Emergency Medical Services Surveillance in Toronto and Beyond Kate

Emergency Medical Services Surveillance in Toronto……and Beyond Kate Bassil, Ph. D 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: i 43 -i 49

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 10 day 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: l l 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. Mortalit y Severit y 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 Environment Canada. 2006.

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.

Mark Toman, Toronto Emergency Medical Services

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 RMI: “Response Master Incident”

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

Number of All EMS Calls, Toronto, 2002 -2005 World Youth Day Rolling Stones Concert Blackout Emergency calls only

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 Most specific Headache M 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: l l 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

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 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 l ¡ Breathing, fainting TEMS data for other syndromes: l l Cold-related illness Influenza-like illness

Future Work ¡ Multi-city study: l l l ¡ Niagara, Vancouver, Montreal, Toronto Focus on EMS and geospatial analysis Emergency department data Vulnerability assessment: l l 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