A guest lecture on informatics for disease surveillance, looking at a number of new new technologies. Delivered at the School of Health and Related Research.

1. Informatics for Disease
Surveillance – New
Technologies
Guest lecture for HAR655/HAR6059 on
Public Health Informatics
Wasim Ahmed (BA, MSc)
Tuesday 28th of February, 2017

2. 28/02/2017 © The University of Sheffield
2
Contents
• Final year PhD student from the Health Informatics
Research Group
• PhD looks at infectious disease outbreaks on Twitter
such as Swine Flu and Ebola
• Worked on a number of industry and academic projects
related to social media and analytics

3. Overview
• Part 1 - What is surveillance?
• Part 2 - Why do we need surveillance for
disease outbreaks?
• Part 4 - Traditional methods of disease
surveillance
• Part 5 - Disease surveillance and new
technologies

4. Learning Outcomes
• Why do we need disease surveillance?
• How has disease surveillance data
traditionally been collected?

5. Learning Outcomes
• What role can new technologies can play
in disease surveillance?
• Current and future role of social media for
disease surveillance (my research area).

6. Part 1:
What is surveillance?

7. Definition
• The Centres for Disease Control (1998)
have described surveillance as:
• “the on-going, systematic collection,
analysis, interpretation, and
dissemination of data regarding a
health-related event”

8. Why do we need surveillance?
• Tracking outbreaks
• Allocating resources
• Analysis/Interpretation - Monitoring
initiatives

9. Why do we need surveillance?
• Monitoring individuals
• Keeping people informed

10. Traditional surveillance
• WHO – International Health
Regulations (IHR 2005) and Global
Health Statistics
• UK – Surveillance via GP (routine and
out of hours), Emergency
Departments and NHS 111/NHS 24

11. NHS Syndromic Surveillance
• NHS 111/NHS 24 - telephone helpline calls to a triage system
with call back from a nurse or a signpost to further information
• Syndromic calls as a % of all calls
• Syndromic indicators: Cold/flu; fever; cough; difficulty
breathing; rash; diarrhoea; vomiting; eye problems; lumps;
double vision
• Concern raised when numbers higher than previous
years
• Also face to face syndromic surveillance – GP
consultations/ED/Out of hours and walk in centres.

12. Public Health England Weekly
National Influenza Report

14. New Technology
• Using the infrastructure we already have
• Internet, mass media, smart phones,
wireless technology.
• May not reach all of the world, but more
than a traditional surveillance system

15. Example - HealthMap

16. HealthMap
• @healthmap
• www.healthmap.org/

17. False Positives
• Many people may diagnose themselves with
an infectious disease
• Methods of assessing the validity and
reliability of the self-diagnosis i.e., looking at
locations

18. Surveillance for gauging public
sentiment
• It is possible to use social media and online
user-generated content in order to gauge
public opinion
• See Chew & Eysenbach (2010): “H1N1 pandemic-
related tweets on Twitter were primarily used to
disseminate information from credible sources to the
public, but were also a rich source of opinions and
experiences.”

19. Using Network Analysis

20. [Divided]
Polarized Crowds
[Unified]
Tight Crowd
[Fragmented]
Brand Clusters
[Clustered]
Community Clusters
[In-Hub & Spoke]
Broadcast Network
[Out-Hub & Spoke]
Support Network
6 kinds of Twitter social media networks

21. [Divided]
Polarized Crowds
[Unified]
Tight Crowd
[Fragmented]
Brand Clusters
[Clustered]
Community Clusters
[In-Hub & Spoke]
Broadcast Network
[Out-Hub & Spoke]
Support Network
6 kinds of Twitter social media networks

22. • Most frequently shared URLs, Domains, Hashtags,
Words, Word Pairs, Replied-To, Mentioned Users, and
most frequent tweeters
• Produces metrics overall and by group of users (users
are grouped by tweet content)
• By looking at different metrics associated with different
groups (G1, G2, G3 etc) you can see the different topics
that users may be talking about
NodeXL produces other metrics

23. #WorldMentalHealthAwarenessDay

24. 28/02/2017 © The University of Sheffield
24
Produces a number of metrics

25. 28/02/2017 © The University of Sheffield
25
Produces a number of metrics

26. 28/02/2017 © The University of Sheffield
26
Produces a number of metrics

27. Foods Standards Agency and Norovirus

28. Centrality
– Centrality measures help address the
question: who the most important or central
person in this network?
– Centrality measures include:
• Degree centrality
• Closeness centrality
• Betweenness centrality
• Eigenvector centrality
• PageRank centrality

29. Betweenness Centrality
From Richard Ingram’s blog post visualising
Data: Seeing is Believing
http://www.richardingram.co.uk/2012/12/visualis
ing-data-seeing-is-believing/

30. Degree Centrality
From Richard Ingram’s blog post visualising
Data: Seeing is Believing
http://www.richardingram.co.uk/2012/12/visualis
ing-data-seeing-is-believing/

31. Patterns are left behind
31
When users engage online
Like, Link, Reply, Rate, Review, Favorite,
Friend, Follow, Forward, Edit, Tag, Comment, Check-in…

32. How does it work?
• Is a tweet relevant to Norovirus?
• #sickbug, norovirus, tummy ache, the
runs, vommed, and chunder etc

33. What is so great about it?
• Quick access to data
• Real time updating
• Large volume of data – the majority of
which is openly accessible
• Linked to locations

34. Practical Element
• Looking at the NodeXL graph gallery
• http://nodexlgraphgallery.org/Pages/Graph
.aspx?graphID=55322
• Search for other topics that may also be of
interest

36. Google Flu Trends
• Similar methodology to Twitter syndromic
surveillance
• Developed by comparing flu related
queries to physician visits and working out
how these queries were mapped to visits
for a specific US state.
• Then using these findings to predict
current and future outbreaks

37. Questions?

38. Further reading and references
• Paul M et al (2015) Worldwide Influenza Surveillance through Twitter. 2015 AAAI workshop.
Available from https://www.aaai.org/ocs/index.php/WS/AAAIW15/paper/viewFile/10161/10255
• Lamb A et al (2012) Separating Fact from Fear: Tracking Flu Infections on Twitter. Available from
http://cmci.colorado.edu/~mpaul/files/naacl13flu-final.pdf
• PHE Weekly National Influenza Report. Available from
https://www.gov.uk/government/statistics/weekly-national-flu-reports
• Houses of Parliament Parliamentary Office of Science & Technology. PostNote. Available from
http://researchbriefings.parliament.uk/ResearchBriefing/Summary/POST-PN-462
• Guardian (2014) Google Flu Trends is no longer good at predicting flu
http://www.theguardian.com/technology/2014/mar/27/google-flu-trends-predicting-flu
• The Parable of Google Flu; Traps in Big Data Analysis (2014). Science. Volume 343. Available
from http://gking.harvard.edu/files/gking/files/0314policyforumff.pdf
• CDC Definition of Surveillance available from
http://www.cdc.gov/mmwr/preview/mmwrhtml/00025629.htm

39. Further reading and references
• Al Rodhan, N et al. (2006) Definitions of globalization: A comprehensive overview and
a proposed definition. Available from
http://www.wh.agh.edu.pl/other/materialy/678_2015_04_21_22_04_13_Definitions%2
0of%20Globalization_A%20Comprehensive%20Overview%20and%20a%20Propose
d%20Definition.pdf
• CDC Framework for Evaluating Public Health Surveillance Systems for Early
Detection of Outbreaks. 2004. Available at
http://www.cdc.gov/mmwr/preview/mmwrhtml/rr5305a1.htm
• Chan EH, Brewer TF, Madoff LC, Pollack MP, Sonricker AL, Keller M, Freifeld CC,
Blench M, Mawudeku A, Brownstein JS. Global capacity for emerging infectious
disease detection. Proc Natl Acad Sci U S A. 2010 Dec 14;107(50):21701-6.
doi:10.1073/pnas.1006219107. Epub 2010 Nov 29. PubMed PMID: 21115835;
PubMed Central PMCID: PMC3003006.

40. To
Discover
And
Understand.