Physician Views: Qualitative Analysis of LLM Adoption

Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents. According to Pokharel et al. (2026) in PLOS ONE, this study protocol describes a planned descriptive qualitative study of primary care physicians' perceptions of large language model (LLM) adoption. According to Pokharel et al. (2026) in PLOS ONE, the protocol was published on August 7, 2026 and lists a recruitment timeline from August 1, 2025 to January 31, 2026. Researchers and UX teams who plan AI-enabled qualitative research need clear methods, reproducible coding plans, and tools to manage transcripts and cross-segment analyses, as Pokharel et al. (2026) recommend thematic analysis using deductive and inductive coding. Pokharel et al. (2026) in PLOS ONE also emphasize ethical protections, including deidentification and protocol approval from the UBC Behavioral Research Ethics Board (H25-01658).

Key Takeaways

According to Pokharel et al. (2026) in PLOS ONE, this is a protocol for a descriptive qualitative study of family physicians in British Columbia examining perceptions of LLM adoption in clinical practice.

Pokharel et al. (2026) published the protocol on August 7, 2026 and plan participant recruitment between August 1, 2025 and January 31, 2026.
Pokharel et al. (2026) will purposively sample physicians by self-reported LLM usage categories: 50+ uses, 5–50 uses, and fewer than 5 uses, to reach sample sufficiency.
Pokharel et al. (2026) state that they "aim to use thematic analysis drawing on deductive and inductive approaches to describe physicians’ perceptions, " supporting mixed coding strategies.

What happened and how the study works

Answer: Pokharel et al. (2026) in PLOS ONE designed a protocol to interview family physicians in British Columbia about LLM adoption and to analyze transcripts with a theory-guided thematic approach.

According to Pokharel et al. (2026) in PLOS ONE, the study uses a descriptive qualitative design with semi-structured interviews to capture everyday language and perspectives.

According to Pokharel et al. (2026) in PLOS ONE, the research team will use the Technology Adoption Behavior (TAB) framework to structure data collection and will apply both deductive codes from TAB and inductive codes that emerge from the data.

According to Pokharel et al. (2026) in PLOS ONE, interviews will be recorded, transcribed verbatim, deidentified, and analyzed in NVivo with reflexive memos and team discussions to refine the codebook.

Findings snapshot (protocol numbers and implications)

Date	Metric	Value	Implication
Published Aug 7, 2026	Study type	Descriptive qualitative protocol	Signals a planned in-depth interview study rather than completed results
Anticipated Aug 1, 2025–Jan 31, 2026	Recruitment timeline	6 months	Allows for iterative piloting of interview guide and early code refinement
Reported data point in protocol	Patients without family doctor (Canada)	4.5M in 2019 → 6.5M in 2023	Contextualizes pressures on primary care that may shape physician views on LLM adoption
Sampling plan (as described)	LLM usage strata	50+ uses, 5–50 uses, <5 uses	Enables cross-segment comparison of experienced vs novice physician perceptions

Implications for qualitative researchers and health system analysts

Answer: The PLOS ONE protocol (Pokharel et al., 2026) shows how to design theory-guided interviews for rapidly evolving AI tools and how to plan coding to balance prior theory and emergent themes.

According to Pokharel et al. (2026) in PLOS ONE, using the TAB framework helps researchers link perceived usefulness and perceived ease of use to task-technology fit, which helps produce findings that inform policy and implementation decisions.

According to Pokharel et al. (2026) in PLOS ONE, purposive sampling across LLM exposure levels is essential for capturing both early adopters and cautious nonusers, enabling cross-segment thematic comparisons.

According to Pokharel et al. (2026) in PLOS ONE, reflexivity procedures and multidisciplinary coding teams strengthen interpretation and reduce single-researcher bias in qualitative AI studies.

How Evidano helps AI-enabled qualitative research

Problem: Large interview volumes slow synthesis

Answer: Evidano accelerates transcript ingestion and thematic coding so teams can focus on interpretation rather than manual organization.

Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents, and the platform supports automated transcript import and bulk coding exports to NVivo-style workflows.

Feature link: See Evidano features for thematic, frequency, and cross-segment analyses.

Problem: Need to balance deductive framework and inductive discovery

Answer: Evidano supports hybrid coding workflows with preloaded codebooks and iterative code refinement.

Evidano lets teams apply a deductive codebook based on the TAB framework and then surface emergent subcodes, matching Pokharel et al. (2026) recommendations to combine deductive and inductive approaches.

Problem: Security and ethics around sensitive clinical transcripts

Answer: Evidano encrypts project data and provides PII redaction and deidentification features so teams can meet ethics board requirements.

For more on protections, see Evidano data security.

FAQ: qualitative analysis of LLM adoption

What is the study design used in the PLOS ONE protocol?

Answer: The study design is a descriptive qualitative study, according to Pokharel et al. (2026) in PLOS ONE.

Pokharel et al. (2026) in PLOS ONE explain the use of semi-structured interviews, purposive sampling by LLM exposure, and thematic analysis combining deductive TAB-based codes with inductive codes.

How do the authors plan to sample physicians by LLM experience?

Answer: The authors will purposively sample physicians into three LLM usage strata: 50+ uses, 5–50 uses, and fewer than 5 uses, as stated in Pokharel et al. (2026) in PLOS ONE.

This stratified sampling is intended to ensure variant perspectives from frequent, occasional, and rare users and to reach analytical sufficiency.

What analytic approach do the authors recommend for coding?

Answer: The authors recommend thematic analysis using both deductive and inductive coding, according to Pokharel et al. (2026) in PLOS ONE.

Pokharel et al. (2026) explicitly state they "aim to use thematic analysis drawing on deductive and inductive approaches to describe physicians’ perceptions."

How long was recruitment planned to take in the protocol?

Answer: Recruitment was planned for six months, from August 1, 2025 to January 31, 2026, according to Pokharel et al. (2026) in PLOS ONE.

The protocol notes that this timeline allows for pilot interviews and iterative refinement of the interview guide.

Are there ethical safeguards described for recording and transcripts?

Answer: Yes, Pokharel et al. (2026) in PLOS ONE report ethics approval (UBC Behavioral Research Ethics Board; H25-01658) and plans to deidentify transcripts and destroy audio files after transcription.

These steps align with standard qualitative research protections and the protocol emphasizes participant informed consent and the right to withdraw.

Conclusion & Next Steps

Answer: The PLOS ONE protocol by Pokharel et al. (2026) provides a transparent, theory-driven plan for interviewing primary care physicians about LLM adoption and for analyzing those interviews with mixed deductive-inductive thematic methods.

According to Pokharel et al. (2026) in PLOS ONE, the TAB framework and purposive sampling by LLM experience are central design choices that make findings actionable for policy and implementation.

Evidano can operationalize the protocol's analytic steps, from secure transcription to hybrid codebook management and cross-segment comparisons; see Evidano features for details.

For teams ready to pilot an AI-enabled qualitative pipeline aligned to this protocol, Try Evidano for free.

Topics

qualitative analysis of LLM adoption
LLM adoption qualitative research
AI-enabled qualitative research
physician perceptions LLM
thematic analysis LLM

Keep reading

Browse all articles

Physician Views: Qualitative Analysis of LLM Adoption

Key Takeaways

What happened and how the study works

Findings snapshot (protocol numbers and implications)

Implications for qualitative researchers and health system analysts

How Evidano helps AI-enabled qualitative research

Problem: Large interview volumes slow synthesis

Problem: Need to balance deductive framework and inductive discovery

Problem: Security and ethics around sensitive clinical transcripts

FAQ: qualitative analysis of LLM adoption

What is the study design used in the PLOS ONE protocol?

How do the authors plan to sample physicians by LLM experience?

What analytic approach do the authors recommend for coding?

How long was recruitment planned to take in the protocol?

Are there ethical safeguards described for recording and transcripts?

Conclusion & Next Steps

Topics

Keep reading

Company

About

Preview Evidano

Account

Newsletter