Digital health advice: AI tools put to the test – what really helps!

Transparenz: Redaktionell erstellt und geprüft.
Veröffentlicht am

The TU Berlin publishes two studies on the effectiveness of AI-supported health advisors. Discover the results!

Die TU Berlin veröffentlicht zwei Studien zur Wirksamkeit von KI-gestützten Gesundheitsberatern. Entdecken Sie die Ergebnisse!
The TU Berlin publishes two studies on the effectiveness of AI-supported health advisors. Discover the results!

Digital health advice: AI tools put to the test – what really helps!

The increasing use of artificial intelligence (AI) in healthcare raises interesting questions about the quality and effectiveness of digital healthcare advice. A current report from the TU Berlin highlights two studies that address this topic in depth. The aim of the research was to assess the accuracy of health recommendations through various digital tools, including ChatGPT and symptom checker apps.

The studies were published in the journals *Journal Scientific Reports* and *npj Health Systems Journal*. A novel testing method was developed that is based on real patient cases, in contrast to previous approaches that primarily used idealized textbook cases. The results show that symptom checker apps are far more helpful to laypeople than ChatGPT, especially in distinguishing between harmless and serious symptoms.

Digital health advice in the test

A key finding of the research is that ChatGPT classifies almost every patient case as an emergency, while specialized apps can provide more appropriate recommendations. This could lead to an overload of the healthcare system, as many laypeople tend to rush to seek medical help for minor complaints.

User experiences play an important role when using such digital tools. A second study from Berlin examined how patients incorporate digital recommendations into their decisions. The participants compared the information from ChatGPT and symptom checker apps with other sources such as Google searches or friends. It turned out that they were often confronted with overwhelming information, which led to unnecessary visits to the doctor.

The quantitative analysis, based on an observational study of 600 subjects, also found that ChatGPT makes self-care more difficult and increases the number of unnecessary doctor visits. Here, well-functioning symptom checker apps could offer valuable support to encourage users to take more responsibility for their health management.

The role of health apps and chatbots

The relevance of health apps and chatbots is continually growing. Loud Onlim Almost every third citizen in Germany uses a health app. There are around 350,000 health apps worldwide, of which 110 are particularly dominant. The COVID-19 pandemic has significantly accelerated the use of these digital tools.

Healthcare chatbots perform versatile tasks such as scheduling appointments, medication management, and providing real-time information. In the future, they could even be the first point of contact for basic medical care issues.

The most promising health chatbots include:

  • Woebot: Ein digitaler Therapeut, der kognitive Verhaltenstherapie-Elemente nutzt.
  • Buoy Health: Hilft bei der Identifikation gesundheitlicher Probleme und empfiehlt weitere Schritte.
  • Ada: Ein KI-gestützter Symptomprüfer mit personalisierten Einschätzungen.
  • Healthily: Bietet eine kostenlose Plattform für Gesundheitsinformationen.

Establishing a qualified dialogue between health applications and users is crucial. A particular challenge is its integration into everyday medical practice. Like an analysis PMC shows, active use and integration of these tools into care is necessary to ensure the desired benefits in the long term.

In summary, the current studies and reports make it clear that while digital health tools open up many opportunities, a critical approach to the information provided is essential in order to avoid misjudgments and overloads in the health system. A comprehensive understanding of AI-powered healthcare services could significantly improve the quality of medical care.