Healthcare organisations across Europe need frontline insight to shape programmes, technology investments and support for their people. The need is urgent. The World Health Organization's 2025 report on the state of the world's nursing estimated a global shortage of 5.8 million nurses in 2023. Every organisation trying to keep caregivers needs to understand what their work is really like.
Yet the usual research methods reach only a few of them. Scheduled interviews, focus groups and long surveys favour people who can step away from work. Caregivers on rotating shifts, under emotional strain and administrative pressure, are the least able to take part, and often the ones whose experience matters most.
Why caregivers are hard to hear from
Time: shifts change weekly, breaks are short and unpredictable.
Emotional load: after a difficult shift, a scheduled call is the last thing anyone wants.
Language: across the Benelux, caregivers work in Dutch, French and German, and many speak other languages at home.
Candour: people may hesitate to criticise their employer face to face.
A research method that ignores these constraints hears only part of the story.
Conversations that fit the shift
Asynchronous, text-based conversations on a phone change the picture. A caregiver can answer a question during a break, continue after a shift and finish the next day. The conversation waits for them.
Research on survey methods supports the approach. Tourangeau and Yan's review of sensitive questions found that people report more openly when they self-administer questions rather than when an interviewer asks them. For caregivers discussing workload, stress or management, that openness matters.
A well-designed research conversation:
sets context clearly: who is asking, why, and what will happen with the answers;
asks open questions and follows up on what the caregiver raises;
paces itself so it never feels like a long form;
works in the caregiver's language, written natively rather than translated;
notices when someone seems distressed, and offers support and contacts rather than continuing as if nothing happened.
Privacy is the precondition
Caregiver experiences are sensitive, and some answers touch on health and wellbeing, which the GDPR treats as special category data. Trust depends on design:
clear, informed consent before the conversation begins, and the right to stop at any time;
collecting only what the research needs;
encryption in transit and at rest;
pseudonymous identifiers, so analysts never see who said what;
access limited by role, with a full audit trail;
results reported in aggregate, never in a form that could identify an individual.
Participants should know exactly how their words will be used. Without that confidence, they will not speak freely, and the research fails.
From conversations to insight
Hundreds of conversations produce a large body of free text. Without a careful method, organisations either drown in it or reduce it to numbers that lose their meaning.
AI can help structure the material: grouping responses into themes, tagging them by setting or role, and surfacing where experiences differ between teams or sites. The interpretation stays with people. Researchers review the themes, read the underlying responses, challenge the patterns and decide what they mean. A companion piece on conversational research at scale sets out that method in detail.
What leaders can learn
Done well, this kind of research shows things dashboards cannot: which parts of the shift drain people most, where administrative work crowds out care, which support actually helps, and what makes people consider leaving. It also shows differences between units, roles and sites that averages hide.
The value is not only in the findings. Caregivers who are asked, and who see that their answers led to change, feel heard. That in itself can affect whether they stay.
Questions before starting caregiver research
What decision will this research inform?
Which caregivers are hardest to reach today, and why?
In which languages must the conversation work?
How will consent, anonymity and data protection be explained to participants?
Who will interpret the findings, and how will caregivers hear what changed?
How Mitochondria's agentic AI systems support caregiver research
The approach. Mitochondria's research framework for a Benelux healthcare organisation designed multilingual, asynchronous research conversations on familiar mobile channels, with GDPR-first design: consent, data minimisation, encryption, pseudonymisation and audit built in from the start.
What it enables. Caregivers take part on their own time, in their own language. Their responses are structured into themes for researchers to review, with the underlying words always traceable. Leadership sees aggregate insight, never individual identities.
The impact. Organisations hear from the caregivers they most need to understand, and can act on what they learn.
Read more about healthcare administration.
Sources
World Health Organisation (2025). State of the World's Nursing 2025.
Tourangeau, R. and Yan, T. (2007). Sensitive questions in surveys. Psychological Bulletin, 133(5).
Regulation (EU) 2016/679 (GDPR), Article 9 on special categories of personal data.
—
Mitochondria builds agentic AI for evidence-led core operations. Research-led and model-agnostic. ISO/IEC 27001:2022 certified, with controls aligned to the GDPR, UK GDPR, India's DPDP Act and Australia's Privacy Act. Amsterdam and Pune. Request a working session →