If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
§ The first two authors contributed equally to the manuscript.
Eric K.W. Poon
Correspondence
Corresponding author at: Department of Medicine, St Vincent’s Hospital, Melbourne Medical School, The University of Melbourne, Melbourne, Vic, Australia
Footnotes
§ The first two authors contributed equally to the manuscript.
Affiliations
Department of Medicine, St Vincent’s Hospital, Melbourne Medical School, University of Melbourne, Melbourne, Vic, Australia
§ The first two authors contributed equally to the manuscript.
Vassili Kitsios
Correspondence
Corresponding author at: Department of Medicine, St Vincent’s Hospital, Melbourne Medical School, The University of Melbourne, Melbourne, Vic, Australia
Footnotes
§ The first two authors contributed equally to the manuscript.
Affiliations
CSIRO, Oceans & Atmosphere, Melbourne, Vic, AustraliaLaboratory for Turbulence Research in Aerospace and Combustion, Department of Mechanical and Aerospace Engineering Monash University, Melbourne, Vic, Australia
Department of Intensive Care, Alfred Hospital, Melbourne, Vic, AustraliaThe Australian and New Zealand Intensive Care Society (ANZICS) Centre for Outcome and Resources Evaluation, Melbourne, Vic, AustraliaThe Australian and New Zealand Intensive Care Research Centre, School of Public Health and Preventive Medicine, Monash University, Melbourne, Vic, Australia
The Australian and New Zealand Intensive Care Research Centre, School of Public Health and Preventive Medicine, Monash University, Melbourne, Vic, AustraliaDepartment of Critical Care, The University of Melbourne, Melbourne, Vic, Australia
Austin Hospital Clinical School, The University of Melbourne, Melbourne, Vic, AustraliaDepartment of Surgery, University of Melbourne, Melbourne, Vic, AustraliaDeakin University, Geelong & Melbourne, Vic, AustraliaJames Cook University, Townsville & Cairns, Qld, AustraliaUniversity of Illinois, Urbana-Champaign, IL, USA
A robust climate-health projection model has the potential to improve health care resource allocation. We aim to explore the relationship between Australian intensive care unit (ICU) demand and various measures of the long-lived large-scale climate and to develop a future nationwide climate-health projection model.
Methods
We investigated patients admitted to ICUs in Australia between January 2003 and December 2019 who were exposed to long-lived large-scale combined climatic measures of temperature and humidity. We analysed the projected demand for respiratory-related ICU average length of stay (in days) per capita (ICUD/C) with four historical and one future projection dataset. These datasets included: i) Australian and New Zealand Intensive Care Society adult patient database, ii) Socioeconomic Data and Applications Center gridded global historical population, iii) Australian Bureau of Statistics national historical population, iv) Japanese 55-year Reanalysis historical climate (JRA55), and v) the fifth Coupled Model Inter-comparison Project future climate projections.
Results
148,638 patients with respiratory issues required intensive care between 2003 and 2019. The annual growth in the population density-weighted wet-bulb-globe temperature—a combined measure of temperature and humidity—is strongly correlated with the annual per capita growth ICUD/C for respiratory-related conditions (r=0.771; p<0.001). This relationship was applied to develop a model projecting future respiratory-related ICU demand with three possible future Representative Concentration Pathways (RCP). RCP2.6 (lowest carbon emission climate scenario) showed only a 33.4% increase in Australian ICUD/C demand by 2090, while the RCP8.5 (highest carbon emission climate scenario) demonstrated almost two-fold higher demand (66.1%) than RCP2.6 by 2090.
Conclusions
The annual growth in population density-weighted wet-bulb-globe temperature correlates with the annual growth in Australian ICUD/C for respiratory-related conditions. A model based on possible future climate scenarios can be developed to predict changes in ICU demand in response to CO2 changes over the coming decades.
Carbon emissions from human economic development impact Earth’s climate (e.g., global warming). Such climate change may also have health care system implications [
]. Intensive care unit (ICU) admissions represent the most resource-intensive aspect of hospitalisation. Translating climate change related observations into future projections of ICU demand has not been performed.
The climate comprises of persistent trends due to climate change [
IPCC. Climate Change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New York, USA; 2013. Available from: https://www.ipcc.ch/report/ar5/wg1/. Accessed November 22, 2022.
]. All of these combine and interact to produce the year-to-year climatic conditions. One of the very distinct differences between climate and weather is that climate represents the average conditions over multiple years, whereas the weather focusses on shorter time scales (days or weeks).
Previous studies have often focussed on the associations between various weather impacts and medical conditions. For instance, a recent study showed that spatially localised short-term weather, like extreme heat days, had been associated with increased mortality in the United States (US) [
]. A similar study in Hong Kong found that a 1°C increase in the diurnal temperature range (defined as the daily difference in maximum and minimum temperature) increases cardiovascular-related admissions in the elderly by 0.86% [
The association between diurnal temperature range and emergency room admissions for cardiovascular, respiratory, digestive and genitourinary disease among the elderly: a time series study.
Excess emergency department visits for cardiovascular and respiratory diseases during the 2019-20 bushfire period in Australia: a two-stage interrupted time-series analysis.
]. These studies formed the very important groundwork for establishing an association of the weather impacts on the health system.
Our focus is on the impact of spatially large, long-lived persistent climatic conditions. In principle, there may be a variety of associations between the climate and the steady increase in Australia’s ICU admissions over the past decades [
]. These observations provide evidence of a possible correlation between climate and long-term trends in ICU admission. This makes it plausible that robust projections of future ICU demand would better inform ICU resource allocation and investment.
Accordingly, we aimed to explore the correlations between individual climate variables and ICU admissions by studying the increasing or decreasing trend between climate variables and ICU admissions. Further, we aimed to develop a climate-health care (ICU admission) projection model and provide a step-by-step guide on building this estimate of future health care demand.
Methods
Participants
Participants were patients admitted to ICUs in Australia that continuously submitted monthly admission data to The Australian and New Zealand Intensive Care Society (ANZICS) Adult Patient Database during the study period. This is a clinical quality registry run by the ANZICS Centre for Outcome and Resource Evaluation for quality benchmarking purposes and contains deidentified individual patient information, including basic demographics, the diagnosis that caused the ICU admission, illness severity and outcomes, including length of stay in ICU. It presently represents over 95% of admissions to public and private ICUs within Australia, and this data may be used for research under specific governance arrangements.
This research was reviewed and approved as a low-risk study by the Human Research and Ethics Committee of The Alfred Hospital, Melbourne, Australia (HREC number 631-20). Informed consent was obtained from all individual participants included in the study.
Population Datasets
Two (2) population datasets were adopted.
The first dataset was the population count of Australia from the Australian Bureau of Statistics [
]. This dataset recorded Australia’s population growth quarterly between June 1981 and June 2020. The second was the Gridded Population of the World (GPW, v4) dataset from the National Aeronautics and Space Administration Socioeconomic Data and Applications Center (SEDAC, https://sedac.ciesin.columbia.edu/). This dataset enabled us to calculate the population density-weighted averages of the climate variables over a 1° × 1° grid spacing across Australia (or any other country) to understand the spatial population density across the nation.
Climate Datasets
Two (2) climate datasets were adopted.
The first was the Japanese Atmospheric Re-analysis (JRA55) [
], which provides a historical account of the climate from 1955 to the present day. JRA55 uses available historical atmospheric observations to correct a numerical simulation of the atmosphere, enabling access to historically representative gridded maps of the required variables for our analysis, specifically, temperature, humidity, and sea-level pressure.
The second dataset was a set of future climate projections generated by different climate models from multiple research centres worldwide, contributing data to the fifth Coupled Model Inter-comparison Project (CMIP5) [
Taylor KE, Stouffer RJ, and Meehl GA. An overview of CMIP5 and the experiment design. Bulletin of the American Meteorological Society. 2012;93(4):485-98.
]. A set of 20 climate models developed at various institutes around the world, produce climate variable data by simulating three specific Representative Concentration Pathways (RCP) to the year 2100. Each RCP represents a possible future carbon emission trajectory and the associated climate. These pathways have been agreed upon by the international climate community. In order of increasing levels of warming, these three scenarios adopted herein are denoted by RCP2.6, RCP4.5 and RCP8.5. These simulations represent the average response of the Earth to a given prescribed level of carbon concentration. Note, none of these scenarios may necessarily precisely come to pass, but they do provide a means of understanding how the Earth may respond to a range of possible future emissions scenarios.
Statistical Analysis
We analysed the Pearson correlations between various categories of ICU admissions (such as cardiovascular disease, respiratory disease, etc.) and Climate Variables [CVs] (including, but not limited to, surface temperature, humidity, and wet-bulb-globe temperature [WBGT]). All analyses were undertaken with python-v3.7.7. WBGT is the most accepted and effective method of assessing heat stress on the human body by quantifying the thermal interactions between our body and the surrounding environment. WBGT has often been used in macroeconomic labour productivity studies as adverse impacts on the amount of physical labour someone can perform [
]. All data were annually averaged before analysis. In terms of ICU admission data from the Adult Patient database, we have excluded the following for data consistency:
i)
ICU admissions pre-2003 and post-2019. The former exclusion was due to an observed inconsistency in time series data pre-2003; the latter was due to the additional uncertainties with COVID-19-related ICU admissions.
ii)
ICUs that did not continuously submit admission data,
iii)
Data with potential administrative errors in ICU length of stay (e.g., missing ICU length of stay, or a length of stay >1 year).
The location of the hospital sites retained with consistent data streams used in the analysis is illustrated in Figure 1.
Figure 1Hospital locations and Australian population density in 2020.
Red dots identify hospitals that fulfilled the inclusion criteria and were included in this analysis. Contours indicated the Australian population density in 2020.
We calculated the lengths of stay (in days) summed for each individual disease category’s admissions in each year, then divided it by the population of Australia in that year to produce a per capita measure—ICU average days per capita (ICUD/C). We then calculated a time series of the annual growth rates in ICU average days per capita () according to:
The use of annual growth rates helped clearly identify the year-to-year trend over time. These annual growth rates were correlated against the annual time-series growth rate from a suite of Climate Variables averaged over Australia, where the average was weighted by the local Australian population density. This population density weighting is required because changes in the climate to regions with no population is irrelevant to the Australia wide changes to ICU admissions. Thus, the following analysis focussed on the time series of the annual growth rates in the population density-weighted CVs averaged over Australia:
i)
Extracted the CV (e.g., temperature, humidity, WBGT, etc.) from the monthly averaged JRA55 variables.
ii)
Interpolated CV onto the SEDAC population density grid (grey scale contours in Figure 1).
iii)
Calculated the weighted average CV based on the SEDAC population density map in 2020.
The time series of the population density weighted annual CV growth rates were given by
For each CV analysis, we further undertook a non-parametric bootstrapping to estimate the errors in the relationship:
i)
Estimated the slope (m) and intercept () of the regression line ( ) for n = 17 (one sample per year) using ordinary least square method.
ii)
Computed the predicted based on this regression for each .
iii)
Calculated the residuals ().
iv)
Sampled with replacement from the residuals n=17 times to produce the pseudo residual set ().
v)
Constructed a pseudo set of ICU samples ().
vi)
Calculated a new slope, offset and correlation coefficients from the pseudo samples.
The above steps were repeated 10,000 times to produce the probability density functions of the slope, intercept, and correlation coefficient for each CV.
Future ICU Projections
To create a model for future ICUD/C projections, we used data from three separate standardised climate scenarios (RCP2.6, RCP4.5, RCP8.5). We calculated the population density-weighted CVs over the next 80 years, from the output generated by these climate models.
To characterise the uncertainty in the climate itself, we bootstrapped the model mean by sampling with replacement across the climate model ensemble and calculated the average CVs growth rate 10,000 times. This enabled us to probability density functions of the ensemble-averaged annual growth rate in the population density-weighted CVs in the years 2030, 2050, 2070 and 2090 for each of the three scenarios (RCP2.6, RCP4.5 and RCP8.5).
We applied the relationship between the growth in CVs and the growth in ICUD/C to existing climate model simulations of potential future climates from the CMIP5 dataset. Lastly, we bootstrapped both the climate model ensemble average and the ICU model parameters to get probability density functions of the growth in projected average days spent in the ICU.
Results
Between January 2003 and December 2019, there were 2,004,575 admissions to 187 ICUs reported to the ANZICS Adult Patient Database. 584,454 admissions were excluded from sites that did not contribute every year. A further 74,747 admissions were excluded due to potential administrative errors. In other words, 1,345,374 admissions from 81 ICUs (Figure 1) were considered between 2003 and December 2019. Among all ICU admission categories, respiratory-related admissions were significantly correlated with a number of climate variables. In total, 148,638 ICU admissions were respiratory-related (Table 1). The climate variable which showed the strongest correlation was the wet-bulb-globe temperature (WBGT).
Table 1Patients admitted to Australian ICU between 2003-2019.
Characteristic
n (%)
Respiratory related intensive care unit (ICU) Admission
Figure 2a illustrates the significant relationship between the annual growth rate in ICU average days per capita (ICUD/C) and the annual growth rate in the population density-weighted WBGT over Australia (r=0.771; p<0.001). The corresponding probability density functions of the slope, intercept, and correlation coefficient for WBGT are shown in Figure 2b-2d.
Figure 2Relationship between the annual growth rate in ICU admissions per capita and population density-weighted WBGT.
(a) raw samples; and results from non-parametric bootstrapping producing probability density functions of the (b) slope; (c) intercept; and (d) the associated correlation coefficient.
Abbreviations: ICU, intensive care unit; WBGT, wet-bulb-globe-temperature.
Figure 3 illustrates the cumulative growth in the population density-weighted WBGT since 2020 for each of the three scenarios (RCP2.6–blue, RCP4.5–grey, RCP8.5–red). The solid line for each scenario is the mean of all 20 climate models each year, and the shaded range was bounded by the 10% and 90% quantile across the model ensemble in that year. Figure 4 is the probability density functions of the climate variables. It demonstrates the probability density functions of the ensemble-averaged annual growth rate in the population density-weighted WBGT in the years 2030, 2050, 2070 and 2090 for each of the three scenarios. The probability density functions become more distinct as the years progress, demonstrating the significant differences in WBGT in later years as the climate change signal becomes stronger.
Figure 3Percentage growth in population density-weighted WBGT.
Projected percentage growth in population density-weighted WBGT since 2020 for CMIP5 climate scenarios RCP2.6 (blue), RCP4.5 (grey), and RCP8.5 (red). The solid line in each figure is the mean of 20 different climate models and the range is bounded by the 10% and 90% quantile across the model ensemble in that year.
From left to right: Year 2030, 2050, 2070 and 2090. WBGT was predicted for CMIP5 climate scenarios RCP2.6 (blue), RCP4.5 (grey), and RCP8.5 (red). The dashed line represents the mean growth of the individual climate model.
Figure 5 presents the bootstrapped results of the projected respiratory-related ICUD/C, and the projected respiratory-related ICUD/C until 2090 (dash lines). In other words, it depicts the probability density functions of the ICU demand as a result of the WBGT. By 2030, we projected a slight increase (5.0%, 6.7% and 7.7% for RCP2.6, RCP4.5 and RCP8.5, respectively) in respiratory-related ICUD/C. By 2050, respiratory-related ICUD/C will continue to increase under all scenarios. Under scenarios RCP2.6, RCP4.5 and RCP 8.5, the projections were 16.8%, 21.4% and 26.2%, respectively. By 2070 ICUD/C is up by 25.3%, 36.1% and 48.3% for RCP2.6, RCP4.5 and RCP8.5, respectively. By 2090, ICUD/C for both lower temperature models RCP2.6 and RCP4.5 are stabilising with only marginal increases from 2070 (33.4% and 44.5%, respectively). On the other hand, the RCP8.5 scenario projects substantial growth in ICUD/C (66.1%). This is increased to 2.0 times ICUD/C as projected by RCP2.6. Interestingly, the differences in ICUD/C between RCP2.6 and RCP4.5 remained constant, whereas we projected widening differences in ICUD/C for carbon climate scenario RCP8.5 (ICUD/C|RCP4.5/ICUD/C|RCP2.6 = 1.3 and ICUD/C|RCP8.5/ICUD/C|RCP2.6 = 1.5 by 2030; ICUD/C|RCP4.5/ICUD/C|RCP2.6 is 1.3 and ICUD/C|RCP8.5/ICUD/C|RCP2.6 = 2.0 by 2090).
Figure 5Probability density functions of annual growth rates in respiratory-related ICU average stay per capita (ICUD/C) since 2020.
From left to right: Year 2030, 2050, 2070 and 2090. Growth in ICUD/C was predicted for CMIP5 climate scenarios RCP2.6 (blue), RCP4.5 (grey), and RCP8.5 (red). The dashed line represents the mean growth of the individual climate model.
Abbreviations: ICU, intensive care unit; RCP, Representative Concentration Pathways; CMIP5, fifth Coupled Model Inter-comparison Project.
This study presents the first approach to estimating ICU demand projections (sum of the ICU days of stay per capita [ICUD/C]) based on individual climate variables. The novelty of this work lies in the uniqueness of the Australian climate conditions and the potential capability of the discovered correlations between historical respiratory ICU admissions (due to the comprehensive ANZICS database) and individual climate variables for better long-term health care resources management and infrastructure planning. We then applied these relationships to three potential future climate scenarios to estimate the implication of climate change for Australian respiratory disease. Over the next 70 years, our model projected a persistent, monotonic increase in respiratory-related ICUD/C under all three carbon emission climate scenarios, with an increasingly widening gap in projected ICUD/C between the lower and higher carbon-intensive scenarios. Specifically, from the years 2030 to 2090, the ratios ICUD/C|RCP4.5/ICUD/C|RCP2.6 remained unchanged, but ICUD/C|RCP8.5/ICUD/C|RCP2.6 grew from 1.5 to 2.0.
The reported associations projected hospital demand in the near future and helped allocate immediate hospital resources (e.g., staff, beds, etc.) [
]. These associations were not designed to project the influence of climate on hospital demand over multiple decades. Alternate approaches are required for long-term resource management and planning of hospital infrastructure. Our focus here has been on the persistent long-lived nationwide impacts, that the climate has upon the population.
In geophysical sciences, the term projection refers to the response of a climate simulation over multiple decades [
Commonwealth Science Industrial Research Organisation. Guiding Principles In Climate change in Australia: Climate information, projections, tools and data 2020.
Near-term climate change: projections and predictability.
in: Stocker T.F. Qin D. Plattner G.-K. Tignor M. Allen S.K. Boschung J. Climate change 2013 – the physical science basis: Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge2014: 953-1028
]. These numerical simulations are subject to prescribed future carbon emissions, associated with an assumed level of economic development. This is distinct from the term forecast, which refers to the predicted evolution of the weather and short-term climate from currently known conditions over days to a few years at most [
]. Over these shorter time scales any differences in the carbon emissions across the various scenarios, are small in comparison to the natural variability of the climate (e.g., seasons, cycle of El Niño and La Niña). Therefore, the term projection is adopted in the current study when referring to future estimates of the climate and ICU demand, which we made over a period of decades.
Since the COVID-19 pandemic, demand for ICU beds has increased dramatically. In Australia alone, although the potential increase in ICU beds would be tripled [
], ICU capacity remained limited by a constant shortfall in appropriately trained staff, and differences in resources level (between states, regional and private hospitals) to maintain pre-COVID-19 models of care [
]. As a result, there was an urgent call to better coordinate ICU staff and resources across the country with/without the pandemic. The influence of climate across the economy is pervasive, with a growing body of research on its impacts on agriculture, food security [
]. The relationship between WBGT (i.e., humidity and temperature) and labour productivity has also been documented. An increase in WBGT reduces the capacity of the able-bodied to do physical work [
]. We argue that our results are consistent with the notion that an increase in WBGT also increases the risk for those potentially predisposed to respiratory problems to require medical care.
This study provides the necessary analyses and steps to build a robust climate-health care projection model. However, having access to climate data over a historical period means that there is potential to further inspect each of the elements of the Earth system (e.g., weather, seasonality, multi-year variability, climate change) [
], and how they may influence a range of medical conditions. Since different climate phenomena would potentially influence human health in alternate ways [
The association between diurnal temperature range and emergency room admissions for cardiovascular, respiratory, digestive and genitourinary disease among the elderly: a time series study.
], it is important to distinguish between various cycles and trends in the Earth system and incorporate these differences in our future projection models. This will ultimately refine the timeline over which healthcare decisions need to be made.
The present study has three key limitations.
First, we treated the entire Australian ICU population admitted with respiratory issues as one cohort. Future work will involve investigating additional diagnostic groups and whether specific demographics and geographical locations are more susceptible to specific climate variations. For example, while aging is normally associated with worsening ICU outcomes [
]. As such, our projection model should not be extrapolated to other diagnostic groups upon thorough investigations due to the complexity of the data.
Although projections can currently be made for the Australia-wide ICU days per capita, making projections for specific local ICUs based on their geographic locations would require alternate spatio-temporal approaches. Noting that the vast amount of land Australia covers in latitude and longitude could lead to various regions within Australia changing in alternate ways for a given future scenario.
One might also consider the impact of climate on other hospital admission types, such as those in the general wards and emergency departments. In addition, patients’ medical history is expected to play an important role in our projection. As a result, analyses of carefully classified subgroups might yield further discoveries.
The second limitation is related to our current exclusions on the impacts of COVID-19 since 2019. The uncertainties surrounding ICU admissions and related complications from COVID-19 will be the subject of further investigations.
Lastly, our projected ICUD/C represented the percentage growth in ICU per capita. Hence, the growth in actual additional hours will be compounded by the growth in the Australian population between now and the end of the century.
Conclusion
Annual growth in population density-weighted wet-bulb-globe temperature correlates with the annual growth in Australian ICUD/C for respiratory-related conditions. A model based on possible future climate scenarios can be developed to predict changes in ICU demand in response to CO2 changes. Informed decisions can then be made with this kind of data, which has the ability to communicate the long-term consequences of today’s decisions on our future health care system. Our findings imply that a good understanding of the relationship between ICU admissions and climate can be achieved. This also suggests that a robust projection model built from this relationship could play a key role in future climate-related health care adaptation and investment.
Conflict of Interest
The authors declare that they have no conflict of interest.
Acknowledgements
The authors and the Australian and New Zealand Intensive Care Society (ANZICS) Centre for Outcome and Resources Evaluation (ANZICS CORE) management committee would like to thank clinicians, data collectors and researchers at the following Australian contributing sites: Albury Base Hospital ICU, Vic; Alfred Hospital ICU, Vic; Alice Springs Hospital ICU, NT; Austin Hospital ICU, Vic; Ballarat Health Services ICU, Vic; Bankstown-Lidcombe Hospital ICU, NSW; Bendigo Health Care Group ICU, Vic; Blacktown Hospital ICU, NSW; Box Hill Hospital ICU, Vic; Bundaberg Base Hospital ICU, Qld; Caboolture Hospital ICU, Qld; Cairns Hospital ICU, Qld; Calvary Hospital (Canberra) ICU, ACT; Calvary Mater Newcastle ICU, NSW; Calvary Wakefield Hospital (Adelaide) ICU, SA; Canberra Hospital ICU, ACT; Coffs Harbour Health Campus ICU, NSW; Concord Hospital (Sydney) ICU, NSW; Dandenong Hospital ICU, Vic; Epworth Freemasons Hospital ICU, Vic; Flinders Medical Centre ICU, SA; Frankston Hospital ICU. Vic; Gold Coast University Hospital ICU, Qld; Gosford Hospital ICU, NSW; Goulburn Valley Health ICU, Vic; Grafton Base Hospital ICU, NSW; Hornsby Ku-ring-gai Hospital ICU, NSW; Ipswich Hospital ICU, Qld; John Flynn Private Hospital ICU, Qld; John Hunter Hospital ICU, NSW; Knox Private Hospital ICU, Vic; Latrobe Regional Hospital ICU, Vic; Launceston General Hospital ICU, Tas; Lismore Base Hospital ICU, NSW; Liverpool Hospital ICU, NSW; Logan Hospital ICU, Qld; Maroondah Hospital ICU, Vic; Mater Adults Hospital (Brisbane) ICU, Qld; Mater Health Services North Queensland ICU, Qld; Mater Private Hospital (Brisbane) ICU, Qld; Melbourne Private Hospital ICU, Vic; Monash Medical Centre-Clayton Campus ICU, Vic; Mount Hospital ICU, WA; Nepean Hospital ICU, NSW; North Shore Private Hospital ICU, NSW; Northeast Health Wangaratta ICU, Vic; Orange Base Hospital ICU, NSW; Prince of Wales Private Hospital (Sydney) ICU, NSW; Queen Elizabeth II Jubilee Hospital ICU, Qld; Redcliffe Hospital ICU, Qld; Rockhampton Hospital ICU, Qld; Royal Adelaide Hospital ICU, SA; Royal Brisbane and Women’s Hospital ICU, Qld; Royal Darwin Hospital ICU, NT; Royal Hobart Hospital ICU, Tas; Royal Melbourne Hospital ICU, Vic; Royal North Shore Hospital ICU, NSW; Royal Perth Hospital ICU, WA; Royal Prince Alfred Hospital ICU, NSW; Shoalhaven Hospital ICU, NSW; Sir Charles Gairdner Hospital ICU, WA; St Andrew’s Hospital (Adelaide) ICU, SA; St Andrew’s War Memorial Hospital ICU, Qld; St George Hospital (Sydney) ICU, NSW; St George Private Hospital (Sydney) ICU, NSW; St Vincent’s Hospital (Melbourne) ICU, Vic; St Vincent’s Hospital (Sydney) ICU, NSW; Sutherland Hospital & Community Health Services ICU, NSW; Sydney Adventist Hospital ICU, NSW; Tamworth Base Hospital ICU, NSW; The Northern Hospital ICU, Vic; The Prince Charles Hospital ICU, Qld; The Townsville Hospital ICU, Qld; The Valley Private Hospital ICU, Vic; The Wesley Hospital ICU, Qld; Toowoomba Hospital ICU, Qld; Tweed Heads District Hospital ICU, NSW; University Hospital Geelong ICU, Vic; Westmead Hospital ICU, NSW; Westmead Private Hospital ICU, NSW; and Wollongong Hospital ICU. NSW. V. Kitsios is funded by the CSIRO Artificial Intelligence for Missions program.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed Consent
Informed consent was obtained from all individual participants included in the study.
Consent Statement
This research was reviewed and approved as a low-risk study by the Human Research and Ethics Committee of The Alfred Hospital, Melbourne, Australia (HREC number 631-20). The need for informed consent was waived as this study used deidentified data already collected by the ANZICS Clinical Quality Registry.
References
Zhang Y.
Beggs P.J.
Bambrick H.
Berry H.L.
Linnenluecke M.K.
Trueck S.
et al.
The MJA-Lancet countdown on health and climate change: Australian policy inaction threatens lives.
IPCC. Climate Change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New York, USA; 2013. Available from: https://www.ipcc.ch/report/ar5/wg1/. Accessed November 22, 2022.
The association between diurnal temperature range and emergency room admissions for cardiovascular, respiratory, digestive and genitourinary disease among the elderly: a time series study.
Excess emergency department visits for cardiovascular and respiratory diseases during the 2019-20 bushfire period in Australia: a two-stage interrupted time-series analysis.
Taylor KE, Stouffer RJ, and Meehl GA. An overview of CMIP5 and the experiment design. Bulletin of the American Meteorological Society. 2012;93(4):485-98.
Commonwealth Science Industrial Research Organisation. Guiding Principles In Climate change in Australia: Climate information, projections, tools and data 2020.
Near-term climate change: projections and predictability.
in: Stocker T.F. Qin D. Plattner G.-K. Tignor M. Allen S.K. Boschung J. Climate change 2013 – the physical science basis: Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge2014: 953-1028
01233-1&id=gr1.jpg){kind=link}
01233-1&id=gr2.jpg){kind=link}
01233-1&id=gr3.jpg){kind=link}
01233-1&id=gr4.jpg){kind=link}
01233-1&id=gr5.jpg){kind=link}