AI Qualitative Analysis in Healthcare: PROMIS CATs

The PLOS One study found that therapists and patients valued CATs because the adaptive format focused questions on the respondent's condition and could support conversations about symptoms and functioning.

One patient said, "Answering the questions made me reflect on my symptoms and what I want help with" (Patient #5). One therapist said, "The graph helps patients see their progress, even when symptoms are still there" (PT #2). These quotations show how qualitative evidence explains the mechanism behind the survey results.

The PLOS One study identified duplicate legacy PROMs, patient-portal problems, uncertainty about T-scores and cut-offs, limited organizational embedding, and concerns about the relevance of some items as implementation barriers.

The study also reported that only 48% of therapists always or mostly discussed CAT results with patients. A qualitative coding workflow can distinguish barriers caused by the CAT itself from barriers caused by the surrounding digital and administrative system.

The PLOS One study showed why therapist perceptions should not be treated as a substitute for patient evidence. Only 24% of therapists believed patients were satisfied with CATs, while interviewed patients generally described CATs as quick, simple, and relevant.

Cross-segment analysis can preserve this divergence rather than averaging it away. Researchers can compare patients, therapists, practice owners, and implementation leaders by theme frequency, representative quotations, and reported barriers.

The PLOS One study recommended phasing out legacy PROMs where appropriate, improving clinician training, securing reimbursement, integrating PROMIS into clinical guidelines, and evaluating additional domains relevant to primary care physical therapy.

The study also reported a therapist quotation that captures the central implementation condition: "If the CAT can replace NPRS, PSFS, and Quebec while giving the same information, that's a win-win" (PT #6). Researchers should therefore analyze replacement value and workflow fit, not only perceived usability.

Evidano is an AI-powered qualitative data analysis platform that helps healthcare researchers analyze interviews, open-ended survey responses, and research documents.

Researchers can ingest interview transcripts, study papers, reports, and survey spreadsheets, then connect numerical response patterns with qualitative themes. Evidano's features include thematic, content, frequency, and cross-segment analyses that can support a workflow similar to the therapist-versus-patient comparison in the PLOS One study.

Evidano can help researchers search across source documents and analyses to locate participant language associated with concepts such as burden, relevance, score interpretation, portal access, and clinical usefulness.

Researchers should review every AI-retrieved quotation against the original transcript and preserve participant identifiers, context, and analytic decisions in the research record.

Evidano can organize qualitative evidence by segment, such as patients, therapists, practice owners, or implementation managers, so researchers can compare whether the same issue appears differently across groups.

A cross-segment workflow could surface the divergence reported by the PLOS One study between therapists' perceptions of patient satisfaction and patients' own interview accounts.

Healthcare research teams should use privacy-preserving workflows when analyzing interview transcripts and patient feedback. Evidano states that data is encrypted and never used to train third-party models; researchers can review the Evidano data security approach before selecting a platform.

AI-assisted analysis should support, not replace, human review, reflexivity, consent requirements, and the study's approved analytic protocol.