What are the next steps by which researchers can focus their work in modeling of health risks posed by climate change? Speakers and panelist offered a variety of answers to that question, including thoughts on model complexity and context, new paradigms, time scales, decision-making needs, and communication.
“We expect complexity; it’s not remarkable,” said Kristie Ebi, of ClimAdapt, LLC. Health systems and disease transmission are complicated, so the question is what to do with the complexity in modeling, she stated. Modelers need to address complexity both by defining the task at hand and by acknowledging climate change in the context of other drivers of health outcomes, noted Anthony Janetos, of Boston University’s Frederick S. Pardee Center for the Study of the Longer-Range Future. Janetos explained that many factors may be more important than climate change and variability in determining health outcomes. Thus, it is important to know that climate change “may have the capacity to influence only one or a few of the factors in important ways but not all of them,” he added. As an example, Jan Semenza, of the European Centre for Disease Prevention and Control, described a “thought experiment” that he published in 20119 in which he and colleagues identified three broad classes of drivers that are involved in unexpected infectious-disease outbreaks and pandemics (such as SARS and avian influenza): globalization and environmental change, social and demographic change, and public-health systems. He developed plausible scenarios that linked each of those drivers with a health outcome and then compared the results with what happened in real outbreak events. The comparison was “much more complicated than expected” and underlined the idea that when it comes to health, climate change should be considered within the broader context of other health drivers, Semenza said.
One health driver that should be accounted for in models of health risks posed by climate change is urbanization, Ebi said. Unplanned urban environments can exacerbate health risks. For example, the Aedes mosquito that carries dengue is now found “in places it shouldn’t be because it likes to overwinter in sewers and is transported around the world in used tires,” she explained. Models for dengue are growing in their ability to look at weather measures, “but we are not looking at urbanization although we know that it is important.” She stressed that including future scenarios of urbanization in models is an important next step.
Sari Kovats, of the Faculty of Public Health and Policy of the London School of Hygiene and Tropical Medicine, noted that integrated assessment models are good at quantifying effects on gross domestic product but that there is a big gap when it comes to quantifying
9Jonathan E. Suk and Jan C. Semenza. 2011. Future Infectious Disease Threats to Europe. American Journal of Public Health. 1010(11):2068-2079.
nonmarket effects. She suggested that adding health cobenefits to integrated assessment models may be easy to achieve in the short term because there are solid data to draw on; for example, the health risks posed by air pollution are well documented.
John Balbus, of the US National Institute of Environmental Health Sciences, emphasized that predictive modeling that includes social factors should also be developed. “We cannot be relevant without being based on demography and on a solid understanding of what is happening in populations and socioeconomics,” he said.
There is a large and robust health-modeling community, but the people in the community generally do not focus on ecologic drivers, Balbus noted. Complicating the issue is that the health community tends to be reactive and look backward and not to think so much about what may happen in the future, he said. “That mindset has to be changed,” he stressed. Ebi asked, How do we proceed with inadequate information? She said that the use of modeling to design surveillance programs to make it possible to avoid projected health effects of climate change will constitute an important paradigm change for the health community.
Janetos recommended a three-pronged approach. First, integrated modeling at a high level can help researchers to understand long-term trends and the evolution of climate–health linkages over several decades and in the developing world, he said. Second, systems modeling at a more detailed level can shed light on specific cases and perhaps help in assessing the value of particular interventions. Third, empirical research and modeling can help researchers to understand where sensitivities arise and the best methods for reducing vulnerability.
Along the way, everyone will benefit by embracing “getting things wrong”, said Anne Grambsch, of the US Environmental Protection Agency (EPA). “In the first work that we do, we are going to be wrong,” she said, pointing out that EPA’s BenMAP model for estimating the health and economic effects associated with air quality got things wrong at first. “It is when you are wrong that you actually make progress,” she continued.
Ebi noted that policy-makers “do not want to know what is happening about heat or diarrheal disease”. The type of question that they want an answer for is, What will happen in sub-Saharan Africa as water resources decline, crop yields fall, temperatures increases, and diarrheal disease and malnutrition increase? The “health outcome-by-outcome, region-by-region, and country-by-country” approach to modeling is not useful to many decision-makers. Stéphane Hallegatte, of the World Bank, agreed, providing example of the types of questions that the World Bank is exploring in its work on climate change and poverty (Box 5-1).
Janetos works with mayors on short-term and long-term planning for climate change. He pointed out that a mayor’s job is “to protect lives, livelihoods, and property”. Mayors are not necessarily concerned with understanding the scientific complexity of systems; they need a more broad-scale view of how climate change may affect their cities and people. The lack of a broad-scale view may be causing a “systematic underestimate of how big a problem climate change is,” Janetos said.
Interactions Between Climate Change and Poverty: Some Policy Questions
To illustrate the types of questions that decision-makers are asking about climate change and health, Stéphane Hallegatte described the World Bank’s activities regarding interactions between health and poverty. The World Bank has a target to reduce extreme poverty drastically by 2030. That means that fewer than 3% of the world’s population would be in extreme poverty in 15 years. The effect of climate change on the effort, however, is a subject of investigation to inform planning. Some questions about efforts to model the health risks posed by climate change:
- Can health effects of climate change influence whether the 2030 target is reached?
- Do approaches to eradicating poverty need to change in light of climate change?
- If the 2030 target is reached, would that avoid or mitigate health effects of climate change?
- If the 2030 target is not reached, would future efforts to eradcate poverty become more complex as the health effects of climate change materialize?
“Public health is a local matter, so the response [to health effects of climate change] needs to be a local response,” said Benjamin Beard, of the US Centers for Disease Control and Prevention. “We need better predictions at the local level.” Some municipalities, such as Los Angeles, are investing in training to help their public-health employees to understand the effects of climate change projected in their areas, said Richard Jackson, of the University of California, Los Angeles. He has been involved in efforts to train public-health officials in Los Angeles about facets of climate change, including vulnerable populations and adaptation management. Balbus noted, however, that information relevant to policy-makers will come from research on both local and global scales. “Many decisions that we make are at the local level, but many decision-making processes, such as benefit–cost analysis, require aggregation and integration that are large-scale,” he said. Both local and global are important.
A compelling case showing the value to policy-makers of incorporating more health into climate-change modeling would go a long way toward helping researchers to obtain funding for these efforts, Balbus said. Making the case for a long-term preventive strategy is daunting, he said, given that institutions may not be around as long as the timeframe that needs to be considered.
The Canadians’ success in engaging decision-makers, as described by Peter Berry, of Health Canada, is a reason for optimism. “I think we have been successful in changing the level of knowledge of public-health decision-makers” compared with 10 years ago, he said. He credits the institutional mechanisms that his country has put into place to link the various groups that can support model development and the generation of data.
One challenge is that modeling “has been very ad hoc,” Kovats noted. “No one has brought the community together to think about model parameters for a coordinated activity or how to network better among modeling communities,” she observed. However, Janetos cautioned that because modeling of health risks posed by climate change is in a nascent stage, model-comparability studies may yield results that are not understandable. Ebi disagreed: she suggested that a coordinated project with a common set of measures, such as time scale, or a variety of scenarios could be useful. Hallegatte noted that careful
consideration needs to be given to goals of comparability studies and to the types of models to include. He stressed that diversity of models is useful and that restricting some models to a specific set of measures could restrict innovation. “Comparability works for some things but not everything—and only when a field has reached a level of maturity,” he stated.
Balbus concluded that the workshop had made it clear that there are many pathways and communities and that the models needed by different constituencies are at different stages. “We are trying to change the paradigm of how we model risk to apply a systems approach . . . to bring in new teams and new ways of thinking.” The overall picture is heterogeneous and at times confusing, he stressed.
“Everyone is saying that if you want to raise the interest of policy-makers and decision-makers, you need to focus on the time between now and 2030, but we need to fight hard to say that there are messages for the 2080s and even beyond that matter for today’s decisions,” Hallegate argued. Ebi added that 2100 seems far away to some but that “children and grandchildren born now will see the consequences of actions we are taking.”
Institutions might not be around over the timeframes that we are discussing, let alone the people in current political offices, Balbus noted. In addition to building models of health risks posed by climate change, the scientific community needs to work at building a long-term preventive mindset among decision-makers, he concluded.