Primary keyword: AI qualitative analysis. This post shows health researchers and qualitative teams how to translate the PLOS ONE interview study on hyperkalaemia into reproducible, AI-enabled thematic insights. According to the PLOS ONE study, nephrologists described a stepwise clinical logic that prioritises RAASi preservation while citing access barriers to newer potassium binders; this article (published July 30, 2026) offers concrete quotes, numeric markers, and analytic steps you can replicate with AI-assisted tools.
Key Takeaways
According to the PLOS ONE study, 12 practising nephrologists from six Spanish regions were interviewed in October 2025 about hyperkalaemia management and RAASi preservation (PLOS ONE). The study found clinicians used a structured, stepwise approach to preserve RAASi and viewed newer binders (patiromer, SZC) as enablers, while administrative barriers limited access.
- 12 nephrologists were interviewed in October 2025, as reported in PLOS ONE published on July 30, 2026.
- 75% of participants had more than 15 years of clinical experience, representing 9 of the 12 clinicians in the sample (PLOS ONE, 2026).
- Usage in the prior 12 months was evenly split: 50% reported prescribing patiromer and 50% sodium zirconium cyclosilicate (6 of 12 each) (PLOS ONE, 2026).
- Interviews averaged 53 minutes (range 45–60 minutes) and were analysed with a codebook approach in ATLAS.ti, producing 30 codes collapsed into 15 sub-themes and five overarching themes (PLOS ONE, 2026).
What happened: how the PLOS ONE study collected and analysed interviews
According to the PLOS ONE study, researchers conducted semi-structured video interviews in October 2025 with 12 nephrologists across six autonomous communities in Spain and transcribed recordings verbatim (PLOS ONE, 2026).
According to the PLOS ONE study, the analytic approach combined inductive and deductive coding in ATLAS.ti, with two researchers developing a codebook, resolving discrepancies, and stabilising a final framework that produced 30 codes, 15 sub-themes, and five themes (PLOS ONE, 2026).
According to the PLOS ONE study, the five themes were: a stepwise clinical algorithm prioritising RAASi preservation, perceptions of newer potassium binders as enablers, system-level access barriers (including visado), the role of patient engagement and nursing, and the unmeasured burden on patients’ daily functioning and HRQoL (PLOS ONE, 2026).
Direct quotations from clinicians illustrate reasoning and lived experience: “I structure the management in a stepped way: first line, diet and diuretics; second line, bicarbonate or binders; third line, reduction or suspension of drugs that increase potassium, as a last resort.” (ID05, Female, 56 years) (PLOS ONE, 2026).
Direct quotations also show clinical impact: “The arrival of the new binders has been a revolution. I have not had to dialyse patients urgently for hyperkalaemia for years, thanks to these treatments.” (ID09, Male, 64 years) (PLOS ONE, 2026).
Findings Snapshot
| Date | Metric | Value | Implication |
|---|---|---|---|
| October 2025 | Interviews conducted | 12 nephrologists | Qualitative sample across six regions; purposive sampling (PLOS ONE, 2026) |
| July 30, 2026 | Publication | PLOS ONE article DOI: 10.1371/journal.pone.0354854 | Peer-reviewed dissemination of clinician perspectives (PLOS ONE, 2026) |
| October 2025 | Interview length | Average 53 minutes (range 45–60) | Depth supports rich thematic coding and direct quote extraction (PLOS ONE, 2026) |
| 2025 (prior 12 months) | Binder prescribing experience | 50% patiromer, 50% SZC (6/12 each) | Clinicians had hands-on experience with newer binders enabling practical insights (PLOS ONE, 2026) |
| Study analysis | Codes and themes | 30 codes → 15 sub-themes → 5 themes | Structured codebook approach enables reproducibility and AI-assisted mapping (PLOS ONE, 2026) |
Implications for qualitative health researchers
According to the PLOS ONE study, researchers should prioritise transparent codebooks and audit trails when analysing clinician interviews because administrative and contextual system factors (like visado) drive treatment decisions beyond pure clinical logic (PLOS ONE, 2026).
According to the PLOS ONE study, capture and report administrative constraints and regional formulary variation explicitly, because clinicians linked access rules to RAASi discontinuation and patient outcomes (PLOS ONE, 2026).
According to the PLOS ONE study, include HRQoL measurement plans: the authors noted that formal HRQoL assessment was infrequent and recommended routine PROMs to quantify the unmeasured burden clinicians reported (PLOS ONE, 2026).
According to the PLOS ONE study, combine participant quotes with code frequencies and cross-segment analysis (for example by years of experience or region) to show how perspectives vary; the original analysis used code frequencies and subthemes to move from raw text to five actionable themes (PLOS ONE, 2026).
How Evidano helps: map the study’s needs to AI features
Definition: what Evidano is
Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.
Researchers replicating a PLOS ONE–style analysis can ingest verbatim transcripts, preserve a full audit trail, and export codebooks for peer review using Evidano (see Evidano features).
Problem: manual codebook development is slow → Solution: AI-assisted code suggestions
According to the PLOS ONE study, the team iteratively developed a codebook through dual coding and consensus (PLOS ONE, 2026); Evidano accelerates this by proposing candidate codes from transcripts and surfacing high-frequency code clusters for human review.
Evidano supports structured codebooks and versioned audit trails so teams can reproduce the stepwise refinement that produced 30 codes and 15 sub-themes in the PLOS ONE analysis.
Problem: extracting and attributing direct quotes → Solution: quote tagging and segment linking
According to the PLOS ONE study, illustrative quotations were translated and attributed to participant IDs to preserve meaning (PLOS ONE, 2026); Evidano extracts verbatim quotes, links them to codes, and preserves speaker metadata to replicate that approach.
Evidano enables export of quote sets with code frequency matrices so you can show both exemplars and prevalence in the same report.
Problem: transcription and PII protection → Solution: accurate speech‑to‑text with redaction
Researchers replicating October 2025 interviews should transcribe and pseudonymise recordings; Evidano provides speech-to-text with custom dictionaries and PII redaction to match the PLOS ONE workflow.
Evidano keeps transcripts encrypted and maintains an auditable chain from audio to coded dataset for ethics committees and journals.
Problem: showing cross-segment patterns (region, years of experience) → Solution: cross-segment analysis and visualisations
According to the PLOS ONE study, participant characteristics (for example, 75% had >15 years’ experience) informed interpretation (PLOS ONE, 2026); Evidano produces cross-tabulations of codes by participant attributes and visual networks to make those contrasts explicit.
These outputs help researchers convert clinician perspectives into actionable recommendations for policy or implementation research.
FAQ: AI qualitative analysis for hyperkalaemia interviews
How can AI reproduce the PLOS ONE thematic analysis?
Answer: AI can assist but human judgment must steer code development.
According to the PLOS ONE study, the team used a human-led codebook approach in ATLAS.ti to ensure interpretive validity (PLOS ONE, 2026). Use AI to propose candidate codes and surface co-occurrence patterns, then let experienced coders accept, merge, or refine codes to mirror the iterative process described in the paper.
Can AI preserve confidentiality and ethical requirements for interview transcripts?
Answer: Yes, when the platform supports pseudonymisation and encrypted storage.
The PLOS ONE authors withheld full transcripts for confidentiality reasons and provided de-identified excerpts on request (PLOS ONE, 2026). Evidano supports PII redaction, audit logs, and encrypted storage so teams can meet ethics committee requirements (see Evidano data security).
How do I extract direct participant quotes and link them to codes?
Answer: Use quote-tagging with speaker metadata and code assignments.
The PLOS ONE study reported verbatim quotes attributed by participant ID and role (PLOS ONE, 2026). With AI-assisted extraction you can pull all quotes assigned to a code, export them with timestamps and participant attributes, and include them in manuscripts or policy briefs while preserving context.
Should I quantify code frequency in a small qualitative sample?
Answer: Quantify carefully and use frequencies to augment, not replace, interpretive claims.
The PLOS ONE study reported code frequencies as part of a codebook-based thematic analysis (PLOS ONE, 2026). For n=12 samples, report frequencies alongside qualitative exemplars and avoid strong causal claims; use cross-segment comparisons to highlight patterns rather than definitive prevalence estimates.
Conclusion & Next Steps
The PLOS ONE study (published July 30, 2026) demonstrates how structured interviews and a codebook approach reveal clinician priorities, access barriers, and quality-of-life gaps in hyperkalaemia care (PLOS ONE, 2026). AI-enabled workflows can accelerate the same steps the authors used: verbatim transcription, transparent codebook development, quote extraction, and cross-segment analysis.
If you are a qualitative health researcher seeking to replicate or extend this work with reproducible, auditable outputs, adopt a hybrid AI + human approach and capture administrative context as the PLOS ONE study did.
Get started by exploring Evidano features for transcripts, codebooks, and visual analysis and Try Evidano for free.