Evaluating Heat Health in Cities with the cityClimateHealth Tool

A statistical R-package to calculate health outcomes, including the impacts of extreme heat, and to identify vulnerable subgroups using health data in cities.
This tool can assist with analysis of any kind of acute health outcome from any type of weather exposure (i.e. heat, cold, flood, power outage, storm). Let's dive deeper into how we use it to consider the impact of extreme heat on populations.
When considering extreme heat, many gaps make it challenging to determine the true health impact on city residents.
  • Gap 1: Cities undercount the true burden of heat. Most health data use hospital codes to identify heat-related emergency department visits, which undercounts the true impact by not capturing everyone who is present for heat-related health impacts.
  • Gap 2: Cities want to know who is most vulnerable to heat (i.e. old, young, on public insurance). This can lead to interventions and outreach to populations most at risk for health impacts from extreme heat.
  • Gap 3: Cities need data on the scale that they can act on. In order for cities to be able to implement interventions, they need data at the appropriate sub-city scale for those interventions, for example neighborhood.
  • Gap 4: Health data are confidential, and sharing confidential health data for external parties to analyze is challenging. Rules around patient data confidentiality limit how much data can be shared with researchers who have the statistical tools and knowledge to analyze them.

The cityClimateHealth tool addresses these gaps!

What is the tool?

The 'cityClimateHealth' R package makes it easy to estimate any acute health outcome (such as emergency department visits), that are related to a weather exposure (such as extreme heat) by applying state-of-the-art methodology into an easy-to-use format that runs quickly on a user's laptop without the need to share the data externally. 
This tool, developed by a research team from Boston University's Center for Climate and Health and the CATCH project, simplifies the process for estimating detailed and specific health impacts from weather exposures like extreme heat across subgroups, such as demographics like age.  

What expertise is needed to run the tool? 

Familiarity with R, the ability to manage sensitive health data, and background knowledge regarding the calculations of relative risk and attributable numbers will be essential for using the tool to its full potential.

Who can use it?

Anyone with R can test this tool through a series of vignettes on the cityClimateHealth site. Anyone with health data (for example: departments of health or researchers) can work through their own location specific case study with their own exposure and outcome data.

What kind of data do I need to be able to use the tool?

In order to run this tool, you will need time-series data on your daily weather exposure (ex: daily temperatures), as well as time-series data of your health outcome of interest (also measured daily). Although the examples below use daily maximum temperature and daily aggregated Emergency Department visits counts, other daily exposures and outcomes are easy to implement (e.g., daily mortality).

How it works

The 'cityClimateHealth' R package takes these inputs and provides the user with relative risks of your health outcome of interest due to the weather exposure of interest (in our example: emergency department visits due to increasing heat) across different demographics, such as age, as well as an estimate of the number of that acute health impact that the user can attribute to the weather exposure for the location of the data and the demographics the user provided. This allows for understanding of where and who are at greatest risk from weather exposures like extreme heat, informing targeted interventions at the local level.

For more details on how to use this tool, and to get started, click the button below!

A real world example in Massachusetts

The cityClimateHealth tool was used in collaboration with the Metropolitan Area Planning Council (MAPC) Lower Mystic Cool Communications Project, which aimed to build heat resilience in the Lower Mystic communities of Chelsea, Everett, Malden, Revere, and Winthrop with local data and a customized heat communication tool kit. An analysis of the 5 communities revealed disparities in health outcomes across key demographic groups and will aid in future communication and interventions. Figure 1 (below) shows the annual heat-attributable all-cause ED visit number (red) and rates per capita (blue) among the populations of Boston, Chelsea Everett, Malden, Revere, and Winthrop. In this example, the difference in rates across areas can help municipal leaders understand how their health outcomes compare to their neighbors and make key resource decisions based on the size of the impact in their municipalities. One key take away here is that municipalities like Chelsea and Everett, despite their smaller populations, have similar rates compared to urban centers like Boston. You can learn more about the project here. The full analysis can be found here.

Figure 1. City-specific totals of annual heat-attributable ED visits (red) and the annual per-capita rates of heat-attributable ED visits (light blue).
Hear more about the project in action below.
Speakers: 
Senior Research Scientist Chad Milando from Boston University
Mariangeli Echevarria-Ramos from MyRWA
Sharon Ron from MAPC