Research
David Sackett Award for research on evidence of AI in healthcare
Potsdam, 29 June 2026
Christoph Wilhelm and Dr. Felix G. Rebitschek from the Harding Center for Risk Literacy at the Brandenburg Medical School Theodor Fontane (MHB) and the University of Potsdam receive this year’s David Sackett Award from the network of evidence-based medicine (EbM Network) for their research program on “patient-relevant benefits and harms associated with AI-based decision-making and evidence-based health information”. Their studies reveal where Artificial Intelligence employed in support of health professionals improves patient care and where pertinent evidence is still insufficient.
Their findings provide an important contribution to the evidence-based evaluation of Artificial Intelligence (AI) in medicine. The focus is not on the technical capacity of AI systems but on the question whether these benefit patients, avoid harm and support informed decision processes.
According to Felix G. Rebitschek, it is the rule to assess artificial intelligence primarily for technical capacity and accuracy of recommendations. The point is, however, whether the use of AI brings demonstrably more benefits for patients compared to traditional medical practice without AI: “We found that this question has not been addressed sufficiently in research up to now.”
AI is changing decision making in medicine at a rapid pace. AI-related decision-making systems are gaining importance in healthcare, the same as generative language models for the provision of health-related information. Public debate often concentrates on innovative power, speed and technical capacity. But the important point from the perspective of evidence-based medicine is whether such systems have patient-relevant benefits, avoid harm and improve informed decision. The award-winning program addresses precisely this issue. Together with other authors, Christoph Wilhelm and Felix G. Rebitschek explore the question whether and under which conditions AI-based decision-making support and generative language models meet the requirements of evidence-based medicine.
Not much reliable evidence to be found in many AI applications
A systematic survey of AI-based aids for health professionals reveals that these are only rarely assessed on the basis of patient-relevant endpoints. The latter are defined as outcomes of immediate significance for affected individuals, such as mortality, morbidity, quality of life or length of hospitalization. The researchers found indications of positive effects in several application areas, e.g. length of hospital stay, symptom severity or mortality. Available evidence in total, however, is not sufficient for a reliable assessment of patient-relevant added value of AI use. Evaluations often take too little account of potential harmful effects, differences between population groups and the transparency of underlying models.
Generative AI falls short of professional standards
Studies on generative language models show that these frequently produce linguistically plausible health information but do not fully meet the requirements of evidence-based health information. Deficits were found specifically in the balanced presentation of benefits and harm, the communication of absolute risks, and in respect of insecurities and evidence quality. A small revision in the wording of the request – a so called prompting boost – improved the quality of answers. Nevertheless, the results fell short of desired standards. Generative AI is therefore currently not sufficient to independently secure informed medical decision-making processes.
“Today, generative AI can phrase understandable health information quickly. But understandability alone is not enough. Informed decision requires the transparent presentation of benefits, risks, uncertainties and the quality of underlying evidence. There is still considerable need for improvement”, so Christoph Wilhelm.
The award-winning studies are significant for evidence-based medicine as they explore AI systems not just as technical innovations but as medical interventions in the regular practice of decision making. The same basic requirements apply to their use as for other diagnostic, therapeutic or information-based procedures: transparent presentation of evidence, systematic benefit-harm assessment, and orientation towards patient-related benefit. The studies therefore play an important role in the objective debate on chances and limits of AI in healthcare.
The jury underlined the high relevance of the research program for evidence-based medicine now and in future. Findings show that the implementation of AI in the health system is a question of more than technical feasibility. The point is that AI-based systems improve care provision, empower patients and facilitate informed decisions. The program of Christoph Wilhelm and Dr. Felix G. Rebitschek addresses a core issue in evidence-based medicine in the digital age.
The award honors research that sets important standards in the assessment of AI in healthcare. The presentation ceremony is scheduled for the coming annual meeting in Göttingen between 30 September and 2 October, a joint event including the EbM Network and three other German medical associations: the DEGAM (general and family medicine), the DGSMP (social medicine and prevention) and the DGMS (medical sociology).
David Sackett Award honors research program
Title: “Patient-relevant benefits and harms of AI-related algorithmic decision-making systems and evidence-based health information”
Publications
Rebitschek FG, Carella A, Kohlrausch-Pazin S, et al. Evaluating evidence-based health information from generative AI using a cross-sectional study with laypeople seeking screening information. npj Digital Medicine. 2025;8:343. https://doi.org/10.1038/s41746-025-01752-6
Wilhelm C, Steckelberg A, Rebitschek FG. Benefits and harms associated with the use of AI-related algorithmic decision-making systems by healthcare. The Lancet Regional Health – Europe. 2024;48:101145. https://doi.org/10.1016/j.lanepe.2024.101145
Wilhelm C, Steckelberg A, Rebitschek FG. Is artificial intelligence for medical professionals serving the patients? Protocol for a systematic review on patient-relevant benefits and harms of algorithmic decision-making. Systematic Reviews. 2024;13:228. https://doi.org/10.1186/s13643-024-02646-6
Background
The David Sackett Award is the most renowned science prize of the network “Evidenzbasierte Medizin e. V.”. Endowed with €2,000, it has been awarded annually since 2008 in the context of the annual meeting. The award honors outstanding scientific performance, methodological innovations or studies in the field of evidence-based medicine and healthcare provision.
Canadian physician David Sackett (1934 - 2015) is considered one of the world’s leading pioneers and founding fathers of evidence-based medicine. He stipulated that medical interventions should not be based on tradition or intuition alone but on the best scientific findings available (evidence).
Harding Center for Risk Literacy
The main focus of the Harding Center for Risk Literacy, located at the Institute of Research in Health Sciences Education, is on improvement of risk communication and risk competence at national and international level, specifically in the areas of health and digitization. An interdisciplinary team explores risk communication, develops innovative methods of improvement, and offers training courses to physicians, health professionals, judges and journalists.
The Center has studied the error handling culture among managers and government officials and devised mechanisms to reduce defensive decisions and improve risk communication. In collaboration with health insurers and authorities, the Center has developed easily understandable formats of health information such as service boxes. The focus in recent research projects is on artificial intelligence in medicine and health information for vulnerable population groups.
Faculty of Health Sciences
Founded in 2018 as an institution jointly organized by the University of Potsdam, the Brandenburg Medical School Theodor Fontane and the Brandenburg University of Technology Cottbus-Senftenberg, the Faculty of Health Sciences develops new concepts of care provision in medicine, nursing and medical technology as well as innovative study programs. Collaboration with further universities and research facilities serves to advance the provision of medical care in the State of Brandenburg.
Contact
Dr. Felix G. Rebitschek
E-Mail: felix.rebitschek@fgw-brandenburg.de
Web: www.hardingcenter.de
EbM-Netzwerk
Phone: +49 (0)30-30833660
E-Mail: kontakt@ebm-netzwerk.de
Web: www.ebm-netzwerk.de