What Four Clinical Services Teach us About Evidence and AI
Authors: Dr Ana Filipa Cardoso1, Dr António Manuel Fernandes1,2,3, Dr Daniela Cardoso1,2,3, Dr Ana Ferreira4, Dr António Marques1,2,3,4, Dr Beatriz Rodrigues2,3,4,Dr Berta Augusto2,3, Dr Claúdia Rocha2,3,4, Dr Cláudia Seguro2,3,4, Dr Joana Francisco2,3,4, Dr Joana Magalhães2,3,4, Dr Joana Silva2,3,4, Dr João Miguel Calvo2,3,4, Dr Lídia Monteiro2,3,4, Dr Liliana Pita2,3,4, Dr Liliana Escada Ribeiro2,3,4, Dr Lucinda Santos2,3,4, Dr Margarida Antunes2,3,4, Dr Márcia Leitão2,3,4, Dr Maribel Pinto2,3,4, Dr Pedro Cunha2,3,4, Dr Pedro Henriques2,3,4, Dr Ricardo Lopes2,3,4, Dr Sérgio Abrunheiro2,3,4, Dr Teresa Santos2,3,4, Dr Vanessa Anjos2,3,4
1. JBI Portugal Centre for Evidence Based-Practice
2. Nursing School of University of Coimbra (ESEUC)
3. Health Sciences Research Unit: Nursing (UICISA: E)
4. University Hospitals of Coimbra
This year’s World Evidence-Based Healthcare Day theme asks what it means to put people at the centre of evidence and AI. Before we can answer the question of what AI adds to evidence-informed care, it is worth asking what is actually required to change practice when the work is entirely human. Our project to transform oral hygiene care across four clinical services – for people with dependency, swallowing difficulties, cognitive impairment or limited mobility – did not involve AI at all. And we think that is exactly why it belongs in this conversation.
How did it begin?
It began with a simple question: are the people we care for receiving oral hygiene that reflects the best available evidence? Oral hygiene is easy to overlook in a busy hospital ward. It is part of everyday practice and, because of that, can almost become invisible. Yet, for someone who is dependent and has swallowing difficulties, cognitive impairment or limited mobility, it is not a minor issue. Oral hygiene affects comfort and dignity, and it is also relevant to safety and the prevention of complications. Our experience offers important lessons for how AI might support evidence-informed healthcare – and why technology alone will never be enough to change practice.
We wanted to know what was actually happening
Using the JBI approach to evidence implementation, participating clinical teams compared everyday practice with evidence-based recommendations. The initial audits revealed a familiar problem: there was a clear distance between what the evidence recommended and what was routinely done. The evidence was available, and professionals were committed to providing good care. Even so, good intentions and access to knowledge did not ensure that best practice reached every person, every day. Each clinical service represented a distinct microcontext, shaped by its patient population, team dynamics, routines, resources and priorities. Across these different realities, however, the same question emerged: how could better oral hygiene be meaningfully embedded into everyday practice?
Changing more than a practice
The first lesson came early. Telling professionals what the evidence says is not the same as helping them change practice. Approximately 150 healthcare professionals became involved. Although most were nurses, beyond them, physicians, speech and language therapists, healthcare assistants and other members of the healthcare teams also contributed. Once the evidence–practice gaps had been identified, the teams worked together to strengthen their capacity to find, interpret and apply evidence within their own contexts. The intention was not simply to give professionals ready-made recommendations. We wanted them to be able to formulate clinical questions, locate relevant knowledge, understand it critically and translate it into more evidence-informed care.
Training in oral hygiene was important, but it was only one part of a broader process. Communication with patients and caregivers was strengthened through training and supporting educational resources. Evidence-based oral hygiene involves more than performing the right procedures: professionals must also explain their purpose clearly, listen to people’s needs and preferences and involve them, as well as their caregivers, in decisions and, whenever possible, in oral hygiene practices.
Working together also strengthened communication, collaboration and relationships across professional disciplines. Teams reviewed electronic health records to support more systematic oral assessment, care planning and documentation. They reconsidered materials, procedures, workflows and the organisation of care. It quickly became clear that closing an evidence–practice gap required more than changing individual behaviour. It required clinical environments that made evidence-informed oral hygiene possible.
When the numbers started to move
After the implementation strategies had been introduced, the teams audited practice again. The direction of change was consistent, although the nature and magnitude of the improvements reflected the needs and starting points of each microcontext.
In one context, training in oral hygiene assessment and management increased from 21.8% to 100%, while support or instruction for daily toothbrushing rose from 8.5% to 76.6%. In another, overall compliance with seven evidence-based criteria increased from approximately 11% to 56%, and the adaptation of oral hygiene for aspiration or swallowing risk rose from 13.3% to 63.3%. Elsewhere, patient and caregiver education increased from 0% to 36.7%. Improvements were also observed in documentation, oral hygiene after meals and the adaptation of practices to patients’ cognitive and functional needs.
Not every area improved to the same extent. Some gaps persisted, particularly when change depended on resources, workflows or wider organisational processes. This was equally informative. Education can support change, but it is rarely sufficient on its own. Follow-up audits allowed us to move beyond counting how many professionals attended training and ask the more important questions: what actually changed in practice, and what still needed to change?
What did we learn?
One strategy worked particularly well: taking audit results back to the teams. Seeing their own data changed the discussion from ‘we think we do this’ to ‘this is what our practice is showing us’. Professionals were involved in identifying barriers, agreeing on realistic strategies and evaluating what followed. Improvements were easier to achieve when education was combined with changes to communication, documentation, resources and care processes.
The process also built capacity for future change. Professionals learned how to identify gaps, explore their causes, work with evidence, implement strategies and reassess practice. This way of working can be applied far beyond oral hygiene.
The relationship between clinical practice and academia was central. The academic team contributed methodological support, access to evidence and experience in implementation and evaluation. Clinical teams contributed equally essential knowledge of patients, workflows, resources, constraints and everyday care. Learning moved in both directions: questions from practice informed research and education, while close collaboration helped academic partners understand what it takes to move evidence beyond a publication or guideline and into routine care.
Did people feel the difference?
The audits show that professional practice changed, but they do not tell us everything about what those changes meant for patients. People do not experience evidence as a guideline or an audit percentage. They experience it when someone notices that they cannot manage their own oral hygiene; when swallowing risk is recognised; when practices are adapted to what they can do; and when they and their families understand what is happening. This led us to a question we had not asked often enough at the beginning: did people feel the difference?
We would now involve patients and caregivers earlier, create stronger mechanisms for sharing learning between teams and measure more outcomes that matter directly to people. Evidence should not only be implemented for people but also increasingly with them.
What could AI change?
Our experience exposed a challenge extending beyond oral hygiene: evidence may exist, but it is not always available to everyone when it is needed. During a structured implementation project, professionals have time and support to search for evidence, discuss it and translate it into practice. In everyday care, they must often do this while facing time pressure, competing priorities and an overwhelming volume of information.
If used responsibly, AI could make trustworthy evidence easier for professionals to access at the point of care. It could help organisations connect clinical data with evidence, recognise recurring gaps and learn across different microcontexts. For patients and caregivers, it could help translate complex evidence into information that is easier to understand, question and use.
In this sense, AI could extend evidence capacity across the healthcare ecosystem. What implementation helped us build locally – the ability to find, understand and use evidence – AI might help make more widely available. But access is only part of the challenge.
AI-generated information may be incomplete, biased or poorly matched to a person or context. Its use, therefore, requires trustworthy sources, transparency, protection of clinical data, attention to digital literacy and inequity, and meaningful human oversight. Professionals must still judge whether information is credible, relevant and appropriate. Patients and caregivers still need opportunities to understand, question and discuss it. Organisations still need to address barriers in resources, workflows and culture. Someone still needs to lead change.
Our experience taught us that identifying a gap is not the same as understanding it, and having evidence is not the same as changing care. The value of AI may, therefore, lie less in generating more information than in helping trustworthy evidence reach the people who need it, when they need it and in a form they can use.
Take-home messages
- Evidence does not implement itself. Professionals need knowledge, relationships, leadership and supportive systems if evidence is to become everyday care
- Implementation builds capacity for future change. Collaboration between clinical teams, patients, caregivers and academia can strengthen the ability to identify, implement and evaluate improvement
- AI may widen access to evidence, but it cannot replace implementation. Evidence must still be interpreted, contextualised and translated into care that people can understand and experience
References
Porritt, K., McArthur, A., Lockwood, C., & Munn, Z. (2023). JBI’s approach to evidence implementation: A 7-phase process model to support and guide getting evidence into practice. JBI Evidence Implementation, 21(1), 3–13. https://doi.org/10.1097/XEB.0000000000000361
Ribeiro, L. E., Augusto, B., Pinto, M., Santos, T., Magalhães, J., Rocha, C., Santos, D., Cardoso, D., & Cardoso, A. F. R. M. (2026). Nurses’ oral hygiene care for patients with stroke: A best practice implementation project. JBI Evidence Implementation. Advance online publication. https://doi.org/10.1097/XEB.0000000000000596
To link to this article - DOI: https://doi.org/10.70253/WHXA2367
Disclaimer
The views expressed in this World EBHC Day Blog, as well as any errors or omissions, are the sole responsibility of the author and do not represent the views of the World EBHC Day Steering Committee, Official Partners or Sponsors; nor does it imply endorsement by the aforementioned parties.