AI in healthcare: from fighting epidemics to the family doctor

An update in an hour instead of a day: in Africa's fight against Ebola, AI is showing how it can bolster healthcare. In Germany too, it could help but experts warn of poor transparency and unresolved risks. Since May, teams on the ground in the Democratic Republic of Congo (DRC) have been tackling the Bundibugyo Ebola virus. Every day, so-called health workers knock on doors and offer counselling. They record recoveries and deaths and refer people with worrying symptoms to medical facilities. According to a report by AI company Anthropic, the knowledge of these health workers, whose notebooks contain the health status of entire communities, has been digitised. Who has fallen ill? Who have they been in contact with and where did they go before they knew they were infected? When the answers and reports were analysed, Anthropic's Claude model was used. The study involved, among others, CEPI, the WHO Regional Office for Africa and the national biomedical research institute INRB. AI helps to assess the spread of epidemics more quickly "A situation report now comes together within an hour instead of taking an entire working day," explains Dr Marie Roseline Darnycka Belizaire, Director for Emergency Preparedness and Response at the World Health Organization's (WHO) Regional Office for Africa. Artificial intelligence can help to assess the spread of epidemics more quickly and to deploy medical assistance more precisely, the WHO representative said on stage at the World Health Summits. The data analysis made it possible to identify potential outbreak areas and prepare responses, such as positioning stocks of medical supplies at key hubs or high risk areas. That can be critical in an emergency. The earlier health authorities react to a possible outbreak, the better they can prepare and organise care. AI is intended to help evaluate information from different sources more quickly and support decision makers. This positive example of how artificial intelligence is used in healthcare shows the level at which AI can be deployed with relatively low risk. Experts, however, see far greater potential for future applications of artificial intelligence. The technology could improve diagnoses, relieve medical staff of administrative tasks and help to use scarce resources more efficiently. At the same time, it raises fundamental questions: How reliable are the results? Can medical professionals understand how an AI system arrives at a decision? And do patients really benefit? AI as support for overstretched health workers Worldwide, the demand for medical care is rising. At the same time, many places lack staff, time and equipment. For Dr Thomas Zapata, Director of the WHO Academy in Lyon, this is a key starting point for the use of artificial intelligence. "We need to think about the paradox we are facing today. We have more health professionals globally than ever before, and we have unprecedented technologies at our disposal," Zapata said during a panel discussion at the World Health Summit. Yet, he added, there is still an enormous shortage of health workers. "The demand for health services keeps growing." He sees AI as a way to narrow that gap. The crucial point, however, is not to treat it as an end in itself, but as a tool that is integrated into existing health systems. "AI is an enabler, a means to an end. At the moment we are very fascinated by technology and AI, but I think we need to place it in the right spot within the health system context," Zapata warned. For him, artificial intelligence fits in two main areas. On the one hand, AI could transform how healthcare delivery is organised. On the other, training and continuing education for health workers has to adapt to these new possibilities. Ultimately, AI can certainly help reduce workloads, streamline administrative processes, improve clinical decision making and optimise diagnoses. What matters most to him, though, is that the technology frees up time for what cannot and should not easily be automated: caring for patients. The Ebola crisis in the Democratic Republic of Congo, in particular, has shown that trust in local health systems is lacking. This gives doctors more time for their patients It is especially in places where health workers operate under extreme time pressure that automated processes can make a difference. Dr Cecilia Mbaca, Senior Technical Advisor for Health and Emergencies at aid organisation CARE, knows the challenges from more than 15 years of frontline work. For her, three questions are crucial: Can AI support medical decision making? Can it take over repetitive tasks? And does it help tailor care more closely to the individual needs of patients? "If AI can do this work very well, there is more time for the personal aspects of care, so that we can ask questions, listen, examine patients and refer them to the next level of care," said Dr Cecilliah Mbaka, Senior Technical Advisor for Health and Emergencies at aid organisation CARE. She herself has spent more than 15 years working on the front line, with little support, limited resources and a heavy workload. In epidemics, it is not only medical data that matters. The trust of the population also plays a key role. People need to be confident that they will be protected and treated appropriately in health facilities. In affected communities, this trust can be decisive in whether patients seek medical help in time. It is estimated that around 50 percent of deaths in villages in the DRC occur outside medical facilities, according to epidemiologist Maximilian Gertler. The local health system is under pressure and was not prepared for the Ebola outbreak. Thousands of protective suits are needed every day to shield carers from contamination. Germany recently increased its aid by 21 million euros. AI can therefore support the organisation of care. But on its own it cannot ensure that people trust medical facilities. As AI experts themselves point out, the path to a diagnosis is often still opaque. Transparency problem: How does AI arrive at its diagnosis? Modern AI systems can process vast amounts of data and spot patterns that humans might miss. Yet it often remains unclear exactly how they arrive at their conclusions. Dr Thomas Thesen, Professor of Medical Education and Informatics at the Geisel School of Medicine at Dartmouth University in the United States, describes the problem like this: "They are extremely good at recognising patterns, but they are really bad at explaining to us what those patterns actually mean." Chatbots and large language models like Claude or ChatGPT have what he calls an explanation problem. Here, artificial intelligence remains a black box. This is particularly relevant in the health sector. When a system produces a diagnosis or a treatment recommendation, medical professionals need to be able to judge whether it is appropriate in the specific case. A result on its own is not always enough. The way in which it was reached also has to be understandable. "That is the problem of explainability. They cannot explain in language that humans can understand how they arrive at these conditions." This so-called black box problem affects not only transparency, but also accountability: Who is responsible if an AI system gives a wrong recommendation? For Dr Mbaka too, it is crucial that AI does not simply generate results, but genuinely supports decision making. Expert: AI must not replace learning in training Artificial intelligence can also be helpful in training medical professionals, for example by making information more accessible or by assisting with recurring tasks. But when it comes to learning, Thesen, Professor of Medical Education and Informatics, sees risks. He refers to a study that compared searching for information via conventional search engines with using a AI chatbot. Students who used the chatbot ended up with more superficial knowledge that was less firmly anchored. "These so-called desirable difficulties are actually very important for the learning process. And if you outsource all cognitive learning to a chatbot, which may be very efficient, you do not learn anything in the process." Anyone who needs to acquire lasting knowledge should, in his view, not let AI take over all of their thinking. In medicine in particular, it can be vital to make the right decision even when technical support is not available. Thesen therefore also recommends using AI mainly in areas where it reduces workload. Where the aim is to understand knowledge and retain it in the long term, independent learning still needs to have its place.