Page 104 - DP-2-1

Design+ AI’s role in medical history taking

of efficiency, stress reduction, and enhanced diagnostic
accuracy, which strongly motivates the adoption of AI.
Nevertheless, addressing concerns about the doctor–patient
relationship and ensuring the reliability of AI systems are
crucial for its successful implementation. Ultimately, the
intention to use AI in medical history taking is driven by its
potential to significantly improve patient care while managing
the inherent challenges associated with its adoption.
The doctors surveyed had very different personal data,
be it the specialist discipline, the amount of professional
experience, or the employment relationship. Nevertheless,
the interview statements were characterized by a surprising
similarity.

A similarity matrix was created to illustrate the
similarity of the interview data. This was created taking
37
into account the variables of age, workplace/position, and
experience with AI based on the occurrence of the codes.
The “simple match” was selected as the calculation variant,
which evaluates both the presence and absence of codes as
a match. This variant was chosen because all codes were
assigned in the majority of the documents and the absence
Figure 3. Frequencies of codes building the factors influencing the of codes therefore does not play a dominant role compared
intention to use
to the presence of codes. The following results were found
The analysis highlighted that AI could alleviate stress for the nine interviews:
by reducing the workload on medical staff. By automating The matrix in Table 1 illustrates the similarity of the
routine tasks, AI helps manage the stress associated with respondents’ answers, with 1.00 corresponding to 100%
staff shortages and increasing administrative duties, making agreement and 0.00 corresponding to 0% agreement. The
it a valuable tool for improving workplace conditions. data indicates that the interviews of respondents B09 and
Physicians acknowledged AI’s potential to enhance B05 exhibit a high degree of similarity in terms of both
diagnostic accuracy and efficiency in specific medical coding and variables, with a value of 0.88. Conversely, the
fields. The ability of AI to adapt to different specialties and interviews of respondents B02 and B06 demonstrate the
provide precise support was seen as a significant benefit, least similarity, with a value of 0.42. The overall average
driving its intended use across various medical disciplines. similarity of the interviews is 0.68.
Concerns were raised about AI potentially diminishing 4. Discussion and prepositions for further
personal interactions between doctors and patients. research
However, many physicians believe that AI can help
enhance these interactions by handling preliminary data The analysis of interviews with medical professionals
collection, thus allowing more quality time for direct reveals a generally positive reception toward AI-supported
patient engagement. anamnesis, highlighting its potential to enhance efficiency
and precision in patient care. All surveyed doctors
AI’s capability to support medical professionals expressed willingness to incorporate AI tools like Idana
through advanced system functions was another factor into their daily practice, citing time savings and improved
influencing its intended use. AI can improve the quality treatment quality as primary advantages.
of anamnesis, identify red flags, and ensure comprehensive
data collection. However, the need for robust verification One respondent articulated this sentiment succinctly:
mechanisms to ensure the reliability of AI-generated “I think, above all, it saves us time and allows
information remains a critical consideration. patients to fill out their medical history at home.
The combined insights from these factors reveal a And then we have already filled it out, the file, the
cautiously optimistic intention to use AI in medical documents are prepared and then we can also ask
anamnesis. Physicians recognize the substantial benefits the patients specific questions.” (I06, 47 – 50)

Volume 2 Issue 1 (2025) 7 doi: 10.36922/dp.7675