Pilot study evaluating an artificial-intelligence-assisted telephone-based pre-consultation history taking to improve clinical workflow and quality of care

INTERVENTION: Group 1: Intervention group: standard care plus an additional AI‐assisted, voice‐based history taking by telephone prior to the scheduled outpatient appointment. The system collects basic history information in a structured manner only. It does not provide diagnoses, treatment recommendations or clinical decisions. The output consists of a transcript and a structured summary that is made available to the treating physician as preliminary information and is reviewed and validated by the physician during the appointment. Group 2: Control group: standard care without AI‐assisted pre‐consultation history taking. History taking and documentation are performed during the regular outpatient appointment as usual. No additional study‐related examinations are performed. CONDITION: Not disease‐specific. The study investigates a care process rather than a disease. The target population comprises adult patients with a scheduled first presentation at a university outpatient clinic. PRIMARY OUTCOME: Consultation and documentation time in minutes, measured from the start of the physician's preparation before the consultation, through the physician‐patient encounter, until completion of the medical documentation. Assessed once, on the day of the outpatient appointment. SECONDARY OUTCOME: All secondary outcomes are exploratory; no adjustment for multiple testing is planned. Unless stated otherwise, assessment takes place in connection with the outpatient appointment. ; 1. Patient satisfaction after the consultation (ZUF‐8, validated questionnaire; both groups). ; 2. Usability and user experience of the voice‐based pre‐consultation history taking (Chatbot Usability Questionnaire CUQ, German translation; intervention group only; after the telephone session). ; 3. Physician rating of the quality and structure of the AI‐assisted pre‐consultation history taking and its output (structured checklist, adapted from Vidal‐Alaball et al.; intervention group only). ; 4. Short physician survey on overall satisfaction, efficiency, quality of available information and patient preparedness (non‐validated; both groups; directly after the consultation). ; 5. Blinded specialist assessment of documentation quality regarding overall quality, ; completeness, structure and clarity of the further management plan (non‐validated; both groups; after study completion, using pseudonymised documents). ; 6. Descriptive process parameters of the AI‐assisted telephone contact: reachability, number of contact attempts, drop‐out rate, call duration, data completeness (intervention group only). INCLUSION CRITERIA: 1. Age 18 years or older. 2. Scheduled first presentation (first physician contact) at a university outpatient clinic. 3. The scheduled outpatient appointment is at least 7 days in the future, allowing sufficient time for information, consent and the pre‐consultation history taking. 4. Sufficient German language skills to complete the questionnaires. 5. Reachable by telephone and able to take part in a telephone‐based history taking. 6. Capacity to consent and written informed consent obtained.