Better Care Pathways: AI Qualitative Analysis in Healthcare

The PLoS ONE SYNC-RF study protocol describes a three-phase mixed-methods programme to understand and strengthen referral and follow-up for people living with multiple long-term conditions in India. AI qualitative analysis can help research teams organize interviews, compare stakeholder perspectives, and connect themes with quantitative findings, while researchers retain responsibility for interpretation.

The protocol uses a systematic review, a district-level assessment in India, and expert consultation to develop the SYNC-RF model, or System for Navigation, Coordination, Referral, and Follow-up. The protocol authors state that the study will generate "context-specific evidence on gaps, barriers, and enabling factors" affecting referral and follow-up for multiple long-term conditions.

• In the protocol published on November 3, 2026, the SYNC-RF programme is organized into three interlinked phases: evidence synthesis, convergent mixed-methods assessment, and model development.
• The protocol plans to assess 20 Primary Health Centres, including 10 rural and 10 urban facilities, with linked Community Health Centres and a district hospital.
• The protocol targets 600 adults aged 30 years and above with at least two chronic conditions, with 30 participants recruited from each selected Primary Health Centre.
• The qualitative component plans approximately 10 patient interviews, 10 caregiver interviews, 10 healthcare provider interviews, and 4 to 6 interviews with state or district health officials, with final numbers guided by information power and data saturation.
• The protocol expects the final SYNC-RF model to be refined through consultation with approximately 8 to 12 experts and completed by mid-2027.

The SYNC-RF protocol uses sequential mixed-methods research to examine how referrals and follow-up operate across India's primary and secondary healthcare system. According to Dash, Seidu, Welford, Patel, Highton, Khunti, Prabhakaran, and Mohan, the design begins with a global systematic review, continues with district-level facility, survey, and interview research, and ends with development and expert refinement of an integrated model.

The qualitative research component uses semi-structured in-depth interviews with adults living with multiple long-term conditions, caregivers, healthcare providers, and health officials. According to the protocol authors, interviews are expected to last 30 to 45 minutes, will be conducted in English or a local language, and will be audio-recorded with consent.

The protocol authors will analyze interview transcripts through reflexive thematic analysis. The planned workflow includes repeated reading, deductive and inductive coding, grouping codes into sub-themes and broader themes, and team discussion of interpretation.

The protocol integrates quantitative and qualitative evidence through a joint display matrix. According to the authors, the matrix will identify convergence, complementarity, and divergence between facility data, patient survey results, and stakeholder interviews.

The protocol is a research plan rather than a completed evaluation. According to the study timeline, facility assessment was completed by mid-2026, survey and qualitative data collection were expected to continue through the end of 2026, and final integration and reporting were expected by the end of 2027.

AI qualitative analysis can reduce the mechanical workload of healthcare research without replacing methodological judgment. In the SYNC-RF protocol, researchers must handle interviews from four stakeholder groups, transcripts in English and local languages, facility-level evidence, and survey results; an AI-assisted workflow can help organize these materials into a traceable analytical structure.

• AI-assisted thematic analysis can identify recurring topics such as referral initiation, counter-referral, communication, patient navigation, travel barriers, and follow-up tracking across large transcript collections.
• Cross-segment analysis can compare patients, caregivers, healthcare providers, and policymakers while preserving the differences between stakeholder accounts.
• Frequency and content analysis can show which issues recur most often, but recurrence should not be treated as importance without researcher interpretation and contextual review.
• Co-occurrence analysis can reveal relationships between themes such as missing documentation, delayed referrals, transport barriers, and interrupted follow-up.
• AI chat over a controlled research corpus can help researchers locate supporting excerpts, test emerging interpretations, and identify contradictory accounts for manual review.

The PLoS ONE protocol emphasizes that divergent findings should not automatically be resolved in favor of one data source. AI qualitative analysis is therefore most useful when it surfaces contradictions for researchers to investigate, rather than presenting a single unchallenged summary.

Healthcare researchers should also protect participant confidentiality when processing interview transcripts. The SYNC-RF protocol states that data sharing will be subject to ethical and participant confidentiality requirements, so any AI workflow should include access controls, de-identification, and a clear data governance process.

The SYNC-RF protocol shows how qualitative research can explain the operational and human causes of fragmented referral and follow-up systems. The protocol's 20-facility assessment, 600-person survey, stakeholder interviews, and expert consultation create a strong use case for AI-assisted organization and cross-segment synthesis, provided that researchers verify outputs and preserve participant confidentiality. AI qualitative analysis in healthcare is most valuable when it connects transcripts to decisions without hiding disagreement, uncertainty, or context. Healthcare researchers can try Evidano for free to analyze interviews, documents, and survey responses in a single qualitative research workflow.