The Population Health Impacts of Heat Key Learnings from the Victorian Heat Health Information Surveillance System 2011–2013

The population health impacts of heat Key learnings from the Victorian Heat Health Information Surveillance System 2011–2013

The population health impacts of heat

Key learnings from the Victorian Heat Health Information Surveillance System 2011–2013

i To receive this publication in an accessible format phone (03) 9096 000, using the National Relay Service 13 36 77 if required.

Authorised and published by the Victorian Government, 1 Treasury Place, Melbourne.

This work is licensed under a Creative Commons Attribution 4.0 licence (creativecommons.org/licenses/by/4.0). You are free to re-use the work under that licence, on the condition that you credit the State of Victoria as author, indicate if changes were made and comply with the other licence terms. The licence does not apply to any branding including the Victorian Government logo, images or artistic works.

Except where otherwise indicated, the images in this publication show models and illustrative settings only, and do not necessarily depict actual services, facilities or recipients of services. This publication may contain images of deceased Aboriginal and Torres Strait Islander peoples.

Where the term ‘Aboriginal’ is used it refers to both Aboriginal and Torres Strait Islander people. Indigenous is retained when it is part of the title of a report, program or quotation.

ISBN/ISSN 978-0-7311-6740-1 (pdf)

Available at www.health.vic.gov.au/healthstatus

(1506001)

ii Contents

List of tables iv

List of figures iv

Acknowledgements iv

Summary 1

Introduction 3

Heat health temperature thresholds in Victoria 5

Heat Health Information Surveillance System 7

Temperature patterns in Victoria 9

Humidity 19

Acclimatisation 21

Appendix 1: Weather forecast districts by heat health temperature threshold, weather station and local government area, Victoria 23

Appendix 2: Relationship between meteorological indicators and health service use in Melbourne 27

Appendix 3: Methods 29

Abbreviations 34

References 35

iii List of tables

Table 1: Number of days heat health temperature thresholds were exceeded, by weather forecast district and year, November–March 2004–2013 10

Table 2: Number of days heat health temperature thresholds were exceeded, by weather forecast district and month, November–March 2011–2013 12

Table 3: Number of days heat health temperature thresholds were exceeded, by weather forecast district and date, November–March 2011–2013 14

Table 4: Ranking of strength of relationship between meteorological indices and health service use in Melbourne 20

List of figures

Figure 1: Heat health temperature thresholds (mean temperatures) for weather forecast districts in Victoria 5

Figure 2: Heat Health Information Surveillance System model 8

Figure 3: Mean daily temperature, by weather forecast district and heat health temperature threshold, November–March 2011–2013 16

Figure 4: Percentage of occasions when excess heat-related presentations are experienced, by mean temperature and month, Melbourne, November–January, 2006–2013 22

Acknowledgements

This report has been prepared by the Health Intelligence Unit, System Intelligence and Analytics Branch in the Victorian Department of Health & Human Services. It has been prepared with assistance from the Health Protection Branch and staff involved in emergency management in the department.

Special thanks are due to all of the organisations and their representatives who provide data to the Heat Health Information Surveillance System over the summer months. Many of the individuals involved in providing information for surveillance purposes have reviewed drafts of this document. Their efforts are greatly appreciated.

iv Summary

This report provides an overview of heat health temperature thresholds and the Heat Health Information Surveillance System (HHISS) for Victoria. It summarises temperature patterns across Victoria during the 2011–12 and 2012–13 surveillance periods and attempts to answer questions about humidity and acclimatisation and their potential impact on human health.

Heat health temperature thresholds in Victoria A heat health temperature threshold is the lower temperature limit above which heat-related illness and mortality increase. Temperature is monitored in Victoria over the summer months by weather forecast district. The temperature thresholds in each district are based on geographical areas with similar, consistent weather patterns and temperatures. The heat health temperature thresholds for Victoria differ by weather forecast district, with higher thresholds established in the northern districts of Victoria. This reflects the higher temperatures experienced in these areas.

Heat Health Information Surveillance System In 2009 the Department of Health and Human Services established the HHISS to monitor the impacts of extreme heat on human health in Victoria. The HHISS forms part of a coordinated and integrated response to heatwaves in Victoria, as outlined in the Heatwave plan for Victoria (Department of Health 2011a). The HHISS operates over the summer months, between November and March, and is dependent on meteorological data, health service activity and mortality information obtained from different sources. The HHISS generates a surveillance report to assist departmental staff in assessing public health risk from the heat.

Temperature patterns in Victoria On average, each weather forecast district in Victoria exceeds the heat health temperature thresholds between one and three days each summer. At the state level, there are about five days each summer when at least one of the weather forecast districts experiences temperatures at, or above, established heat health temperature thresholds. Most breaches of the heat health temperature thresholds occur in January, the warmest month of the year.

During the 2012–13 surveillance period there were four weather forecast districts (Mallee, North Central, Northern Country and Central) in Victoria that experienced a higher than average number of days when temperatures exceeded the established thresholds. On average, we would have expected three days when temperatures would exceed the established thresholds for the Central district, two days for the Mallee, and one day for the North Central and Northern Country districts. The Central district experienced five days, North Central four days, the Mallee district three days and the Northern Country district experienced two days of extreme heat, when temperatures were above established thresholds.

Extreme heat was less frequent during the 2011–12 surveillance period. There were two days when the heat health temperature threshold was exceeded for the Central district and one day for the South West and North Central districts.

1 Humidity Humidity is a measure of the amount of water vapour, or moisture in the air. It is important to heat health because it can interfere with the body’s attempt to regulate internal temperature. There are a number of indicators used by different weather services around the world that combine measures of humidity and temperature to provide an indication of ‘how hot it actually feels’. The Bureau of Meteorology in Australia combines humidity with temperature and wind speed to calculate ‘apparent temperature’, which provides an indication of ‘perceived temperature’. The analysis undertaken by the department shows a strong relationship between apparent temperature and heat-related health service use in the metropolitan area. Relative humidity and composite indicators that combine temperature with humidity, which are in use in Canada and the United States, proved less sensitive. The analysis suggests that apparent temperature, which is based on a formula adapted to Australian conditions, is suitable as an indicator to predict human health impacts from heat. However, composite indicators from overseas that are adapted to different conditions and indicators based solely on humidity may not be as useful.

Acclimatisation Acclimatisation is best understood as an adaptation to an environmental stressor, such as heat. Acclimatisation in the human context is not clearly understood as it involves a complex interplay between physiological responses and adaptive behaviours. There is a lot more known about plants and animals and the way in which they respond to environmental stressors.

An analysis of heat-related emergency department presentations from metropolitan Melbourne demonstrates an acclimatisation effect between spring and summer. The analysis undertaken by the department showed that excess heat-related presentations were more likely, at a given temperature, in late spring (November) rather than the middle of summer (January).

2 Introduction

Heatwaves are brief periods of unusually high temperatures. They vary in intensity and duration between locations and may also vary in intensity and duration at different times of the year. They are an important issue in public health because of their risk to human health and wellbeing. Although anyone is susceptible to the effects of extreme temperature, research has shown that those at greatest risk of harm include older adults, the very young, those with pre-existing medical conditions, the homeless, those who live alone and others with limited means to escape the more severe extremes of the heat.

There was an extreme heat event in Victoria in late January 2009 (National Climate Centre 2009), associated with a higher than expected level of illness and a higher than expected number of deaths (Department of Human Services 2009). In response, the Department of Health & Human Services) developed the Heatwave plan for Victoria (Department of Health 2011a), a coordinated and integrated response to heatwaves in Victoria. The plan provides a clear understanding of the actions and systems in place to support at-risk groups in the community and includes information about the health impacts of heat and how to stay healthy in the heat.

As a part of the response, the department also established the Heat Health Information Surveillance System (HHISS) to monitor the human health impacts of extreme heat. The surveillance system operates between 1 November and 31 March each year. The department obtains actual temperature data, along with health service activity and mortality information in order to track health service use and mortality in relation to temperature. This information is used to assist in assessing public health risk from extreme heat.

This is the second report on key learnings from heat health surveillance in Victoria. The first report was released in 2011 and covered the period 2009–10 and 2010–11. This report presents temperature patterns during the surveillance periods 2011–12 and 2012–13. It builds on the learnings from the first report by exploring the relationship between meteorological indicators that incorporate humidity and health service use. This report also examines data for evidence of acclimatisation to changes in temperature between spring and summer in Victoria.

3 4 Heat health temperature thresholds in Victoria

Key points • A heat health temperature threshold is the lower temperature limit above which heat-related illness and mortality increase. • Temperature is monitored by weather forecast district in Victoria each summer. With the exception of the North East, these districts are based on geographical areas with similar, consistent weather patterns and temperatures. • The heat health temperature thresholds for Victoria differ by district, with higher thresholds in the northern districts of Victoria reflecting the higher temperatures experienced in these areas.

There is no single, internationally accepted definition for a heatwave or period of extreme heat because similar temperatures can have different impacts on different communities at different times. Factors such as the demographic profile of a population, acclimatisation and urban design all influence the impact of extreme heat events on human health.

The Victorian Department of Health & Human Services has developed a technical definition for extreme heat events. This is based on the lower temperature limit above which there is likely to be an impact on human health. This technical definition is described as the ‘heat health temperature threshold’.

There are three heat health temperature thresholds in Victoria that apply to three broad geographical bands or zones, running horizontally from east to west across the state (Figure 1). The highest threshold (mean 34 °C) applies to the northernmost (warmest) area of Victoria and the lowest threshold (mean 30 °C) applies to the more southern (cooler) areas of the state.

Figure 1: Heat health temperature thresholds (mean temperatures1) for weather forecast districts in Victoria

Mallee 34 °C Swan Hill Mean 34 °C

Hopetoun

Echuca Mean 32 °C Rochester Wodonga Charlton Shepparton Wimmera 32 °C Elmore North Wangaratta St Arnaud Bendigo Country 32 °C North East 32 °C Horsham Bright Stawell Heathcote Seymour Falls Creek Edenhope Maryborough Halls Gap Castlemaine Avoca NorthMt Buller Mean 30 °C Central 30 °C Omeo Lexton Kyneton Ararat Marysville East Gippsland Daylesford 30 °C Ballarat Kalkallo Casterton Skipton Healsville Melbourne Bairnsdale Hamilton Central West & South South West Warburton Lismore 30 °C Bunyip Gippsland 30 °C Winchelsea Mortlake Cressy Pakenham 30 °C Warragul Sale Lorne Portland Port Fairy Colac Korumburra Wonthaggi

Source: Image adapted from the Country Fire Authority’s Find your fire district map (2012).

1. Mean temperature is calculated from the daily maximum temperature and the following overnight temperature, which is the minimum for the following day. Mean temperature = (daily maximum + following overnight minimum)/2.

5 The department operates a heat health alert system for Victoria each summer. This involves monitoring temperature in each of the weather forecast districts shown in Figure 1 and issuing heat health alerts when forecast temperatures are predicted to exceed respective heat health temperature thresholds (Department of Health 2012).

The weather forecast districts are defined geographically by boundaries for local government areas (see Appendix 1) and correspond to the nine fire districts used for total fire bans and fire danger ratings in Victoria (Country Fire Authority 2012; Department of Health 2012). The districts are geographical areas with similar, consistent weather patterns and temperatures. The exception is the North East district, which has considerable variation due mainly to the difference in temperatures experienced in the alpine areas and along the New South Wales border.

The heat health temperature thresholds for Victoria have been derived from a range of information sources including the results of mortality-based threshold research undertaken by Monash University on behalf of the department (Loughnan, Nicholls & Tapper 2009; Nicholls et al. 2008), the Bureau of Meteorology analysis of meteorological conditions, mean temperatures for each weather forecast district, and heat health thresholds adopted in other Australian states and territories. The Monash University research found that daily mean temperature was the best meteorological indicator for predicting increases in mortality.2 Mean temperature takes into account high temperature during the day, followed by lower temperatures at night. The university’s research suggested that a lack of relief from the heat overnight contributed to higher than normal levels of mortality during periods of extreme heat.

2. Daily (24-hour) minimum and maximum temperatures were also considered in the research undertaken by Monash University (Loughnan, Nicholls & Tapper 2009; Nicholls et al. 2008).

6 Heat Health Information Surveillance System

Key points • The HHISS monitors the human health impacts of extreme heat events in Victoria. • The surveillance system operates over the summer months – from November through to the end of March. • The HHISS relies on climatic data, health service activity and mortality information obtained from different sources. • The HHISS generates a surveillance report to help assess public health risk from the heat.

The HHISS was established in November 2009 to monitor the human health impacts of extreme heat events as close to real time as possible. At the time, there was no systematic means of monitoring the extent of the health impacts of extreme heat in Victoria.

The HHISS provides senior managers in the Department of Health & Human Services, including the Victorian Chief Health Officer, with information about health service activity and mortality, between 1 November and 31 March each year. This information is used to help assess public health risk from the heat. During the summer period, the department obtains meteorological data and health service activity and mortality information from:

• Registry of Births, Deaths and Marriages • Coroners Court of Victoria • National Home Doctor Service • Ambulance Victoria • NURSE-ON-CALL • public hospital emergency departments • Bureau of Meteorology • Victorian Institute of Forensic Medicine3 (Appendix 3).

The surveillance system consists of a series of components, including a data storage facility and a processing and analytical function, as well as a reporting function (Figure 2). The department obtains meteorological data, health service activity and mortality data each day in summer. An initial check is performed to ensure the data are in order before they are formatted for analysis. Following analysis, a surveillance report is produced and sent to senior managers within the department who have a role in assessing public health risk and issuing heat health alerts.

3. Data from the Victorian Institute of Forensic Medicine became an important part of the surveillance system in January 2014.

7 Figure 2: Heat Health Information Surveillance System model

Registry of National Victorian Coroners Hospital Births, Home Ambulance NURSE- Institute of Bureau of Court of emergency Deaths & Doctor Victoria ON-CALL Foresnic Meterology Victoria departments Marriages Service Medicine

Data storage and processing

Surveillance report produced

Department of Health and Human Services staff

Source: Department of Health & Human Services

The surveillance report compares daily mean temperature with the state heat health temperature thresholds to identify days of extreme heat. The surveillance report also compares expected levels (derived from data from recent years) of health service activity and mortality against actual levels to identify days with higher than expected service activity or mortality. During periods of warmer weather, or when an extreme heat event is forecast and the department issues a heat health alert, surveillance efforts are intensified and activity levels monitored more closely.

8 Temperature patterns in Victoria

Key points • Temperature data from 2004–2011 shows that Victoria has an average of five days each year where at least one weather forecast district experiences temperatures at or above established heat health temperature thresholds. • On average, each weather forecast district in Victoria exceeds the heat health temperature thresholds between one and three days each year. • The heat health temperature thresholds are more likely to be exceeded in January, the warmest month of the year. • There were seven days during the 2012–13 surveillance period when at least one weather forecast district in Victoria experienced an extreme heat event, higher than the expected five days.

Temperature patterns from recent years The department obtained daily (24-hour) minimum and maximum temperature data from the Bureau of Meteorology for November–March 2004–2013. The data were used to calculate mean temperatures for weather forecast districts in Victoria. The heat health temperature thresholds for weather forecast districts were used to define and identify days of extreme heat around the state from November to March each year.

Table 1 shows that Victoria has an average of five days each year when at least one weather forecast district experiences temperatures above the established heat health temperature thresholds. The table shows the Central weather forecast district, which has the largest population as it includes Melbourne, has an average of three days each year when temperatures are expected to exceed the established threshold (mean 30 °C). The South West, North East and Mallee districts have an average of two days, while the North Central, West & South Gippsland, East Gippsland, Wimmera and Northern Country districts have an average of one day each year when the mean temperature is expected to be higher than established thresholds.

Table 1 shows there were three days during the 2011–12 surveillance period and seven days during the 2012–13 surveillance period when at least one weather forecast district in Victoria experienced an extreme heat event. Therefore, there were fewer days than expected during 2011–12 and more days than expected during 2012–13 when at least one district in the state experienced an extreme heat event.

9 Total Total 5 7 3 0 8 8 8 6 6 1 Victoria 2 3 0 0 4 5 2 0 3 0 East Mallee North 2 1 0 0 3 6 0 1 1 0 Country Northern 1 2 0 0 2 3 1 1 1 0 1 1 0 0 1 5 2 0 2 0 East 1 1 0 0 0 3 0 Gippsland Wimmera 1 2 0 South West & West 1 1 0 0 0 3 0 Gippsland 0 1 0 West South 2 0 1 0 3 3 3 1 3 1 North Central 1 4 1 0 1 2 1 1 1 0 Central (Mean 30°C) (Mean 30°C) (Mean 30°C) (Mean 30°C) (Mean 30°C) (Mean 32°C) (Mean 32°C) (Mean 32°C) (Mean 34°C) Expected days 3 2012–13 5 2011–12 2 2010–11 0 2009–10 2 2008–09 4 2007–08 7 2006–07 5 2005–06 4 Year 2004–05 1 Table 1: Number of days heat health temperature thresholds were exceeded, by weather forecast district and year, district and year, exceeded, by weather forecast were thresholds 1: Number of days heat health temperature Table 2004–2013 November–March Data source: Bureau of Meteorology Bureau Data source: 2004–2011. November–March expected to be exceeded, based on data from are Expected days is the number of that thresholds

10 January is the warmest month of the year (Department of Health 2011b), when Victoria is most likely to experience a period of extreme heat. Table 2 shows that January was the most common month for heat health temperature thresholds to be exceeded in 2011–12 and 2012–13. There have been no breaches of heat health temperature thresholds between April and October for the years 2004–2013.

11 Total Total Victoria 0 0 0 1 0 1 0 0 0 2 2 4 0 0 0 1 0 0 1 1 0 0 0 0 0 1 0 0 0 0 East Mallee North 0 0 0 0 0 1 0 0 0 1 Country Northern 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 East Gippsland Wimmera 0 0 0 0 0 1 0 0 0 0 South West & West Gippsland 0 0 1 0 0 0 0 0 0 0 West South 0 0 0 0 1 2 0 1 0 1 North Central 0 1 1 0 1 2 0 1 0 1 Central (Mean 30°C) (Mean 30°C) (Mean 30°C) (Mean 30°C) (Mean 30°C) (Mean 32°C) (Mean 32°C) (Mean 32°C) (Mean 34°C) March March February February 2011–12 January January December December November 2012–13 November Table 2: Number of days heat health temperature thresholds were exceeded, by weather forecast district and month, exceeded, by weather forecast were thresholds 2: Number of days heat health temperature Table 2011–2013 November–March Data source: Bureau of Meteorology Bureau Data source:

12 Temperature patterns November–March 2011–2013 Table 3 shows the actual dates when heat health temperature thresholds were exceeded in each weather forecast district, for November–March 2011–2013.

The table shows that the Central district experienced two days when the mean temperature exceeded the established threshold during the 2011–12 surveillance period, followed by the North Central and South West districts with one day. There were more extreme heat events during the 2012–13 surveillance period, with the Central district experiencing five days of extreme heat, followed by the North Central district with four days and the Mallee district with three days of extreme heat.

13 3 3 3 3 3 3 3 Total Total Victoria 2 5 3 7 0 3 3 3 33 3 3 2 1 0 3 East Mallee North 1 2 0 3 3 Country Northern 1 1 0 3 1 1 0 3 East Gippsland Wimmera 1 1 0 3 South West & West Gippsland 2 0 1 3 West South 1 4 1 3 3 3 3 3 North Central e threshold was exceeded for the weather forecast district on this date. was exceeded for the weather forecast e threshold 3 3 3 3 3 3 3 Central (Mean 30°C) (Mean 30°C) (Mean 30°C) (Mean 30°C) (Mean 30°C) (Mean 32°C) (Mean 32°C) (Mean 32°C) (Mean 34°C) The heat health temperatur 2011–12 Expected days 3 2012–13 29 Nov 2012 23 Dec 2012 4 Jan 2013 5 Jan 2013 7 Jan 2013 17 Jan 2013 12 Mar 2013 2012–13Total 5 2 Jan 2012 3 Jan 2012

25 Feb 2012 2011–12Total 2 Table 3: Number of days heat health temperature thresholds were exceeded, by weather forecast district and date, exceeded, by weather forecast were thresholds 3: Number of days heat health temperature Table 2011–2013 November–March Data source: Bureau of Meteorology Bureau Data source: 3 2004–2011. November–March 2011–2013, based on data from expected to be exceeded for November–March were Expected days is the number of that thresholds

14 Figure 3 shows mean daily temperature by weather forecast district and heat health temperature threshold for November–March 2011–2013. The graphs include the heat health temperature threshold (in red) for each weather forecast district. A day of extreme heat was experienced where the mean temperature bar (in orange) intersects with the heat health temperature threshold.

15 Figure 3: Mean daily temperature, by weather forecast district and heat health temperature threshold, November–March 2011–2013

Central Nov 2011–Mar 2012 Central Nov 2012–Mar 2013 50 50

40 40

30 30

20 20 Temperature °C Temperature Temperature °C Temperature 10 10

0 0 1-Nov-11 1-Dec-11 1-Jan-12 1-Feb-12 1-Mar-12 1-Nov-12 1-Dec-12 1-Jan-13 1-Feb-13 1-Mar-13

North Central Nov 2011–Mar 2012 North Central Nov 2012–Mar 2013 50 50

40 40

30 30

20 20 Temperature °C Temperature Temperature °C Temperature 10 10

0 0 1-Nov-11 1-Dec-11 1-Jan-12 1-Feb-12 1-Mar-12 1-Nov-12 1-Dec-12 1-Jan-13 1-Feb-13 1-Mar-13

South West Nov 2011–Mar 2012 South West Nov 2012–Mar 2013 50 50

40 40

30 30

20 20 Temperature °C Temperature °C Temperature 10 10

0 0 1-Nov-11 1-Dec-11 1-Jan-12 1-Feb-12 1-Mar-12 1-Nov-12 1-Dec-12 1-Jan-13 1-Feb-13 1-Mar-13

West & South Gippsland Nov 2011–Mar 2012 West & South Gippsland Nov 2012–Mar 2013 50 50

40 40

30 30

20 20 Temperature °C Temperature Temperature °C Temperature 10 10

0 0 1-Nov-11 1-Dec-11 1-Jan-12 1-Feb-12 1-Mar-12 1-Nov-12 1-Dec-12 1-Jan-13 1-Feb-13 1-Mar-13

East Gippsland Nov 2011–Mar 2012 East Gippsland Nov 2012–Mar 2013 50 50

40 40

30 30

20 20 Temperature °C Temperature °C Temperature 10 10

0 0 1-Nov-11 1-Dec-11 1-Jan-12 1-Feb-12 1-Mar-12 1-Nov-12 1-Dec-12 1-Jan-13 1-Feb-13 1-Mar-13

Mean temperature Heat health temperature threshold

16 Figure 3: Mean daily temperature, by weather forecast district and heat health temperature threshold, November–March 2011–2013 (continued)

Wimmera Nov 2011–Mar 2012 Wimmera Nov 2012–Mar 2013 50 50

40 40

30 30

20 20 Temperature °C Temperature Temperature °C Temperature 10 10

0 0 1-Nov-11 1-Dec-11 1-Jan-12 1-Feb-12 1-Mar-12 1-Nov-12 1-Dec-12 1-Jan-13 1-Feb-13 1-Mar-13

Northern Country Nov 2011–Mar 2012 Northern Country Nov 2012–Mar 2013 50 50

40 40

30 30

20 20 Temperature °C Temperature Temperature °C Temperature 10 10

0 0 1-Nov-11 1-Dec-11 1-Jan-12 1-Feb-12 1-Mar-12 1-Nov-12 1-Dec-12 1-Jan-13 1-Feb-13 1-Mar-13

North East Nov 2011–Mar 2012 North East Nov 2012–Mar 2013 50 50

40 40

30 30

20 20 Temperature °C Temperature Temperature °C Temperature 10 10

0 0 1-Nov-11 1-Dec-11 1-Jan-12 1-Feb-12 1-Mar-12 1-Nov-12 1-Dec-12 1-Jan-13 1-Feb-13 1-Mar-13 50 Mallee Nov 2011–Mar 2012 Mallee Nov 2012–Mar 2013 50 50 40

40 40 30

30 30 20

20 20 Temperature °C Temperature Temperature °C Temperature 10 10

0 0 1-Nov-11 1-Dec-11 1-Jan-12 1-Feb-12 1-Mar-12 1-Nov-12 1-Dec-12 1-Jan-13 1-Feb-13 1-Mar-13

Mean temperature Heat health temperature threshold

Data source: Bureau of Meteorology

17 18 Humidity

Key points • Humidity is a measure of the amount of water vapour, or moisture in the air. It is important to heat health because it can interfere with the body’s attempt to regulate internal temperature. • The Bureau of Meteorology combines humidity with temperature and wind speed measures to produce apparent temperature – an estimate of how hot it actually feels. • There is a good relationship between apparent temperature and health service use in metropolitan Melbourne. • Similar composite indicators from overseas include the Canadian Humidex and the Heat Index from the United States. Their relationship with health service use in the metropolitan area is not as strong as the relationship with apparent temperature.

Background Humidity refers to the amount of water vapour, or moisture in the air. It is important to heat health because humidity can interfere with thermoregulation, or the body’s attempt to regulate internal temperature. When it is hot, the body produces sweat, which cools the body as it evaporates. When it is humid, the sweat that is produced does not evaporate readily enough to cool the body, and once humidity reaches about 90 per cent, the evaporation and body cooling process ceases altogether (Canadian Centre for Occupational Health and Safety 2013). The more humid the conditions, the more difficult it becomes for the body to cool itself and the greater the risk of ill health.

Relative humidity is the amount of moisture in the air expressed as a percentage of the total amount of moisture the air can actually hold at a given temperature. It is an important metric used in forecasting rain, dew and fog, as well as calculating apparent temperature, or how hot or cold it actually feels. In hot weather, increases in relative humidity lead to increases in apparent temperature, or how hot it actually feels, as higher relative humidity hampers body cooling by impeding the evaporation of sweat.

Apparent temperature is a metric calculated from measures of temperature and humidity, as well as factors like wind speed and solar radiation. Apparent temperature is commonly classified according to windchill or a heat index, depending on the season. A heat index consists of ranked categories based on escalating apparent temperature levels. It provides a guide as to how hot the weather actually feels and the level of health risk associated with the heat (Bureau of Meteorology 2013; Environment Canada 2013; Meteociel 2013; NOAANWS 2013).

An index of apparent temperatures, based on a combination of temperature and humidity measures, was first developed in the 1970s by RG Steadman (Steadman 1979a; 1979b). The Steadman model has since been refined, tested and further modified (Steadman 1984; 1994), and there are now a number of complex algorithms in use around the world, in different climate zones, for estimating how hot or cold it actually feels. Anderson et al. (2013) reviewed 21 of the algorithms currently in use and found that most incorporate the theoretical concepts that underpinned the early work by Steadman. Consequently, most algorithms have been shown to produce values close to Steadman’s values, although some algorithms perform poorly, with inconsistent results, especially with low temperature and low humidity. Thus, Anderson et al. advise caution in the selection and use of different formulae to calculate apparent temperature as the utility of the indicator may be dependent on local climate.

19 Results of analysis The relationship between health service use in the metropolitan area and humidity was explored with linear regression modelling. Four separate meteorological measures that incorporate humidity, along with mean temperature, were entered into bivariate models with health service use data from the metropolitan area of Victoria. Ambulance Victoria emergency records with a heat code, public hospital emergency department presentations (total presentations and those with a heat-related diagnostic code), National Home Doctor Service after-hours locum consultations with a heat-related diagnosis and NURSE-ON-CALL heat-related telephone calls were treated as dependent variables in the analysis. The relationship between each dependent variable was tested against the Canadian Humidex, the Heat Index from the United States, mean temperature, relative humidity and apparent temperature from the Bureau of Meteorology in Australia (Appendix 2).

There were significant linear relationships between most of the meteorological measures and health service use data (Appendix 2). However, mean temperature and apparent temperature produced the strongest relationships with the health service data (Table 4). The Canadian Humidex ranked third, followed by relative humidity and the Heat Index from the US.

Although the Humidex and the Heat Index performed well overall, they did not perform as well as mean temperature and apparent temperature. Part of the issue may be the fluctuations and extremes in meteorological conditions in the metropolitan area of Victoria compared with the climate in the countries these indicators have specifically been designed for.

Table 4: Ranking of strength of relationship between meteorological indices and health service use in Melbourne

Meteorological indicator Ranking

Mean temperature 1

Apparent temperature 2

Humidex (Canada) 3

Relative humidity 4

Heat index (US) 5

See Appendix 2 for more detailed information.

20 Acclimatisation

Key points • Acclimatisation is best understood as an adaptation to an environmental stressor, such as heat. • The mechanisms leading to human acclimatisation are not fully understood as they involve a complex interplay between physiological responses and adaptive behaviours. • An analysis of heat-related emergency department presentations demonstrates an acclimatisation effect in the metropolitan area between November and January.

Background Acclimatisation is best understood as an adaptation to an environmental stressor, such as heat. There are numerous studies of physiological reactions to changes in temperature in animals and plants, in light of concerns about species adaptation to changing weather patterns. However, acclimatisation in the human context is much more complex as it involves an interplay between physiological responses and adaptive behaviours. Human studies of acclimatisation tend to focus on associations between temperature change and health impacts or human behaviour or, alternatively, they focus on physiological responses to specific environmental stressors (Banwell et al. 2012; Hansen et al. 2014; Havenith & van Middendorp 1990; Hori, Ihzuka & Nakamura 1976; Medina- Ramon & Schwartz 2007; Seretakis et al. 1997). The interplay between factors is not well addressed and, as a consequence, the mechanisms leading to human acclimatisation are not fully understood.

In Australia, studies of behaviour and attitudes towards heat reveal what has been described as a serious lack of concern and knowledge about extreme heat as a health issue, especially among vulnerable groups such as older people and those who work or exercise outside (Banwell 2012; Department of Health 2013; Hansen et al. 2014). This has generated calls, both in Australia and internationally, for a greater emphasis on heat protection and heat health promotion activities, such as promoting acclimatisation as an issue, to reduce the burden of heat-related illness (Singh, Hanna & Kjellstrom 2013; Stoecklin-Marois et al. 2013).

Concerns have been raised previously that as we move from winter into spring people are more likely to become ill from a rise in temperature than they are during mid-summer (VCOSS 2013). Periods of extreme heat that occur in spring and early summer have been reported to have a far greater impact on human health than those that occur in late summer (Vaneckova et al. 2008). The impact may be compounded in cooler climes, where people may be physiologically more sensitive to the heat or less likely to have access to air-conditioning and engage in, or have knowledge of, other heat- adaptive behaviours (Hansen et al. 2014; Nitschke et al. 2011).

Results of analysis Figure 4 suggests there is an acclimatisation effect in the metropolitan area because excess heat- related presentations appear to be more likely, at a given temperature, in late spring (November) than in the middle of summer (January). The data in the graph show the percentage of occasions (based on the number of times) there have been excess (over and above what we would normally expect) heat-related presentations to Melbourne public hospital emergency departments. The data are presented for the months of November, December and January, by mean temperature.

21 Figure 4: Percentage of occasions when excess heat-related presentations are experienced, by mean temperature and month, Melbourne, November–January, 2006–2013

100 x

90 In November, on days where the mean temperature in Melbourne has been 80 27–28 °C, we have always (100%) experienced excess heat-related 70 emergency department presentations.

60 In contrast, during January, on 50 days where the mean temperature x in Melbourne has been 27–28°C,

Per cent we have only experienced excess 40 heat-related emergency department presentations half the time (50%). 30

0 13–14 15–16 17–18 19–20 21–22 23–24 25–26 27–28 29–30 31–32 33–34 35–36 Mean temperature °C

Nov Dec Jan

Data sources: Department of Health & Human Services and the Bureau of Meteorology

Heat-related emergency department presentations from hospitals in Melbourne include presentations with any diagnostic ICD-10-AM code ‘E86’, ‘T67’ or ‘X30’. This includes presentations for volume depletion/dehydration, heat exhaustion, heat syncope, heat cramp, heat stroke, heat fatigue, heat oedema and exposure to excessive natural heat - sunburn.

22 Local government area Buloke Shire Gannawarra Shire Mildura Rural City Swan Hill Rural City Road) (north of Galaquil East & West Yarriambiack Hindmarsh Shire Horsham Rural City Northern Grampians Shire WimmeraWest Shire Road) (south of Galaquil East & West Yarriambiack Ararat Rural City Colac Otway Shire Shire Pyrenees Corangamite Shire Glenelg Shire Moyne Shire Southern Grampians Shire City Warrnambool Weather station Weather sites Heat health temperature threshold Mean 34 °C Mildura Weather Weather district forecast Mallee Wimmera Mean 32 °C Horsham South West Mean 30 °C Hamilton Appendix 1: Weather forecast districts by heat health temperature districts by heat health temperature forecast Appendix 1: Weather weather station and local governmentVictoria threshold, area,

23 Local government area Campaspe Shire Bendigo City Greater Shepparton City Greater Loddon Shire Moira Shire Strathbogie Shire Central Goldfields Shire Lake Mountain Alpine Resort (Unincorporated) Mitchell Shire Mount Alexander Shire Murrindindi Shire Alpine Shire Indigo Shire Shire Towong Rural City Wangaratta City Wodonga Benalla Rural City Mansfield Shire Alpine Resort (Unincorporated) Falls Creek Mount Buller Alpine Resort (Unincorporated) Mount Hotham Alpine Resort (Unincorporated) Mount Stirling Alpine Resort (Unincorporated) Weather station Weather sites Bendigo Shepparton Heat health temperature threshold Weather Weather district forecast Northern Country Mean 32 °C North Central Mean 30 °C Seymour North East Mean 32 °C Wodonga

24 Moorabool Shire Moreland City Moreland Mornington Peninsula Nillumbik Shire Port Phillip Shire Queenscliffe Borough Queenscliffe Stonnington City Surf Coast Shire Whitehorse City Whittlesea City Wyndham City Yarra City Yarra Yarra Ranges Yarra Greater Geelong City Greater Hepburn Shire Hobsons Bay City Hume City Kingston City Knox City Macedon Ranges Shire Manningham City Maribyrnong City Maroondah City Maroondah Melton Shire Monash City Moonee Valley City Moonee Valley Local government area Baw Shire City Latrobe Mount Baw Alpine Resort (Uincorporated) South Gippsland Shire Shire Wellington Ballarat City Banyule City Bass Coast Shire Bayside City Boroondara City Boroondara Brimbank City Cardinia Shire Cardinia Casey City Darebin City Darebin Frankston City French Island (Unincorporated)French Melbourne City Glen Eira City Golden Plains Shire Greater Dandenong City Greater Weather station Weather sites Melbourne Ballarat Geelong Heat health temperature threshold Mean 30 °C Sale Mean 30 °C Weather Weather district forecast East Gippsland Mean 30°C & South West BairnsdaleGippsland East Gippsland Shire Central

25 26 2 4 5 Ranking 2 Adj R Emergency presentations department all p -value 2 Adj R Emergency department heat- p -value related presentations related 2 Adj R phone calls heat-related heat-related NURSE-ON-CALL p -value 2 Adj R consultations National Home hours heat-related hours heat-related p -value Doctor Service after- 2 Adj R with a heat code 0.000 0.52 0.001 0.09 0.000 0.27 0.000 0.33 0.000 0.04 1 0.000 0.33 0.000 0.06 0.000 0.26 0.000 0.34 0.002 0.02 3 0.000 0.16 0.001 0.03 0.011 0.02 0.003 0.25 0.264 ns 0.000 0.22 0.001 0.02 0.000 0.04 0.000 0.04 0.954 ns emergency records emergency records Ambulance Victoria p -value values show the proportion of variation in the data that is explained by model. values show the proportion 2 Meteorological Meteorological indicator Mean temperature Apparent temperatureApparent 0.000 0.51 0.000 0.10 0.000 0.23 0.000 0.38 0.292 ns Humidex Relative humidity Heat index Appendix 2: Relationship between meteorological indicators and Appendix 2: Relationship between meteorological health service use in Melbourne p ≥ 0.05). relationship indicators and health service in the table unless otherwise stated (ns: non-significant between all meteorological was a significant ( p < 0.05) linear relationship There with each health service indicator. of their relationship to the strength and then ranked according scored indicators were Meteorological Data sources: Ambulance Victoria, National Home Doctor Service, Medibank Health Solutions, Department of Health & Human Services, Bureau of Meteorology Ambulance Victoria, National Home Doctor Service, Medibank Health Solutions, Department of & Human Services, Bureau Data sources: indicator is due to chance. between the health service activity and meteorological that the linear relationship The p -value shows the level of probability The adjusted R

27 28 Appendix 3: Methods

Data sources Ambulance Victoria – total emergency and emergency records with a heat code Ambulance Victoria provides daily de-identified information about emergency ambulance responses to the department between November and March each year. The data provided include emergency response records that have a heat code assigned. This report includes the outcome from analysing metropolitan emergency records with a heat code.

National Home Doctor Service – after-hours heat-related consultations The National Home Doctor Service provides after-hours medical care in the Melbourne metropolitan, Geelong and Mornington Peninsula areas of the state. Patients seeking care from their general practitioner when a practice is closed are directed to contact the service for a doctor to make a home visit. The notes from each home visit by a locum are recorded on a database and made available to the patient’s usual doctor. De-identified daily counts of locum consultations are extracted for heat health surveillance purposes each day during summer from the service database using keywords for heat-related diagnoses. This report includes the outcome from analysis of after-hours locum consultation records for the metropolitan area that include a heat-related diagnosis.

NURSE-ON-CALL – heat-related phone calls NURSE-ON-CALL is a telephone service that provides immediate, expert health advice from a registered nurse, 24-hours a day, seven days a week. The details of telephone calls to the service are collected, including the symptoms or health concern prompting each call. The service is run by Medibank Health Solutions in Victoria, which provides de-identified heat-related call data to the department each day during summer for heat health surveillance purposes. This report includes the outcome of an analysis of telephone call records for the metropolitan area that included a reference to a heat-related condition.

Victoria Emergency Minimum Dataset – total and heat-related presentations The Victorian Emergency Minimum Dataset is a data collection containing de-identified administrative and clinical information on presentations to public hospital emergency departments in Victoria. De-identified daily counts of total and heat-related presentations are extracted from this dataset for heat health surveillance purposes during summer. Heat-related presentations included in the analysis in this report include those records containing an ICD-10-AM code ‘E86’, ‘T67’ or ‘X30’. This includes people who are diagnosed in a Victorian public hospital emergency department with volume depletion/dehydration, heat exhaustion, heat syncope, heat cramp, heat stroke, heat fatigue, heat oedema and exposure to excessive natural heat–sunburn. This report includes the outcomes from an analysis of both total and heat-related emergency department presentations from public hospitals in the metropolitan area.

Victorian Registry of Births, Deaths and Marriages – total (registered) deaths Victorian mortality data are collated by the Victorian Registry of Births, Deaths and Marriages. Death certificates are required to be completed by a registered medical practitioner within 48 hours of a death and then forwarded to the registry. De-identified daily counts of registered deaths (by date of death) are forwarded to the department for heat health surveillance purposes during summer each year.

29 Deaths reported to the Coroners Court of Victoria – total deaths reported Certain deaths are required by law to be reported to the Coroners Court of Victoria. This includes deaths:

• that are unexpected or appear to have resulted directly or indirectly from accident or injury • where the identity of a person is not known, or • where a medical practitioner has been unable to provide a death certificate.

De-identified daily counts of deaths eportedr to the Coroners Court of Victoria are forwarded to the department for heat health surveillance purposes during summer.

Deaths reported to Victorian Institute of Forensic Medicine – total death admissions The Victorian Institute of Forensic Medicine became an important part of the Heat Health Information Surveillance System in January 2014. De-identified daily counts of death admissions to the institute for forensic investigation are forwarded to the department for heat health surveillance purposes.

Bureau of Meteorology – temperature, humidity and wind speed The Bureau of Meteorology provides a range of meteorological statistics for Victoria on its website (see www.bom.gov.au), including minimum and maximum temperatures, relative humidity and wind speed readings from weather stations around the state at different time points during the day. The data are obtained daily between November and March for surveillance purposes from key weather stations around Victoria. The analyses in this report are based on actual temperature data (as opposed to forecast temperature data) from weather stations around Victoria.

Definitions Heatwave A heatwave is a period of unusually hot weather or extreme heat that may vary in intensity and duration between locations and may vary in intensity and duration at different times of the year. The department uses the heat health temperature threshold to define an extreme heat event or heatwave.

Heat health temperature threshold A heat health temperature threshold is the lower temperature limit above which there is likely to be an impact on human health. There are three heat health temperature thresholds that apply to three broad geographical bands or zones, running horizontally from east to west across Victoria (Figure 1). The highest threshold (mean 34 °C) applies to the northernmost (warmest) area of Victoria and the lowest threshold (mean 30 °C) applies to the southernmost (coolest) area of the state.

Weather forecast district Temperature is monitored by weather forecast district in Victoria each summer. With the exception of the North East, these districts are based on geographical zones with similar, consistent weather patterns and temperatures. Weather forecast districts can be defined by local government area boundaries (Appendix 1).

30 Mean temperature Daily maximum and minimum temperature data were obtained from the Bureau of Meteorology (www.bom.gov.au). Mean temperature was calculated from the daily maximum temperature and the following overnight temperature, which is the minimum for the following day.

Mean temperature = (daily maximum + following overnight minimum)/2

Mean temperature is used to define heat health temperature thresholds in Victoria.

Statistical significance Statistical significance provides an indication of how likely a result is due to chance. It does not infer clinical significance, the relative importance of a particular finding or the actual magnitude of difference between values. Statistical differences were deemed to exist in the analysis presented in this report where the probability of a result being due to chance was less than five per cent (p < 0.05).

Relative humidity (Australian Bureau of Meteorology) Relative humidity is the amount of moisture in the air, expressed as a percentage of the amount of moisture the air can actually hold. Warmer air can hold more moisture than cooler air, which means that for a given amount of atmospheric moisture, relative humidity will be lower if the air is warm than if the air is cool (Bureau of Meteorology 2014).

Heat Index (United States) The National Oceanographic and Atmospheric Administration developed the Heat Index used by the National Weather Service in the United States. The Heat Index combines temperature and relative humidity into a single value to reflect perceived temperature (NOAANWS 2014):

Heat Index = –42.379 + 2.04901523 * T + 10.14333127 * RH – 0.22475541 * T * RH – 0.00683783

* T * T – 0.05481717 * RH * RH + 0.00122874 * T * T * RH + 0.00085282 * T * RH

* RH – 0.00000199 * T * T * RH * RH

Where:

T: Temperature (in Fahrenheit)

RH: Relative humidity (in per cent)

If the relative humidity is less than 13 per cent and the temperature is 80–112 °F, then the following adjustment is subtracted from the Heat Index:

Adjustment = [(13 – RH) / 4] * SQRT {[17 – ABS (T – 95.)] / 17}

Where:

ABS: Absolute value

SQRT: Square root

If the relative humidity is greater than 85 per cent and the temperature is 80–87 °F, then the following adjustment is added to the Heat Index:

Adjustment = [(RH - 85) / 10] * [(87 – T) / 5]

31 The Rothfusz regression is not appropriate when conditions of temperature and humidity warrant a Heat Index value less than 80 °F. Where this occurs, a simpler formula is applied to calculate values consistent with Steadman’s results:

Heat Index = 0.5 * {T + 61.0 + [(T – 68.0) * 1.2] + (RH * 0.094)}

This regression is not valid for extreme temperature and relative humidity conditions beyond the range of data considered by Steadman.

Humidex (Canada) The Humidex was first developed by Canadian meteorologists to describe how hot and humid the weather feels to the average person. It is similar to the Heat Index, as it also combines temperature and relative humidity into a single value to reflect perceived temperature, but is derived using a different formula (Environment Canada 2013):

Humidex = T + (0.5555 * (e – 10))

Where:

T: Temperature (in Celsius) e: Vapour pressure (in millibars)

Apparent temperature (Australian Bureau of Meteorology) The Bureau of Meteorology calculates apparent temperature values derived from a formula adapted from Steadman’s Norms of apparent temperature in Australia (Bureau of Meteorology 1994). The formula takes temperature, humidity and wind speed into account, but not solar radiation, so is actually an approximation of apparent temperature (Bureau of Meteorology 2013):

Apparent temperature = Ta + 0.33 * e – 0.70 * ws – 4.00

Where:

Ta: Dry bulb temperature (in Celsius) e: Vapour pressure (humidity) (hPa) ws: Wind speed (m/s) at an elevation of 10 metres

Data analysis The health service data used in the analysis are data that have been used for heat health surveillance in Victoria. The data may not have been final or complete at the time they were provided to the department. Although the data are fit for surveillance purposes, they may contain missing or duplicate records or contain other minor errors. Any changes to the data following finalisation are expected to be minor changes and would not be expected to impact on patterns and trends or the direction of an effect in the analysis presented in this report.

The meteorological data used in analysis to produce this report are based on actual observations (rather than forecast observations) from weather stations across Victoria (Appendix 1). The data were obtained from the Bureau of Meteorology website (www.bom.gov.au).

32 The Melbourne metropolitan area was selected as the geographical unit of analysis in this report because of the size of the population and the large number of health service records generated compared with other areas of the state.

All analyses were conducted using Stata statistical software version 12.1 (StataCorp, College Station, Texas, USA) and SPSS Statistics version 20 (IBM SPSS Statistics).

Humidity Ordinary least squares linear regression models were used to explore the relationship between various meteorological indicators, which incorporate humidity, and health service use in the metropolitan area. Mean temperature, which does not incorporate humidity, was included in the analysis as a comparator. Mean temperature has previously been shown to have a good relationship with health service use (Department of Health 2011b; Loughnan, Nicholls and Tapper 2009; Nicholls et al. 2008) and is currently used in heat health surveillance in Victoria.

The regression modelling was performed on data from the 2011–12 and 2012–13 summer surveillance periods (November–March). All data were checked for normality and transformed, where required, prior to modelling. Each model received a ranked score from 1 to 5 based on whether statistical significance was attained (p-value) and the proportion of variation in the data explained by the model, after adjustment for the number of observations and variables in the model (adjusted R2). Meteorological indicators were then ranked from 1 to 5 based on the sum of final scores.

Acclimatisation The acclimatisation analysis involved calculating the number of times excess heat-related emergency department presentations occurred in the metropolitan area, at different mean temperature increments, between November and January, for the years 2006–2013. The temperature data for the acclimatisation analysis was obtained from the Melbourne regional weather station (#086071). Excess presentations were based on the difference between observed and expected presentations. Expected presentations were based on the average for each month, excluding outliers.

33 Abbreviations e Vapour pressure

HHISS Heat Health Information Surveillance System hPa Hectopascal

ICD-10-AM International statistical classification of diseases and related health problems, 10th revision, Australian modification

RH Relative humidity

T Temperature

Ta Dry bulb temperature ws Wind speed

34 References

Anderson GB, Bell ML, Peng RD 2013, ‘Methods to calculate the heat index as an exposure metric in environmental health research’, Environmental Health Perspectives, doi:10.1289/ehp.1206273.

Banwell C, Dixon J, Bambrick H, Edwards F, Kjellstrom T 2012, ‘Socio-cultural reflections on heat in Australia with implications for health climate change adaptation’, Global Health Action, doi:10.3402/ gha.v5i0.19277.

Bureau of Meteorology 2013, Thermal comfort observations, viewed 24 October 2013, .

Bureau of Meteorology 2014, Average relative humidity, viewed 9 April 2014, .

Canadian Centre for Occupational Health and Safety 2013, Humidex rating and work, viewed 24 October 2013, .

Country Fire Authority 2012, Find your fire district, viewed 2 June 2014, .

Department of Human Services 2009, January 2009 heatwave in Victoria: an assessment of health impacts, State Government of Victoria, Melbourne.

Department of Health 2011a, Heatwave plan for Victoria: protecting health and reducing harm from heatwaves, State Government of Victoria, Melbourne.

Department of Health 2011b, The population health impacts of heat: key learnings from the Victorian heat health information surveillance system, State Government of Victoria, Melbourne.

Department of Health 2012, Heat health alert system 2012–2013, State Government of Victoria, Melbourne.

Department of Health 2013, Adaptive capabilities in older people during extreme heat events in Victoria: a population survey, State Government of Victoria, Melbourne.

Environment Canada 2013, Humidex, viewed 24 October 2013, .

Hansen A, Nitschke M, Bi P, Pisaniello D, Newbury J, Kitson A, Dal Grande E, Avery J, Zhang Y, Fraser-Adams M, Kelsall L, Tucker G 2014, ‘Heat-health behaviors of older persons in two Australian states’, Australasian Journal on Ageing (in press).

Havenith G, van Middendorp H 1990, ‘The relative influence of physical fitness, acclimatization state, anthropometric measures and gender on individual reactions to heat stress’, European Journal of Applied Physiology & Occupational Physiology, vol. 61, pp.419–427.

Hori S, Ihzuka H, Nakamura M 1976, ‘Studies on physiological responses of residents in Okinawa to a hot environment’, Japanese Journal of Physiology, vol. 26, pp. 235–244.

Loughnan M, Nicholls N, Tapper N 2009, Temperature thresholds associated with increased mortality in ten major population centres in rural Victoria, Australia, Monash Climate, School of Geography and Environmental Science, Monash University, Melbourne.

Medina-Ramon M, Schwartz J 2007, ‘Temperature, temperature extremes, and mortality: a study of acclimatisation and effect modification in 50 US cities’, Occupational and Environmental Medicine, vol. 64, pp. 827–833.

35 Meteociel 2013, Humidex, viewed 24 October 2013, .

National Climate Centre 2009, The exceptional January–February 2009 heatwave in south-eastern Australia, Special Climate Statement 17, Bureau of Meteorology, Melbourne.

National Oceanographic and Atmospheric Administration National Weather Service (NOAANWS) 2013, Heat: a major killer, viewed 24 October 2013, .

National Oceanographic and Atmospheric Administration National Weather Service (NOAANWS) 2014, The heat index equation, viewed 9 April 2014, .

Nicholls N, Skinner C, Loughnan M, Tapper N 2008, ‘A simple heat alert system for Melbourne, Australia’, International Journal of Biometeorology, vol. 52, pp. 375–384.

Nitschke M, Tucker G, Hansen A, Williams S, Zhang Y, Bi P 2011, ‘Impact of two recent extreme heat episodes on morbidity and mortality in Adelaide, South Australia: a case-series analysis’, Environmental Health, doi:10.1186/1476-069X-10-42.

Seretakis D, Lagiou P, Lipworth L, Signorello LB, Rothman KJ, Trichopoulos D 1997, ‘Changing seasonality of mortality from coronary heart disease’, Journal of American Medical Association, vol. 278, pp. 1012–1014.

Singh S, Hanna EG, Kjellstrom T 2013, ‘Working in Australia’s heat: health promotion concerns for health and productivity’, Health Promotion International, pp. 1–12.

Steadman RG 1979a, ‘The assessment of sultriness. Part I: a temperature-humidity index based on human physiology and clothing science’, Journal of Applied Meteorology, vol. 18, pp. 861–873.

Steadman RG 1979b, ‘The assessment of sultriness. Part I: effects of wind, extra radiation and barometric pressure on apparent temperature’, Journal of Applied Meteorology, vol. 18, pp. 874–885.

Steadman RG 1984, ‘A universal scale of apparent temperature’, Journal of Climate and Applied Meteorology, vol. 23, pp.1674–1687.

Steadman RG 1994, ‘Norms of apparent temperature in Australia’, Australian Meteorology Magazine, vol. 43, pp. 1–16.

Stoecklin-Marois M, Hennessy-Burt T, Mitchell D, Schenker M 2013, ‘Heat-related illness knowledge and practices among California hired farm workers in The MICASA Study’, Indigenous Health, vol. 51, pp. 47–55.

Vaneckova P, Hart MA, Beggs PJ, de Dear RJ 2008, ‘Synoptic analysis of heat-related mortality in Sydney, Australia, 1993–2001’, International Journal of Biometeorology, doi:10.1007/s00484- 007-0138-z.

Victorian Council of Social Service (VCOSS) 2013, Feeling the heat: heatwaves and social vulnerability in Victoria, VCOSS, Melbourne.