Primary care teams and qualitative researchers need clear, actionable findings about how clinicians will adopt large language models, and the primary keyword for this brief is "llm adoption in primary care". According to the PLOS ONE protocol, the study by Pokharel et al. (2026) will use semi-structured interviews to explore family physicians' views on LLMs in British Columbia, Canada, and the protocol was published on August 7, 2026 (PLOS ONE). This post refracts that protocol through the lens of AI-enabled qualitative research and explains concrete steps research teams and health system leaders can take to accelerate trustworthy synthesis.
Key Takeaways
According to the PLOS ONE protocol by Pokharel et al. (2026), this planned qualitative study will use semi-structured interviews and the Technology Adoption Behavior (TAB) framework to examine how primary care physicians perceive LLM adoption in clinical practice (PLOS ONE).
- The study protocol was published on August 7, 2026 and states, in the authors' words, "This qualitative study aims to employ a descriptive qualitative design to explore how primary care physicians perceive the adoption of LLMs in the context of their clinical practice." (Pokharel et al., 2026).
- Pokharel et al. (2026) plan to recruit physicians in British Columbia with purposive sampling and expect recruitment and data collection to run from August 1, 2025 to January 31, 2026, a six-month window described in the protocol (PLOS ONE).
- The protocol highlights a national access trend used as context: the number of Canadians without a family doctor rose from about 4.5 million in 2019 to about 6.5 million in 2023, a statistic cited in the study to justify focusing on primary care (Pokharel et al., 2026).
What Happened / How the PLOS ONE protocol works
Answer: The PLOS ONE protocol describes a descriptive qualitative study using purposive sampling, semi-structured interviews, and thematic analysis guided by the TAB framework to interrogate physician perceptions of LLMs.
According to Pokharel et al. (2026) in PLOS ONE, the research team will purposively sample family physicians in British Columbia across experience and self-reported LLM use categories (frequent: 50+ uses, occasional: 5–50 uses, and nearly none: <5 uses) to ensure varied perspectives (PLOS ONE).
According to Pokharel et al. (2026) in PLOS ONE, interviews will be recorded, transcribed verbatim, deidentified, and analyzed using a combined deductive and inductive thematic coding approach, with reflexive memos and a multidisciplinary team guiding interpretation.
The protocol notes ethical approval from the UBC Behavioral Research Ethics Board (H25-01658), and the authors report the manuscript was received July 18, 2025 and accepted July 22, 2026 before publication on August 7, 2026 (PLOS ONE).
Findings Snapshot
| Date | Metric | Value | Implication |
|---|---|---|---|
| 2019 | Canadians without a family doctor | 4.5 million | Used in protocol as context for primary care pressures (Pokharel et al., 2026) |
| 2023 | Canadians without a family doctor | About 6.5 million | Used in protocol to justify focus on family medicine (Pokharel et al., 2026) |
| Aug 1, 2025 to Jan 31, 2026 | Planned recruitment and data collection window | 6 months | Timeline the authors expect to complete interviews and initial transcripts (Pokharel et al., 2026) |
| Aug 7, 2026 | Protocol publication date | Published in PLOS ONE | Methodological transparency before data collection begins (Pokharel et al., 2026) |
Implications for qualitative researchers and primary care leaders
Answer: The PLOS ONE protocol signals that theory-grounded, transparent qualitative work is needed now to shape safe LLM deployment in primary care.
According to Pokharel et al. (2026) in PLOS ONE, using the TAB framework ties user perceptions, task fit, and environment together and can help transform interview findings into policy-relevant recommendations by showing which perceived barriers affect perceived usefulness and ease of use.
According to Pokharel et al. (2026) in PLOS ONE, purposive sampling across levels of prior LLM use strengthens analytic contrast and helps researchers differentiate concerns tied to unfamiliarity from those tied to observed harms or benefits.
Primary care leaders should note that the protocol connects system pressures (for example the rise from 4.5 million in 2019 to 6.5 million in 2023 of patients without a family doctor) to the potential role of LLMs in administrative relief, but the authors caution that benefits must be weighed against bias, hallucination, and workflow disruption (Pokharel et al., 2026).
How Evidano helps with studies like the PLOS ONE protocol
Problem: Large interview volume and slow synthesis
Answer: Evidano automates core qualitative workflows so teams can go from recordings to thematic insight faster.
Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.
Evidano's transcription pipeline can ingest recorded interviews and produce verbatim, deidentified transcripts with a custom dictionary and PII redaction to match the PLOS ONE protocol’s requirement to destroy original audio and work with deidentified transcripts, and the platform's transcription details are described on the Evidano speech-to-text page (speech-to-text).
Problem: Balancing deductive (TAB) and inductive coding
Answer: Evidano supports hybrid coding so teams can apply deductive codebooks and iteratively surface inductive themes.
Evidano's thematic and subcode visualizations let teams import a TAB-aligned deductive codebook, run automated coding suggestions, and record reflexive memos to match the protocol's plan for monthly codebook refinements, and the platform features are documented at Evidano Features.
Problem: Multi-stakeholder dissemination
Answer: Evidano produces exportable tables, co-occurrence networks, and frequency analyses to translate physician themes into policy briefs.
Evidano's cross-segment and frequency analyses help show where concerns are concentrated (for example by level of prior LLM use), which mirrors the protocol's purposive sampling intent to compare frequent, occasional, and nearly no-use clinician groups.
FAQ: llm adoption in primary care
What methods will the PLOS ONE study use to understand physician perceptions?
Answer: The study will use semi-structured interviews, purposive sampling, and thematic analysis guided by the TAB framework.
According to Pokharel et al. (2026) in PLOS ONE, interview data will be transcribed verbatim, deidentified, and coded using both deductive codes from the TAB framework and inductive codes that emerge from the data.
What is the TAB framework and why did the authors choose it?
Answer: The TAB framework integrates perception-based models (ease of use and usefulness) with task-technology fit to explain adoption behavior.
According to Pokharel et al. (2026) in PLOS ONE, TAB unifies prior models like TAM and task-technology fit so researchers can link perceived fit to perceived usefulness when analyzing clinician accounts.
Has the PLOS ONE study finished data collection?
Answer: No, at the time of protocol publication the study had not yet begun data collection.
According to Pokharel et al. (2026) in PLOS ONE, the protocol states recruitment and data collection were anticipated between August 1, 2025 and January 31, 2026, and no datasets had been generated at publication on August 7, 2026.
How can AI-enabled qualitative tools reduce bias when analyzing interviews about LLMs?
Answer: AI-enabled platforms can standardize transcription, surface code co-occurrence, and make coder decisions auditable to reduce procedural bias.
According to best-practice methods cited in the PLOS ONE protocol, reflexive memos, multidisciplinary teams, and transparent codebooks are central to trustworthy interpretation, and AI tools can make those artifacts searchable and reproducible (Pokharel et al., 2026).
Conclusion & Next Steps
Answer: The PLOS ONE protocol by Pokharel et al. (2026) provides a transparent, theory-driven plan to capture how primary care physicians perceive LLM adoption, and AI-enabled qualitative tools can make that plan faster and more auditable (PLOS ONE).
Researchers conducting similar studies should preregister deductive codebooks, use purposive sampling to contrast experience levels, and preserve reflexive memos as the protocol recommends to support nuanced interpretation (Pokharel et al., 2026).
If you are running interviews about clinician experiences with AI, a platform that combines accurate transcription, hybrid coding, and cross-segment visualization can cut synthesis time and improve reproducibility; see Evidano for those capabilities (Evidano Features).
Next step: pilot your interview guide, build a TAB-aligned codebook, and use a reproducible pipeline for transcription and deidentification before analysis. Try our platform to streamline that work: Try Evidano for free.
Topics
- llm adoption in primary care
- qualitative analysis of LLMs
- physician perceptions of AI
- AI-enabled qualitative research
Keep reading
- Commentary on NewsPhysician Perceptions of LLMs: Adoption in Primary CarePractical explainer of physician perceptions of LLMs from a PLOS One protocol (7 Aug 2026). Learn methods, timelines, and how AI-enabled qualitative research accelerates insight.
- Commentary on NewsPhysician Perceptions of LLMs: Qualitative AnalysisWhat primary care physicians think about LLM adoption and how AI-enabled qualitative research can speed insight. Learn methodological takeaways and next steps.
- Commentary on NewsProtocol: Physician perceptions of LLMsHow primary care physicians view large language models, per a PLOS ONE protocol (Aug 7, 2026); learn methods, numbers, and how AI-enabled qualitative research accelerates insights.