When everyday work makes the difference

What benefits does AI offer doctors?  

Which problems could it genuinely help to solve? And what risks do professionals see? The discussion often takes place at organisational leadership level, yet the solutions ultimately reshape the daily work of healthcare professionals who care for patients and clients.

I spoke with Jari Kankaanpää, Chief Physician of Neurology at Seinäjoki Central Hospital in the Finnish wellbeing services county of South Ostrobothnia, about the expectations that healthcare professionals have for AI and in which areas it would be most helpful.

Kankaanpää describes everyday clinical work directly: “We spend a lot of time searching for the same information repeatedly because it’s scattered across different places. AI could automatically compile that information and clearly show the sources.”

Especially for patients with multiple chronic conditions, reviewing all the relevant details can take hours — and even then, clinicians cannot be fully certain that everything essential has been considered.

Because Kankaanpää also works part-time as a data analyst at Seinäjoki Central Hospital, he sees the potential value of automated summaries: “Humans can make mistakes. AI would be able to highlight inconsistencies and help us to ensure that nothing important is overlooked.”

Making documentation lighter, not heavier

A large share of healthcare work consists of record-keeping. Kankaanpää believes that AI could help with this: “If I write that a patient has hypertension, the system could automatically add that to the diagnosis. We shouldn’t have to repeat the same information several times.”

Existing information could be reused more effectively — for example, when preparing statements and referrals. Guidelines from different organisations and authorities could be integrated more closely into documentation so that essential details would be recorded and communicated correctly and at a high standard the first time. AI could also help to flag missing data in referrals or statements.

However, Kankaanpää emphasises the need to simplify processes: “The main thing is that the number of steps we need to complete goes down, not up. We don’t need a new Clippy, the old MS Windows paperclip assistant, popping up to remind us of things every five seconds.”

Support that strengthens patient safety

Healthcare work often involves situations where even a small oversight can become a significant risk. AI could act as an extra layer of assurance, alerting professionals to critical omissions. Kankaanpää uses thrombosis prophylaxis for bedridden patients as an example: “If the medication is accidentally missing, AI could point it out. That’s not the system being pushy. That’s patient safety.”

Support could even be tailored to a professional’s level of experience. A medical student might receive more guidance, while an experienced specialist could hide the alerts they don’t need.

Supporting diagnostics, not replacing the doctor

AI is at its best when it highlights key findings and supports professional decision-making. For example, in radiology, AI is already used to assist in interpreting X-ray images. In the future, AI could prioritise urgent findings for the radiologists working in the emergency services, helping patients’ care to follow the correct order of urgency even more effectively.

AI could also help professionals to identify rare diseases and accelerate diagnostics. In such cases, diverse symptoms may be misleading, resulting in multiple appointments and examinations. AI could speed up the diagnostic process by recognising combinations of symptoms and findings across the patient’s data. Kankaanpää notes that AI-generated insights must be easy to verify at source, and that the final responsibility for diagnosis and treatment decisions always must remain with the doctor.

Wellbeing at work — not just efficiency

Discussions around AI in healthcare often focus on efficiency. Kankaanpää challenges this emphasis and the metrics behind it: should we purely focus on door-hinge metrics and the number of visits, or should we look more qualitatively the impact of the issues resolved during appointments?

Kankaanpää sums up the core idea: “Good AI brings quality, not just efficiency.” Automation can reduce time pressure and ethical stress, allowing more time for thinking and for meaningful patient encounters.

My hope is that AI will support professionals in managing their workload and working hours, ensuring that the time available can be spent on tasks that genuinely matter for both the patient and the doctor.

Professional voices are heard when they are given space

A clear theme emerged throughout our discussion: AI solutions cannot be developed without the involvement of doctors. A poorly designed system goes unused, while good solutions arise from genuine co-creation.

“A good system can only be built if professionals are involved. We have clinical knowledge and many of us also have technical expertise. It should be utilised,” Kankaanpää stresses. In Finland, it would be wise to share experiences across organisational boundaries and avoid overlapping development.

Professionals’ hopes for AI are practical: fewer routine tasks, smoother documentation, better patient safety, and more time to think.

“Remove unnecessary steps and bring essential information to the forefront. That’s where AI belongs,” Kankaanpää concludes.

Photos Christoffer Björklund
In the picture Chief Physician of Neurology Jari Kankaanpää from the Finnish wellbeing services county of South Ostrobothnia (left) and Head of Medical  Wilhelmiina Sorvali, Tieto Caretech.