Apps designed to facilitate people identify potential illnesses based on their symptoms are gaining popularity, but a recent evaluation by Stiftung Warentest reveals significant differences in their accuracy.
- Using symptom-checker apps, individuals can attempt to identify illnesses via an app
- Stiftung Warentest conducted testing of several of these apps
- Two apps performed well in the evaluation
The number of users of these so-called Symptom-Checker-Apps has doubled in Germany over the past five years. This increasing reliance on digital health tools highlights the need for reliable and accurate assessments.
What are Symptom-Checker-Apps?
Symptom-Checker-Apps, according to Stiftung Warentest, offer an alternative to searching for disease information online and are apps and websites where individuals can input their symptoms to have an algorithm determine a possible diagnosis.
These apps function in various ways. Generally, they compare entered symptoms against a medical database and identify potential diagnoses. AI-based apps utilize advanced methods such as neural networks or machine learning to recognize patterns from large amounts of medical data and user information. Symptom-checkers without AI typically rely on decision trees created by medical professionals, based on pre-programmed rules and correlations between symptoms and possible diagnoses.
To arrive at a diagnosis, various data points are collected, including the duration and severity of symptoms, demographic information, medications, pre-existing conditions, and allergies. Information about family medical history, fitness levels, dietary habits, and recent travel may similarly be relevant.
The Stiftung Warentest Findings
Stiftung Warentest thoroughly examined ten free, German-language Symptom-Checker-Apps. The evaluation focused on five common conditions: herniated discs, urinary tract infections, depression, angina pectoris (chest pain due to coronary artery disease), and restless legs syndrome. The privacy policies of the websites and apps were also assessed, as these applications collect highly personal information.
The apps’ performance varied considerably. The ratings ranged from “good” to “insufficient.” Notably, the organization found that none of the apps collected excessive user data. While none of the ten tested apps received a “very good” rating, the findings suggest that some symptom checkers can be a useful tool for preliminary self-assessment. Detailed test results are available for download as a PDF from Stiftung Warentest.
