FSH - School of Health and Human Performance header
School of Health & Human Performance
Medical professional with patient looking at screen

Tech is Simplifying Medical Info, But Is Anyone Healthier?

A recent systematic review suggests that while Artificial Intelligence (AI) is rapidly being deployed to simplify complex medical documents, translate health materials, and answer patient questions, there is little to no evidence yet showing these tools actually improve patient health literacy.

The review, published by DCU researchers, analysed 18 studies on AI's role in health literacy over the past decade (2014-2024).

Researchers found that AI techniques—including traditional machine learning and advanced deep learning models—are being actively used in various ways:

  • Simplifying Text: Breaking down difficult medical jargon in articles and online materials.
  • Evaluating Complexity: Assessing how hard a document is to read.
  • Translation and Q&A: Providing answers and translating materials for non-English speakers.

However, the study identified a shortcoming: The field is failing to measure outcomes.

The authors noted that only seven of the reviewed studies even bothered to define health literacy (the ability to obtain, understand, and use health information), and only one study used an established tool to measure whether the AI intervention had improved an individual’s actual health literacy. Measurement of organisational-level impact was also largely overlooked.

The findings highlight a disconnect: AI models are being evaluated based on their technical performance (like translation accuracy), but not on their real-world impact on patients.

The authors conclude that future research must prioritise more rigorous measurement of health literacy outcomes and focus on the practical, robust implementation of these AI tools in actual hospital and community settings to ensure they are fulfilling their potential to close communication gaps in healthcare.