AI as a Rater: The Potential of Large Language Models in Assessing Transcribed Semi-Clinical Interviews

AI as a Rater: The Potential of Large Language Models in Assessing Transcribed Semi-Clinical Interviews - AI-D GRID-HAMD-17

Status

Active, not recruiting

Phases

Unknown

Study type

Observational

Source

DRKS

Registry ID

DRKS00036483

Enrollment

Registered

2025-03-26

Start date

2023-05-21

Completion date

Unknown

Last updated

2026-08-03

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

F32

Interventions

Group 1: The non-experimental cross-sectional design of the study includes two measurement time points, which exist mainly due to different querying modalities. The time point itself is not a variable

Williams et al., 2008]) via an anonymous video conferencing platform, while interviewers simultaneously rate the symptoms on the GRID-HAMD-17 scale. Interviewers were trained by a professional under t

Eligibility

Sex/Gender

All

Age

18 Years to No maximum

Inclusion criteria

Inclusion criteria: The ability to take part in an interview in German

Exclusion criteria

Exclusion criteria: Aborting the survey

Design outcomes

Primary

Measure	Time frame
Predictive accuracy of the AI (measured as the correlation between AI-predicted depressive symptoms and interviewer-rated depressive symptoms).	—

Secondary

Measure	Time frame
Accuracy of the AI judgment.	—

Countries

Germany

Contacts

Public ContactYoussef Shiban

PFH Göttingen

shiban@pfh.de+49 551 54700 477

Outcome results

None listed

Source: DRKS (via WHO ICTRP) · Data processed: Aug 10, 2026