AI-enabled qualitative analysis: hospital pharmacy crisis

According to the study Crisis management in Finnish hospital pharmacies during the COVID-19 pandemic (Latonen et al., 2025), hospital pharmacies in Finland reported specific operational changes, uneven information distribution, and clear gaps in preparedness during the COVID-19 waves.

• A national survey was sent to all hospital pharmacy heads (n = 21) in October–November 2020 and achieved a 57% response rate (n = 12), according to Latonen et al. (2025).
• Latonen et al. (2025) report that risk perception about the pharmaceutical supply chain rose from 58% to 100% following the pandemic onset, measured in the October–November 2020 survey.
• In March–May 2021 Latonen et al. (2025) conducted six semi-structured interviews to triangulate findings, and the dataset was made available on 5 July 2025.
• Latonen et al. (2025) conclude that "Preparedness of hospital pharmacies could be improved with pre-established crisis teams and plans, and data management systems providing easily accessible information to support decision-making."

The study directly surveyed and interviewed hospital pharmacy leaders to map crisis responses and gaps, answering the question of what changed during COVID-19 in Finland.

According to Latonen et al. (2025), the authors developed a national cross-sectional survey based on crisis management process models and sent it to all hospital pharmacy heads (n = 21) in October–November 2020, and they later triangulated survey results with six purposive interviews in March–May 2021.

According to Latonen et al. (2025), the survey measured preparedness (plans, teams), operational changes (communication, clinical services, supply and procurement), and collaboration with other actors, and the authors used deductive content analysis for open-ended responses.

Latonen et al. (2025) reported concrete operational changes: internal communication and management changes in 92% of responding pharmacies, clinical pharmacy service changes in 67%, medicine supply adjustments in 58%, procurement changes in 42%, and pharmaceutical production operations changes in 25%.

The primary implication is that small national studies with mixed methods need AI-assisted workflows to scale thematic coding and triangulation quickly.

Latonen et al. (2025) show that deductive content analysis anchored to crisis management process models yields actionable categories, so qualitative researchers should map their codebook to established process models before coding.

Latonen et al. (2025) also show that limited survey response (57% in October–November 2020) and purposive interviews (six in March–May 2021) require transparent reporting of sample frames and analytic steps to support transferability.

• Design: Use process-model-based instruments to ensure coverage of preparedness, communication, supply, and collaboration as Latonen et al. (2025) did.
• Sampling: Report invited n and achieved n, as the Finnish study reported n = 21 invited and n = 12 respondents in October–November 2020.
• Triangulation: Combine survey open-texts with interviews, and record interview dates, as Latonen et al. (2025) documented interviews in March–May 2021.

Latonen et al. (2025) provide concrete dates, counts, and quotations that qualitative researchers can emulate when analyzing crisis management data: a national survey in October–November 2020 (n invited = 21, n responded = 12), interviews in March–May 2021, and dataset publication on 5 July 2025.

Applying AI-enabled qualitative analysis reduces time spent on coding and synthesis while preserving methodological transparency needed for small mixed-methods studies like the Finnish hospital pharmacy project.

If your team needs to scale deductive coding, transcript production, and cross-segment frequency analysis for crisis preparedness research, consider an AI workflow with reproducible outputs and secure data handling. Try Evidano for free.