The NHS Spends Less on Staff—So Where Does the Money Go?

 

The NHS Spends Less on Staff—So Where Does the Money Go?

In my previous post, I showed that the National Health Service allocates a smaller share of its budget to staff than comparable healthcare systems. ( https://successinhealthcare.blogspot.com/2026/04/healthcare-staff-are-you-paid-well-no.html )

That’s not controversial—it’s visible across OECD data.

But it leads to a more interesting question:

If the NHS spends less on staff, where does the money go?

The instinctive answer is usually wrong.

There isn’t a single category absorbing the difference.
What the data shows is more structural than that.

---

The constraint most discussions ignore

Every healthcare system divides spending into:

• Staff (labour)
• Everything else

That “everything else” includes:

• Medicines
• Clinical supplies
• Estates and infrastructure
• Contracted services
• Administration
• Capital investment

There is nowhere else for the money to go.

So if staff take a smaller share, everything else must take a larger one.

---

The size of the gap

Across international comparisons:

• NHS: ~45–50% on staff
• Comparable systems: ~55–70%

At UK scale:

• Total health spending ≈ £240–260 billion

👉 That implies:

~£30–60 billion per year less going to staff

This is a large structural difference.

---

 

This chart shows the entire argument in one image:
the UK has a clearly smaller staff share.

---

Where the money appears instead

Using UK health accounts (~£200bn NHS England scale), spending looks broadly like:

• Staff: ~45–50%
• Medicines: ~12–15%
• Clinical supplies: ~8–12%
• Outsourced services (incl. agency): ~8–15%
• Estates & maintenance: ~3–6%
• Administration: ~5–8%
• Capital: ~3–5%

Compared to systems that spend more on staff:

• No single category stands out as unusually large
• The difference is spread across multiple areas

👉 The key point:

The gap is distributed across the system, not concentrated in one place.

---

 

What this shows:

• Staff is clearly lower in the UK
• Other categories are slightly higher—but none dominates

---

An important nuance: how spending is recorded

Some of what appears as “non-staff” spending is still labour—just classified differently.

Examples:

• Agency staff recorded as procurement
• Outsourced services including clinical labour within contracts
• Support functions not always recorded as administration

In other systems, similar activity may be counted as staff costs.

👉 So part of the difference reflects:

• Real structural choices
• Accounting and classification differences

---

What is actually different

Taking this into account, the consistent differences are:

• A lower share of spending on staff
• Lower workforce capacity indicators (e.g. doctors, beds per capita)
• Historically lower capital investment

Other categories vary, but none dominate.

---

This makes the scale tangible:

• ~£36bn difference in staff spending
• Same total budget, different distribution (UK-NHS: Staff £94 bn, Non-staff £106; Comparators: Staff 106 bn non-staff £70 bn)

---

Interpreting the pattern

What emerges is not a system with one unusually large alternative cost category.

Instead:

A smaller share going to staff means a larger share is spread across the rest of the system.

This reflects how healthcare systems are structured and accounted for.

What it implies for efficiency, outcomes, and value is a separate question.

---

The core conclusion

The NHS is distinctive not because it clearly overspends in a single area,
but because a smaller proportion of total resources is directed to healthcare staff.

Everything else follows from that.

---

Why this matters

How much a system allocates to staff affects:

• Workforce size
• Pay levels
• Capacity
• Ability to meet demand

The UK already operates with:

• Fewer doctors per capita than the OECD average
• Fewer hospital beds per capita

So differences in spending structure are likely to matter in practice.

---

Final thought

If you start with:

“The NHS spends less on staff”

Then the logical follow-on is:

“So the rest of the system must take a larger share.”

And that is what the data shows.

Not a single dominant category— but a system where spending is distributed differently, with less going to people.

---

References

1. OECD – Health at a Glance
   https://www.oecd.org/health/health-at-a-glance/
2. Office for National Statistics – UK Health Accounts
   https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem
3. UK Parliament – NHS funding
   https://commonslibrary.parliament.uk/research-briefings/sn00724/
4. The King's Fund – International comparisons
   https://www.kingsfund.org.uk/publications
5. OECD – System of Health Accounts
   https://www.oecd.org/health/health-systems/health-accounts.htm
6. NHS England – Annual reports
   https://www.england.nhs.uk/publication/annual-report/

 

My LinkedIn page:  https://www.linkedin.com/in/m-hemadri-819681a/ 

My X handle: @HemadriTweets 

 

Surgical Policy Analysis

The Gallbladder Paradox: Why National Policy is Failing the Surgical Patient

Registry and population-level evidence reviewed against NICE CG188 and GIRFT

In the modern NHS, “Index Admission Cholecystectomy” is more than a clinical recommendation; it is a performance metric. Driven by NICE CG188 and reinforced by the Getting It Right First Time (GIRFT) programme, the mandate is clear: operate within one week of diagnosis. If that window is missed, NICE Quality Standard QS104 suggests a fallback: delay until the acute episode has subsided (>4 weeks).

However, an analysis of real-world operational data suggests this “within-a-week” framing is too blunt. By prioritising throughput over individual technical safety, policy risks normalising surgery in higher-risk parts of the admission timeline—while failing to reliably deliver a protected early-delayed pathway that many surgeons recognise as safer in practice.

1

The NICE Window: 0–3 days “Golden” vs. Days 4–7 “Hazard Zone”

NICE advocates surgery “within 1 week.” The problem is that this single target spans two very different phases of operative risk.

● The Golden Window (Days 0–3)

In a population-based cohort study of 43,870 emergency cholecystectomies drawn from the NHS Hospital Episode Statistics database, Wiggins et al. (Surgical Endoscopy, 2019) demonstrated that outcomes are optimised when surgery occurs within the first three days of admission. Bile duct injury (BDI) rates were lowest in this window at 0.6%—rising to 1.0% for surgery on days 4–7, and 1.8% for surgery after day 8. For context, the BDI rate for routine elective laparoscopic cholecystectomy in a non-inflamed gallbladder is approximately 0.24% (Klos et al., Czech national register, n=76,345)—establishing the true baseline against which acute surgery should be judged. The relatively lower risk in the days 0–3 window is consistent with the recognised surgical principle that early oedema around the gallbladder can facilitate dissection planes before fibrosis and organised inflammation set in; this window closes rapidly as the acute episode progresses.

⚠ The Hazard Zone (Days 4–7)

The evidence shows that BDI rates rise when surgery is delayed into the latter half of the week. This is consistent with Blohm et al. (GallRiks, 2017), who found across 87,108 cholecystectomies that bile duct injury, 30- and 90-day mortality, and intra- and postoperative adverse events were all significantly higher when time-to-surgery exceeded approximately 4 days—the point at which inflammation becomes organised and dissection planes less reliable.

By lumping “Day 1” and “Day 7” into a single target, the system turns timing into a biological gamble, implicitly accepting late-within-week surgery as “good performance,” despite evidence that risk increases with delay across that same week.

2

The Fallback Failure: “>4 weeks” is a minimum, not a pathway

NICE QS104 suggests a fallback delay of more than 4 weeks if the index window is missed. But real-world registry data suggest there is a meaningful difference between “just over 4 weeks” and a reliably protected early-delayed slot.

The Evidence for the “Sweet Spot” (a protected 6–8‑week pathway)

In the Popowicz et al. (2023) register study (n=8,532), a statistically significant reduction in perioperative complications and cystic duct leakage was only seen when surgery was performed more than 30 days after discharge.

From a policy perspective, a 6–8‑week pathway represents a pragmatic, safety‑weighted target: it sits comfortably beyond the >30‑day inflection point where surgical risk settles, while remaining short enough to avoid the heavy recurrence burden that accumulates with the current 5-month median wait.

3

The Three Pathways: What the Data Actually Supports

The table below maps the three real-world pathways against key safety metrics. Two important distinctions underpin the recurrence figures: Lucocq et al. (NHS Scotland, n=261) reported cumulative readmission risk of approximately 48% by 10 months in biliary colic patients awaiting elective cholecystectomy—a distinct population from Helenius et al. (GallRiks/Swedish national registry, n=909), who reported a 38.3% recurrence rate at a median of 82 days in patients managed non-operatively for acute cholecystitis. These figures measure different populations, different index presentations, and different timepoints; they are cited separately and should not be aggregated.

Metric	Pathway A: Emergency (Index)	Pathway B: NHS Wait (~5 months)	Pathway C: Protected 6–8w Pathway
Timing Signal	Outcomes worsen as delay increases within admission (Days 0–3 best; later worse)	Patients re-present with more complicated disease	Surgery after hazard zone, before long-wait recurrence
Short-term Risk	1.4% mortality (AMBROSE International Benchmark)	Risk increases due to recurrences and interventions	~0.2% mortality (AMBROSE benchmark for delayed elective surgery; see note)
Recurrence Risk	None	Biliary colic: ~48% readmission by 10 months (Lucocq) Acute cholecystitis NOM: ~38.3% recurrence at median 82 days (Helenius)	Minimized by compressed wait
Bile Duct Injury (BDI)	~0.6% (Days 0–3); rising to 1.0–1.8% beyond Day 3 cf. ~0.24% for elective LC (Klos et al.)	Blended ~0.5–1.0%; higher in patients with prior recurrent episodes (OR 2.44 vs first-episode; GallRiks)	~0.3–0.5%; safe zone confirmed >30 days post-discharge (Popowicz/GallRiks)

Note on the 0.2% mortality benchmark: The AMBROSE figure for delayed cholecystectomy represents the best available real-world benchmark for the delayed elective setting. AMBROSE’s “delayed” category encompasses surgery performed at varying intervals after the acute admission—not exclusively within a protected 6–8 week window. A specifically protected 6–8 week pathway, performed before any recurrent biliary event and in the optimal surgical window identified by Popowicz et al., would be expected to perform at least as well on mortality, given that the Popowicz data confirm statistically significant reductions in perioperative complications for surgery more than 30 days post-discharge.

Pathway B — The NHS Reality

Pathway B is the worst of all worlds. Patients face a readmission risk of approximately 48% by 10 months for biliary colic (Lucocq) and a 38.3% recurrence risk for conservatively managed acute cholecystitis (Helenius), and eventually undergo surgery with anatomy significantly distorted by repeated inflammatory episodes—with documented increases in bile duct injury risk (OR 2.44 for recurrent versus first-episode cholecystitis, GallRiks), conversion rates, and operative time.

4

Conclusion: Administrative Triage vs. Patient Safety

The operational evidence supports three uncomfortable truths:

1

Earlier within admission is safer than later: A blunt “within 1 week” target obscures the rising risk after 72 hours.

2

Delay requires maturity: Technical risk only settles significantly beyond 30 days post-discharge.

3

Waiting is not benign: Long waits drive recurrence and complexity, turning elective surgery into a higher-risk journey.

If the NHS wants to “Get It Right First Time,” it must stop treating timing guidance as a throughput metric. We must build a protected “hot-to-cold” pathway that reliably delivers surgery beyond the high-risk subacute period—before the burden of recurrence is allowed to build.

Measurable Audit Standards for a Protected Early-Delayed Pathway

✓  30-day mortality: ~0.2%

✓  Bile duct injury rate: ~0.3–0.5%

✓  Patients experiencing recurrent biliary event before surgery: <20%

These are measurable standards and should be used as audit criteria for any trust implementing an early-delayed cholecystectomy pathway.

Validated Reference List

1. NICE. Gallstone disease: diagnosis and management (CG188). (2014, last reviewed 2018).
2. NICE. Quality Standard QS104: Gallstone disease. (2015).
3. AMBROSE Collaborative (Wong GYM et al.). (2025). Annals of Surgery. “30-day Morbidity and Mortality after Cholecystectomy for Benign Gallbladder Disease (AMBROSE).” (Emergency mortality 1.4%, delayed 0.2%).
4. Mytton J, et al. (2021). Annals of Surgery. “Outcomes following an index emergency admission with cholecystitis: a national cohort study.” (n=95,523; 51.1% of patients did not receive cholecystectomy within 1 year; 1-year mortality 12.2% non-operated vs 2.0% operated.)
5. Wiggins T, et al. (2019). Surgical Endoscopy. “Optimum timing of emergency cholecystectomy for acute cholecystitis in England.” (NHS HES n=43,870; BDI 0.6% Days 0–3; 1.0% Days 4–7; 1.8% ≥8 days.)
6. Popowicz A, et al. (2023). World Journal of Surgery. “Timing of Elective Cholecystectomy After Acute Cholecystitis: A Population-Based Register Study.”
7. Blohm M, et al. (2017). Journal of Gastrointestinal Surgery. “The Sooner, the Better? Data from the National Swedish Registry for Gallstone Surgery, GallRiks.” (n=87,108; BDI, adverse events and 30/90-day mortality significantly higher when time-to-surgery exceeds 4 days.)
8. Lucocq J, et al. (2023). World Journal of Surgery. “Readmission Rates While Awaiting Cholecystectomy for Biliary Colic.” (Biliary colic patients, NHS Scotland, n=261; cumulative readmission risk ~48% by 10 months.)
9. Helenius L, et al. (2025). BMJ Open Gastroenterology. “Relapse in gallstone disease after non-operative management of acute cholecystitis.” (Acute cholecystitis NOM, GallRiks/Swedish national registry, n=909; 38.3% recurrence at median 82 days.)
10. Wiggins T, et al. (2018). Surgical Endoscopy. (1-year mortality benefit for operated elderly.)
11. Riall TS, et al. (2010). Journal of the American College of Surgeons. “Failure to Perform Cholecystectomy… Associated with Increased Morbidity, Mortality, and Cost.”
12. Klos D, et al. (2023). Langenbeck’s Archives of Surgery. “Major iatrogenic bile duct injury during elective cholecystectomy: a Czech population register-based study.” (n=76,345; elective laparoscopic cholecystectomy BDI rate 0.06–0.24%.)

Healthcare Staff - are you paid well? No is the simple answer

What the Numbers Actually Show About NHS Spending 

There are many claims about how the NHS uses its money—too many managers, inefficient structures, or excessive reliance on public provision. When you compare the data internationally, a more precise and limited set of conclusions emerges. This post focuses on what can be clearly supported by comparative evidence. 

 

Overall Spending: 

 

Broadly Similar to Peers The UK spends around 10–11% of GDP on healthcare, which is close to comparable countries:  Scandinavia: ~10–12%  Western Europe: ~11–13% UK: ~10–11% [1][2] This places the UK within the normal range for high-income countries. The differences are therefore less about total spending, and more about how that spending is distributed. 

 

 Staffing: Fewer People, Lower Pay 

 

A consistent finding across international comparisons is that the NHS has: 

• Fewer doctors per capita than many comparable countries 

• Fewer nurses per capita 

• Lower average pay for both groups [3][4][7] 

The gap is particularly noticeable for nurses. In several comparable systems, nurses are both more numerous and better paid. This contributes to a larger share of spending being directed toward nursing care.

 Doctors in the UK are also generally paid less than in many other high-income countries, although the gap varies depending on the comparator [7]. 

 

Wage Share: 

 

Lower Than Comparable Systems Because of these differences in staffing levels and pay: 

• The NHS spends around 45–50% of its budget on staff [5][6] 

• Comparable systems often spend 55–70% 

This difference can be explained by:  Fewer staff overall, and lower pay levels. There is no evidence that higher staffing explains the lower wage share; the available data indicates the opposite.

 

Nursing as a Key Difference 

 

Across countries, spending on doctors as a proportion of total healthcare budgets tends to fall within a relatively similar range. The larger variation is in nursing: • Some systems allocate a greater share of resources to nursing staff • This reflects both higher staffing levels and higher pay This is one of the more important structural differences between the NHS and its peers. 

 

 Management and Administration

 The NHS has: • A smaller share of spending on administration (around 5–8%) [8] • A lower proportion of managerial and administrative staff than some comparable countries However, this should be understood in context: • The NHS also has fewer clinical staff, including doctors and nurses • Lower administrative spending reflects overall staffing patterns, rather than a distinct structural feature in isolation Comparative systems—especially those with insurance-based models—tend to have higher administrative costs due to billing, contracting, and regulatory complexity [9]. 

 

Use of the Private Sector 

The UK also differs in how healthcare is delivered. Approximate share of publicly funded care delivered by private providers: 

• UK: ~7–10% [10] 

• Scandinavia: ~10–20% 

• Western Europe: ~30–50%+ [11] 

This indicates that the NHS relies less on private sector delivery than many comparable systems. It is important to distinguish between Public funding (which remains dominant across these systems) and who delivers care (public vs private providers) The NHS is more heavily weighted toward public provision. 

 

What Can Be Concluded From these comparisons 

 

• The UK spends a similar share of GDP on healthcare as comparable countries [1][2]

 • The NHS has fewer doctors and nurses per capita [3][4] 

• These staff are paid less on average [7] 

• A smaller proportion of spending goes to wages, largely due to these factors [5][6] 

• Administrative spending and staffing are also lower, in line with overall staffing levels [8] 

• The NHS makes less use of private sector delivery than many other systems [10][11] 

 

The Bottom Line 

 

The clearest, evidence-based interpretation is this: Compared to similar countries, the NHS allocates a smaller share of its resources to healthcare staff, reflecting both lower staffing levels and lower pay—particularly in nursing—while also relying less on private sector provision. 

 

References 

 

[1] OECD (2024), Health expenditure and financing (OECD Health Statistics) – https://stats.oecd.org/index.aspx?DataSetCode=SHA 

[2] Office for National Statistics (2023), UK Health Accounts – https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem 

[3] OECD (2023), Health at a Glance: Europe – https://www.oecd.org/health/health-at-a-glance-europe/ 

[4] Nuffield Trust (2023), The NHS workforce in numbers – https://www.nuffieldtrust.org.uk/resource/the-nhs-workforce-in-numbers 

[5] The King’s Fund (2024), Key facts and figures about the NHS – https://www.kingsfund.org.uk/insight-and-analysis/data-and-charts/key-facts-figures-nhs 

[6] Institute for Fiscal Studies (2022), UK health spending – https://ifs.org.uk/publications/uk-health-spending 

[7] OECD (2023), Remuneration of doctors and nurses – https://stats.oecd.org 

[8] The King’s Fund (2017), How much does the NHS spend on administration? – https://www.kingsfund.org.uk/blog/2017/06/how-much-does-nhs-spend-administration 

[9] Commonwealth Fund (2020), U.S. health care from a global perspective – https://www.commonwealthfund.org [10] The King’s Fund (2023), The role of the private sector in the NHS – https://www.kingsfund.org.uk/publications [11] OECD (2023), Health system characteristics / provider ownership data – https://www.oecd.org/health/ 

 

 ©M HEMADRI


Follow me 

M HEMADRI on X @HemadriTweets

Healthcare learning from the armed forces

 

In UK, healthcare was told that we should learn from aviation. Healthcare was told we should learn from other fields such as industry.

Healthcare was told we should learn from the armed forces.

The concept of healthcare learning from other fields can be confusing or exciting or both.

The core activity of the armed forces is fighting an adversary. All other activities are in preparation for fighting the enemy.

How often does the British Army, Navy or Air force face active combat or deploy a weapon against an adversary? Best reasonable estimates are a maximum of 20% of soldiers will face active combat over a lifetime and about 10% would deploy a weapon (give a command, press a button or fire a personal weapon) at an adversary.

In theory or even in practice is it possible for someone to join the defence forces as a soldier (not a back office person) and retire without ever fighting the enemy? The answer is an obvious yes.

What does a soldier do when not fighting? Training, preparing and similar. Essentially a lifetime or a career full of training and repeated training.

During that training, taking the case of marksmanship the ‘pass standard’ can be as low as 40% for long range and 80% for short range.

In actual combat the accuracy of marksmanship falls to as low as 5% but thought to be no more than 25%. The forces have a great explanation for this – apparently the poor accuracy is due to most of the firing being suppressive (‘suppressive firing’). Okay, lets exclude the suppressive firing and look at the data – no great data exists (and to some extent expected due the understandable chaos in active combat)

After a lifetime of repeated training.

Now look at whether training is safe for a soldier – ‘Between 2014 and 2022, 34 UK service members died during training exercises—more than were killed in hostile action during that period’ (Ian Overton, AOAV, 18 Oct 2024).

 

Versus Healthcare

Healthcare clinical staff do not spend most of the time in training or preparation – the nature of their work means they spend their time in ‘active duty’ of delivering healthcare to their patients – every day, every week, all through the year.

If you are an A&E staff, you will deal with trauma every hour, if you are an on-call surgeon you could deal with operating on someone every day of your on-call. If you are a ward doctor you could see a patient whose physiology changes adversely every hour.

Active duty at all working times – unremitting.

Some groups such as senior doctors get a portion of their time for development (which could be as low as less than 4 hours a week in self directed personal development) and most doctors have a study leave in their contracts which is about 10 days in a year with such limited budgets allocated to the study which probably is enough for about 3 days worth of study leave.

Given the status of ‘active duty’ all through the year – healthcare seems to function very effectively. Taking one of the core activities of healthcare – saving lives – as an example and looking at the sharp end (active combat comparison) the NHS has a mortality rate of 5% to 8% for emergency admissions with nearly 70% of those who died being above the age of 70 years. Well, even for major emergency surgical operations the overall  mortality seems less than 10% and again predominantly in the elderly.  Given the fact that death is one of the consequences of acute illness needing emergency admission to a hospital especially among the elderly – we may assume that the NHS is doing really well.

 

 

Let us look at culture

The recent NHS staff survey shows that staff face bullying harassment rates of 9% from managers and 18% from colleagues; with BAME staff facing higher levels of B&H

For the armed forces it seems to be 12%. There seems no significantly higher rate for BAME staff in the forces.

Overall bullying rates - Not a great difference

Rate of sickness and absence in NHS 5.3% (2025); the armed forces don’t seem to have a well published sickness absence but discharged for medical reasons seem to be 1.4% - again on balance perhaps similar. Healthcare frontline staff are increasingly from an older age group.

 

Money

Given that UK spends about £300000 per employee for the armed forces and £125000 per employee on the NHS, with the NHS delivering millions of active engagements of core duties per day – the NHS compares well.

 

The point of training is to deliver your core role in an effective and efficient manner. The information above suggests that the NHS seems to be doing fine. We have in the NHS people from the airlines and military talking to us on how it is done there; we actually need more of people from the NHS to be talking to other industries on how we do it given the constraints and complexities.

 

Having said all this, the NHS undeniably needs to do better, far better. The NHS also needs to learn from non-healthcare sources. The crux is what should the NHS learn and from whom – the NHS does not get that right though (we will look at this hopefully in a future post).