Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


"The global distribution and burden of dengue" by Prof. Simon I. Hay, University of Oxford

A plenary talk given at the Third International Conference on Dengue and Dengue Haemorrhagic Fever: Global Dengue: Challenges and Promises in Bangkok, Thailand on 21 October 2014.

Prezi Training

on 6 November 2013

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of "The global distribution and burden of dengue" by Prof. Simon I. Hay, University of Oxford

Correction factors for comparison derived from literature – results extremely plausible
The Third International Conference on Dengue and Dengue Haemorrhagic Fever
Global Dengue: Challenges and Promises
Bangkok, Thailand

The global distribution and burden of dengue
i) comprehensive collection of occurrence data
ii) evidence-based consensus
iii) bespoke environmental covariates data
Disease occurrence mapping
Why map dengue?
Larvae and adults thrive in urban areas
Urbanization trends are good for dengue
spreads readily on boats
Dengue virus spreads rapidly in humans on aircraft
~ as malaria declines, dengue rises
Anders & Hay (2012). , (12): 977-984
Pair probability of occurrence with cohort studies to infer inapparent (n=54) and apparent (n=39) incidence per pixel
Then pair with population surfaces for 2010 to sum up global totals
Consistent global estimates for BMGF, GAVI and surfaces for GBD2013
Bhatt (2013). , (7446): 504-507.
The International Research Consortium on Dengue Risk Assessment, Management and Surveillance (IDAMS: http://www.idams.eu) is funded by the European Commission Seventh Framework Programme
Green open-access with European PubMed Central ID: PMC3651993

Prof. Simon I. Hay
Monday 21 October 2013
University of Oxford

Ecological theory
What is disease occurrence mapping?
Dengue as an example
Why map dengue?
Disease occurrence mapping: how does it work?
Not just pretty maps
From probability of occurrence to new burden estimations
“Borrows” directly from rich ecological theory
assume disease observations approximate the realized niche of pathogen
correlate with covariates where known and use statistical model to predict extensively
It is clearly not perfect
i) sampling – do we capture enough of the realized distribution?
ii) biotic interactions, biogeography and human impact – especially important at the global scale
iii) environment – is our covariate suite representative?
Mitigated from the outset
Uncertain distribution and burden
Mild to fatal disease with no cure
Day-biting vectors hard to avoid
Control is extremely difficult
A disease of the future
From dengue risk to burden
(a) evidence consensus;
(b) occurrence;
(c) pseudo-absence;
(d) environmental / epidemiological covariates which together predict
(e) the probability of occurrence of an infectious disease
Occurrence mapping
ITHG disagree on endemic status of 34 territories

Massive systematic search and treatment of evidence of transmission

Brady (2012). , (8): e1760

Gold open-access with European PubMed Central ID: PMC3413714
(a) evidence-based consensus
(b) occurrence data
Formal literature searches were conducted
-PubMed, ISI WoS and PROMED

All papers identified manually

All abstracts read manually

All occurrence data geo-positioned manually

This yielded 8309 points for mapping from 106 countries (1963-2012)

An additional 1622 points were provided by HealthMap (2007-2012)

N = 8309 from literature searches and N = 1622 from HealthMap
90% occurrence points from the tropics
Africa relatively data poor – a focus of future research

(b) occurrence data
All niche modelling techniques
These data are generated in proportion to definitive extent certainty
We explore the full effect through ensemble approaches

(c) pseudo-absence data
(d) environmental datasets
Lots of useful maps available, from climatologies to GDP
Synoptic precipitation http://www.worldclim.org
From 0 mm (yellow) to 800 mm (blue) per month
(d) environmental datasets
Synoptic mean monthly temperature http://www.worldclim.org
Converted to daily dengue transmissibility suitability for
Covariate is the annual number of days suitable for dengue transmission
(d) environmental datasets
Lots of opportunity to use “epidemiological” covariates
Geographically-based economic data
Pixel-based equivalent of gross domestic product (GDP)
BRT mapping with (i) evidence base; (ii) modern cartography and (iii) uncertainty
Statistically accurate and geographically plausible (definitive extent; ++occurrence data)
Biologically plausible: covariates selected have epidemiological relevance: precipitation (37%), temp. suitability (20%), G-econ (9%)
(e) probability of occurrence
Thank You
Talk Aims
More data to inform evidence-based consensus; Africa
More occurrence data; everywhere there are gaps
More cohort studies everywhere; most critical in high-transmission areas

Future Work
More data is crucial
Use similar conceptual approach to infer future risk of dengue
Strong focus on future risk to Europe
Use BRT methodology (similar to burden paper)
Project to 2020, 2050, 2080 using future vector, climate, economic, demographic scenarios
In accordance with other metrics but at a much finer spatial scale and with detailed uncertainty estimates
Previous burden estimates

Actually reported:
≈ 20,000 cases

Comparison with national reporting

Worked example of Sri Lanka if we assume an ideal reporting system
Seven-fold underreporting is plausible in a suboptimal system
Full transcript