Implementing Nurse-Based Follow-Up: Qualitative Insights

Implementing nurse-based follow-up is a practical challenge for clinical teams and researchers who want to scale nursing interventions into routine care. The primary source for this post is a qualitative implementation study published in PLOS One by Hjorth and Forsberg (2026) that interviewed 29 clinicians and managers between 2018 and 2022 and reported implementation barriers and enablers. This post explains what the PLOS One study found, translates those findings into research and operational decisions for clinical teams, and shows how AI-enabled qualitative research methods speed synthesis and improve reproducibility for implementation projects.

Key Takeaways

The PLOS One study by Hjorth and Forsberg (2026) found that nurse-based follow-up succeeds when interventions are simple, clearly mandated, and seen as valuable by the clinic, and fails when staffing shortages, unclear roles, or ethical tensions from trial randomization exist (PLOS One).

29 clinicians and managers were interviewed across six Swedish hospitals between 2018 and 2022, representing 881 minutes of interview data (mean 40 minutes) according to Hjorth and Forsberg (2026).
The randomized trial that embedded the intervention planned to recruit 500 patients from November 2016 to December 2022 but enrolled 167 patients, the study authors reported in 2026.
Hjorth and Forsberg (2026) concluded that "implementation of nurse-based interventions is facilitated by simple procedures that can be performed independently by RNs, " and that "an adequate number of benefits are needed to drive the success of an implementation."
Managers, physicians and RNs in the PLOS One study identified 23 implementation factors, of which 11 reduced and 12 increased the probability of success (Hjorth and Forsberg, 2026).

What happened and how the study measured implementation

Answer-first: Hjorth and Forsberg (2026) used directed content analysis of 22 semi-structured interviews with 29 informants to identify 23 factors that affected implementation of a nurse-led outpatient model for liver cirrhosis.

The PLOS One study (Hjorth and Forsberg, 2026) recruited RNs, physicians and managers from six Swedish hospitals and conducted interviews in two waves from November 15, 2018 to May 20, 2022 to capture changing organisational conditions.

The PLOS One analysis (Hjorth and Forsberg, 2026) applied the PARiHS framework to sort statements into evidence, context and facilitation elements and used NVivo for coding.

The PLOS One team (Hjorth and Forsberg, 2026) reported concrete process metrics: a planned sample of 500 trial patients (NCT02957253) versus 167 actual recruits by December 2022, and 881 total interview minutes across 22 interviews.

Findings snapshot

Date / Period	Metric	Value	Implication
Nov 2016–Dec 2022	Planned trial recruitment	500 patients	Large trial goal increased resource and recruitment demands (Hjorth and Forsberg, 2026)
Dec 2022	Actual enrolled patients	167 patients	Recruitment shortfall prolonged study and limited generalisability (Hjorth and Forsberg, 2026)
Nov 15, 2018–May 20, 2022	Interview dataset	22 interviews, 29 informants, 881 minutes	Rich qualitative data capturing evolving context (Hjorth and Forsberg, 2026)
Aug 21, 2026	Publication date	PLOS One article published	Peer-reviewed dissemination of implementation lessons (Hjorth and Forsberg, 2026)

Implications for clinical researchers and implementation teams

What should trial designers change when testing nurse-led models?

Answer-first: Trial designers should reduce ethical friction from randomization and build recruitment plans that match local capacity, because Hjorth and Forsberg (2026) found that the randomized design created a moral dilemma for RNs when patients in need could not access the intervention.

Hjorth and Forsberg (2026) reported that four of six hospitals struggled to establish recruitment routines, which slowed recruitment and stretched timelines.

Practical steps include co-designing consent and allocation procedures with bedside RNs and piloting recruitment flows in the same organisational units where the intervention will be delivered, consistent with the Medical Research Council process evaluation guidance (Medical Research Council guidance).

How should managers prioritise resources during implementation?

Answer-first: Managers must explicitly protect RN time and mandate the role, because Hjorth and Forsberg (2026) found that RN mandates were fragile and easily lost when reorganisations or staffing shortages occurred.

Hjorth and Forsberg (2026) noted that staffing shortages and reorganisations were recurring barriers across sites, and that manager engagement helped stabilise RN roles when present.

What do qualitative teams need to audit during rollout?

Answer-first: Teams should track role clarity, recruitment flow, and clinician attitudes as leading indicators, because Hjorth and Forsberg (2026) identified unclear roles, low physician awareness, and variable clinic culture as modifiable implementation determinants.

Hjorth and Forsberg (2026) documented 23 specific factors and advised using frameworks like PARiHS to map evidence, context and facilitation activities over time.

How Evidano helps implementation research teams

Problem: large qualitative datasets slow synthesis

Answer-first: Teams lose weeks when transcribing and coding interviews manually, which slows feedback loops and weakens implementation decisions.

Hjorth and Forsberg (2026) used 22 interviews totalling 881 minutes, a dataset that benefits from automated ingestion and structured coding to accelerate analysis.

Solution: Evidano accelerates trustworthy synthesis

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

Evidano automates verbatim transcription, supports PII redaction and custom dictionaries useful for clinical terms, and generates thematic, content, frequency and cross-segment analyses to map factors such as those identified by Hjorth and Forsberg (2026).

Evidano integrates with research workflows to produce extractable outputs: counts by code, time-series of sentiment or recruitment barriers, and quotable excerpts for reporting, helping teams operationalise the PARiHS elements reported in the PLOS One study.

For implementation studies that collect interviews across years, Evidano features such as automated transcript alignment and hierarchical code→subcode visualisations reduce manual NVivo-style effort; see the Evidano features page for details.

Problem: decision-makers need short, evidence-linked summaries

Answer-first: Managers and clinicians need concise implementation briefs with source-linked quotes and counts, because Hjorth and Forsberg (2026) emphasised the role of clear evidence and leadership for adoption.

Evidano produces exportable executive summaries with named-source quotes and frequency tables so teams can show, for example, that '11 factors reduced success and 12 improved success' as reported by Hjorth and Forsberg (2026).

FAQ: implementing nurse-based follow-up

What were the main barriers to implementation identified in the PLOS One study?

Answer-first: The main barriers were staffing shortages, unclear roles, fragile RN mandates, recruitment difficulties and ethical tensions from randomization, as reported by Hjorth and Forsberg (2026).

Hjorth and Forsberg (2026) list 11 factors that reduced implementation likelihood, including high workload and managers prioritising financial outcomes over nursing research.

Which enablers increased the probability of implementation success?

Answer-first: Enablers included simple RN-led procedures, engaged managers, team collaboration, structured training and perceived clinical value, according to Hjorth and Forsberg (2026).

Hjorth and Forsberg (2026) observed that tutorial sessions and external facilitation created consensus and professional growth among RNs.

How should qualitative data be collected to capture changing context over years?

Answer-first: Use repeated interviews and time-stamped data collection, because Hjorth and Forsberg (2026) collected interviews at two time points per site from 2018 to 2022 to capture evolving conditions.

The Medical Research Council process evaluation guidance recommends time-sensitive process indicators and the PARiHS framework for mapping evidence, context and facilitation over time (Medical Research Council guidance).

Can AI tools be used ethically on clinical interview data?

Answer-first: Yes, when data are de-identified and platform governance protects participant confidentiality, consistent with the ethics described by Hjorth and Forsberg (2026).

Hjorth and Forsberg (2026) noted that participant confidentiality constrained public data sharing; researchers should apply PII redaction and follow local ethics approvals when using AI platforms.

Conclusion & Next Steps

Answer-first: The PLOS One implementation study by Hjorth and Forsberg (2026) shows that clear RN mandates, simple procedures and manager engagement are decisive for scaling nurse-based follow-up.

Hjorth and Forsberg (2026) documented 23 implementation factors and concluded that "an adequate number of benefits are needed to drive the success of an implementation, " which teams should measure and report.

If you run implementation research or are piloting nurse-led follow-up, use structured qualitative methods and AI-assisted synthesis to shorten the feedback loop and produce decision-ready evidence.

To try a workflow that ingests interviews, transcribes, codes and generates executive summaries with quotable source links, Try Evidano for free.

Topics

implementing nurse-based follow-up
nurse-based follow-up implementation
qualitative implementation study
AI-enabled qualitative analysis

Keep reading

Browse all articles