TL;DR
Researchers studying the intersection of climate change and health face significant data and infrastructure challenges. A large-scale survey revealed that the most desired resources are analysis-ready datasets and educational materials, while lack of time is a major barrier to data sharing. Investing in professional data creation services, knowledge sharing, and robust infrastructure could significantly accelerate progress in this critical field.
What Problem Does This Paper Address?
The world's climate is changing, leading to a cascade of health impacts, from extreme heat and weather events to food insecurity and the spread of infectious diseases. Understanding and mitigating these impacts requires interdisciplinary research, bringing together experts from earth science, public health, and data science. However, this collaboration is hampered by the complexity of data, methods, and software, as well as challenges in data accessibility and sharing. This paper aims to identify the specific barriers and needs of researchers working in this domain to guide future investments and infrastructure development.
Methods
The researchers conducted a large-scale online survey of individuals who have published research at the intersection of climate and health. They identified potential participants by querying PubMed for relevant articles published between 2021 and 2023 and then scraped email addresses of authors. A total of 9195 authors were invited to participate in the survey, which took approximately 7 minutes to complete. The survey collected information on participants' demographics, career stage, research focus (US vs. non-US populations), data and software needs, and barriers to data and code use and sharing. The response rate was 11.8%, yielding 1041 usable responses. Statistical analysis was performed on quantitative data, and free-text responses were analyzed qualitatively to identify emergent themes.
Datasets
Not discussed in the paper. The paper focuses on a survey of researchers, not on the analysis of specific climate or health datasets.
Key Insights and Ideas
- The "Data Divide" is Real: Researchers across all career stages and geographic focuses struggle with data availability, particularly concerning spatial and temporal resolution. This highlights a fundamental need for more accessible and appropriately scaled data.
- Beyond Data Availability: Expertise Matters: Simply having data isn't enough. A significant challenge is the lack of software and technical expertise to process and analyze climate and health data, indicating a need for enhanced training and mentorship.
- Global Perspective is Crucial: The majority of authors publishing in this field are not solely focused on US data, underscoring the necessity for global data efforts and infrastructure to support international collaboration.
- Time is the Ultimate Barrier: Personal constraints, especially lack of time, are the most frequently cited barriers to sharing both data and code. This suggests that making data and code sharing easier and more efficient is paramount.
- Willingness to Share is High, but Infrastructure is Lacking: A majority of respondents expressed a willingness to share their data and code in the future, but current practices for data and code deposition are low, with many researchers relying on less standardized methods like personal websites or direct requests.
- "Analysis-Ready" is the Holy Grail: The most desired resource by far is "analysis-ready" climate and health datasets, indicating a strong demand for curated, pre-processed data that researchers can use directly.
Key Results
- Response Rate: 11.8% response rate from 9195 invited authors, resulting in 1041 usable responses.
- Geographic Diversity: Respondents from over 75 countries, with only 16.4% primarily working with US populations.
- Career Stages: Evenly represented across early (38.1%), mid-career (30.2%), and established career (31.7%) professionals.
- Most Desired Resources: Analysis-ready climate/health-related datasets (96% interested), live webinars (88%), and video tutorials/workshops (87%).
- Barriers to Data Use/Sharing:
- Data not in the needed spatial/temporal resolution (31% of all respondents).
- Lack of software or expertise (19%).
- Personal constraints (e.g., no time to curate data) (23% for data sharing).
- IRB or DUA constraints (28% for data sharing).
- Data Sharing Practices: Only 22% of respondents deposit their data to a data sharing platform.
- Code Sharing Practices: Similar to data, only 15% deposit their code on a sharing platform.
- Willingness to Share: A majority are willing to share data (55%) and code (53%) in the future.
Limitations and Future Work
The authors acknowledge that the survey response rate was low (11.8%), which may limit the generalizability of the findings to the entire climate and health research community. However, they argue that the large number of responses (1041) provides a valuable snapshot. They also note that their PubMed-based recruitment strategy might over-represent health-focused researchers compared to those primarily from climate science.
Future work should focus on addressing the identified barriers. This includes:
- Investing in Data Creation as a Service: Developing professional services for data curation and harmonization to alleviate the burden on individual researchers.
- Enhancing Knowledge Sharing and Collaboration: Creating platforms and opportunities for interdisciplinary communication and learning.
- Building Robust Research Infrastructure: Developing and promoting accessible, FAIR-compliant data repositories and code-sharing platforms.
- Developing Educational Resources: Providing more training on data management, analysis, and the use of tools like GitHub.
- Addressing Structural Barriers: Exploring solutions for IRB/DUA constraints and advocating for academic systems that value data and code sharing.
- Adopting a Global Lens: Ensuring that future data initiatives are inclusive and address data gaps in low- and middle-income countries.
Summary updated