Ambulance call-outs and response times in Birmingham and the impact of extreme weather and climate change

Transcription

1 Ambulance call-outs and response times in Birmingham and the impact of extreme weather and climate change John Edward Thornes, 1,2 Paul Anthony Fisher, 3 Tracy Rayment-Bishop, 4 Christopher Smith 4 1 Health Protection Agency, Centre for Radiation, Chemical and Environmental Hazards, Oxfordshire, UK 2 Department of Geography, University of Birmingham, Birmingham, UK 3 Real-time Syndromic Surveillance Team, Health Protection Agency, Birmingham, UK 4 West Midlands Ambulance Service, NHS Trust, Regional Ambulance Headquarters, West Midlands, UK Correspondence to Professor John Edward Thornes, Health Protection Agency, Centre for Radiation, Chemical and Environmental Hazards, Chilton, Didcot, Oxfordshire OX11 0RQ, UK; john.thornes@hpa.org.uk, j.e.thornes@bham.ac.uk Received 13 August 2012 Revised 6 December 2012 Accepted 3 January 2013 Published Online First 28 February 2013 To cite: Thornes JE, Fisher PA, Rayment- Bishop T, et al. Emerg Med J 2014;31: ABSTRACT Although there has been some research on the impact of extreme weather on the number of ambulance call-out incidents, especially heat waves, there has been very little research on the impact of cold weather on ambulance call-outs and response times. In the UK, there is a target response rate of 75% of life threatening incidents (Category A) that must be responded to within 8 min. This paper compares daily air temperature data with ambulance call-out data for Birmingham over a 5-year period ( ). A significant relationship between extreme weather and increased ambulance callout and response times can clearly be shown. Both hot and cold weather have a negative impact on response times. During the heat wave of August 2003, the number of ambulance call-outs increased by up to a third. In December 2010 (the coldest December for more than 100 years), the response rate fell below 50% for 3 days in a row (18 December 20 December 2010) with a mean response time of 15 min. For every reduction of air temperature by 1 C there was a reduction of 1.3% in performance. Improved weather forecasting and the take up of adaptation measures, such as the use of winter tyres, are suggested for consideration as management tools to improve ambulance response resilience during extreme weather. Also it is suggested that ambulance response times could be used as part of the syndromic surveillance system at the Health Protection Agency. INTRODUCTION Research into the impact of climate change and extreme weather on human health in the UK has primarily focused on predicted changes in morbidity and mortality. 1 The impact on NHS infrastructure and ambulance services has not yet been examined in any detail in the UK and this paper attempts to help redress this balance by assessing the impact of extreme weather on ambulance response times in Birmingham. It is hoped that useful recommendations to help ambulance trusts to adjust to extreme weather and adapt to climate change can be made. As well as suggesting operational improvements for ambulance trusts using real time weather forecast information, it might also be possible, in the future, to use real time ambulance response data to feed back timely emergency warnings. For example, the number and type of ambulance emergency calls could provide valuable early morbidity information for heat- or cold-related illnesses or water borne diseases during floods. This could allow useful additional time for appropriate interventions and could be built in to the current Health Protection Agency syndromic surveillance system. Within England there are 12 National Health Service (NHS) organisations that provide ambulance services and more than 8 million emergency call-outs were received in 2010/11, of which more than 80% required an emergency response. From April 2011 these calls were placed into two categories: Category A: Immediately Life Threatening (Target: 75% within 8 min) Chest pain with, for example: abnormal breathing; not alert; or cardiac history Unconscious/fainting: not alert; ineffective breathing; or multiple episodes Breathing problems: severe respiratory distress; not alert; or ineffective breathing All Other Calls: Not Life Threatening In there were 1.67 million Category A incidents with a response rate, arriving at the scene of the incident, of 74.9% within 8 min. The total number of emergency patient journeys was 4.29 million and 1.76 million patients were treated at the scene. 2 The total cost of the NHS ambulance service is close to 2 billion of which about 1.5 billion is spent on emergency services and the rest on ambulatory (prearranged) services. The average cost per emergency incident is between about 200 and On a given day a number of factors can affect ambulance response times including: Number of ambulances available Number of staff members available Congestion in hospital accident and emergency departments increasing turnaround times Road conditions and the weather Call volume (eg, increase due to swine flu in July 2009) Rurality (average distance to patient). The variability in response times from day to day will primarily be driven by changes in call volume, the weather and consequent road conditions. The other factors are mostly constant and can be planned for in advance. Ambulance services normally plan up to several weeks ahead for the likely number of ambulance call-outs, taking into account call-out figures for the same week in the previous year, the season, the day of the week, staff holidays, national holidays and so on. Very little research has been published on the impact of extreme weather on ambulance call-outs and response times. 4 There is a growing literature on the relationship between temperature and the number of ambulance emergency call-outs during 220 Thornes JE, et al. Emerg Med J 2014;31: doi: /emermed

2 heat waves 5 10 but there is a lack of data on the impact of coldwaves. 11 Nitschke et al show 12 that during the extreme heat waves of 2008 and 2009 in Adelaide, ambulance call-outs increased by 10% and 16%, respectively, compared with 4.4% during previous heat waves, suggesting that the climate is changing and becoming more extreme in summer. Wong and Lai 13 show that in Hong Kong there is a predictable relationship between temperature and the number of ambulance call-out sand conclude that real time, short term weather forecasts would help to manage extreme weather events. A University of Sheffield report 14 examined the health benefit of the 8 min response time. For out-of-hospital cardiac arrest (around 1% of 999 calls) not witnessed by an ambulance crew, they found that with other factors taken into account a 1 min reduction in response time was estimated to increase the odds of survival by 24% which could save approximately 149 lives per year in England and Wales. They also found that the arrival of the ambulance crew before cardiac arrest increases the chance of survival sevenfold. Pell et al 15 conclude that reducing ambulance response times from 8 to 5 min could almost double survival rates for cardiac arrests not witnessed by ambulance crews. However, the cost of additional ambulances and staff would no doubt be prohibitive. A Canadian study 16 found a steep decrease in the first 5 min of the out of hospital cardiac arrest survival curve with the odds of survival decreasing by 0.77 for each additional minute. A number of other studies have also reported improvements in survival when the cardiac arrest is witnessed by paramedics therefore supporting the need for a rapid ambulance response Variations in cardiac arrest survival rates between ambulance trusts are discussed by Perkins and Cooke. 20 Hence if extreme weather increases ambulance call-outs and response times, this may adversely affect survival rates for cardiac arrest and other serious health problems such as strokes. More research is required to assess the impact of extreme weather on these and other Category A call-out incidents. Excess winter mortality and illness are mostly caused by diseases linked to circulation problems causing heart attacks and strokes. Respiratory illness can also be exacerbated by cold weather. There is often a lag between the coldest day and the peak numbers of heart attacks (2 days), strokes (5 days) and respiratory problems (up to 12 days). It can take up to a month for death rates to return to normal after a cold period of weather Excess summer mortality and illness, especially in heat waves, are mostly respiratory and cardiovascular diseases. There are a number of additional heat-related illnesses including heat cramps, heat rash, heat oedema, heat syncope, heat exhaustion and heat stroke. 23 Any lag between the hottest temperatures and mortality and morbidity is normally less than 5 days. 11 It is not surprising that the number of ambulance call-outs and response times will increase when extreme weather occurs. Vulnerable groups include: old people; those with chronic and severe illnesses; those who struggle to adapt to cold or hot weather for example, because of Alzheimer s, a disability, being bed-bound, being very young, are homeless or have an outdoor job. Even fit and healthy individuals can become vulnerable if the weather is extreme enough. However, there is very little quantitative research that enables accurate estimates of likely increases in call volume and subsequent increases in response times, depending upon the severity of the weather. AMBULANCE CALL-OUTS, RESPONSE TIMES AND EXTREME WEATHER IN BIRMINGHAM This study, as well at looking at the number of daily emergency ambulance calls in Birmingham, also examines the daily average ambulance response times. Due to the epidemiological challenges from the large number of factors that influence ambulance response time detailed above, this study examines the association between factors rather than direct causality. Response times are also directly affected by difficult road conditions caused by extreme weather such as ice, snow, fog, gales and flooding. Currently there is very little liaison between local authorities responsible, for example, for salting the roads and ambulance trusts. Response times are also affected by the total number of emergency calls made on an individual day. During a heat wave or cold wave, for example, the total number of emergency ambulance calls significantly increases and response times suffer as a consequence. Accurate timely weather forecasts may help to predict such events in advance. Daily ambulance response data for Birmingham have been supplied by the West Midlands Ambulance Trust and the data for the dates 1 April November 2011 have been analysed (1705 days including emergency calls of which were Category A). These data have been compared with daily climate data extracted from the University of Birmingham Weather Station and Birmingham Airport (snow) archives for almost the same period (1 April August 2011). Clear weather-related signals are present in the ambulance response data as discussed below. For this study, the Birmingham ambulance response data analysed comprise four different metrics for each day: total number of 999 calls for ambulance services; total number of Category A (life threatening) calls; the percentage of Category A calls responded to within 8 min; and the average travel time to arrive at 999 incidents. (Several other metrics are archived such as the percentage of 999 incidents taken to hospital and average turnaround time but are not analysed here. Limited call-out data for the heat wave of August 2003 were also made available). Each ambulance trust collects similar standardised data in real time but changes in data classification (meta data) need to be accounted for. Also daily data (1 July August 2011) for the total daily coded medical condition (36 categories for example, breathing problems, cardiac arrest) have been provided and compared with the daily weather data (781 days). The accuracy of the initial pathology of the patients, made on the spot by ambulance crews, does mean that there are likely to be a small proportion of incorrect diagnoses. Nevertheless, by aggregating diseases it is possible to get a good idea of the main morbidity classes from the 999 ambulance calls. Table 1 shows the percentage of identified aggregated call-out morbidity classes for each season over the 2-year period. The seasonal differences are not great but winter is marginally worse for most categories. The Table 1 Seasonal breakdown (%) of Birmingham calls to West Midlands Ambulance Service ( ) in comparison with December 2010 and 17 and 22 December 2010 Breathing Chest Fall Sick Unconscious Other Spring Summer Autumn Winter December December December Thornes JE, et al. Emerg Med J 2014;31: doi: /emermed

3 Figure 1 Total daily 999 ambulance calls 1 April November cold month of December 2010 (the coldest December for over 100 years with a mean temperature of 1 C) clearly shows increased breathing problems and falls and also an increased number of sick patients. Due to widespread snow and ice on the 17 December the number of falls rose dramatically to 23.8% but the peak in breathing difficulties was 5 days later on the 22 December when 17.2% of calls were breathing-related and 16.5% were related to falls. The climate data initially comprise of daily minimum, mean and maximum air temperature plus the number of days with snow falling. From a climate point of view, this time period (April 2007 to November 2011) contains a range of extreme weather including the floods of July 2007 and the very cold Decembers of 2009 and 2010 (however, the exceptional rainfall in the summer of 2007 did not cause a major flooding problem in the Birmingham area). Figure 1 shows that there is a gradual rise in the daily number of ambulance call-outs in Birmingham from an average of just over 400 per day in 2007 to just over 500 per day in This represents an increase of 23% over a period of less than 5 years. This increase in calls is caused by a range of factors (eg, reduced opening times of doctor s surgeries and increased availability of mobile phones 24 ) of which the weather is just one. The number of calls was clearly affected by extreme weather, however, with significant peaks in December 2009 (the coldest December since 1995) and December 2010 (the coldest December since 1890). Figure 2 shows the number of life threatening Category A calls for a similar time period (up to 20 September 2011) showing a comparable clear cold weather signal in the data. Figure 3 displays the average daily ambulance response times with the highest peak of over 20 min during the swine flu pandemic of A total of 607 calls (a 29% increase compared with the 2009 daily average of 472) were received on the 2 July of which 214 were considered life threatening (Category A). In December 2010, the response rate fell below 50% for 3 days in a row (18 December to the 20 December 2010) with a mean response time of 15 min. The air temperature fell to 13.5 C on the 20th December. For every reduction of air temperature by 1 C there was a reduction of 1.3% in performance. Thus, the cold weather of December 2010 had almost as big an impact as the swine flu pandemic of July Figure 4 and 5 examine the number of patients classified as having Breathing Difficulties as the reason for calling the ambulance. Figure 4 shows that these numbers peak in cold winter weather and figure 5 shows that there is a clear inverse relationship between breathing problems and the daily mean temperature. Figure 6 and 7 show the relationship between Falls and Back Injuries and the daily mean temperature. As the mean daily temperature goes below zero, so the number of Falls and Back injuries increases sharply. Unprecedented ice and snow on and before the 17 December 2010 led to 745 emergency 999 calls (50% increase on November 2010 average) of which 177 related to falls and back injuries (150% increase on November 2010 average). 222 Thornes JE, et al. Emerg Med J 2014;31: doi: /emermed

4 Figure 2 Total daily Category A incidents 1 April November Figure 3 Daily average travel time in minutes to Category A incidents 1 April November 2011 (target 8 min). Thornes JE, et al. Emerg Med J 2014;31: doi: /emermed

5 Figure 4 Number of ambulance patients with breathing problems. Figure 5 Daily number of ambulance patients with breathing difficulties versus daily mean temperature in Birmingham Thornes JE, et al. Emerg Med J 2014;31: doi: /emermed

6 Figure 6 Daily number of ambulance patients with falls/back injuries traumatic Ambulance chiefs today took the unprecedented step of urging people not to travel unless they have to, following a sharp rise in the number of 999 callouts. West Midlands Ambulance Service spokesman said: A huge number of these calls are reports of slips, trips and falls as people venture out into the snow and onto icy conditions. Clearly many of these could have been avoided if the journey was not essential. (Birmingham Mail 17 th December 2010) Figure 8 shows that during the 2003 heat wave in Birmingham there was a positive linear relationship between the number of 999 ambulance calls and the daily maximum air temperature. On the hottest day (9 August 2003), the air temperature reached 33 C and there were 568 emergency calls compared with the July 2003 average of 420 which represents a 35% increase. (Response times are not available for 2003 but it is most likely that response times also suffered in the heat.) Figure 9 clearly shows that there is an inverse U-shaped relationship between daily mean temperature and the percentage of Category A calls achieved within 8 min (national target is 75%). Both hot and cold weather can cause significant delay problems for the ambulance service. After fitting a polynominal best fit line to the data, the temperature associated with the best response time (peak of the curve) was 5.05 C. Figure 7 Daily number of falls versus minimum daily temperature degrees Celsius. Thornes JE, et al. Emerg Med J 2014;31: doi: /emermed

7 Figure 8 Number of daily 999 ambulance calls versus daily mean maximum temperature 1 27 August 2003 Birmingham. Figure 10 (supplied by West Midlands Ambulance Trust) shows that there is a clear inverse relationship between the number of emergency calls and the percentage of Category A calls reached within 8 min. Days with snow falling clearly result in more emergency calls and coupled with adverse road weather conditions this leads to a delayed emergency response. Figure 11 shows that for the period April 2007 to December 2010, the Category A 75% target was met but that a seasonal breakdown shows the lowest response rate in winter and summer. DISCUSSION Nearly all of the health studies using ambulance data have been looking at the impact of heat waves but this study has shown that cold waves can also significantly affect ambulance call-out requirements and response times. The average daily number of call-outs in December 2010 was 590 compared with 495 in November 2010 which represents a 20% increase. Nitschke et al 12 found an increase of 10% and 16% for the heat waves in Adelaide, Australia, in 2008 and There are no comparable published cold weather call-out data but Turner et al working in Brisbane 11 and Wong and Lai 13 working in Hong Kong found a much smaller cold weather impact, which is not surprising as winter temperatures in Brisbane and Hong Kong are well above freezing. This study also suggests that ambulance daily coded medical condition and increased ambulance call-out and response data could be a valuable early indicator of the impact of extreme weather on health in a region. These results are comparable with findings in other countries, for example, in Australia, where it is also recognised that ambulance data can give a useful early outlook: In terms of public health policy, examining ambulance attendances rather than admission or mortality data would potentially help to pick up early signs of temperature effects and also the Figure 9 Percentage of daily Category A responses within 8 min versus mean daily temperature. 226 Thornes JE, et al. Emerg Med J 2014;31: doi: /emermed

8 Figure 10 Daily number of incidents versus Category A % performance with snow days at Birmingham airport marked with a cross. effects on health conditions that cannot be examined through analysis of mortality and hospital admissions data. 11 Recent research in Italy also confirms the possible use of ambulance data for syndromic purposes: In conclusion, the use of emergency ambulance dispatches as an indicator of health effects of heat episodes other than mortality is important in order to achieve a wider view of the effects of biometeorological discomfort on human health and may be suitable to establish real-time surveillance systems. 24 CONCLUSIONS This is the first study of its type in the UK and has clearly shown that extreme weather has an adverse impact on the number of emergency call-outs and the ambulance response times in Birmingham. The increase in ambulance response times is both related to the increased number of emergency calls in adverse weather and, in winter, the adverse road conditions encountered. Better prediction of adverse weather and road conditions should be used in real time by ambulance trusts to better plan ambulance deployment. Although some factors such as road conditions are likely to be related to the temperature on a given day, other factors, such as call volume may exhibit a lag (health impacts may result a few days later in summer and potentially a few weeks later in winter) which will be factored in future work. 25 Future work will also examine how the type and spatial location of call-outs varies between seasons and during cold and heat waves. Information on call volume and type during the early, middle and later stages of a heat wave/cold wave would aid understanding of the health effects and allow for more refined early warning systems Potential factors that impact on call volume could be controlled for, Figure 11 Percentage of Category A responses in 8 min or under by season. Thornes JE, et al. Emerg Med J 2014;31: doi: /emermed

9 including the number of ambulances on duty on a given day and other factors such as air pollution. The impact of climate change will primarily be seen through more incidents of extreme cold and hot weather and therefore it will be useful to examine adaptation measures such as the use of winter tyres and more flexibility in the number of ambulances available for operation during extreme weather episodes. Furthermore, with colder winters and hotter summers the challenges associated with achieving summer/winter response times, particularly due to increases in call volume, are likely to increase. A trial in the use of daily Birmingham ambulance call-out numbers, response times and the coded medical condition is currently being carried out with the Health Protection Agency Real-time Syndromic Surveillance Team based in Birmingham. Contributors JET had the initial idea for the research and wrote the first draft. PAF contributed to the text and helped to edit the text. TR-B and CS provided the ambulance data and commented on the text. Funding The study was funded by the Health Protection Agency and carried out within the Air Pollution and Climate Change Group. Competing interests None. Provenance and peer review Not commissioned; externally peer reviewed. Data sharing statement The daily weather data for Birmingham are available from JET, the corresponding author. REFERENCES 1 Hames D, Vardoulakis S. Climate change risk assessment for the health sector. Department for Environment, Food and Rural Affairs, uk/environment/climate/government/risk-assessment/ (Accessed 3 Dec 2012). 2 National Audit Office. Transforming NHS ambulance services. 2011, HC NHS Information Centre. Ambulance Services England Speakman D. Ambulances adrift: the impact of the snowstorms of the winter of 1990/91 upon the services of the City of Birmingham. Weather 1994;49: Dolney TJ, Sheridan SC. The relationship between extreme heat and ambulance response calls for the city of Toronto, Ontario, Canada. Environ Res 2006;101: Cerutti B, Tereanu C, Domenighetti G, et al. Temperature related mortality and ambulance service interventions during the heat waves of 2003 in Ticino (Switzerland). Soz Praventivmed 2006;51: Wolf T. Integrated assessment of vulnerability to heat stress in urban areas (PhD unpublished). Kings College London, Bassil KL, Cole DC, Moineddin R, et al. The relationship between extreme heat and ambulance response calls for the city of Toronto, Ontario, Canada. J Epidemiol Community Health 2011;65: Schaffer A, Muscatello D, Broome R, et al. Emergency department visits, ambulance calls, and mortality associated with an exceptional heat wave in Sydney, Australia, 2011: a time-series analysis. Environ Health 2012;11:3. 10 Williams S, Nitschke M, Sullivan T, et al. Heat and health in Adelaide, South Australia: Assessment of heat thresholds and temperature relationships. Sci Total Environ 2012;414: Turner LR, Connell D, Tong S. Exposure to hot and cold temperatures and ambulance attendances in Brisbane, Australia: a time-series study. BMJ Open 2012;2:e Nitschke M, Tucker GR, Hansen AL, et al. Impact of two recent extreme heat episodes on morbidity and mortality in Adelaide, South Australia: a case-series analysis. Environ Health 2011;10: Wong HT, Lai PC. Weather inference and daily demand for emergency ambulance services. Emerg Med J 2012;29: O Keeffe C, Nichol J, Turner J, et al. Role of ambulance response times in the survival of patients with out-of-hospital cardiac arrest. Emerg Med J 2011;28: Pell JP, Sirel JM, Marsden AK, et al. Effect of reducing ambulance response times on deaths from out of hospital cardiac arrest: cohort study. BMJ 2001;322: De Maio V, Stiell I, Wells G, et al. Optimal defibrillation response intervals for maximum out of hospital cardiac arrest survival rates. Ann Emerg Med 2003;42: Kette F, Sbrojavacca R, Rellini G, et al. Epidemiology and survival rate of out of hospital cardiac arrest in North East Italy: the F.A.C.S. study. Fruili Venezia Guilia Cardiac Arrest Co-operative Study. Resuscitation 1998;36: Herlitz J, Engdahl J, Svensson L, et al. Factors associated with an increased chance of survival among patients suffering out of hospital cardiac arrest in a national perspective in Sweden. Am Heart J 2005;149: Finn J, Jacobs I, Holman C, et al. Outcome of out of hospital cardiac arrest patients in Perth, Western Australia, Resuscitation 2001;51: Perkins GD, Cooke MW. Variability in cardiac arrest survival: the NHS Ambulance Service Quality Indicators. Emerg Med J 2012;29: Donaldson GC, Keating WR. Early increases in ischaemic heart disease mortality dissociated from, and later changes associated with, respiratory mortality, after cold weather in south-east England. J Epidem Comm Health 1997;51: National Health Service. Cold Weather Plan for England National Health Service. Heatwave Plan for England Wu O, Briggs A, Kemp T, et al. Mobile phone use for contacting emergency services in life-threatening circumstances. J Emerg Med 2012;42: Alessandrini E, Zauli Sajani S, Scotto F, et al. Emergency ambulance dispatches and apparent temperature: a time series analysis in Emilia-Romagna, Italy. Environ Res 2011;111: Vardoulakis S, Heaviside C, eds. Health effects of climate change in the UK. Chilton: Health Protection Agency, (Accessed 3 Dec 2012). 27 Thornes JE, Rennie M, Marsden H, et al. Climate change risk assessment for the transport sector. Department for Environment Food and Rural Affairs, (Accessed 3 Dec 2012). 228 Thornes JE, et al. Emerg Med J 2014;31: doi: /emermed

10 Ambulance call-outs and response times in Birmingham and the impact of extreme weather and climate change John Edward Thornes, Paul Anthony Fisher, Tracy Rayment-Bishop and Christopher Smith Emerg Med J : originally published online February 27, 2013 doi: /emermed Updated information and services can be found at: References alerting service These include: This article cites 19 articles, 7 of which you can access for free at: Receive free alerts when new articles cite this article. Sign up in the box at the top right corner of the online article. Topic Collections Articles on similar topics can be found in the following collections Press releases (32) Notes To request permissions go to: To order reprints go to: To subscribe to BMJ go to: