
to speak you have to look at how
information goes you know goes through
many of these networks how that spreads
fear panic you know all of those
socio-cultural aspects and of course you
have to know the disease dynamics you
have to know how the contagion actually
spreads and how how infection spreads
and how do you model all that and so for
the u.s. with about 300 million
population it turns out that the if you
actually want to
look at it from an evidence base that
actually integrates all of this
information it would involve something
like a hundred gigabytes of data right
and so this is where you know it has to
be done with you know I think serious
professional help from from computer
scientists data scientists and so on who
can who can get involved in this and
this is why I believe that this has to
be in some ways a multidisciplinary
interdisciplinary effort that that comes
together to take care of this I just
want to also bring out that
another aspect in here is prediction how
do you actually predict how can you
predict ahead of time that there’s going
to be an outbreak or high incidence of
some disorders and so this is the work
by an undergraduate student in
mathematics at at ISC
Atlanta Chatrapathi who just wrote her
thesis on this where you know again
using time series data and she was able
to actually build a model that gives a
reasonable prediction over the next six
months what the incidence of dengue will
be in Singapore and this you know she
was able to build this model and train
it using modern machine learning
techniques and so on partly because you
had excellent data from Singapore right
so the the you know you actually have
historical data on which you can train
these these predictive models and now
she’s going on to actually add
temperature and rainfall day
you know to improve the velocity and the
predictive power of these models and so
I you know when she started doing this
work we tried very hard to get data like
this in India and that was a challenge
so so I think you know we we need to put
together efforts now to to pull together
better data to be able to drive this
kind of evidence driven epidemiology I
think that’s that’s a list of the kinds
of models that one could build for
different types of epidemics and and how
how you could you could use those to
design things now you know I think the
nipa story actually made me feel very
hopeful because you know it seems like
we are getting into position to do
things a lot better but you know let’s
face it I think we are actually in a
situation where we need to do a lot
right The Lancet ranking of public
health preparedness in in India and our
ability to reach the Sustainable
Development Goals the SDGs we were
ranked in the 140 s and now we’ve
actually slipped down to the 150 50 150
fourth or something like that in in in
the rankings worldwide rankings that
Lancet puts out so so there’s a lot of
work to be done but and this is where
you know we I think we have to bring all
the tools that that we have at our
disposal to to try to improve things I
think
the the three major factors that are
listed there are things like hygiene
malaria so I think the combination of
basic development approaches and
epidemiological approaches are needed if
you want to really reach goals that that
are aspirational goals for us I mean the
Prime Minister’s of course declared that
we will stop TB by 2025 and that one of
those sustainable development goals
which from 2030 it’s been pulled back to
2025 and in order to get anywhere close
to that I think we need we need a lot of
different approaches that that can help
us there I mentioned
sepi which is the coalition and this
Coalition has been pulled together to
improve preparedness for epidemics and
and it’s driving innovation and there’s
there’s a there’s a large fund that has
been pulled together I think Gates
Foundation Wellcome Trust the World
Economic Forum several agencies have got
involved and and the interesting thing
is that India is now a nodal centre
because of our ability to manufacture
vaccines today you know every third
vaccine given to a child in the world he
has actually made in India so so we have
that ability to to to manufacture at
scale and you know and I think the
research and development to actually
come up with vaccines that are relevant
to dealing with epidemics I think is is
what is being built
of course the main activity today has
started in
the National Capital Region where you
see th STI and several Institute’s being
brought together to work on this i cg b
it’s another stat Center but of course
the various you know the entire sort of
medical establishment in India will
obviously need to be involved in in
taking these innovations and diffusing
them and validating the new technologies
and so so let me just conclude then with
this idea that we are at a situation now
we really need to focus on building
something that I am calling the health
heat map of India and I think in order
to build something like this we you know
I again I’m repeating myself but you
really need to bring data scientists
epidemiologists public health agencies
stakeholders together and and to create
an integrated platform where where we
would be in a position to to respond to
health emergencies but also just improve
our general levels of of healthcare
delivery planning and and and I think
that’s the idea that I would like to
leave you to and and again being the
Mahalanobis day today Indian statistics
day let’s actually think about how all
of these ideas in data science and
computation and health
health expertise can be brought together to bear on this so thank you very much
data science can actually play an
enormous role in healthcare and more
broadly I think digital technologies and
healthcare has a very good meeting point
that probably India has an opportunity
to leapfrog and also you know greatly
improve the delivery of healthcare with
with the use of some of these
technologies I think using the word
epidemiology and for most of you who are
connected with the medical profession
but I I teach at the unit’s Geoscience
and and one of the courses that I’ve
just introduced there is a course on
digital epidemiology and and the of
course the word epidemiology comes from
epi and demos and which means about
populations or upon populations now you
know the New England Journal of Medicine
holds a spring meeting it’s actually
called the Massachusetts Medical Council
it it holds a meeting in the spring
every year and this year’s meeting was
dedicated to epidemics and and the key
lecture was given by Bill Gates and and
it was it’s called the Shattuck lecture
and in his and if you get the New
England Journal of Medicine they made
the first issue the the lecture has
reproduced there so I wrote you to read
it
and in that he talks about how in many
ways you know we’ve learned how to deal
with
the big challenges of HIV we’re now your
people you can manage a patient with HIV
with even malaria is now slowly coming
under control but this whole idea that
they could be a pandemic because all it
takes is for one person to sit in a
plane and move an infection from one
country to another
in fact the other passengers in the
plane or whatever there is now an
increasing sense of threat that the
entire world is sensing and and you know
and of course what you see on the slide
there is a computer simulation of what
would happen if there was a global
epidemic of influenza for which you had
no ability to control and how within six
months or so you would have about 32
million people seriously affected and
there have been many many such studies
carried out and so on
so there’s actually a coalition now for
looking at innovations to address
epidemics
it’s called sepi C epi we will talk more
about it later and can sepi has been
funded by a group of agencies and they
put about 630 million dollars or the
last hour and they’re actually going
after various strategies of how to
address epidemics and pandemics right
now interestingly their first focus on
vaccine development has been on not what
you normally think would think of it
well so anybody wants to guess what
vaccines are being developed for for
to control pandemics yeah it sorry I
couldn’t hear that Zika you know could
be one but right it it turns out that
the first three that they’re going after
our Lhasa MERS and NEPA okay so that
really caught my attention because you
know we faced the challenge with NEPA
right not so long ago and in fact my
friend dr. Arun Kumar who at Manipal
runs the virus research group there was
instrumental in detecting it early right
and and if you look at that timeline
there I’m not sure how visible it is but
the samples were sent to him and by I
think May 18th
he had a diagnosis right now the reason
he was prepared with the diagnosis was
that much earlier on when we had I think
one of the flu epidemics that was
starting to threaten I think it was SARS
and then you know after that swine flu
avian flu and so on the ICMR decided to
set up Sentinel sites around the country
and the money Powell Center for virus
research is sort of overseeing all of
these centers and so arune’s group was
prepared to diagnose nipa because they
were looking at
I think Tripura and various states in
Northeast where they were expecting that
there could be an outbreak because
there’d been outbreaks in West Bengal in
in Bangladesh and they were so they were
prepared for that so they were able to
run the tests and quickly diagnose it
but then it took another couple of days
for for the National Institute of
biology to actually confirm the
diagnosis so they didn’t announce it
till the May 20th I think but the
Kettler public health authorities moved
into action and actually Arun and I did
a webinar a couple of nights back to a
group of people who are taking a course
on epidemiology through the Takshashila
Institute of course epidemiology and
public health there’s is there’s a lot
of challenges and and we know how to how
to get to those challenges but you know
this is all theoretical how do you
actually get to those contagion models
how do you actually know how you’re
going to contain and what are the
strategies and how can you make all this
evidence driven and not a minute driven
right as often happens so you know there
are there are approaches now that are
starting to get very very sophisticated
and this is where data science comes in
and essentially you have to look at
behavioral networks social networks so
to speak you have to look at how
information goes you know goes through
many of these networks how that spreads
fear panic you know all of those
socio-cultural aspects and of course you
have to know the disease dynamics you
have to know how the contagion actually
spreads and how how infection spreads
and how do you model all that and so for
the u.s. with about 300 million
population it turns out that the if you
actually want to
look at it from an evidence base that
actually integrates all of this
information it would involve something
like a hundred gigabytes of data right
and so this is where you know it has to
be done with you know I think serious
professional help from from computer
scientists data scientists and so on who
can who can get involved in this and
this is why I believe that this has to
be in some ways a multidisciplinary
interdisciplinary effort that that comes
together to take care of this I just
want to also bring out that
another aspect in here is prediction how
do you actually predict how can you
predict ahead of time that there’s going
to be an outbreak or high incidence of
some disorders and so this is the work
by an undergraduate student in
mathematics at at ISC
Atlanta Chatrapathi who just wrote her
thesis on this where you know again
using time series data and she was able
to actually build a model that gives a
reasonable prediction over the next six
months what the incidence of dengue will
be in Singapore and this you know she
was able to build this model and train
it using modern machine learning
techniques and so on partly because you
had excellent data from Singapore right
so the the you know you actually have
historical data on which you can train
these these predictive models and now
she’s going on to actually add
temperature and rainfall day
you know to improve the velocity and the
predictive power of these models and so
I you know when she started doing this
work we tried very hard to get data like
this in India and that was a challenge
so so I think you know we we need to put
together efforts now to to pull together
better data to be able to drive this
kind of evidence driven epidemiology I
think that’s that’s a list of the kinds
of models that one could build for
different types of epidemics and and how
how you could you could use those to
design things now you know I think the
nipa story actually made me feel very
hopeful because you know it seems like
we are getting into position to do
things a lot better but you know let’s
face it I think we are actually in a
situation where we need to do a lot
right The Lancet ranking of public
health preparedness in in India and our
ability to reach the Sustainable
Development Goals the SDGs we were
ranked in the 140 s and now we’ve
actually slipped down to the 150 50 150
fourth or something like that in in in
the rankings worldwide rankings that
Lancet puts out so so there’s a lot of
work to be done but and this is where
you know we I think we have to bring all
the tools that that we have at our
disposal to to try to improve things I
think
the the three major factors that are
listed there are things like hygiene
malaria so I think the combination of
basic development approaches and
epidemiological approaches are needed if
you want to really reach goals that that
are aspirational goals for us I mean the
Prime Minister’s of course declared that
we will stop TB by 2025 and that one of
those sustainable development goals
which from 2030 it’s been pulled back to
2025 and in order to get anywhere close
to that I think we need we need a lot of
different approaches that that can help
us there I mentioned
sepi which is the coalition and this
Coalition has been pulled together to
improve preparedness for epidemics and
and it’s driving innovation and there’s
there’s a there’s a large fund that has
been pulled together I think Gates
Foundation Wellcome Trust the World
Economic Forum several agencies have got
involved and and the interesting thing
is that India is now a nodal centre
because of our ability to manufacture
vaccines today you know every third
vaccine given to a child in the world he
has actually made in India so so we have
that ability to to to manufacture at
scale and you know and I think the
research and development to actually
come up with vaccines that are relevant
to dealing with epidemics I think is is
what is being built
of course the main activity today has
started in
the National Capital Region where you
see th STI and several Institute’s being
brought together to work on this i cg b
it’s another stat Center but of course
the various you know the entire sort of
medical establishment in India will
obviously need to be involved in in
taking these innovations and diffusing
them and validating the new technologies
and so so let me just conclude then with
this idea that we are at a situation now
we really need to focus on building
something that I am calling the health
heat map of India and I think in order
to build something like this we you know
I again I’m repeating myself but you
really need to bring data scientists
epidemiologists public health agencies
stakeholders together and and to create
an integrated platform where where we
would be in a position to to respond to
health emergencies but also just improve
our general levels of of healthcare
delivery planning and and and I think
that’s the idea that I would like to
leave you to and and again being the
Mahalanobis day today Indian statistics
day let’s actually think about how all
of these ideas in data science and
computation and health
health expertise can be brought together to bear on this so thank you very much
Please follow and like us: