Emergency Medical Services
Surveillance in Toronto……and Beyond
Kate Bassil, PhD
June 13, 2008
QPHI Meeting
Outline



Part 1: EMS Data and Surveillance
Part 2: EMS Surveillance in Toronto
for heat-related illness
Part 3: Future directions,
opportunities for collaboration
Part 1:
EMS Data and Surveillance
Advantages of Using EMS Data for
Surveillance

Timeliness: in the capture and process of the data

Simplicity: use of pre-existing data



Acceptability: willingness of stakeholders to
contribute to data collection and analysis
Portability: system could be duplicated in another
setting
Cost: could be done with no significant software or
hardware requirements
Surveillance principles from Public Health Agency of Canada and CDC Surveillance Principles
Added value



EMS data provides geospatial information
about the location where the individual has
become ill.
Differs from many other traditional medical
data sources that use place of residence.
Important for syndromes where place
matters e.g. outdoor recreation areas like
heat illness.
NYC EMS Surveillance






NYC 911 receives 1 mill calls/year
Implemented in 1998
Particularly useful for ILI and HRI
surveillance
ILI codes: RESP, DIFFBR, SICK, SICPED
Use up to 3 years of baseline data
Alarm generated when the ILI rate
exceeds the upper confidence limit
Mostashari F, et al. 2003. Use of ambulance dispatch data as an early warning
system for communitywide influenza like illness, New York City. J Urban Health 80:i43-i49
World Youth Day
July 12-28, 2002: Canada

Biannual international celebration of the
Catholic Church

For youth ages 17-35 years old

Past attendance 2 million +

Days in the Dioceses across Canada followed by
the major celebrations in Toronto (overall, a 10day event)
Findings from EMS Data
World Youth Day, 2002



From July 15-Aug 7 a total of 11,250 calls
were logged
39% met a syndrome definition
Most useful EMS call-code cluster was for
heat-related illness (HRI)
Part 2:
EMS Surveillance in Toronto
for Heat-related Illness
Health Impacts of Hot Weather
Klinenberg E. 2003. A Social Autopsy of Disaster in Chicago.
Heat-related illness (HRI)



Europe, 2003: > 70,000
excess deaths
Chicago, 1995: > 700 excess
deaths
Historical analysis of
Canadian cities:


Toronto: 120 annual heatrelated deaths
Projected that in the future
these values will more than
double by 2050 and triple by
2080
Pengelly LD, et al. Anatomy of heat waves
and mortality in Toronto: Lessons for
public health protection. Can J Public
Health. 2007 Sep-Oct;98(5):364-8.
Mortality
Severity
of Effect
Hospital
admission
Medical seeking behaviour:
ER, physicians office, 911,
Telehealth, clinic
Heat cramps, heat exhaustion, heat stroke
Mild symptoms, discomfort, subtle effects
Proportion of Population
Canadian Urban Areas






Urban Heat Island
Continued urbanization
Vulnerable population
Aging population
Lack of acclimatization in temperate
zones
Future projections of increasing
temperature means and variance
Urban Heat Island Profile
Natural Resources Canada
http://adaptation.nrcan.gc.ca/perspective/health
Temperature Trends, Toronto
Toronto Annual Temperature
(1878-2005)
Temperature (°C)
16
Maximum
12
Mean
8
Minimum
4
0
1878
1898
1918
1938
Year
Environment Canada. 2006.
1958
1978
1998
Heat Health Warning Systems (HHWS)

“A system that uses meteorological forecasts
to initiate acute public heath interventions
designed to reduce heat-related impacts on
human health during atypically hot weather”
Koppe C, et al. Heatwaves: impacts and responses. Copenhagen: World
Health Organization, 2003.


Surprisingly few countries and cities have a
HHWS
Implementation of interventions at
municipal/national level
Heat Interventions
Bassil et al. 2007. What is the evidence on applicability and effectiveness of public health interventions in reducing morbidity and
mortality during heat episodes? A review for the National Collaborating Centre in Environmental Health.
Gaps


Heat/Health Warning Systems are
based on mortality….what about
indicators of morbidity (e.g.
syndromic surveillance sources)?
Interventions are not currently
targeted geographically
Toronto Emergency Medical Services (EMS):
Communications Centre

Single-provider EMS system

Annual call volume - approx. 425,000

Fully computerized system

Uses the Medical Priority Dispatch System
(MPDS), a widely used EMS call sorting
algorithm, to classify calls.
Aren’t The Paramedics Coming?
Emergency
Call
Address
Police & Fire may be sent as well
Unit
Assigned
Situational
Information
Patient
Information
Unit
Responds
Paramedics
Arrive
Other
Location
Information
Mark Toman, Toronto Emergency Medical Services
Patient Care
Instructions
Patient
goes to
Hospital
MPDS Code Categorization
Entry Questions
Key Questions:
1. Is s/he completely awake?
2. Is s/he breathing normally?
3. Is s/he changing colour?
4. What is her/his skin temperature?
Dispatch Codes:
20-D-1 Heat/Cold Exposure, not alert
20-C-1 Heat/Cold Exposure, cardiac history
20-B-1 Heat/Cold Exposure, change in skin colour
20-A-1 Heat/Cold Exposure, alert
Medical Priority Dispatch System, Priority Dispatch Corp., Salt Lake City, Utah
EMS Variables in Data Set
Data Variable
RMI_ID
Description
Unique identifier for each
call
RMI_MPDSDeterminant Full MPDS determinant
code
RMI_ResponseDate
Time of the call
RMI_Location_Type
Kind of location of the call
RMI_Call_City
Municipality in which call
pick-up is located
Latitude of the call location
Longitude of the call
location
Additional text information
about the call
RMI_Call_Latitude
RMI_Call_Longitude
Comment
RMI: “Response Master Incident”
Example
854700
06C03
3/8/2002 1:39:39
PM
Park/Playground,
Firehall
Toronto
43.7969
-79.3133
M 81 SOB PALE
CLAMMY
Toronto EMS Data




Daily call information for all emergency calls to
EMS between 2002-2005 (approximately
850,000 calls)
Excludes cancelled calls and inter-facility
transfers.
Microsoft Access database format
Quality assurance of MPDS assignation – 98%
agreement, call assignation to US National
Academy of Emergency Medicine standards
Emergency calls only
1000
900
800
700
600
500
400
300
200
100
0
2004
8/1/2003
9/1/2003
10/1/2003
11/1/2003
12/1/2003
8/1/2005
9/1/2005
10/1/2005
11/1/2005
12/1/2005
1000
900
800
700
600
500
400
300
200
100
0
7/1/2003
2005
7/1/2005
6/1/2003
2002
6/1/2005
5/1/2003
4/1/2003
3/1/2003
2/1/2003
1/1/2003
Number of Calls
1000
900
800
700
600
500
400
300
200
100
0
5/1/2005
4/1/2005
3/1/2005
2/1/2005
Number of Calls
12/1/2002
11/1/2002
10/1/2002
World Youth Day
1/1/2005
12/1/2004
11/1/2004
10/1/2004
9/1/2002
8/1/2002
7/1/2002
6/1/2002
5/1/2002
4/1/2002
3/1/2002
2/1/2002
1/1/2002
Number of Calls
1000
900
800
700
600
500
400
300
200
100
0
9/1/2004
8/1/2004
7/1/2004
6/1/2004
5/1/2004
4/1/2004
3/1/2004
2/1/2004
1/1/2004
Number of Calls
Number of All EMS Calls, Toronto, 2002-2005
2003
Rolling Stones Concert
Blackout
Defining HRI with EMS Data
Unknown trouble (man down)
Most sensitive
C
A
L
L
V
O
Sick person
Cardiac
Abdominal pain
Unconscious/fainting
L
U
Headache
M
Most specific
E
Heat/cold exposure
Developing the case definition
i) Clinical process:
- Approx 500 medical dispatch call categories
reviewed.
- Series of expert focus groups
ii) Empirical process:
-Each call category was assessed graphically with
daily mean temperature
- 4 groups of call categories were selected as ones
which may represent HRI:




Heat/cold exposure,
Breathing problems,
Unconscious/fainting,
Unknown problem/“man down”
MPDS Card: Heat/cold exposure
Solid line: proportion of calls
Dotted line: Daily average temperature
MPDS Card: Unknown problem/”man down”
Solid line: proportion of calls
Dotted line: Daily average temperature
Call categories that most clearly
represent HRI
20D01
20C01
20B01
20B02
20A01
Heat/Cold exposure- Not alert
Heat/Cold exposure – Cardiac history
Heat/Cold exposure – Change in skin colour
Heat/Cold exposure – Unknown status (3rd party caller)
Heat/Cold exposure - Alert
1.40
40
1.20
35
1.00
30
25
0.80
20
0.60
15
0.40
10
0.20
5
0.00
0
6/1
6/8 6/15 6/22 6/29 7/6 7/13 7/20 7/27 8/3 8/10 8/17 8/24 8/31 9/7 9/14 9/21 9/28
Date
Proportion
Max temp
Mean temp
Temperature (c)
% Heat-related 911 calls
Heat Illness Calls & Temperature, Toronto, Summer 2007
Time Series Analysis



On average, for every one degree increase
in mean or maximum temperature there
was a 30% increase in EMS calls for HRI
(p<.0001).
Lag effect of 1 day (ranged from a 7 to
18% increase in calls for max temp,
p<.0001)
Ozone: positive but statistically insignificant
Public Health Challenges




Technical issues – several days when data
was not sent, so occasionally sent in
batches every few days.
Timing with current heat health warning
system
Requires daily person time – not a fully
automated system
Limited demographic information
Public Health Advantages



Additional data source to support
decisions around declaring heat alerts
New geospatial information to assist
in intervention targeting
Situational awareness
Part 3:
Future directions
Future Work

Further exploration of call codes


Breathing, fainting
TEMS data for other syndromes:


Cold-related illness
Influenza-like illness
Future Work

Multi-city study:




Niagara, Vancouver, Montreal, Toronto
Focus on EMS and geospatial analysis
Emergency department data
Vulnerability assessment:


Multiple data sources (EMS, ED)
Focus on heat-related illness
Acknowledgements




Toronto Public Health – Effie Gournis, Elizabeth Rea,
Marco Vittiglio, Eleni Kefalas
University of Toronto – Donald Cole, Wendy Lou,
Rahim Moinnedin
Toronto EMS – Dave Lyons, Alan Craig
Sunnybrook Basehospital – Brian Schwartz,
Sandra Chad
Comments and questions?
For more information:
kate_bassil@sfu.ca