AI Thematic Analysis for Healthcare Research

The PLOS ONE study, published on October 28, 2026, found that healthcare providers at Tanzania’s Ocean Road Cancer Institute perceived three main barriers to psychosocial cancer care: inadequate training, unsupportive organizational structures, and difficulty building trusting relationships with patients.

• The qualitative study recruited 18 healthcare providers, including 15 nurses and three social workers, through purposive sampling in 2022.
• Researchers conducted face-to-face interviews between April 20 and May 19, 2022, with each interview lasting between 30 and 48 minutes.
• None of the 18 participants reported attending a formal course specifically focused on psychosocial care.
• The researchers identified three themes through thematic analysis: inadequately trained staff, an unsupportive organizational environment, and challenges establishing patient-provider trust.
• AI thematic analysis can help healthcare researchers code transcripts, compare themes across participant groups, and trace findings back to source quotations, but researchers must verify AI-generated interpretations.

The October 28, 2026 PLOS ONE study used an exploratory qualitative design to examine healthcare providers’ perceptions of barriers to psychosocial cancer care at Ocean Road Cancer Institute in Dar es Salaam, Tanzania.

The study team recruited 18 providers until information saturation was reached, including 15 registered nurses and three social workers. According to the study, participants ranged in age from 28 to 60 years and had between 3 and 34 years of experience caring for patients with cancer.

The researchers conducted semi-structured interviews in Kiswahili, transcribed the interviews verbatim, and translated material into English during coding and theme development. The researchers manually coded the data in Microsoft Excel, compared codes through peer debriefing, and grouped codes into sub-themes and broader themes.

The study identified staffing and training constraints as a central barrier. One nurse explained, “I have not received training on psychosocial care; most of the training I have been receiving is on pain management.”

The study also identified organizational fragmentation. Participants described the absence of a dedicated psychosocial care unit, unclear referral pathways, limited coordination, and a lack of guidelines or protocols.

The third barrier involved communication and trust. One participant said, “I personally don’t get too much involved in asking patients about psychosocial things for fear of not being able to find good words to tell them.”

The authors concluded that training, coordination systems, multidisciplinary protocols, and the inclusion of psychosocial care competencies in healthcare curricula could help address the reported barriers. The study was conducted at one cancer institute, so its findings should not be assumed to represent every cancer-care setting.

Healthcare researchers can use AI thematic analysis to make interview evidence easier to audit, compare, and communicate, while retaining human responsibility for interpretation.

The Tanzania study shows why a single overall theme count is not enough. The finding about inadequate training connects to several underlying issues, including limited psychosocial-care education, insufficient confidence, heavy workloads, and competing clinical responsibilities. An AI-assisted workflow can organize these relationships into codes, sub-codes, and supporting quotations for researcher review.

The study also shows why cross-segment analysis matters. Nurses and social workers may experience the same organizational system differently, and comparing professional groups can reveal where responsibilities, referral pathways, or communication processes diverge.

Healthcare teams should treat AI output as an analytical aid rather than clinical guidance. Researchers should review source passages, check translations, test alternative interpretations, document coding decisions, and protect participant confidentiality before reporting findings.

• Use thematic analysis to identify recurring barriers and the meanings attached to those barriers.
• Use content or frequency analysis to show how often specific concerns appear, without treating frequency as proof of importance.
• Use cross-segment analysis to compare nurses, social workers, managers, patients, or other defined groups when the dataset supports those comparisons.
• Use quotation retrieval to connect every reported theme to the original participant language.
• Use redaction and secure data handling when transcripts contain names, clinical details, or other personally identifying information.

The October 28, 2026 PLOS ONE study shows how qualitative interviews can expose interconnected barriers to psychosocial cancer care, from missing training and staff capacity to fragmented coordination and limited trust.

AI thematic analysis can help healthcare researchers manage similar evidence more efficiently by linking themes, sub-themes, participant segments, and quotations in a traceable workflow.

The strongest approach combines AI assistance with human validation, ethical data management, reflexive analysis, and transparent reporting of study limitations.

Researchers who want to accelerate evidence-grounded qualitative synthesis can try Evidano for free.