Supervisory team
LSHTM
- Lead supervisor: Hugh Sharma Waddington ([email protected])
- Supervisor: Shakoor Hajat ([email protected])
Nagasaki University
- Supervisor: Lina Madaniyazi ([email protected])
Project
Climate change has intensified the frequency and severity of extreme weather events.1,2 Whilst there are clear relationships between health and climate hazards such as heatwaves, wildfires and extreme precipitation events (floods and thunderstorms), there is a shortage of rigorous, actionable evidence about what works to protect populations from these events. Population level interventions such as weather warning systems provide a route for health protection.
There is also concern that compound weather events can exacerbate physical health conditions to a much greater degree than individual events among vulnerable groups.3 For example, recent extreme summers have demonstrated that the co-occurrence of such events can also overwhelm public health and other support services.4 Understanding the potential for interventions like weather warnings to help address these impacts is therefore critical for strengthening health system preparedness to current and future climate change.
Natural experiments (NEs) apply statistical approaches to analyse existing climate and health data to determine the effectiveness of ‘real-world’ interventions. There is great potential for NEs to infer causality of weather warning interventions because decision makers use thresholds on exogenous variables like weather metrics to issue them. In the UK, for example, heat health warnings are issued largely on temperature whereas in Japan it is the Wet Bulb Globe Temperature prediction (a combination of temperature, humidity, wind speed and radiant heat). When combined with big data on climate and health, there is an opportunity to conduct discontinuity analysis to evaluate the impacts of these interventions on health rigorously. These methods have been shown to produce the same findings as RCTs in comparable populations and settings – that is, causal inference without randomisation.5
This study aims to develop and test approaches for the rigorous evaluation of weather warnings on mortality and morbidity, in the UK and Japan, including compound weather events, using natural experimental methods. These two countries provide contrasting contexts—both experience extreme weather events but differ in climate, health infrastructure, housing stock and demographic profiles, and use different methods for the issuing of weather warnings—making them ideal for a comparative study of effectiveness. Japan’s long-standing disaster management strategies can be extended and adapted to enhance climate resilience, ensuring the lessons learned from earthquakes, typhoons, and tsunamis translate into better responses to climate events. These transferable strategies offer valuable insights to inform resilience-building efforts against compound climate events in both countries.
The study will leverage daily data from national health databases and integrate them with weather data in both countries. While the initial focus is on the UK and Japan, the findings will have broader implications for other countries facing similar risks. If the student has sufficient time, the analysis may be expanded to other countries, including low- and middle-income countries, as relevant data are available through the supervisors' collaborative network. The study will offer insights into health system vulnerabilities across diverse socioeconomic and climatic contexts, further contributing to global efforts to mitigate the health risks of climate change.
References
1. Hajat S, Proestos Y, Araya-Lopez J-L, Economou T, Lelieveld J. Current and future trends in heat-related mortality in the MENA region: a health impact assessment with bias-adjusted statistically downscaled CMIP6 (SSP-based) data and Bayesian inference. Lancet Planetary Health; 2023: doi: 10.1016/S2542-5196(23)00045-1.
2. Madaniyazi L, Armstrong B, Tobias A, et al. Seasonality of mortality under climate change: a multicountry projection study. The Lancet Planetary Health. 2024 Feb 1;8(2):e86-94.
3. AghaKouchak A, Chiang F, Huning LS, Love CA, Mallakpour I, Mazdiyasni O, Moftakhari H, Papalexiou SM, Ragno E, Sadegh M. Climate extremes and compound hazards in a warming world. Annual Review of Earth and Planetary Sciences. 2020 May 30;48(1):519-48.
4. Howarth C et al. (2024) Turning up the heat: Learning from the summer 2022 heatwaves in England to inform UK policy on extreme heat. London: Grantham Research Institute on Climate Change and the Environmental, London School of Economics and Political Science.
5. Sharma Waddington H, Villar PF, Valentine JC. (2022) Can Non-Randomised Studies of Interventions Provide Unbiased Effect Estimates? A Systematic Review of Internal Replication Studies. Eval Rev. 2023 Jun;47(3):563-593. doi: 10.1177/0193841X221116721. Epub 2022 Sep 1. PMID: 36047928; PMCID: PMC10186563.
The role of LSHTM and NU in this collaborative project
The student will be based primarily in LSHTM, with a minimum of six months at Nagasaki for training, data analysis, and supervisory guidance.
HSW will supervise the natural experimental evaluation design, conduct and reporting, including toolkit development and production of case studies. LM and SH will facilitate data collection for Japan and the UK on environmental exposures and health outcomes, while also offering in-depth subject and methodological supervision.
The student will also benefit from related work being undertaken within the LSHTM HPRU project on Climate Security and Health, including access to data and decision-making at UK Health Security Agency and the Met Office.
Skills we expect a student to develop/acquire whilst pursuing this project
The student will develop expertise in environmental epidemiology, natural experiments using advanced statistical modelling techniques, and climate-health research, with a particular focus on the health impacts of climate-related disasters across different countries. They will also gain valuable experience working with large environment and health datasets. Finally, they will gain policy facing experience working with decision making bodies such as UKHSA and the Met Office.
Particular prior educational requirements for a student undertaking this project
The student should possess a Master’s level qualification in public health, economics, epidemiology, or medical statistics. However, candidates with another Master’s level qualification, as long as it includes a robust epidemiological and/or statistical component, will also be considered.
Additionally, experience working with programming languages such as Stata or R is highly desirable.