Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-1821
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-18T10:43:05.323051Z
status: published
lens: g_space
sector: health
headline: AI-supported first consultation preparation in oncology: a real-world clinical usability study
dek: Abstract Large Language Models (LLMs) may support clinicians by facilitating information management, but their usability in real-world oncology practice remains understudied. This study evaluated the workflow performance of Artificial Intelligence-based First Consultation Preparation (AI-FCP), generating structured summaries from unstructured clinical documentation, co-created with end users. In a prospective single‑group pre‑post design, nine clinicians completed a person-level baseline, followed by 33 consulta…
gain_title: Clinicians using AI-FCP to prepare for first oncology consultations reported lower perceived workload and some self-reported time savings per case.
problem_title: (none)
trace_subject: (none)
gain_reading: Clinicians using AI-FCP to prepare for first oncology consultations reported lower perceived workload and some self-reported time savings per case.
gain_evidence: overall perceived workload (NASA-TLX) was 28.9% lower than baseline | self-reported average time savings of 3.26 min (SD = 2.65) per case
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers tested an AI tool called AI-FCP that turns unstructured clinical documentation into structured summaries for first oncology consultations. The tool was co-created with end users and tried in real clinical practice. Nine clinicians first completed a baseline measure and then used it across 33 consultations.

The result suggests that language models could ease some of the information-management burden clinicians face before meeting new cancer patients. The evidence comes from a very small, single-group study based on self-reported measures, and the authors call the findings exploratory and say generalizability still needs to be established.
limitation: The study is small and exploratory, with a single-group design and case numbers that varied by scheduling and enrollment, which the authors say may introduce survivorship bias. Generalizability is not established.
tag: Evidence-backed gain
key_points: The design was a prospective single-group pre-post study with nine clinicians and 33 consultations. | Perceived workload on the NASA-TLX scale was 28.9% lower than baseline, with a Hodges-Lehmann difference of -12.3 (95% CI -21.7 to -3.3). | Clinicians self-reported an average time saving of 3.26 minutes per case. | Preparation succeeded in 84.9% of consultations, and 63.6% of the summaries were complete.
rundown: AI-FCP generates structured summaries from unstructured clinical documentation to help clinicians prepare for first consultations in oncology. It was developed with input from end users rather than imposed on them.

Usability was assessed with validated questionnaires plus study-specific instruments. The workload comparison used a person-level baseline against measurements taken after the tool was used.

Process measures show the tool did not always work fully. Preparation succeeded in about 85% of consultations, but fewer than two thirds of the summaries were complete, which indicates remaining quality gaps.
sources:
- peer_reviewed | npj Digital Medicine | https://doi.org/10.1038/s41746-026-03396-6 | 2026-10-17
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

b8e4aa7fada578c3ef25373ab28cc00a469785196be81351ba333a428abc9c82

previous

0000000000000000000000000000000000000000000000000000000000000000