A July 9, 2026 PLOS One study of 1, 217 Latvian women shows nearly equal preference for clinician-collected (45.9%) and self-sampling (42.6%) for HR‑HPV screening, with lack of confidence (81%) the top reason to prefer clinician-collected care. This post refracts those findings through an AI-enabled qualitative research lens: how to extract the emotional themes (confidence, embarrassment, trust), compare subgroups (colposcopy vs GP), and turn open-text reasons into targeted interventions using Evidano.
Key Takeaways
Evidano is an AI-powered qualitative data analysis platform that ingests questionnaires and free-text responses, auto-extracts themes, and creates cross-segment visuals and exemplar quotes. This post shows how to turn a 2026 Latvian HPV self-sampling study (n=1, 217) into prioritized interventions by coding short open-text reasons and combining them with Likert emotion items.
- Study snapshot: Feb 2021–Apr 2022; published 9 Jul 2026; n = 1, 217, with 769 GP-recruited and 448 colposcopy participants.
- Main quantitative split: 45.9% preferred clinician collection, 42.6% preferred self-sampling, 11.5% undecided (reported with 95% CIs in the paper).
- Top reported driver of clinician preference: lack of confidence performing self-sampling correctly, cited by 81% of those preferring clinician collection.
- Practical move: thematic coding of brief open-text reasons plus cross-segment analysis (e.g., recruitment setting, nationality, BMI) surfaces targeted communications to boost self-sampling uptake.
Fast take + source
The fast take is: the Latvia study shows self-sampling availability alone is not sufficient, confidence is the pivot that programs must address. Key source: PLOS One (published 9 Jul 2026).
- Study period: Feb 2021–Apr 2022; sample size for analysis: n = 1, 217 (769 GP-recruited, 448 colposcopy).
- Preference split: clinician-collected 45.9% (95% CI: 43.1–48.7), self-sampling 42.6% (95% CI: 39.9–45.4), no clear preference 11.5% (95% CI: 9.8–13.4).
- Top qualitative signal: lack of confidence performing self-sampling correctly, reported by 81.0% of participants who preferred clinician collection.
Findings snapshot
| Metric | Value | Notes / Source |
|---|---|---|
| Study period | Feb 2021 – Apr 2022 | PLOS One (published 9 Jul 2026) |
| Sample size (analysis) | n = 1, 217 | 769 GP-recruited; 448 colposcopy |
| Preference: clinician-collected | 45.9% (95% CI: 43.1–48.7) | Tabled in results |
| Preference: self-sampling | 42.6% (95% CI: 39.9–45.4) | Tabled in results |
| No clear preference | 11.5% (95% CI: 9.8–13.4) | Tabled in results |
| Top reported reason for clinician preference | Lack of confidence performing self-sampling correctly, 81.0% | Open-ended responses summarized |
Qualitative analysis of HPV self-sampling: methods & what was collected
The study captured short free-text reasons, Likert-style emotional items, and recruitment context, enabling mixed-format qualitative work. Self-sampling was performed unsupervised in a private clinic space, which is an important contextual detail when interpreting reported emotions.
- Data types available for qualitative work: short free-text reasons, categorical emotional responses (embarrassment, discomfort, confidence, intrigue), and recruitment context (GP vs colposcopy).
- Analytic opportunity: these mixed-format data are suitable for thematic coding, sentiment mapping, and cross-segment comparison (for example, non-Latvian vs Latvian; BMI categories; recent gynecologist visit frequency).
- Caveat: the paper is cross-sectional and not primarily qualitative, so open-text responses are brief, but they are actionable when aggregated and coded systematically.
What this means for researchers & UX teams
For UX / design researchers
UX and design researchers should treat lack of confidence as a product problem that maps to instruction clarity and guidance. Use thematic coding on open-text reasons to prioritize microcopy updates and to decide whether video, stepwise prompts, or live chat are needed.
For public health teams
Public health teams should segment messaging because women with recent gynecologist visits and colposcopy referrals skew toward clinician preference. Offer reassurance about accuracy and clear post-test counselling pathways, and prioritize mailed kits and confidence-building materials for under-screened groups who show interest in self-sampling.
For policy & implementation
Policy and implementation teams should measure emotional readiness, not only uptake, by including brief open-text prompts and emotion scales in pilots to surface barriers beyond logistics. Plan multilingual and BMI-sensitive materials because the study found higher indecision among non-Latvian women and those with BMI ≥25.
FAQ: HPV self-sampling
What were the main preference percentages for clinician versus self-sampling in the Latvia study?
The main preference split was nearly even: 45.9% preferred clinician-collected samples, 42.6% preferred self-sampling, and 11.5% reported no clear preference. These values are reported with 95% confidence intervals in the paper.
Why did participants prefer clinician collection over self-sampling?
The top reported reason for preferring clinician collection was a lack of confidence performing self-sampling correctly, cited by 81% of those who preferred clinician collection. Open-ended responses summarized this confidence gap as the dominant qualitative signal.
What qualitative data did the study collect that teams can code and analyze?
The study collected short free-text reasons for preference, categorical Likert-style emotion items (embarrassment, discomfort, confidence, intrigue), and recruitment context tags (GP vs colposcopy), which together support thematic coding and cross-segment analysis.
How can teams use these qualitative findings to increase self-sampling uptake?
Teams can code open-text reasons to prioritize interventions that build confidence (for example, clearer instructions, instructional videos, or clinician reassurance) and then test those interventions in pilots while measuring change in mentions of 'confidence' and preference shifts.
Do more, faster with Evidano (map to this use case)
Problem: short, scattered open-text reasons → Solution: thematic synthesis
Evidano ingests questionnaires and free-text reasons, auto-extracts themes such as 'confidence', 'accuracy doubts', and 'privacy', and returns frequency-ranked themes with example quotes. This is useful when each respondent writes one or two brief lines.
Problem: mixed-format data (Likert + text) → Solution: cross-segment analysis
Evidano combines categorical emotion items with free-text themes to produce cross-tabbed thematic frequency and significance testing across segments, for example GP vs colposcopy, nationality, and BMI, so teams can see whether high embarrassment also co-occurs with accuracy concerns.
Problem: multilingual responses & hard-to-reach groups → Solution: translation + secure handling
Evidano handles translations with custom dictionaries and transcription for audio replies, and the platform encrypts data and does not use customer data to train third-party models, which is useful for sensitive health data from non-Latvian speakers.
Problem: stakeholder buy-in → Solution: visual narratives
Evidano produces shareable outputs such as word clouds, co-occurrence networks, hierarchical code trees, and ready-to-use slide text to show concrete intervention levers like improved instructions or added clinician reassurance.
Checklist: 7-step workflow to reproduce and extend these insights
Run this 2-week pilot workflow to reproduce and extend the study's insights using Evidano.
- 1) Import the dataset: questionnaire fields, free-text reasons, and recruitment tags.
- 2) Normalize demographics and create segment flags (colposcopy vs GP; nationality; BMI bin).
- 3) Auto-extract themes from open-text and merge them with Likert emotion scores.
- 4) Run cross-segment frequency and co-occurrence analyses to surface top drivers (for example, confidence × recruitment setting).
- 5) Export exemplar quotes and build a short stakeholder brief showing the top three actionable changes.
- 6) Iterate messaging prototypes (instructions, images, short videos) and A/B test with small cohorts.
- 7) Re-run thematic analysis post-intervention to measure change in 'confidence' mentions and self-sampling preference.
Conclusion, convert the PLOS findings into targeted action
The Latvia study (published 9 July 2026) shows confidence is the pivot that determines whether self-sampling will be accepted at scale. If teams designing communications, materials, or pilots combine the study's quantitative splits with systematic qualitative coding, those teams can target interventions precisely and equitably.
- Next move: upload the study's questionnaire and open-text responses into Evidano to generate a prioritized theme list, segment comparisons, and stakeholder-ready visuals in hours, not weeks.
- Try Evidano for free to run the workflow above on your screening pilot data and produce targeted materials that close the confidence gap.