The final “basket” of eight indicators rated the highest for quantifying health outcomes related to climate change in Canada based on the five evaluation criteria used in this report. Half of the indicators are modeled, which offers both benefits and challenges. Modeled indicators are defined, in the context of this research, as indicators that require statistical calculations and modeling based on collected data. For example, a city may have the data for daily mortality rates and temperature measurements. In order to calculate excess mortality due to heat, modeling on past rates of mortality needs to be performed to determine the expected mortality rates, which can then be analyzed against temperature data to determine the excess mortality due to heat. Modeled indicators are valuable in quantifying the health impacts of climate change because they are better defined in terms of their specificity (i.e., their relationship to climate change), because they can incorporate emerging scientific findings about exposures to climate hazards, and because they are particularly useful in describing projected climate change effects on health. However, modeled indicators are often less comparable than non-modeled indicators and they vary in quality depending on calculation methods; they also vary in quality due to the greater uncertainty associated with necessary assumptions that must be made to calculate them. Also, problematic for public health adaptation applications, the modeled indicators identified here are not currently available in for many communities in Canada. Considering the value of these indicators, this paper makes a strong case for their development.
The non-modeled indicators identified are more readily available for immediate use by Canadian public health authorities. They are particularly valuable in the assessment of climate change-related public health outcomes for local/municipal governments and health units because they are easy to use and analyze. Despite the relative simplicity of the four non-modeled indicators, however, daily all-cause mortality and daily non-accidental mortality due to heat and air pollution require time-series analysis and statistical processing including adjustment for confounders before they can be used for climate change and health studies. For those reasons, daily mortality data may be more suitable for use on an annual basis at provincial/territorial or federal agencies that have greater capacities for statistical support and analysis. At the local level, the most readily available data from the basket of indicators are West Nile and Lyme Disease incidence in humans. Throughout Canada, these diseases are reportable diseases to public health, although issues with underreporting exist (Public Health Agency of Canada 2005).
In addition to the proposed basket of eight indicators, local health units may find others important to use to monitor climate change impacts on health over time or to inform assessments. In Table 10, all 77 indicators rated are listed in descending order based on their individual scores; health authorities may use this list to identify other indicators tailored to their specific needs. For example, a certain health unit may have local data on asthma incidence and prevalence, which rated well and, therefore, would be useful for their jurisdiction. Despite the efforts of this study, rating scores may not be the same for each level of government in varying areas in Canada. We encourage the users to incorporate local knowledge in using this list of indicators.
The indicators involved in this study were complex due to technical and methodological differences. Even though indicators may have similar names in different reports and articles, for instance, various researchers may use different collection and calculation methods. Consequently, two reports may refer to different relationships while referring to the same indicator. At this current stage, we acknowledge that this project was not able to address this issue. In our rankings of quality, much like other criteria, we attempted rankings based on general problems with indicators, relative to each other, rather than attending to methodological distinctions. For example, for mortality indicators due to certain causes, we have rated them lower than mortality indicators not due to specific causes, based on the reasoning that identifying cause of death is often difficult and contains greater room for error than identifying a death without specifying a cause. Each user of the indicators may consider these issues for their jurisdiction and be fully aware of potential difficulties in calculation, definition, and maintenance of the indicators. As indicators are adopted for use, we will understand more clearly what issues are the most prominent and find solutions accordingly.
The multifaceted relationship between climate change, public health, and health indicators adds another layer of complexity to this research. The relationship is complex because most climate-sensitive health impacts have many causal factors in addition to climate and because the causal relationships between climate change and public health can be indirect and non-linear. In some instances there is also a lag time between exposure and response, which can affect certain groups differently and can result in varying degrees of confidence when trying to link climate change and health (McMichael 1995; Ebi 2008; Vineis 2010; Xun et al. 2010; Forsberg et al. 2012). Indentifying useful indicators is also difficult because of the unpredictability of climate change itself. As current scientific findings indicate, the future health impacts from climate change will likely worsen and Canadians may face, in certain regions or communities, new health threats that have not yet been experienced (e.g., exotic disease) (Seguin 2008). Additionally, as the body of research surrounding climate change grows, health effects currently unknown or considered of less concern may receive more attention from researchers. For example, there may be a need to consider both the acute and chronic health outcomes to fully understand the risks facing Canadians. While this paper does not clearly separate the health effects into acute and chronic, we have included indicators that can belong to both categories based on current knowledge. For example, while heat and air pollution related health effects may be immediately visible in the data (acute), it may take a number of years to identify a changing trend in the pattern of infectious diseases such as Lyme borreliosis (chronic). The relationship between climate change and health is dynamic and a rapidly expanding area of science; this current basket of indicators will need to be re-evaluated and modified on a regular basis based on new scientific findings.
Last, the project reported here did not investigate all possible health effects influenced by climate change. Some health outcomes, such as the health effects of harmful algal blooms and Arctic glacial melting, have not been considered through the literature review and rating of indicators. We have limited the scope of this report to the health effects identified in Fig. 1 due to resource restrictions. We were also not able to address special population effects within health outcomes, i.e. vulnerable groups and special populations. Further research can expand the scope of this project to include a larger range of health effects and address the differential health effects of climate change on different population groups.
In conclusion, in proposing this basket of eight indicators for quantifying climate change health effects, we make two overarching observations. The first is that we currently lack modeled indicators and models for evaluating the health effects of climate change even though modeled indicators are the best at informing action; consequently, we argue that there is a need to develop these indicators. The second main observation regarding the use of these indicators is that each jurisdiction may need to customize the use of them depending on a variety of local factors including resources, data availability, and applicability to the specific group served; each jurisdiction will also need to continuously evaluate and modify their analysis.
In responding to the threats of climate change, this project helps in efforts to monitor, assess, and project human health effects through quantifying such effects, as well as through the evaluation of existing and planned adaptation efforts. This is a promising step for assisting Canadians in climate change adaptation, and the process may also be applicable in other countries. We propose the following research needs as next steps to help further adaption to the effects of this national and global problem:
modeled indicators in Canada and globally (including extreme heat and air pollution-related excess and premature deaths).
Streamlined and consistent definitions and calculations for modeled indicators.
Investigation of other health outcomes from climate change reflecting the changing importance of each health outcome and ;
expansion of the indicators to include other useful areas such as differential effects of climate change on different groups (vulnerabilities and adaptive capacity).