[Health] Hospital beds in Europe — 10 questions about healthcare capacity by region (1993-2024)

What this dataset contains

eurostat_hospital_beds_nuts2 (DataCivicLab, source Eurostat) measures available beds in hospitals — the total number of hospital beds per 100,000 inhabitants — by NUTS2 region, for all EU countries, from 1993 to 2024.

Coverage: 25,888 rows, 13 columns, 1993–2024
Granularity: regional (NUTS2), annual
Metrics: beds per 100,000 inhabitants (P_HTHAB), number of beds (NR), inhabitants per bed (HAB_P)

---

Questions the data can answer

1. Why does Poland have 6× more hospital beds per capita than Sweden?

Zachodniopomorskie (PL): 1,236 beds/100k
Västsverige (SE): 184 beds/100k

A 6× gap between two EU countries. Is Poland over-hospitalised or Sweden under-provided? Or does it reflect different models of care — more inpatient in Eastern Europe, more outpatient/day care in Nordic countries?

2. Are Swedish regions systematically low?

4 out of the bottom 10 regions in Europe are Swedish (Västsverige, Sydsverige, Norra Mellansverige, Mellersta Norrland). All below 210 beds/100k. Sweden has one of the most outpatient-oriented healthcare systems in the world. Is this efficiency or underinvestment?

3. Which countries have the highest hospital bed density?

Poland dominates the top tier (Zachodniopomorskie 1,236, many others above 700), followed by Romania (București 1,047), Hungary (Budapest 1,012), Bulgaria, and Czechia. These are all countries with legacy Soviet-style healthcare systems built on inpatient care. Is the model shifting?

4. Are Western European countries converging?

Germany: most regions between 550-750 beds/100k
France: 350-550
Netherlands: 300-400
UK: 250-350

There's a clear West-East gradient in Europe, with Germany being an outlier — high density by Western standards. What explains Germany's hospital-heavy model?

5. Has hospital capacity been declining everywhere?

Across Europe, the trend is downward: fewer beds, shorter stays, more outpatient care. But the pace varies. Which countries have reduced capacity fastest? And did COVID reverse the trend?

6. Did COVID expose the low-bed countries?

Sweden, the UK, and Spain — all with relatively low bed density — struggled during COVID peaks. But so did Italian regions with higher density (Lombardia ~325 beds/100k). Is bed count a good proxy for pandemic preparedness, or does it depend on ICU capacity and staffing?

7. Are Turkish regions closing the gap?

Turkey (Ağrı, Kars): 199 beds/100k is at the bottom, but Western Turkish regions (Tekirdağ, İstanbul) are at 250-350. Turkey has been investing heavily in health infrastructure (the "Health Transformation Program"). Is the investment paying off?

8. Why do former Soviet countries have such high bed counts?

Poland, Romania, Bulgaria, Hungary, Czechia, Slovakia — all former Eastern Bloc countries — consistently rank above Western Europe. The Soviet healthcare model was hospital-centric. 30+ years after the transition, the legacy persists. But is it being maintained or reduced?

9. Does hospital bed density correlate with health outcomes?

Not necessarily. Spain has relatively few beds (250-350) but high life expectancy. Poland has many beds (600+) but lower life expectancy. Quality of care, staffing levels, and disease prevention matter more than bed counts. Is bed density a vanity metric?

10. Which Italian regions have the best and worst hospital capacity?

Valle d'Aosta: 375 beds/100k (best)
Calabria: 220 beds/100k (worst)

The gap is 1.7× within the same country. And Italian regions are all in the middle of the European pack — no extremes. Does Italy's regional health system deliver uniform quality despite the capacity differences?

---

How to explore

Available via MCP: eurostat_query(slug='eurostat_hospital_beds_nuts2', sql='...')

-- Regions with most hospital beds per capita (2021)
SELECT geo_label_en, ROUND(value, 1) AS beds_per_100k
FROM data
WHERE unit='P_HTHAB' AND year=2021 AND nuts_level='NUTS2' AND value IS NOT NULL
ORDER BY value DESC LIMIT 10;

-- European average by country (2021)
SELECT LEFT(geo, 2) AS country, ROUND(AVG(value), 0) AS avg_beds
FROM data
WHERE unit='P_HTHAB' AND year=2021 AND nuts_level='NUTS2' AND value IS NOT NULL
GROUP BY LEFT(geo, 2) ORDER BY avg_beds DESC;

Note: COVID caused temporary increases in some regions (field hospitals). The long-term trend is declining. Bed definitions vary slightly between countries — some include psychiatric beds, others don't.

[Gabrymi93]

Answer to Q1: Poland has ~3× more beds than Sweden — here's the data

I queried the national average across all NUTS2 regions (2021):

Average hospital beds per 100,000 inhabitants by country

Country	Beds/100k	vs Sweden (×)
🇧🇬 Bulgaria	776.5	3.8×
🇷🇴 Romania	730.8	3.6×
🇦🇹 Austria	678.9	3.4×
🇭🇺 Hungary	676.5	3.3×
🇵🇱 Poland	658.5	3.3×
🇨🇿 Czechia	666.5	3.3×
🇩🇪 Germany	~600 (estimated)	~3×
...	...
🇮🇹 Italy	315.0	1.6×
🇪🇸 Spain	300.0	1.5×
🇸🇪 Sweden	202.5	1×

Why the gap?

1. Historical model: Eastern European countries (PL, BG, RO, HU) inherited a Soviet-style healthcare system centred on inpatient hospital care. Ambulatory/outpatient care was less developed. This legacy persists.

2. Day surgery vs inpatient: Sweden and the Nordic countries have aggressively shifted to day surgery and outpatient care. A Swedish patient with a hernia repair goes home the same day. In Poland, they might stay 2-3 days. Both are fine — the data just tracks beds, not outcomes.

3. Ageing population: Sweden's population is older than Poland's (median age 41 vs 42 — similar). But Sweden has more elderly care infrastructure outside hospitals (nursing homes, home care), shifting the bed mix.

4. Efficiency ≠ bed count: Sweden has more doctors and nurses per bed, shorter stays, and better outcomes on many metrics. Germany has many beds, high spending, and good outcomes. Bulgaria has many beds, low spending, and poorer outcomes. Bed count alone doesn't tell you about healthcare quality.

5. Poland vs Sweden specifically: The gap is 3.3×, not 6× (I overestimated in the question). Polish regions (Zachodniopomorskie 1,236) score much higher than the national average, while Swedish regions (Västsverige 184) are at the bottom — the extremes make the gap look bigger than the average.

Want to explore Q5 (COVID preparedness — does bed count matter)?

[Gabrymi93]

Answer to Q5: the decline is general, Italy cuts harder than Europe — and COVID didn't reverse it

Coverage caveat first: the per-capita series (P_HTHAB) at NUTS2 has a 10-year gap (2011-2020 — no values), so the long-term comparison uses 2000 vs 2023 on the 14 countries that have both, and the COVID window (2019-2021) is computed on the absolute bed count (NR), which is complete over 2015-2023. Averages are unweighted across NUTS2 regions.

1. Yes — the decline is everywhere. 13 of 14 countries reduced beds per 100k between 2000 and 2023 (unweighted country average: 633 → 492, −22%). The only exception is Bulgaria (+15%, 700 → 806).

2. Who cut the most: Finland (−66%), Sweden (−50%), Italy third (−35%). Finland's figure deserves caution — a −66% drop in 23 years is large enough to suspect a definition break, worth checking before citing it. Italy went from 469 to 305 beds/100k: a cut deeper than the EU average, from a starting point already in the middle of the pack.

3. Italy: every single region cut, but not equally. Bolzano and Calabria −46%, Lazio −45%, FVG −45%; at the other end Umbria −24% and Valle d'Aosta −11% (still the best endowed, 335 beds/100k in 2023). The absolute bed count confirms the slide: 9,241 (2015) → 8,542 (2023), −7.6% in eight years.

4. Did COVID reverse the trend? No. In the 2019→2020→2021 window, most countries ended 2021 below their 2019 level. Italy shows the single most "COVID-shaped" blip: 2020 was the only non-negative year since 2015 (+0.2%), immediately absorbed in 2021 (−2.2% vs 2019), and the decline resumed (8,542 by 2023). The only countries with more beds in 2021 than 2019 — Turkey, Bulgaria, Spain, Poland, Romania, Greece, Portugal — are the same ones that were already not cutting before the pandemic.

Bottom line: the European bed decline is structural, not cyclical — the pandemic pause, where visible, was a pause, not a reversal. Italy sits on the aggressive side of the trend: third-highest relative cut among the 14 comparable countries, and every region reduced capacity.

Limits: unweighted averages; 14-country window for the long-term comparison; NR is an absolute count (no population normalization); Finland flagged as possible series break; P_HTHAB gap 2011-2020 means the mid-period trajectory is not observable at regional level.

Want to explore Q6 next (did COVID expose the low-bed countries — Sweden, UK, Spain — or does bed count not predict pandemic performance)?

Query (DuckDB, public clean parquet):

SELECT LEFT(geo, 2) AS country,
       ROUND(AVG(CASE WHEN year = 2000 THEN value END), 0) AS y2000,
       ROUND(AVG(CASE WHEN year = 2023 THEN value END), 0) AS y2023,
       ROUND((AVG(CASE WHEN year = 2023 THEN value END) / NULLIF(AVG(CASE WHEN year = 2000 THEN value END), 0) - 1) * 100, 0) AS chg_pct
FROM data
WHERE unit = 'P_HTHAB' AND nuts_level = 'NUTS2' AND year IN (2000, 2023) AND value IS NOT NULL
GROUP BY LEFT(geo, 2)
HAVING AVG(CASE WHEN year = 2000 THEN value END) IS NOT NULL
   AND AVG(CASE WHEN year = 2023 THEN value END) IS NOT NULL
ORDER BY chg_pct;