AI and digital health technologies map to pre-, peri-, and post-travel diagnostics
The study reports a narrative review evaluating how artificial intelligence and digital health tools are applied across the pre-travel, peri-travel, and post-travel phases of patient care. Investigators highlight machine learning for individualized risk prediction, AI-assisted border screening that detected outbreaks up to nine days earlier than conventional methods, and the MALrisk model, which achieved an AUC of 0.98 with 100% sensitivity and 72% specificity for imported malaria. In hospitalized returned travelers, ChatGPT-4o correctly identified diagnoses in 68% of cases, while metagenomic next-generation sequencing improved diagnostic yield by 24.2% over routine testing. The review concludes that while these technologies show strong potential, the field must shift from developing isolated applications to prospective, multicenter implementation studies focused on clinical integration.
The original study
Artificial Intelligence and Complementary Digital Health Technologies Across the Travel Medicine Continuum: A Narrative Review.
- Authors
- van Genderen PJJ, van Sprang ENM
- Journal
- Journal of travel medicine
- Type
- Journal Article
- PMID
- 42713903
Original abstract
BACKGROUND: Artificial intelligence (AI) and complementary digital health technologies are increasingly being applied to improve prevention, diagnosis and surveillance in travel medicine. This narrative review evaluates current applications of these technologies across the pre-travel, peri-travel and post-travel phases of the travel continuum. METHODS: A narrative literature review was conducted using a clinically oriented three-phase framework encompassing pre-travel preparation, peri-travel monitoring and post-travel diagnosis and surveillance. Machine learning, clinical decision-support systems, large language models, wearable technologies, telemedicine, outbreak surveillance, precision diagnostics and interoperable digital health infrastructure were reviewed. RESULTS: Machine learning improved individualized risk prediction before travel and supported diagnostic decision-making after travel, while clinical decision-support systems enabled more personalized preventive and therapeutic recommendations. During travel, AI-assisted border screening detected SARS-CoV-2 outbreaks up to nine days earlier than conventional surveillance, and wearable Internet of Things (IoT) technologies enabled continuous physiological monitoring in travelers and mass gatherings. The MALrisk model predicted imported malaria with an area under the receiver operating characteristic curve of 0.98, achieving 100% sensitivity and 72% specificity. In hospitalized returned travelers, ChatGPT-4o identified the correct diagnosis in 68% of patients and included the correct diagnosis among its three leading differential diagnoses in 78%, while metagenomic next-generation sequencing increased the diagnostic yield by 24.2% beyond routine investigations. CONCLUSIONS: AI and complementary digital health technologies have the potential to improve travel healthcare throughout the travel continuum. The principal challenge is no longer the development of individual AI applications, but determining how complementary technologies can be effectively integrated into routine travel medicine. Current evidence remains largely retrospective, highlighting the need for prospective multicenter implementation studies evaluating clinically meaningful outcomes.