One in three European doctors is over 55. According to an estimate based on minimum thresholds for universal health coverage, the EU had an approximate deficit of 1.2 million doctors, nurses, and midwives in 2022. Demand for care increases as the population ages, while the medical workforce approaches retirement age: the two curves are intersecting right now, and training capacity is not keeping pace.
The essentials
- The EU is short 1.2 million doctors, nurses, and midwives as of 2022, according to OECD Health at a Glance: Europe 2024.
- More than one-third of European doctors are over 55; one-quarter of nurses as well, with risks of insufficient growth in nursing graduates and declining interest among young people in these professions.
- Demographics is aggravating both sides of the problem simultaneously: more elderly patients, an effective supply of caregivers at risk of not keeping up with demand, notably due to retirements and reduced working hours.
- Some Nordic countries and the Netherlands are experimenting with task-sharing models between professions that free up doctor time without opening new medical schools.
- Without decisions about training, work organization, and use of digital tools, the gap risks widening.
A deficit that precedes the wave
The figure of 1.2 million represents the shortage recorded in 2022, before the baby-boomer generation had fully transitioned into care dependency. European health systems are already operating below full staffing. The crisis is already underway.
The OECD has been documenting this phenomenon over several editions of its biennial report. Staffing levels have grown in many countries, while aging simultaneously increases demand for care. The pandemic triggered early exits from the profession, particularly among nurses, intensifying tensions. Tensions vary by profession and indicator: Italy and Bulgaria stand out notably for aging physicians, while Germany and Greece face specific difficulties. Nordic countries and the Netherlands are faring relatively better, for reasons worth understanding.
The age structure of physicians is the most concerning signal. A high proportion of professionals over 55 increases the risk of numerous retirements in the coming years. We know how many doctors will leave. The question is how many will enter. And there, training figures do not suggest rapid catch-up: it takes ten to twelve years to train a doctor from the start of studies to the end of specialization.
First-time medical graduates can emerge approximately six years after the actual start of their curriculum, while specialists arrive later.
Aging drives demand at the most unfavorable moment
Europeans live longer. This is a collective success, documented and measurable. But living longer does not mean living in good health longer in the same proportions. WHO indicators on years of healthy life—what epidemiologists call HALE, for Healthy Life Expectancy—show that the life expectancy gained over the past thirty years has been partly accompanied by chronic conditions: diabetes, heart failure, dementia, musculoskeletal disorders. These illnesses do not kill quickly.
They consume medical time over years, sometimes decades.
A patient with dementia requires caregiver attention nearly continuously. A type 2 diabetic visits multiple times a year, sometimes multiple specialists. A patient with chronic heart failure generates repeated hospitalizations. Multiply these trajectories by the number of people entering these categories each year in Europe: pressure on medical staffing is arithmetic, not hypothetical.
Italy and Greece are among the fastest-aging EU countries, facing significant healthcare workforce challenges. According to Destatis scenarios, Germany could face a shortage of approximately 90,000 to 350,000 nursing staff within about a decade, in a country already struggling to fill open positions in rural hospitals. Geography aggravates the problem: caregivers concentrate in cities, elderly people often live in areas where supply is thinnest.
The limits of funding alone
The most common political response is budgetary: raise salaries, create positions, fund training. These measures are necessary but insufficient.
Raising salaries improves retention but does not create additional caregivers in the short term. Opening medical schools produces visible effects in ten years. And funding more nursing training alone does not solve the question of attractiveness in a profession where some leave before age forty, exhausted by working conditions.
Research on the comparative effectiveness of European health systems, notably work synthesized in PMC literature on hospital performance, points to a structural problem: work organization has not evolved at the same pace as needs. Tasks that could be delegated to specialized nurses or pharmacists remain reserved for doctors by regulation. According to the OECD, up to 30 percent of certain repetitive administrative tasks could be at least partially automated.
The problem is both quantitative and organizational. Some countries have begun acting on this second lever.
Models that free up doctor time
The Netherlands has developed, since the 2000s, a system of nurse practitioners—nurses with two to three years of additional training who can prescribe, diagnose, and manage chronic conditions without systematic recourse to doctors. The model has reduced waiting times for routine consultations and freed up specialized time for complex cases.
Finland and Sweden have expanded nurse competencies in rural areas, where low physician density made healthcare access intermittent. Nurses there provide initial contact, triage and manage a portion of consultations, with a protocol for medical teleconsultation for situations exceeding their delegation level. This reorganization is designed to be sustainable.
The United Kingdom, outside the EU but closely observed by its neighbors, has deployed “physician associates,” intermediate health professionals trained in two years after a science degree. Deployment has sparked tensions with general practitioners over the question of responsibility boundaries, showing that reorganization is not socially neutral and requires serious sectoral negotiation.
These models share a principle: delegate downward what can be, reserve upward what requires the longest training. Medicine is not a monolithic block. Much of routine consultations—prescription renewals, follow-up of stable chronic conditions, preventive check-ups—can be handled by professionals trained differently, more quickly, and in numbers more easily adjustable.
Digital tools as amplifier, not substitute
Artificial intelligence is entering European health systems through several doors: medical imaging analysis, diagnostic assistance, patient record management, prediction of hospital readmissions. Documented time gains in radiology and pathology departments are real. A radiologist assisted by an image analysis system processes more cases per day, with attention maintained on complex cases the system flags as priority.
But digital does not cure. It helps cure more efficiently. The distinction matters because it delimits what technology can contribute to the staffing problem. It can amplify the capacity of each caregiver. It does not replace human contact in long-term care, clinical diagnosis of complex situations, or end-of-life support.
The challenge for the coming years is whether European health systems will be able to combine multiple levers: reorganization of competencies, training of intermediate profiles, investment in caregiver attractiveness and working conditions, and digital tools for existing professionals. Each lever taken separately is insufficient. Together, they could narrow the gap without waiting for medical schools to produce an entire generation of additional graduates.
Work organization as a lever through 2035
This staffing deficit raises a fundamental question that goes beyond budget and training: an aging medical team must find organizational forms allowing it to serve an aging population.
Two trajectories are emerging, though we cannot yet say which will dominate.
In the first, health systems increasingly resort to international recruitment, while the evolution of working hours varies by country and profession. Caregivers trained in Romania, Poland, or the Philippines fill vacant positions in Germany or the Netherlands. This solution works short-term. It deepens care inequalities between origin and host countries: a Romanian nurse who leaves for Berlin is one fewer nurse in Cluj hospitals. International recruitment can weaken origin countries when their health personnel are already insufficient, as illustrated by data in the OECD Health at a Glance 2024 report on health professional mobility.
The question of the collective cost of this circulation deserves a European debate that has not yet taken place.
In the second trajectory, systems invest in retraining their own workforce and creating new professional profiles. Nursing assistants become nurses on an accelerated basis. Nurses become specialized practitioners. Training duration is compressed through alternating curricula and prior learning recognition. This path is slower to launch but more sustainable, because it builds internal capacity instead of drawing it from elsewhere.
Both trajectories can coexist. What is less tenable is to do neither and hope the problem resolves itself. OECD data show a significant deficit in 2022 and persistent pressures. Tensions vary by country and sector. It closes through choices, not inertia.
A signal to watch in the years ahead: the evolution of healthy life expectancy indicators. If compression of morbidity—remaining healthy longer before becoming dependent—advances faster than expected, demand for intensive care could be delayed in time, giving health systems additional room for maneuver. WHO data on this point are worth monitoring with each update. For now, gains remain modest and unequal across countries.
The trade-offs governments still avoid
Training caregivers costs. Enhancing the status of nurses and nursing assistants to make them attractive costs. Reorganizing competencies requires negotiating with professional bodies defending their scope. And funding all this in countries where public debt is already constrained, as detailed in Agnès Verdier-Molinié’s reading on French budgetary trade-offs, requires decisions that few governments have had the courage to formulate clearly.
The European Commission has put healthcare workforce numbers on the agenda since 2021, with the EU4Health program and several calls for projects on training and mobility. But organization of health systems remains a national competency. Brussels can fund, coordinate, recommend. It cannot decide in place of Paris, Berlin, or Rome how many medical school places to open, nor how to reform regulated professions.
This division of competencies is not absurd; health is sensitive to national contexts and local care organization. But it can slow adjustment in a field where training takes a long time and the effects of today’s decisions are visible only a decade later.
Each European health system will have to decide whether to undertake now the retraining, task delegation, and training investment that OECD figures make necessary, or whether to wait until the situation becomes unsustainable. Some countries have already made this choice. Others remain in wait-and-see mode. The gap between the two groups will be visible, in ten years, in emergency waiting times and in healthcare access in the least densely populated territories.
Sources
- OECD Health at a Glance: Europe 2024, OECD, 2024
- OECD Health at a Glance: Europe, 2020-2022 editions, data on health professional mobility and staffing by country
- WHO, HALE indicators (Healthy Life Expectancy), European Health Observatory database
- PMC literature on comparative effectiveness of European hospital systems, PubMed Central
