Boost Recall: Qualitative Analysis of Clinical Consultations

Clinician validation during recalled consultations increases incidental recall of spoken health messages, the effect was measured as ~19% higher odds per item in the published study. Evidano is an AI-powered qualitative data analysis platform that ingests transcripts and audio, runs thematic coding, and exports visuals for stakeholders. Use validation as a codable variable, measure recall or adherence outcomes, and reproduce the analytic pipeline across cohorts.

• Lee et al. (Published July 20, 2026) tested autobiographical recall of validating vs. invalidating consultations and incidental memory for 20 audio health messages (n = 245).
• Those who described a validating consultation had about 19% higher odds of recalling each health message (item-level binomial GLMM).
• Data and code for the study are available on OSF; automated scoring agreed with human raters at 97.5%.
• The study collected data 8–30 Oct 2024 and used attention checks and AI-text review to ensure data quality.

Lee et al. (Published July 20, 2026) tested whether autobiographical recall of validating versus invalidating consultations affects incidental memory for 20 spoken health messages.

• Design: online quasi-randomised autobiographical writing task plus audio encoding and an incidental recall test.
• Sample: 245 adults with chronic pain, data collected 8–30 Oct 2024.
• Key result: those who described a validating consultation had approximately 19% higher odds of recalling each health message (item-level binomial GLMM).
• Mediation: the effect was direct, it was not explained by short-term changes in pain-related fear or cognitive anxiety.

This section summarizes the study procedures and measures. Participants were recruited via Prolific (pre-screened), wrote for at least three minutes about a validating or invalidating healthcare consultation, then listened to 20 brief health messages in audio format, and incidental recall was measured. Analysis used item-level binomial GLMMs with random intercepts for participant and message.

• Autobiographical reactivation is a standard affect-manipulation technique and it was used here to approximate real-world consultation effects.
• Messages were balanced: 10 action-oriented versus 10 restriction-oriented; mean message length = 4.25 words.
• Data quality steps included attention checks, manual review for AI-generated text, and automated scoring validated against human raters.

This section gives a step-by-step reproducible workflow teams can run in Evidano to test whether validation language predicts recall.

• 1) Collect: record one-to-one consultations or scripted simulated encounters, capture audio and clinician notes.
• 2) Transcribe: upload audio to Evidano and apply a custom dictionary and PII redaction.
• 3) Tag: import or create a validation/invalidation codebook and run automated coding across transcripts.
• 4) Link outcomes: attach recall tests, adherence metrics, or survey responses as participant-level metadata.
• 5) Analyze: run thematic frequency and cross-segment analyses, such as validated versus invalidated segments and pain-duration buckets.
• 6) Visualize: generate co-occurrence networks and hierarchical code maps to show which clinician behaviours co-occur with validation.
• 7) Act & iterate: export tables and figures for training, update clinician scripts, and rerun the pipeline on subsequent cohorts. Need a fast start? Sign up or request a demo at Evidano.

When analysing clinical narratives, obtain informed consent, redact personally identifiable information, and follow local data-protection and ethics approvals before storing or processing recordings; this study is research-focused and non-diagnostic.

Lee et al.'s July 20, 2026 study (n = 245) provides a clear, replicable signal: perceived clinician validation relates to better recall of health advice, about 19% higher odds per item. For teams studying consultations, the actionable move is to treat validation as a codable variable and to measure its downstream impact on recall or adherence.

Ready to operationalise this? Run the 7-step workflow above on your transcripts and audio in Evidano, request a demo, start a trial at Try Evidano for free, and link your analyses back to the original paper: PLOS ONE.