AI synthesis: qualitative analysis of LARC/PM uptake

Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents. This post applies an AI-enabled qualitative research lens to the PLOS ONE study "Understanding the low uptake of long-acting reversible and permanent contraceptive methods in rural Bangladesh" by Shoma and Barden-O’Fallon, with the primary keyword qualitative analysis of LARC/PM uptake. According to the PLOS ONE article (published Aug 25, 2026), the authors used a socio-ecological framework and in-depth interviews to explain why LARC/PM use remains low in rural Cumilla, Bangladesh. This write-up highlights the study's methods, concrete numbers and quotes, practical implications for qualitative researchers, and how AI tools accelerate trustworthy thematic synthesis.

Key Takeaways

According to the PLOS ONE study by Shoma and Barden-O’Fallon (published Aug 25, 2026) PLOS ONE, low LARC/PM uptake in rural Cumilla is driven by intertwined individual fears, family and community norms, provider practices, and health system constraints.

The study conducted 28 in-depth interviews and 9 key informant interviews between Jan 21 and Feb 20, 2023, with 37 participants consenting out of 46 approached (9 refusals), as reported in PLOS ONE on Aug 25, 2026.
According to PLOS ONE, Cumilla district had LARC/PM use of 13.6% in 2022 and national LARC/PM share declined to 8% by 2022, underscoring a programmatic gap.
The PLOS ONE results identify recurrent themes: (1) knowledge gaps and embodied fear of side effects, (2) male-dominated decision-making and mother-in-law gatekeeping, and (3) provider bias, brief counselling, privacy deficits, and workforce shortages.
The PLOS ONE study documents direct experience-based quotes such as: "Look, sister, we are uneducated people. We don’t know or understand these methods" (W 02), and a provider reflection, "Most women who come for IUD removal complain of bleeding problems" (P 04), which anchor the qualitative claims in participants’ voices.

What happened and how the study was done

Answer: The PLOS ONE study used purposive sampling and thematic analysis to map multi-level barriers to LARC/PM uptake in Cumilla, Bangladesh.

According to the PLOS ONE article (published Aug 25, 2026), the authors carried out 28 in-depth interviews with married women and men plus 9 key informant interviews with family planning providers and an Imam, collecting data between Jan 21 and Feb 20, 2023.

According to PLOS ONE, transcripts were transcribed in Bengali, translated to English, and analyzed using Braun and Clarke’s thematic analysis guided by McLeroy et al.’s socio-ecological model, using a hybrid inductive-deductive coding approach.

According to PLOS ONE, the analytic sample included 22 women, 6 men, 8 family planning service providers and 1 religious leader, and the team assessed saturation iteratively during fieldwork.

Findings snapshot

Date	Metric	Value	Implication
Aug 25, 2026	Publication	PLOS ONE	Peer-reviewed qualitative evidence available for program design
Jan 21–Feb 20, 2023	Data collection window	28 IDIs + 9 KIIs	In-depth perspectives from demand and supply sides
2022	LARC/PM prevalence in Cumilla	13.6%	Locally low uptake within Chattogram Division
1991→2022	Trend in LARC/PM share	Declined from 12.1% (1991) to 8% (2022)	Program delivery and method mix shifts require investigation
Sample consent	Approached vs consented	46 approached, 37 consented, 9 refusals	Refusals concentrated among men, indicating recruitment constraints

Implications for qualitative researchers and program evaluators

Answer: The PLOS ONE study shows that qualitative methods must capture multilevel context, embodied experiences, and power relations when evaluating contraceptive uptake.

According to PLOS ONE, researchers should design interview guides that probe embodied side effects, family power dynamics, and religious meaning because participants cited lived bodily experiences and community narratives as decisive.

According to PLOS ONE, program evaluators should pair qualitative findings with facility readiness data (for example, workforce shortages and counselling quality) to avoid blaming clients for system failures.

Practical research decisions informed by the PLOS ONE study include oversampling male partners and influencers where possible, scheduling longer interviews for women to elicit embodied narratives, and triangulating provider accounts with client quotations to detect provider bias.

How Evidano helps: AI tools mapped to the study’s barriers

Problem: Large interview corpus, slow manual synthesis → Solution: Rapid thematic coding

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

Evidano can ingest transcriptions (including custom dictionaries and PII redaction) and auto-suggest inductive themes so teams replicate Braun and Clarke workflows at scale; this directly accelerates tasks performed in the PLOS ONE study such as hybrid inductive-deductive coding.

See the platform features page for relevant capabilities: Evidano features.

Problem: Cross-cutting socio-ecological themes are hard to compare → Solution: cross-segment analysis

Evidano supports cross-segment analyses that compare themes by participant type (for example, women vs providers vs religious leader) so researchers can quantify how frequently an issue like "fear of infertility" appears across groups and dates.

This mirrors the PLOS ONE need to show how knowledge gaps and stigma vary across individual, interpersonal, community, and system levels.

Problem: Quotes and evidence need traceability → Solution: searchable transcripts + AI chat

Evidano preserves verbatim quotations and links them to source timestamps and participant meta-data, enabling transparent citation of participant voices such as "Look, sister, we are uneducated people" (W 02) as reported in PLOS ONE.

Evidano’s AI chat over your project data helps produce extractable evidence summaries for reports, grant applications, or policy briefs while maintaining an audit trail.

Problem: Iterative probing and translator checks → Solution: integrated transcription and translation

Evidano offers transcription and translation with custom dictionaries, which helps teams that, like the PLOS ONE authors, transcribe in Bengali and translate to English while preserving technical terms and local phrasing.

Accurate translation keeps quotations defensible and reduces meaning loss during cross-language analysis.

FAQ: qualitative analysis of LARC/PM uptake

What were the main barriers to LARC/PM uptake identified by the PLOS ONE study?

Answer: The PLOS ONE study found barriers at individual, interpersonal, community, institutional, and health-system levels.

According to PLOS ONE, individual barriers included knowledge gaps and fear of side effects; interpersonal barriers included male-dominated decision making and mother-in-law influence; community barriers included son preference, stigma, and religious interpretations; institutional barriers included provider bias, brief counselling, and privacy deficits; health-system barriers included workforce shortages and low incentives.

How was data collected and analyzed in the PLOS ONE study?

Answer: The PLOS ONE study used purposive sampling, in-depth semi-structured interviews, and thematic analysis.

According to PLOS ONE, interviews were conducted from Jan 21–Feb 20, 2023 (28 IDIs and 9 KIIs), transcribed in Bengali, translated into English, and coded using a hybrid inductive-deductive approach with Braun and Clarke’s framework and the socio-ecological model as an organizing lens.

Can AI tools change interpretation or bias in such qualitative studies?

Answer: AI tools can accelerate coding and surface patterns but must be used with careful human oversight to avoid reinforcing bias.

According to best practices and the PLOS ONE emphasis on reflexivity, researchers should use AI to propose codes and frequencies, while retaining human review for context-sensitive interpretation, verbatim quote selection, and ethical judgment.

Are the PLOS ONE findings generalizable beyond Cumilla district?

Answer: The PLOS ONE authors state the findings are context-specific and analytically transferable rather than statistically generalizable.

According to PLOS ONE, the purposive sample and qualitative design aimed for depth in Cumilla, and the authors caution about regional variation, limited male sample size, and predominance of Muslim participants when applying the findings elsewhere.

Conclusion & Next Steps

Recap: The PLOS ONE qualitative analysis (published Aug 25, 2026) documents how fear, social norms, provider practices, and health-system constraints combine to suppress LARC/PM uptake in rural Bangladesh; the reports’ 28 IDIs and 9 KIIs provide concrete, attributable evidence.

Next steps for researchers include pairing these qualitative insights with facility readiness metrics and designing migration- and gender-sensitive interventions that address embodied concerns and provider accountability, as recommended by PLOS ONE.

If you want to accelerate transparent, auditable thematic synthesis of interviews like those used in the PLOS ONE study, try an AI-first workflow: Try Evidano for free.

Topics

qualitative analysis of LARC/PM uptake
LARC uptake Bangladesh qualitative
AI-enabled qualitative research
thematic analysis family planning

Keep reading

Browse all articles