Digital Health: From Online Evidence to Serious Games - "Stop the Bugs"
1. 1
Digital Health: From Online
Evidence to Serious Games
“Stop the Bugs”
Patty Kostkova
UCL
London, UK
2. Healthcare at the end of 20th century: Internet
“The impact of the Internet has largely been
unforeseen, and it may have a revolutionary role in
retooling the trillion-dollar health care industry in the
United States” (June Forkner-Dunn 2003).
2
3. Healthcare in 21st century: mobile
31% (up from 17% in 2010) of cell phone
users have used their phone to look up
health information
3
4. Information Overflow
The information you have
is not the information you
want
The information you
want is not the
information you need
The information you
need is not known
The information that’s
known can’t be found in
time
4
5. 5
Projects
hosted by
the NeLI
portal
National electronic Library of Infection (NeLI, formerly NeLCD) – professionals oriented
Bugs and Drugs on the Web (AR DL) – public oriented
Training in Infection Manual (TII) – for trainees
National Resource for Infection Control (NRIC) – for professionals
eBug – EU DG SANCO project – education pack and games for children about AR
SeaLife – semantic web browser for life sciences
Medicines Support Unit for Optometrists – in collaboration with Department of Optometry, City Uni.
WHO Lab Resources evaluation project
IFH project – International Scientific Forum for Home Hygiene
10. Google, msn searches that led to the AR site between 1 Sep & 31st Jan
38883 searches
10
7161 included the word alcohol
12949 antibiotic
6515 chest
3467 bacteria
2085 virus
1852 acne
1149 flu
…. And 250 included "cat"
• Some users have an addiction problem:
• Quit antibiotics
• Quit working
• Many users are concerned about animal
welfare:
• Antibiotics in disadvantaged animals
• information on cat colds
• Others have an international approach:
• swedish acne remedy
• mrsa chinese formula
11. Professional vs public Needs: Media
Influence or Outbreak Early Warning
3 Datasets:
Professional Information Needs: The
NeLI/NRIC webserver logs
Public Information Needs: The Google
Trends data
Media coverage: Guardian Open Platform
API
12.
13.
14. Knowledge and attitude Change: Pre and Post Questionnaires: Bugs and
14
Drugs Site for Public
Pre and Post Use Questionnaires
– Used to evaluate differences in patient knowledge and attitudes
before and after using the site
– Users were free to browse the site between questionnaires
– First study took place in the Science Museum London as part of
‘live science’
– 227 visitors took part of which 177 completed both questionnaires
– Study repeated at Nottingham City Hospital Open Day and also
Oxford University Medical School
18. FEM wiki – social networks and training
wiki resource for epidemiologists
18
19. FEM wiki Launch 2010 and beyond
300+ users registered at
ESCAIDE event.
Now: over 600 users
A red bar represents a single user’s period of activity.
19
32. Edugames4all
DG SANCO funded project to develop web games teaching pupils
about microbes, hand and respiratory hygiene, and appropriate use
of antibiotics
33. Project Background
European wide, DG SANCO funded, antibiotic
and hygiene teaching resource for junior and
senior school children.
17 countries are involved covering 62% of the
European population.
34. Learning Objectives…
Microbes are all shapes
and sizes:
Microbes exist in various locations:
(inside the body)
(in the kitchen)
(on the skin’s surface)
35. Teaching by playing:
Good and Bad Microbes
• Player transports lactobacillus to make yogurt
DEMO
36. Teaching by playing:
Food Hygiene
The food sorting game is used to show
hand, respiratory and respiratory hygiene.
DEMO
37. Antibiotics and Vaccine Use
• Player throws white blood cells
• Player delivers full course of antibiotics to
infection
• DEMO
38. Junior Game Evaluation:
Statistically significant knowledge gain
Level 3, Q1: “We use good microbes to
make things like bread and yogurt” (p <
0.001, chi = 14.46)
Level 1, Q1: “If you cannot see a microbe it
is not there” (p = 0.02, chi =5.60)
Level 2, Q2: “Soap can be used to wash
away bad bugs” (p =0.02, chi=5.28)
39. Learning through Mechanics or
Text? - using these LO:
Statistically significant
◦ LO1: “Soap can be used to wash away bad
bugs”
◦ LO3: “We use good microbes to make things
like bread and yogurt”
Statistically not significant
◦ LO2: “Our bodies have nature defences that
protect us”.
◦ LO4: “All microbes and bad for us”.
40. Demographics
60 children
Age: 6-11 years old
Recruited on a voluntary basis
30 children in each group
Group 1:
◦ Text taught LO1 and LO2
◦ Mechanics taught LO3 and LO4
Group 2:
◦ Text taught LO3 and LO4
◦ Mechanics taught LO1 and LO2
41. Results
The results do not show a statistical significant for
either method of teaching the LOs – delivering
each LO remain subjective to the topic
“We use good microbes to make things like bread
and yogurt” was the only statistically significant
LO and the one where mechanics was more
successful than teaching through text. Also, it was
the most successful in the large evaluation using
combined text+mechanics approach - > this
mechanics works
Work in progress: larger studies, more LOs and
combination of methods
42. Seamless Evaluation Framework
LOs Scoring
Plot
Outline
Quest
Set
Conversa
tion
Nodes
Assets
Rules
Story
Rules
Game
Mechanics
Conversati
on Rules
Animation
s
+
+
+
+
Education
Layer
Mission
Layer
Quest
Layer
Conversation
Layer
+ Presentation
Layer
43. Study: Integrated assessment vs
external assessment
49 kids age 6 – 13
Two groups:
◦ Integrated assessment: How to be a Millionaire Show?
◦ Non integrated assessment: survey in the game
44. Results
Game perceived as
“fun” (p=0.34, CI=95%)
“having fun by playing the game” (p= 0.27,
CI=95%)
“nice to play” (p=0.65, CI=95%)
45. Results continued
63% with integrated evaluation preferred it
20% reported that the integrated evaluation increased
their game experience
30% said game more interesting, another 30% did not
affect them in any way, 15% stated the quiz did not
affect their game experience but prefer not to have it
integrated in the game
5% do not want the quiz as part of the game
60% did not notice that their knowledge was evaluated
through the integrated evaluation.
46. IDS Detective Game
Interactive Digital Storytelling Game
Target group: senior school children (12-16 years)
Several Missions
◦ Bad Bacteria at BBQ – importance hand hygiene
◦ When Bugs go Wild – prudent antibiotics use
◦ Gambling Never Pays - prudent antibiotics use
54. Seamless Evaluation – the Evaluation
Aims to assess:
Player’s perception of the assessment (21 participants – that did
not leave the survey incomplete)
Effectiveness of the game in delivering the LOs (145 participants
– participants who finish playing the game)
Participants:
A mix of persons playing the game online and in a
controlled environment
First the persons played the game and then they were
asked to fill a survey
55. SE Evaluation
Players perception of the integrated assessment:
Preferred assessment method: 94% preferred the evaluation
integrated in the game flow
56. Training mission: “Nathan – clean
up the locker room!”
Based on “when bugs go wild”
mission
Nathan, non-player character,
failed to clean the locker room
Dirty socks in footballers’ cocker
room
Nathan claims” these are clean but
smell off “deodorant”
-> microbision shows presence of
microbes
E-Copter lab tests shows these are
fungi
Nathan admits the failure and
clears the room
57. Quantifying it: Usability Evaluation
Evaluation in a “London School”
15 kids participated
◦ 8 played with training mission
◦ 7 without training mission
◦ Choosing non-control evaluation by selecting only
those who declared that they have never played IDS
type of game would demonstrate improvement but
we wanted to know “how much” and is it “usable”
◦ Small numbers: practical issues in working at schools,
difficulties engaging enough target group children,
asking the right questions etc.
58. Usability Evaluation
SUS (system Usability
Scale) adapted to games
system -> game
use -> play
Results: avg SUS score
Tutorial group = 61.25
(δ=8.95)
Non-Tutorial group = 60
(δ=19.94)
Not statistically significant
p=0.84) considering a
confidence interval of
95%.
# Question p
1 I think I would like to play this game frequently.
0.78
2 I found the game unnecessarily complex.
0.18
3 I thought the game was easy to play.
0.20
4 I think I would need the support of a teacher or other
expert to be able to play this game.
0.47
5 I found various functions in this game were well
integrated.
0.13
6 I thought there was too much inconsistency in this
game.
0.37
7 I would imagine that most people would learn to play
this game very quickly.
0.45
8 I found the game very awkward to use.
0.04
9 I felt very confident playing the game.
0.13
10 I needed a lot of help before I could get to play this
game.
0.16
59. SUS Results Comparison
The only statistically significant difference is for question
eight – I found the game awkward to use (p=0.04) for a
confidence interval of 95%.
60. Summary
Training mission needed to reduce drop out
rates in children unfamiliar with the IDS
games
It worked: 1st evaluation confirmed
satisfaction, were not put off, ease of use
2nd evaluation confirmed
◦ USABILITY with/without training similar (SUS
score)
◦ Kids who played with training found the game less
awkward to use
NOT SURPRISING RESULTS – evidence we
achieved our goal
62. Number of website visitors, page views and
resource downloads per month since January 2008
63. Media Coverage
Invited Speaker – “Idea Champion” at BMJ Panel:
◦ "The idea most likely to make the biggest impact on healthcare by 2020" (the NHS Innovation
Expo 2011, Excel, London, March 2011)
◦ ECDC Invited Speaker European meeting on “Epidemic Intelligence” (Stockholm, October 2010)
BMJ feature
◦ Can Twitter predict disease outbreaks?
◦ BMJ 2012; 344 doi: 10.1136/bmj.e2353 (Published 17 May 2012
Media Coverage
◦ “Epidemic Inteligence”Scientific Film
◦ BMJ: Medical Innovation: Sabreena Malik
◦ Social networking sites can help prevent pandemics:
(AFP) (15th December 2010) and replicated across more than
30 online news outlets
63
“Patty Kostkova and her colleagues at City ehealth Research
Centre, City University, London, showed that the 2009 H1N1 flu
outbreak could have been identified on Twitter one week before
it emerged in official records from general practitioner reports
“E-Bug uses
interactive
games to teach
children about
infection control”
64. CeRC staff – THANKS!
CeRC
Gawesh Jawaheer
Gayo Diallo
Sue Wiseman
David Farrell
Steve D’Souza
Gemma Madle
Jane Mani-Saada
Anjana Roy
Julius Weinberg
Mike Catchpole
Faiza Hansraj
Nancy Lai
Sandy Beverage
John Lawrenson
Martin Szomszor
Lisa Lazareck
DasunWeerasinghe
Ros Nyugi
Christina Daly
Ed de Quincey
Helen Oliver
David Fowler
Simon Hammond
UCL: Stephan Garbin
Justin Moser, Wendy Pan
64
65. Contact
Thank you to CeRC, ISI Foundation and
UCL
p.kostkova@ucl.ac.uk
Sites:
◦ www.neli.org.uk
◦ www.edugames4all.org
◦ www.nric.org.uk
◦ www.femwiki.com (hosted by ECDC)