Nurse

The profession today

Hôpital du Valais / Spital Wallis — formerly Réseau Santé Valais — brings together six hospital sites: Sion, Visp, Brig, Martigny, Monthey, Sierre. Beyond these, dozens of EMS (residential care facilities) and a network of CMS (community health centres) provide home care throughout the canton, including in the most remote mountain communes. Nurses work across radically different settings: acute care, long-term care, home support, psychiatry, emergency — often contending with distances and isolation that major urban centres never face.

Daily work spans a wide spectrum:

• Clinical assessment and monitoring: vital signs, tracking patient condition, early detection of deterioration
• Care and treatment delivery: medication, dressings, injections, IV insertion, technical procedures
• Documentation in the electronic patient record (EPR): traceability of every act, shift handovers, recording clinical observations
• Coordination with multidisciplinary teams: doctors, physiotherapists, social workers, family carers
• Patient and family relations: information, psychological support, delivering difficult news, end-of-life care
• Triage and prioritisation: in A&E, care homes, community settings — deciding who to see first and why
• Care planning: developing, adapting and re-evaluating individualised care plans

Swiss studies place between 30 and 40% of nursing time on administrative and documentation tasks¹ — time that is not direct patient care.

What AI is preparing

The most immediate — and best-documented — gain concerns clinical documentation. Today, writing handover notes, updating the EPR, completing admission and discharge forms is a burden that often falls outside care time, during breaks or after shifts end. Voice input tools and automatic EPR structuring will reduce this to a quick validation of a generated summary. This is not magic — it is contextual voice recognition trained on clinical vocabulary, coupled with a model that knows the record's structure. The nurse dictates, validates, corrects where needed. Manual entry drops sharply; time shifts towards validation, correction and traceability.

Monitoring and early detection. Embedded sensors — connected beds in care homes, monitoring bracelets, fall-detection devices — feed predictive models capable of flagging deterioration several hours before it becomes clinically visible. AI does not diagnose; it produces an alert and a probability. The clinician decides what to do with it.

Care planning support. Drawing on EPR data, medical history, current protocols and available resources, an AI assistant can propose a structured care plan for each patient at the start of care. A working first draft — not a final document. The nurse adapts, refines and validates against what data cannot show: the patient's emotional state that morning, the family's unspoken request, the team's fatigue.

Medication preparation and interaction checking. Prescription support systems — already partially deployed in some Swiss hospitals — will be extended to real-time verification of drug interactions, contraindications and dosing. The nurse remains the last human check before administration. This automated verification layer does not replace that check — it reinforces it by flagging what a tired eye at 3 a.m. might miss.

What rises in judgement

When documentation largely takes care of itself, when alerts arrive before symptoms are visible, and when care plans are proposed automatically — what remains irreplaceable becomes proportionally sharper in value.

Holistic clinical assessment. Sensors measure vital signs. They do not see that the patient has refused to eat since yesterday, that his gaze has shifted, that the way he answers questions is no longer quite the same. This composite reading — integrating verbal cues, non-verbal signals, the person's history and family context — is the core of nursing judgement. It does not automate.

The therapeutic relationship. Presence, touch, voice, the fact of being there — these are full nursing acts, documented as such in nursing science. AI can schedule a visit; it cannot be present. In a mountain care home where some residents no longer have family nearby, this presence is sometimes the only real social bond. It carries clinical and human weight that nothing replaces.

Ethical situation management. How far to continue care? How to arbitrate between the patient's wishes and the family's? When to request urgent psychiatric support? These decisions are made under high pressure, with incomplete information, against the clock. AI can inform; the decision belongs to the nurse and the team.

Coordination under uncertainty. When multiple things happen simultaneously — a deteriorating patient, a family in crisis, a short-staffed team — the capacity to prioritise, delegate, reassure and maintain consistency of care draws on a situational intelligence that language models do not reproduce.

Who has the final say?

Composite illustration. A night nurse in an Upper Valais care home receives an alert at 2:30 a.m.: a slight rise in heart rate and unusual restlessness in an 84-year-old resident. Vital signs are within normal range. The algorithm classifies the situation as requiring enhanced monitoring. She enters the room — and immediately understands that something is wrong: the man is conscious but not responding normally, his gaze is evasive and he is lightly perspiring. She suspects a stroke, activates the protocol and calls the duty physician. Treatment begins within 40 minutes. The AI alert set the nurse in motion. The clinical assessment was made in the light of her torch and her experience. (Fictional situation, composite of cases encountered in geriatric settings.)

Job profile 2030

Three new competencies will need to be explicitly included in job profiles — they are not acquired through initial training today.

The first is clinical validation of assisted alerts: knowing how to interpret a signal produced by a monitoring system or predictive algorithm in the patient's overall context, deciding on its relevance, and documenting the decision taken. This is not a technical competency — it is augmented clinical judgement that presupposes a sound understanding of algorithmic bias in healthcare.

The second is governance of the augmented patient record: mastering data flows between different tools (EPR, monitoring, prescription support systems), ensuring the consistency and traceability of clinical information, and detecting errors or omissions in automatically generated summaries. The nurse becomes, in part, an auditor of clinical information quality.

The third extends a role already present but set to grow: ethical and relational mediation in a technologically dense environment. Explaining to a patient or family why an alert triggered an intervention at 2 a.m. Maintaining the human dimension of care in a context of continuous monitoring. Being the visible guarantor that behind the sensors, there is a person who holds responsibility and presence.

What AI does not do

AI does not fix the nursing shortage. It replaces neither sufficient staffing, nor sustainable working hours, nor a retention policy. Badly deployed, it can even add a layer of monitoring, alerts and validation that increases workload instead of lightening it. Its value only appears if it genuinely removes administrative work from nurses without shifting the burden towards an invisible digital responsibility.

Going further. The deeper question is not only how AI helps the nurse who is already there — it is whether Valais should create an intermediate role between nurse and doctor, augmented by AI, to cover areas without a family physician. This model already exists in Quebec, the United States, the Netherlands, and since 2017 in the canton of Vaud. Valais has no equivalent legal framework yet. → Nurse practitioners in Valais: the missing link between medical shortage and clinical AI

The Valais context demands that this point be named clearly. According to Obsan projections commissioned by the Canton of Valais, care staffing grew by 17% between 2012 and 2019 — and growth of 42% would be needed between 2019 and 2030 to meet demand driven by an ageing population⁴. No technology bridges a gap of this magnitude. What AI can do is reshape the structure of work so that every nurse present is more where their presence counts — and less absorbed by tasks the machine can handle.

Territorial anchoring

Valais is ageing faster than the Swiss average³. Its mountain communes have high proportions of isolated elderly people, poorly served by public transport, with access times to care that remain a geographical reality. The nursing shortage is felt more acutely than in Lausanne or Geneva: recruiting for a care home at 1,400 metres altitude, in a village of 600 people, is nothing like filling a post at the HUG.

If AI reduces a significant share of the documentation burden — even partially — that is potentially recovered direct-presence time: with patients, with families, in the moments where care quality is really decided. In chronically understaffed facilities, this gain is not marginal. It can make the difference between a care home that maintains sufficient human presence and one that outsources, merges or closes.

The question for Valais is not whether AI will transform nursing — that is already under way. It is whether Valais institutions, large and small, will be equipped and trained to make use of it before the demographic constraint becomes a crisis with no solution.

What decision-makers must do now

For hospital or care home management (Hôpital du Valais / Spital Wallis, private institutions)

Launch an audit by 2026 of the real documentation burden in care units: how many hours per FTE per week are spent on EPR data entry, written handovers, administrative forms? This figure is the basis for calculating the potential gain from a clinical documentation AI tool — and the main argument for convincing care teams that deployment is not additional surveillance, but a return of time.

For an HR manager in the health sector

Integrate the three new competencies into evaluation frameworks and development reviews — without waiting for them to feature in HES-SO curricula. Nurses who already have a reflective practice around digital clinical tools are levers of internal transformation. Identify them, train them, give them responsibility for supporting their teams.

For a cantonal authority (Public Health Service)

Require that any AI deployment in care on Valais territory be accompanied by a data governance framework compliant with the FADP and publicly documented. Patient — and staff — trust in these tools rests on transparency: who has access to the data, within what perimeter, with what traceability. This framework cannot be left to the discretion of individual institutions.

---

¹ Data from Swiss Nursing Association (SBK/ASI) studies on nursing administrative workload, 2022–2024. ² Federal Act on Data Protection (FADP), AS 2022 491, in force since 1 September 2023. ³ Federal Statistical Office — Demographic trends by canton, projections 2025–2050. ⁴ Canton of Valais / Obsan — Healthcare staffing projections, Valais 2019–2030.

Jérôme Deshaie is the founder of MCVA Consulting SA, an agency specialising in AI transformation of organisations in Valais, and author of the Bisse Cognitif.