AI technology is crucial for identifying gaps in global AMR research, promoting a unified One Health approach, and guiding future efforts to combat the AMR threat effectively.
A new study co-lead by Newcastle’s Professor David W Graham demonstrates how AI can highlight the urgent need for better coordination in AMR research methods across different sectors and regions.
Contents:
- What is AMR?
- What is One Health?
- The issue at hand
- The research requirement
- What did the research aim to do?
- How was the research carried out?
- What were the findings?
- Next steps
What is AMR?
Antimicrobials are medicines, such as antibiotics, antivirals, antifungals, and antiparasitics. They’re used to prevent and treat infections in humans, animals, and plants.
Antimicrobial resistance (AMR) happens when bacteria, viruses, fungi, and parasites stop responding to these medicines. This makes antibiotics and other treatments ineffective, so infections become harder or sometimes impossible to treat. Because of this, diseases can spread more easily, and the risk of severe illness, disability, and death can increase.
Antimicrobial resistance is a growing threat to global health. Bacterial pathogens that display AMR traits are causing more untreatable infections and more prolonged infections, therefore affecting humans, animals, our food supply, the environment, and the economy.
An estimated 1.27 million extra deaths were directly linked to AMR in 2019 and almost five million indirect deaths - numbers that could exceed 10 million annually by 2050. AMR also affects work and productivity - it’s suggested that AMR causes an extra 3.5 million sick days per year, due to waterborne antimicrobial resistance.
What is One Health?
One Health is an approach that sees the health of humans, animals, plants, and the environment as interconnected.
Beginning in the early 2000s, the approach aims to address health threats by considering these connections. In 2022, the United Nations Environment Programme (UNEP) joined the World Health Organization’s (WHO) effort, creating a joint plan to tackle global health issues like AMR by integrating evidence from multiple sectors.
The issue at hand
AMR can spread from the environment (like air, water, and soil), animals, and agricultural products, to humans. Waste from healthcare, communities, and animals can also mediate AMR transfer into the environment. These pathways increase the chance of new AMR genes and strains developing.
Socioeconomic factors, such as healthcare quality and sanitation, water quality, and hygiene infrastructure - especially in low-income countries where population sizes and the demand for antibiotics is rising - also play a role in increasing the spread of AMR.
As part of the new study by the Chinese Academy of Sciences and Newcastle University, researchers discovered that the methods and terms used in AMR research varied greatly between medical, veterinary, food safety, plant agriculture, and environmental sectors. These differences reduce cross-sectoral collaboration and communication between fields, and inconsistent messages to decision-makers as a result.
The research requirement
To tackle AMR effectively, researchers need a holistic approach that includes studying humans, animals, plants, food, and ecosystems in tandem.
In the past, studies have often focused on specific sectors or regions. One of the main problems is that the volume of previous studies and data are huge – therefore requiring better methods for integrating knowledge, increasing the value of previous research, and prioritising new activities.
Artificial Intelligence (AI) technologies - like Natural Language Processing (NLP) and Large Language Models (LLM) – can help by organising and analysing the vast amount of existing and rapidly growing research.
What did the research aim to do?
Co-leading the study, Chinese Academy of Sciences’ Professor Yong-Guan Zhu, and Emeritus Professor of Engineering at Newcastle University, David W Graham, aimed to shed light on global patterns of AMR research.
The research looked at using AI to analyse a large body of AMR literature. By using NLP, the study mapped global patterns in AMR research, identifying gaps in knowledge and encouraging more integrated One Health research.
AI could quickly and accurately categorise information from thousands of studies, helping them to understand global AMR research trends and guide future work.
How was the research carried out?
The team created a new tool and database, examining 254,738 AMR publications from 178 countries between 2003 and 2023.
Using AI methods, they were able to sort and analyse global AMR research patterns, producing detailed maps of AMR studies around the world. This information will therefore be crucial for developing solutions based on a One Health approach.
.jpg?width=600&height=243&name=global-mapping-amr-research-publications%20(1).jpg)
Global mapping of publication numbers related to MRSA.
What were the findings?
The study has provided valuable insights into global AMR patterns based on research trends. Not only this, but it has demonstrated how AI-based research methods can be used in a beneficial and socially constructive way to combine data from various cross-disciplinary and cross-sectoral studies. Professor David W Graham, said:
“The findings highlight the urgent need for greater coordination in research methods across sectors and regions. For instance, the medical and veterinary communities need information about living AMR infectious pathogens to prioritise decisions, whereas environmental researchers often focus on genetic targets.”
It allows for equitable data sharing across sectors and countries, highlighting the importance of combining culture-based and genomic AMR analysis across all sectors.
Next steps
The database developed in this study, which is editable and open for collaborations, and AI methodology can be used to create new tools and practises, as well as enhance information sharing.
“Our paper's findings support key messages from the UN Environment Programme and World Health Organization that emphasise the best way to mitigate AMR is through prevention and integrated surveillance, which is key to prioritising solutions.”
The new research will likely help shape policies and collaboration efforts to combat the threat of AMR. The AI methods used are the first of their kind to collect and analyse global AMR research and can be updated in the future as new research emerges.
Both scientists recommend continued research and increased investment in capacity development, particularly in low-income countries to address the pressing AMR challenges in regions where the burden is often highest.
You might also like
- discover full details of the research, in the paper: Cai Chen, Shu-Le Li, Yao-Yang Xu, Jue Liu, David W. Graham, Yong-Guan Zhu, Characterising global antimicrobial resistance research explains why One Health solutions are slow in development: An application of AI-based gap analysis, Environment International, Volume 187, 2024, 108680, ISSN 0160-4120, https://doi.org/10.1016/j.envint.2024.108680
- learn more about Professor David Graham, Professor of Ecosystems Engineering at Newcastle University
- read the press release
- explore Newcastle University’s School of Engineering
- find out more about antimicrobial resistance via the World Health Organization
