Awareness, adoption, and predictors of artificial intelligence use among public health researchers in Bo District, Sierra Leone: a cross-sectional study
Despite high awareness of artificial intelligence among public health researchers in Bo District, Sierra Leone, actual adoption remains low and is not significantly predicted by individual factors like training or experience, suggesting that structural and infrastructural barriers are the primary determinants of this awareness–adoption gap.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the vast and often quiet world of public health research, scientists work to understand how diseases spread, why people get sick, and how to keep communities healthy. For decades, this work has relied on careful observation, surveys, and statistical analysis to turn raw numbers into life-saving knowledge. Recently, a powerful new tool has entered the scene: artificial intelligence. This is not a robot that thinks like a human, but rather a sophisticated computer system designed to find patterns, learn from data, and make predictions much faster than a person could. In wealthy nations, these systems are already helping researchers spot disease outbreaks early and design better treatments. But in many parts of the world, where resources are tight and internet connections can be unreliable, the story is different. The big question for scientists in these regions is not just whether they know about this technology, but whether they can actually use it to improve their work.
A recent study set out to answer this question in Bo District, Sierra Leone, a major hub for health research in the country. The researchers wanted to see if local public health experts were aware of artificial intelligence, if they had received training on how to use it, and whether they were actually applying it to their daily work. They also wanted to understand what stopped them from using these tools, even when they knew they existed. The team surveyed two hundred people working in universities, hospitals, government offices, and non-profit organizations across the district. These were the people directly involved in gathering health data, analyzing disease trends, and making decisions based on evidence.
The results revealed a striking gap between knowing about a tool and using it. Nearly three out of every four researchers in the district said they were aware of artificial intelligence. They knew the term, they understood the basic idea, and they recognized its potential. However, when the researchers asked who was actually using these tools in their work, the number dropped significantly. Only about one in three researchers reported using artificial intelligence. Even more telling was the finding on training: only about a quarter of the participants had ever received any formal instruction on how to use these systems. This suggests that while the idea of artificial intelligence has reached these researchers, the practical ability to use it has not.
The study looked closely at whether having training made a difference in whether someone used the technology. It is a common assumption that if you teach someone how to use a new tool, they will use it. In this group, those who had received training were slightly more likely to use artificial intelligence than those who had not. However, the difference was small enough that it could have happened by chance. The data showed that training alone did not guarantee adoption. Similarly, the researchers checked if using these tools made people feel their work was better. They found that using artificial intelligence did not statistically predict whether a researcher felt their work had improved. Some users felt it helped, while others were unsure, and the same was true for non-users.
The study also examined whether the type of workplace a person worked in—such as a university versus a hospital—or how many years of experience they had, influenced their use of the technology. The analysis found no clear link between these factors and the use of artificial intelligence. This points to a deeper issue. The researchers concluded that the barriers to using artificial intelligence in this setting are not simply about a lack of knowledge or individual skills. Instead, the problem appears to be structural. In a place where internet connections can be unstable, computers might be old or slow, and there may be no official policies supporting the use of new technology, even a well-trained researcher will struggle to use these tools effectively.
The findings suggest that simply raising awareness or holding a one-off training workshop is not enough to bring artificial intelligence into routine use in places like Bo District. The gap between knowing about the technology and using it is wide, and it is held open by the physical and organizational environment. To bridge this gap, the researchers argue that investment must go beyond teaching people how to use software. It must include building reliable digital infrastructure, ensuring affordable access to the necessary tools, and creating clear rules within institutions that support the ethical and effective use of these systems. Until these foundational pieces are in place, the potential of artificial intelligence to transform public health research in low-resource settings will remain largely untapped, waiting for the conditions to catch up with the knowledge.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.