Better Care: AI Qualitative Analysis for Healthcare

According to PLOS ONE, a qualitative study published on November 5, 2026, clinical supervision in Lesotho’s public hospitals combined quality assurance, resource management, professional development, and psychosocial support. The study shows how AI qualitative analysis for healthcare research can help researchers compare these interconnected themes across interviews while preserving the context of participants’ accounts.

• The PLOS ONE study collected eight in-depth, face-to-face interviews with nurse managers between May 1, 2024, and July 31, 2024.
• The PLOS ONE researchers identified three themes and eight sub-themes through thematic analysis conducted by a researcher and an independent co-coder.
• The PLOS ONE findings link clinical supervision with staffing, equipment availability, performance monitoring, quality assurance, skill development, conflict management, and staff well-being.
• The PLOS ONE participants described nurse managers as needing to be accessible and responsive; one participant said, “We have to be readily available for registered nurses when assistance is needed.”
• The PLOS ONE study identified resistance to supervision, staff skill gaps, limited leadership preparation, and a lack of standardized supervision frameworks as continuing challenges.

The PLOS ONE study found that nurse managers viewed clinical supervision as a broad practice that supports both nursing standards and the conditions required for safe care. The study was designed as exploratory descriptive qualitative research because the authors sought detailed accounts of nurse managers’ experiences in public hospitals in Lesotho.

The PLOS ONE researchers selected eight nurse managers purposively, with one participant from each of eight public hospitals across the districts of Mohale’s Hoek, Mafeteng, Maseru, Berea, Leribe, and Butha-Buthe. Interviews lasted between 45 and 60 minutes, were recorded with consent, and were transcribed verbatim.

The PLOS ONE researchers and an independent co-coder used Braun and Clarke’s six-step thematic analysis process. The team familiarized itself with the transcripts, generated initial codes, developed candidate themes, reviewed and named themes, and selected participant quotations to ground the final interpretation.

The first PLOS ONE theme concerned supervisory dimensions that facilitate healthcare quality. The theme included human resource support, material resource support, supervisory methods, and supervision intended to improve patient outcomes.

The second PLOS ONE theme concerned practices that equip and support registered nurses. Participants described hands-on assistance, mentoring, advocacy, emotional support, demonstrations, and targeted training.

The third PLOS ONE theme concerned challenges with clinical supervision. Participants discussed staff attitudes, accountability, communication gaps, incomplete appraisals, resistance to new systems, and nurse managers’ limited access to supervision training.

One PLOS ONE participant summarized the training gap by saying, “We need workshops or training focused on clinical supervision… the absence of such training leaves us at a disadvantage.”

Healthcare qualitative researchers should treat AI as an analytical aid rather than a substitute for methodological judgment. The PLOS ONE study demonstrates that credible interpretation depends on purposive sampling, verbatim transcription, co-coding, reflexivity, participant quotations, and a documented analytical process.

AI qualitative analysis for healthcare research is most useful when the research team has a defined question and a traceable coding structure. Researchers can use AI to identify recurring references to staffing, equipment, supervision methods, patient outcomes, emotional support, and training, then review the supporting excerpts against the original transcripts.

Cross-segment analysis can also help healthcare teams compare accounts by hospital, role, ward, shift, professional experience, or interview period when those fields are available and ethically appropriate. The PLOS ONE study did not establish statistical generalizability, so AI-assisted comparisons should likewise preserve the difference between thematic patterns and population-level claims.

Healthcare researchers should protect confidentiality when analyzing interview data. The PLOS ONE researchers used participant identification codes, removed personally identifying information from transcripts and reports, and obtained ethics approval from the North-West University Health Research Committee and Lesotho’s Ministry of Health.

A practical workflow is to transcribe and de-identify interviews, import the documents, generate provisional themes, inspect representative quotations, compare themes across segments, and document the human decisions that led to the final interpretation.

The November 5, 2026 PLOS ONE study shows that clinical supervision in Lesotho’s public hospitals involves quality assurance, staffing and resource support, professional development, and psychosocial care. The study also shows why healthcare researchers need workflows that preserve participant context while making patterns easier to identify and compare.

AI qualitative analysis for healthcare research can reduce manual synthesis time, support transparent theme development, and help teams explore differences across interviews without removing human oversight. Evidano brings these capabilities together for interview transcripts, open-ended responses, and research documents.

Researchers can try Evidano for free to organize qualitative healthcare data and begin analyzing themes more efficiently.