This post explains how to design and analyze LLM adoption qualitative research focused on primary care physicians, and how AI-enabled tools can speed transcription, coding, and thematic synthesis. The primary keyword for this post is "LLM adoption qualitative research" and the target audience is qualitative researchers and health services teams planning interview studies. The protocol we summarize is published in PLOS One and was received on July 18, 2025 and published on August 7, 2026, so the methods and timelines described below reflect that source.
Key Takeaways
According to PLOS One, the protocol by Pokharel et al. (2026) proposes a descriptive qualitative study using semi-structured interviews and the Technology Adoption Behavior (TAB) framework to examine how primary care physicians perceive LLM adoption in clinical practice (PLOS One).
- Pokharel et al. (2026) published the protocol in PLOS One on August 7, 2026 and reported the study design but had not started recruitment as of that date.
- Pokharel et al. (2026) plan purposive sampling by LLM usage: 50+ uses, 5–50 uses, and fewer than 5 uses, and an anticipated recruitment window from August 1, 2025 to January 31, 2026.
- Pokharel et al. (2026) cite a Canadian context where the number of people without a family doctor rose from 4.5 million in 2019 to about 6.5 million in 2023, motivating a primary-care focus.
- Pokharel et al. (2026) state the analytic approach will include thematic analysis using both deductive and inductive coding and reflexive memoing.
What happened and how the protocol works
The PLOS One protocol by Pokharel et al. (2026) proposes a descriptive qualitative study that uses semi-structured interviews and the TAB framework to map physicians' perceptions of LLMs, and the authors expect recruitment and data collection to complete within six months when begun (Pokharel et al., 2026).
According to Pokharel et al. (2026) in PLOS One, the study will purposively sample practicing family physicians in British Columbia to capture variation in LLM experience, years in practice, and clinical context.
According to Pokharel et al. (2026) in PLOS One, interviews will be audio recorded, transcribed verbatim, deidentified, and analyzed using NVivo with a combined deductive code set drawn from the TAB framework and inductive codes created from the data.
According to Pokharel et al. (2026) in PLOS One, the TAB framework integrates perceived ease of use, perceived usefulness, and task-technology fit to explain adoption decisions.
Findings snapshot
| Date | Metric | Value | Implication |
|---|---|---|---|
| August 7, 2026 | Protocol published | PLOS One, Pokharel et al. (2026) | Method details and timelines publicly available |
| July 18, 2025 → July 22, 2026 | Manuscript timeline | Received July 18, 2025; Accepted July 22, 2026 | Reflects extended peer review and protocol refinement |
| Aug 1, 2025 → Jan 31, 2026 (planned) | Recruitment window | 6 months anticipated | Targets completion of data collection within one year of start |
| 2019 → 2023 | Primary-care access | 4.5 million to about 6.5 million without a family doctor | Supports focus on primary care in LLM adoption research |
| LLM experience categories | Usage thresholds | 50+, 5–50, <5 lifetime uses in clinical context | Enables analytic contrasts by familiarity |
Implications for qualitative researchers studying LLM adoption
Answer: The protocol implies careful sampling, reflexivity, and mixed deductive-inductive coding are central to credible LLM adoption qualitative research.
According to Pokharel et al. (2026) in PLOS One, purposive sampling across experience levels is required to surface both adoption drivers and resistance reasons.
According to Pokharel et al. (2026) in PLOS One, using a theory-guided framework such as TAB helps structure interviews and supports deductive coding while leaving room for emergent themes via inductive coding.
- Plan for audio recording, verbatim transcription, and deidentification as required by the PLOS One protocol (Pokharel et al., 2026).
- Include reflexive memoing and monthly codebook review when multiple coders are involved, as recommended by Pokharel et al. (2026).
- Anticipate ethics approvals and informed consent procedures similar to the UBC Behavioral Research Ethics Board process described by Pokharel et al. (2026).
How Evidano helps
Problem: Slow transcription and messy qualitative data
Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.
Feature mapping: Use Evidano's speech-to-text transcription with custom dictionaries and PII redaction to convert recorded interviews into deidentified, analysis-ready transcripts, matching Pokharel et al.'s requirement to transcribe verbatim and deidentify (Pokharel et al., 2026).
Problem: Balancing deductive and inductive coding at scale
Feature mapping: Evidano supports hierarchical codebooks and thematic synthesis to apply deductive TAB-based codes and then expand with inductive codes, aligning with the NVivo-led coding plan described in Pokharel et al. (2026).
Feature mapping: Evidano's cross-segment and frequency analyses let researchers compare the protocol's usage groups (50+, 5–50, <5 uses) and produce tables like the snapshot above for rapid reporting.
Problem: Reproducible synthesis and stakeholder-ready outputs
Feature mapping: Evidano exports reproducible codebooks, thematic summaries, and visualizations so teams can produce policy-facing briefs and conference slides, supporting the dissemination aims in Pokharel et al. (2026).
Learn more about capabilities on our features page.
FAQ: LLM adoption qualitative research
What is the Technology Adoption Behavior (TAB) framework used in this protocol?
Answer: The TAB framework integrates perceived ease of use, perceived usefulness, and task-technology fit to explain adoption decisions.
According to Pokharel et al. (2026) in PLOS One, TAB unifies classic models such as TAM and task-technology fit to guide interview design and deductive coding.
How should I sample clinicians for an LLM adoption study?
Answer: Purposive sampling across experience levels, years of practice, and clinical roles is recommended.
According to Pokharel et al. (2026) in PLOS One, the protocol stratifies by LLM lifetime uses (50+, 5–50, <5) and seeks variability in years of practice to enable analytical contrasts.
Can Evidano replace NVivo for thematic analysis in this protocol?
Answer: Evidano can perform the same deductive and inductive coding workflows and adds AI-enabled automation for tagging and cross-segment frequency analysis.
According to Pokharel et al. (2026) in PLOS One, the team planned to use NVivo; Evidano offers comparable coding tools plus integrated transcription and AI chat over documents for rapid synthesis.
What ethical steps are essential when studying physicians' views on LLMs?
Answer: Obtain institutional ethics approval, informed consent, audio destruction plans, and transcript deidentification.
According to Pokharel et al. (2026) in PLOS One, the UBC Behavioral Research Ethics Board approval guided recruitment materials, consent forms, audio destruction, and use of deidentified transcripts.
Conclusion & Next Steps
Pokharel et al. (2026) in PLOS One provide a concrete, theory-informed protocol for exploring primary care physicians' perceptions of LLM adoption, with purposive sampling, TAB-guided interview questions, and combined deductive and inductive analysis.
For teams preparing similar studies, plan recruitment strata by prior LLM use and build reflexivity and codebook review into monthly workflows as the protocol recommends (Pokharel et al., 2026).
If you want to accelerate transcription, deidentification, and thematic synthesis for LLM adoption qualitative research, consider tools that combine speech-to-text and AI-enabled coding.
Get started by running your transcripts through an AI-enabled platform and compare extraction workflows, or Try Evidano for free.
Topics
- LLM adoption qualitative research
- qualitative research LLM adoption
- physician perceptions of LLMs
- technology adoption behavior framework
- AI-enabled qualitative analysis
Keep reading
- Commentary on NewsPhysicians and LLMs: Qualitative analysis of LLM adoptionHow primary care physicians view LLMs and how to run qualitative analysis of LLM adoption: methods, timelines, direct quotes, and tools to act with Evidano.
- Commentary on NewsStudy Protocol: Physician Perceptions of LLMsReview of a PLOS One protocol on primary care physician perceptions of LLMs, with dates, sample metrics, and AI-enabled qualitative research guidance.
- Commentary on NewsPhysicians and LLMs: qualitative analysis of adoptionPractical guide to the PLOS ONE qualitative protocol on LLM adoption for researchers. Learn methods, timelines, and how AI-enabled qualitative analysis speeds insight.