The crucial role of early diagnosis for patients and the nation, understanding the costs of late-stage cancer diagnosis from a large district general hospital in England

Abstract

Background

Early cancer diagnosis is crucial in improving the survival. The main goal in NHS in cancer diagnosis is detection rate of 75% by 2028. Our study presents the economic analysis of impact of early versus late-cancer diagnosis on healthcare resources use and costs within our Trust also exploring the influence of deprivation index.

Methods

We retrospectively analyse the cost-of-care and patient-level data for 4596 patients across nine cancer groups who fully completed their cancer pathway between April 2020 and September 2024. Costs were compared between early (stage 1 and 2) versus late (stage 3 and 4) diagnosis.

Results

Significant variations in costs were determine across cancer types, with colorectal and haematological malignancies being most costly. Early-stage diagnosis averaged £11,2K, significantly lower than late- stage £23,8K with largest differences seen in haematological, colorectal and breast cancers. A hypothetical 75% early detection rate could save the trust £14.7 million over four years. Successful treatment yielded an average 10.74 years of healthy life expectancy, further increased by early detection.

Conclusions

Late cancer diagnosis dramatically increases healthcare costs underscoring the importance of early detection and advanced screening methods. Extrapolating a 75% early detection rate across the NHS could yield substantial financial savings, highlighting its impact on healthcare efficiency.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12962-025-00657-1.

Introduction

Cancer poses a significant global health challenge and a leading cause of death [1, 2]. This condition is characterized by uncontrolled growth of cells within a certain type of tissue and can expand in the surrounding structures or migrate to other organs [3]. When cancer cells spread to lymph nodes or distant organs, this process is called metastasis [3]. Staging describes the size and extent of a lesion as well as spread within other organs [4]. A general staging description is as follows: stage one represents when a small lesion is localized within a certain organ; stage two, when the lesion is bigger but has not grown larger than the organ wall; stage three, when the malignancy is spread to the lymph nodes or surrounding structure; and stage four, when the cancer has metastasized within another organ [4]. Stages one and two indicate early-stage cancer, whereas stages tree and four are generally considered advanced or late-stage cancers [4]. The most important objective of cancer treatment is to achieve a cure that leads to better survival rates and positive prognostic outcomes [5].

The prognosis and survival rates vary significantly depending on the type of cancer being diagnosed [6]. While over 95% of individuals diagnosed with breast, prostate, or skin cancer between 2015 and 2019 survived one-year post-diagnosis, manifesting a slow progression, less than 50% of those with stomach, oesophageal, lung, liver, or pancreatic cancer achieved similar survival rates, which manifest fast progression [7]. Cancer Research UKs evidence characterizing eight of the most prevalent cancer types indicates that nearly 80% of patients achieve a survival rate of 10 years when diagnosed at stages one and two. On the other hand, when diagnosed at advanced stage (three and four), their survival declines to 25% [8]. The largest intergovernmental organization for economic cooperation and development, comprising 38 member countries, reports that the cost of cancer treatment will increase exponentially by 2050, with an add-on of £14.4 billion to the UKs expenses [9]. This financial impact on National Health Services represents a significant challenge, highlighting the need for innovative and creative solutions. The variability in the progression of cancer, potential prognosis that depends on stage during the diagnosis, and a very high economic impact of cancer underscores the requirement for targeted actions within cancer management in the UK. These actions require three strategic domains: Innovations within cancer diagnosis and treatment through research, 2. prevention, and 3. early detection.

Interventions such as immunization and behavioural modifications targeted at risk-modifiable factors such as smoking, alcohol use, and obesity aim to decrease the incidence of the disease and act as the main plan of action for cancer prevention [10, 11]. According to data from 2015, Cancer Research UK reported that nearly 40% of cancers are preventable, with smoking recognized as the leading cause, accounting for 15%, based on data from 2023, and obesity accounting for 6% [11, 12]. The Action on Smoking and Health estimates that the NHS spends over £ 1.9 billion annually on smoking-related diseases, including cancer [11]. By contrast, cancer diagnosis and treatment innovations through research targets can improve the outcomes of patients who already have the disease. Patients are usually divided into two main groups: symptomatic and asymptomatic patients. Symptomatic patients require campaigns to increase their awareness of cancer and red-flag symptoms by optimizing diagnostic pathways and addressing different socioeconomic barriers [13]. Therefore, asymptomatic populations will benefit from the development of advanced screening programs [14, 15].

Although numerous measures and initiatives have been implemented in the UK, the UK has the highest cancer rate compared to other countries in the G7 group [6]. In England, approximately one in three people is estimated to receive a cancer diagnosis during their lifetime, and approximately three million English citizens are expected to be living with cancer by 2030 [16]. While the government, coupled with the NHS, is pursuing the prevention and enhancement of early diagnosis through many strategies, active public awareness campaigns, and improved healthcare access, achieving a 75% early diagnosis target (stages 1 and 2) by 2028 remains a significant challenge [17].

Achieving this ambitious goal necessitates major changes within the NHS, and the economic implications of these changes are the main factors driving early cancer diagnosis. Like many other healthcare systems worldwide, the NHS faces huge financial pressure, especially in the post-COVID-19 pandemic period, in trying to deliver high-quality care at the lowest possible cost. The current healthcare model, represented graphically in Fig. 1, requires readjustments to enhance these complex processes. Based on strong evidence provided by the experience of breast, prostate, and lung cancer diagnostics and treatment, we would like to reinforce the idea that the key to success is emphasising the need for early and advanced methods of diagnosis. Key factors influencing early diagnosis include variations in screening and the time taken for patients to seek care and for referrals to specialists [18–21].

Fig. 1.

The world health organization (WHO) early diagnosis and treatment essential elements

Progress in the early diagnosis and treatment of cancer is significantly impacted by financial resources, which include funding for research, infrastructure, and the workforce. Understanding the costs associated with cancer diagnosis and treatment is essential for the advancement of this field. Within the literature, there is a gap in the reporting of the costs associated with cancer care pathways in England and the UK. Most of the recorded and reported data are based on episodic healthcare visits and do not describe the entire pathway from diagnosis to follow-up [16]. In this study, we address this gap and provide patient-level cost data that quantify the entire cancer care pathway within our large district general hospital. We compared the economic impact of the early stages (1 and 2) and late stages (3 and 4) on the direct utilization of resources within our NHS Trust. We also analysed the socioeconomic value characterized by the deprivation status of patients diagnosed with and treated for cancer within the catchment area.

Methods

Data sources

Our study included data from five primary sources: the Somerset Cancer Register (SCR), Hospital Episode Statistics (HES), Patient-Level Information Costing System (PLIC), Office of National Statistics (ONS), and National Disease Registration Service (NDRS) data.

SCR provides information on the clinical characteristics of patients, including tumour site, date of cancer diagnosis, length of the treatment pathway, and date of death. Using the treatment start and end dates we were able to extract the total costs and activity for cancer patients across all hospital settings, inpatients, outpatients and A&E from the single multiyear consolidated dataset.

The HES analyses data on patients’ utilization of hospital inpatient and outpatient care for all NHS patients across England including our Trust.

The Index of Multiple Deprivation (IMD), developed by ONS serves as a tool for quantifying deprivation across England. To calculate this index ONS integrates data from various sources, including administrative records, regional and national surveys, and official statistics. Based on patient postcode a IMD was generated in order perform deprivation profile analysis [22, 23].

Finally, PLIC provides in-depth insights into resource utilization at the patient level, encompassing personnel, pharmaceuticals, diagnostic examinations, definitive treatment (surgery/chemotherapy/radiotherapy), and rehabilitation. All NHS hospitals are mandated to report the costs of every service delivered to their patients at the end of the financial year. Combining these four data sources facilitated the disaggregation of treatment costs by resource type, enabling a comprehensive analysis of treatment expenditures for each patient across the entire care pathway, from the start of treatment to discharge from secondary care. Combining this information with ONS & NDRS data also enables the assessment of increased socioeconomic value from early cancer detection.

Patients

We considered all individuals with a recorded diagnosis of breast, colorectal, gynaecological, haematological, head and neck, lung, skin, upper gastrointestinal, and urological SCR between the 1st of April 2020 to 31st and March 2024. To ensure a valid comparison of resource utilization and associated costs between early- and late-diagnosed cancer, our patient sample included only patients above 18 years of age who had completed cancer pathways and were treated in our hospitals. The detailed exclusion criteria are shown in the Supplementary Material (Appendix 1). Complex cancers were defined as patients requiring advanced treatment and highly trained specialists (e.g., multiorgan resection or a recurrent disease requiring a multi-specialist surgery or oncology team). In addition, the study team divided the cohort into two major groups: early-stage patients diagnosed with stages 1 and 2 and late-stage patients diagnosed with stages 3 and 4.

Outcome

The primary outcome of the study was an assessment of the economic benefit of detecting cancer early and shortening the treatment length, both through the analysis of differences in average costs between early-stage cancer diagnosis (stages 1 and 2) and late stages of the disease (stages 3 and 4).

The secondary aim of this study was to examine the potential impact of deprivation on cancer pathway costs.

Ethical considerations

Institutional Approval was granted before the collection of anonymized data. This study was approved by the Divisional Governance Committee and adhered to all the local information and research governance guidelines. The analysis used anonymized and de-identified patient-level data to ensure compliance with NHS data management regulations and the Health Insurance Portability and Accountability Act (HIPAA). Administrative permission was granted for access to data. As the study utilized secondary, retrospective, and fully de-identified data, it was exempted from formal ethics review by the NHS Health Research Authority decision tool [24].

Study setting

Our Trust includes two acute district general hospitals in England that serve a large and diverse population of East London. Although we are a district general hospital (DGH), we serve a large population of patients across three boroughs: Barking & Dagenham, Havering, and Redbridge, and we have tertiary referral centres for some specialties across Essex. The patient population is diverse and affected by high levels of low pay, income deprivation, homelessness and premature mortality [25, 26].

Deprivation profile analysis

To analyse the socio-economic context of the patient population, the Index of Multiple Deprivation (IMD) was used [22]. The IMD represents a comprehensive measure of deprivation in England that characterising it across several domains: income, employment, education, health, crime, access to services, and living environment. An overall IMD score is generated for each Lower Layer Super Output Area (LSOA) by combining the scores from these 7 domains though a weighted average. This score provides a clear picture of the relative level of deprivation in the area where the patients’ lives compared to other areas in England.

Deprivation profiles were assigned to individual patients based on first 3 letter of their residential postcode. The IMD is indicated in the decile (from most deprived to least deprived: 1–2, 3–4, 5–6, 7–8, and 9–10) [22, 23]. A lower IMD rank means an area has a higher proportion of its population experiencing multiple forms of deprivation. The IMD allows for comparison between areas, but it doesn’t quantify the absolute level of deprivation in a specific area. For example, it can tell you that area A is more deprived than area B, but not necessarily how much more deprived. Deprivation index analysis was performed to describe the relationship between the population diversity and cancer stage at diagnosis and detailed within Supplementary Material (Appendix 1).

Socio-economic value analysis

The socio-economic-value analysis aimed to quantify the welfare gain associated with changes in life expectancy and healthy life expectancy (HLE) within our patient cohort [27, 28]. This analysis involved 3 main steps: measurement of life years gained, applying cancer survival rates and estimating economic value.

Measurement of life years gained

Life expectancy (LE) and HLE for the patient population in our study were obtained from the ONS. Specifically, data was sourced from “Life expectancy local areas (Life expectancy for local areas of Great Britain: between 2001 to 2003 and 2021 to 2023) (released 4 Dec 2024) “and “Healthy life expectancy England and Wales (Healthy life expectancy in England and Wales by age and sex: between 2011 to 2013 and 2021 to 2023) (released 12 Dec 2024) [27, 28]. Using each patient’s gender, residence, and age as at the date of diagnosis, and 2021–23 values from those tables.

Applying cancer survival rates by stage of diagnosis

The measured LE and HLE values were then adjusted using cancer survival rates obtained from the National Disease Registration Service (NDRS) [29, 30]. Data were extracted from “Cancer Survival in England: Cancers diagnosed 2016 to 2020, followed up to 2021: Adults.” Specifically, Fig. 2 from the document, which provides 1–5 year survival estimates for 23 common cancers by stage of first diagnosis, was utilized [30]. These 23 cancers were mapped to the 10 cancer groups defined in our study, and survival rates were averaged based on the patient numbers provided in the NDRS table to derive estimates for our 10 categories. The survival rates were then linked to each patient by cancer group, stage at diagnosis, and gender. The 5-year overall survival rates served as a multiplier to adjust the ONS’ LE and HLE. In cases where the life expectancy was less than 5 years, the corresponding survival rate for that shorter period (rounded down) was applied.

Fig. 2.

A. Study population flow chart; B. Trend of yearly stage (stage 1 and 2) diagnostic rate within study population

Estimating economic value

For economic value estimation, two distinct approaches were employed. Initially applied, HM Treasury’s 2022 Green Book value for a Quality-Adjusted Life Year (QALY) of £70,000 (at 2020/21 prices) was applied, discounted at health discount rates of 1.50% (or 1.29 beyond 30 years) [31]. Subsequently this value was substituted with a lower range of £20,000–£30,000, which is commonly used by the National Institute for Health & Care Excellence (NICE) to evaluate the cost-effectiveness of medicines and clinical interventions. This lower value was chosen on grounds of prudence [32]. The Green Book discount rates were then applied to this substituted value. The calculation initially used LE as the duration for economic value, but subsequently applied a scaling factor, implicit between LE and HLE, to derive corresponding economic values for HLE.

Data analysis

Data were analysed using basic descriptive statistics, including measures of central tendency (mean and median) and dispersion (standard deviation, range). Subgroup analyses were not performed in the present study. All statistical analyses were performed using Microsoft Excel (Redmond, WA, USA).

Data linkage

Our study used multi-source data to create the analytical dataset. Data management was facilitated by skilled workers in our hospital finance department who possess expertise in SQL programming. The dataset was compiled using the Microsoft Access (Redmond, WA, USA). SCR provided statistical parameters, such as tumour site, diagnostic date, staging, treatment pathway duration, and dates of follow-up cure or death. The HES dataset provides information on hospital stay, intensive care unit requirements, operation costs, and outpatient follow-ups. The ONS provided a population deprivation index and life expectancy estimates that made socioeconomic analysis possible. NDRS provided cancer survival rates used in combination with ONS life expectancy. Overall, the PLIC provided data on surgery, chemotherapy, radiotherapy, and other oncological treatment requirements, as well as rehabilitation costs.

Results

Study population

Between the 1st of April 2020 to 31st and March 2024 in the SCR 20,535 patients were diagnosed with breast, colorectal, gynaecological, haematological, head and neck, lung, skin, upper gastrointestinal, and urological malignancies within our Trust (Fig. 2A). To achieve homogeneous 2 groups and enable a similar comparison, we applied several exclusion criteria. We initially excluded 5440 patients who were not receiving any active treatment. Subsequently we excluded 4075 patients with incomplete oncological treatment who still required intervention. An additional 2711 patient were excluded because their cancer care was transferred to another hospital for further treatment, interrupting the continuity of care within our institution. We also excluded 2471 patient that lacked complete staging information. Lastly 1242 patient with multiple, recurring, or complex cancers requiring advanced treatment and highly trained specialists were excluded.

Our final sample included 4596 patients, with near-equal distribution between early and late-stage diagnoses with 2,227 diagnosed at stage 1 and 2 and 2,319 diagnosed at stages 3 and 4. The breakdown of cancer site and stage is presented in Table 1. When analysing the trend of early cancer diagnosis per year we can note a positive trend within the Fig. 2B.

Table 1.

Detailed costs split involved in cancer treatment of our cohort divided by cancer type

	BREAST n = 1075		COLORECTAL n = 854		GYNAE n = 162		HAEMATOLOGY n = 261		HAED&NECK n = 188		LUNG n = 484		SKIN n = 319		UPPER GI n = 390		UROLOGY n = 863
	Early Stage	Late Stage	Early Stage	Late Stage	Early Stage	Late Stage	Early Stage	Late Stage	Early Stage	Late Stage	Early Stage	Late Stage	Early Stage	Late Stage	Early Stage	Late Stage	Late Stage	Late Stage
Avg Cost of total care pathway/per patient in £	14,035	26,243	23,515	33,026	14,100	17,866	16,912	36,538	19,014	26,159	9,153	18,074	3,267	8,421	20,321	23,401	3,354	6,732
Total pathway length Per patient in days	196	278	227	283	183	201	272	283	202	245	147	108	98	175	193	129	204	154
Treatment length difference between early and late stage in days (%)	82 (41.84)		56 (24.67)		18 (9.84)		21 (7.72)		43 (21.29)		−39 (−26.53)		77 (78.57)		−64 (−33.16)		−50 (−24.52)
Average treatment length in days	212		271		190		278		226		112		107		137		186
AE cost /per patient in £	245	463	322	514	239	542	459	991	233	486	517	456	104	251	311	491	112	250
IP (DC & EL) cost/per patient in £	5,349	6,468	12,872	12,227	6,819	3,387	2,036	3,214	9,520	4,839	637	1,127	1,282	3,401	5,708	6,141	1,124	1,282
Avg NEL cost/per patient in £ including palliative care	935	2,154	3,933	7,462	2,369	4,447	4,164	10,003	1,776	4,123	4,220	6,213	807	2,653	7,647	8,212	454	1,902
CMDT cost/per patient in £	669	998	598	715	554	500	673	715	691	823	502	469	272	491	584	592	340	481
OP appointments cost /per patient in £ including palliative care	3,466	6,027	1,884	2,797	2,357	3,328	3,355	4,002	2,320	3,359	1,419	2,403	641	1,194	1,305	1,731	704	1,380
Avg Critical care cost /per patient in £	95	244	2,242	2,772	205	147	760	2,050	1,266	1,649	534	432	72	-	1,092	1,848	303	185
Regular day admission cost /per patient in £	527	1,598	104	1,427	48	964	597	1,960	34	568	41	417	31	-	971	1,391	8	92
Radiotherapy/per patient in £	701	1,775	1,157	3,030	911	777	1,324	359	3,174	8,650	636	1,885	37	425	1,977	1,071	260	286
Chemo cost/per patient in £	2,049	6,518	403	2,082	599	3,774	3,544	13,244	-	1,661	648	4,672	20	6	726	1,925	48	875
Proportion patients presented to A&E during cancer treatment pathway in %	20.77	35.68	31.49	45.47	28.28	52.38	32.46	54.42	30.59	47.57	34.55	40.79	9.89	33.33	28.89	50.43	13.02	20.65
Average number of A&E attendances during cancer treatment pathway (n=)	2	2	2	2	2	2	2	3	2	2	2	2	1	1	1	2	2	2
Proportion of patients that required Critical Care admissions during cancer treatment pathway in %	2.20	3.29	37.57	37.74	7.07	4.76	5.26	8.84	12.94	11.65	3.64	3.26	0.35	0.00	17.78	16.23	0.72	1.94
Avg LoS in Critical Care Unit per patient in nights (n=)	3	5	3.8	4.6	2.3	2.7	8.7	12.6	5.3	8.2	8.0	7.9	10.0	0.00	4.3	6.5	20.0	6.0
Proportion of patients that required Elective admissions for procedure in %	70.88	76.53	64.64	52.30	67.68	33.33	11.40	16.33	76.47	32.04	3.64	6.76	1.77	13.89	22.22	19.13	3.25	4.52
Avg LoS for patients that underwent elective procedure (n=)	2	2	6.1	8.1	2.0	6.9	9.2	11.9	2.9	2.6	3.0	3.6	1.6	1.0	4.2	9.0	2.2	5.1
Proportion of patients that required NEL in %	13.57	28.64	28.18	47.85	16.16	49.21	31.58	53.06	23.53	45.63	41.82	57.34	5.65	25.00	55.56	55.36	7.05	17.42
Avg LoS for patients requiring NEL Non Elective (n=)	8	8	10.9	12.7	13.9	8.6	13.2	18.0	8.6	9.3	10.9	12.7	16.9	14.2	13.7	15.3	8.2	13.1
Avg number of CMDT discussions/per patient (n=)	4	6	3.3	4.0	3.1	2.9	3.7	4.0	3.8	4.6	2.8	2.6	1.5	2.6	3.2	3.3	1.9	2.7
Avg number of OP clinics for follow up per patient (n=)	11	20	9	13	6	10	16	20	9	13	6	8	3	8	7	10	4	6
Avg number of regular day admissions (apart of chemotherapy treatment)/per patient (n=)	1	3	0.2	3.8	0.2	2.4	1.3	3.3	0.1	1.4	0.1	0.7	0.1	-	2.7	3.4	0.0	0.2
Avg number of regular day admissions for patient requiring chemotherapy (n=)	1	4	0.3	2.1	0.3	3.3	1.9	5.7	-	1.0	0.3	1.2	0.0	0.0	1.0	1.4	0.1	0.9
Proportion of patient requiring palliative team input in %	2.90%	7.51%	3.87%	17.83%	5.05%	30.16%	4.39%	14.29%	3.53%	21.36%	23.64%	42.66%	3.18%		24.44%	46.09%	0.36%	8.39%
Avg palliative Care days (n=)	19	23	12.3	18.3	15.4	20.6	9.2	11.0	9.7	25.1	27.5	17.2	5.8		15.9	19.2	12.0	15.6

Legend: Avg- average; A&E- Accident and Emergency; BHRUT- Barking, Havering, and Redbridge University Hospitals NHS Trust IP - Inpatient admission; DC - Day case admission; EL - Elective admission; NEL - Non elective admission; CMDT - Cancer Multi-disciplinary team meeting; OP - Outpatient attendance; AE - Accident and Emergency

Impact of deprivation profile on early and late cancer diagnosis [22, 23]

Our analysis revealed a strong association between socioeconomic deprivation, healthcare needs, and treatment costs. According to Geographic Data Service, 53% of the whole population our Trust serves comes from higher-deprived areas (Deprivation Indices 1–5), with nearly half residing in Barking and Dagenham, an area with 99% deprivation (Fig. 3A).

Fig. 3.

A. Deprivation profile B. Comparison of average pathway cost per patient based on deprivation group (low deprivation group- IMD deciles 6 to 10 and high deprivation group- IMD deciles 1 to 5)

When we stratify our patient cohort by IMD and cancer stage we found that patients diagnosed with cancer at a later stage were more likely to come from higher-deprived areas (56%) compared to those with lower-deprivation backgrounds (44%), indicating a 12% difference (Fig. 3A). Additionally, Fig. 3B shows that the average treatment costs for patients from deprived areas were 15% higher than those from more affluent areas.

Further analysis of deprivation profiles across different cancer sites showed that a higher proportion of patients from deprived areas are diagnosed at later stage, which is often associated with poorer outcomes. Specifically, the proportion of people whose cancer is diagnosed at an early stage is around 9% lower in the most deprived areas compared to the most affluent (Appendix 1). These findings suggest that deprivation significantly affects both the timing of cancer diagnosis and subsequent cost of treatment.

Differences in cost of care between early and late-stage cancer diagnosis

While performing a cancer site analysis we noted that breast and skin cancers demonstrated a higher proportion of early-stage diagnoses, whereas colorectal, lung, and upper gastrointestinal cancers were more frequently detected in the advanced stages (Fig. 2A). This difference in stage at diagnosis had a direct impact on treatment costs (Table 1). Figure 4 illustrate the variation in patient care costs across distinct tumour groups, comparing early- and late-stage diagnoses. A significant cost disparity was observed across all tumour groups, with early diagnosis resulting in lower average costs per patient (£11.2 K) compared to later stages (£23.8K) (Fig. 4B). The most substantial cost differences were evident within the haematological, breast, and colorectal tumour groups, exhibiting increases of £19.6K, £12.2K, and £9.5K, respectively, when diagnosed at later stages.

Fig. 4.

A. Average (Avg) and difference cost per patient pathway between early vs late diagnosis by cancer type (£) B. Box plot illustrating the differences of average(Avg) cost of patients diagnosed at later stages of the disease (Stage 3&4) vs earlier stage (1&2)

The highest average costs of pathway being noted in colorectal (31 K) and haematological cancer types (36 K), which also showed the most extended treatment cycles (Fig. 4A and Table 1). Skin cancers showed the lowest overall average cost (£3.8K) and shortest pathway (107 days on average). The most significant drivers of cost across all cancer types are elective and non-elective admission costs and length of stay in critical care (Table 1). We identified that hospital admissions, both planned (elective) and unplanned (non-elective), coupled with extended lengths of stay in both general and intensive care units, constitute the primary factors contributing to the overall cost of cancer care. These findings align with the cost figures presented in Table 1, demonstrating a clear correlation between resource utilisation and the cost per cancer site.

Impact of early detection on treatment length and cost of care

The average treatment duration was 190 days for early-stage cancers and 204 days for late-stage cancers, with a wider interquartile range observed in the latter.

Within Table 1 we can note that the most substantial decreases in treatment duration were evident for breast cancer (82 days), skin cancer (77 days), and colorectal cancer (56 days). Conversely, treatment duration exhibited increases in lung cancer (−39 days), upper gastrointestinal cancers (−63 days), and urological cancers (−50 days). The shorter pathway for later diagnosis in the latter groups may be attributed to a shift in treatment provision towards other National Health Service (NHS) providers as well as more aggressive types of diseases facilitating higher mortality rates, particularly in the lung, upper gastrointestinal (UGI), and colorectal cancer. Urology cases being referred to other units.

The analysis revealed a strong correlation between the treatment pathway length and the number of outpatient follow-up appointments. Patients with late-stage haematological, colorectal, and breast cancers demonstrated significantly longer treatment cycles and required more frequent outpatient visits than their early-stage counterparts (Table 1). Specifically, early-stage breast cancer patients averaged 11 outpatient appointments, whereas late-stage patients required 20. Similarly, patients with early-stage haematological cancer had 16 outpatient appointments compared to 20 patients with late-stage cancer. For colorectal cancer, early-stage patients underwent 9 outpatient appointments, whereas late-stage patients required 13. This increased treatment burden was reflected in the higher associated costs for outpatient visits: £6,027 for late-stage breast cancer vs. £3,466 for early-stage cancer, £4,002 for late-stage haematological cancer vs. £3,355 for early-stage cancer, and £2,797 for late-stage colorectal cancer vs. £1,884 for early-stage cancer (Table 1).

On average, the total cost per patient pathway of late cancers is more than double than of cancers detected early, and our average early detection rate over the past four years has been around 50% (Fig. 2B, Fig. 4B).

Overall, 83% of cancer patients were diagnosed and treated within a year, with 48% completing treatment within three months. However, 18% of patients had treatment cycles exceeding one year. Notably, 17–18% of patients across all cancer stages experienced prolonged treatment.

For example, skin cancer has the shortest treatment duration. 97% Of skin cancer patients completed treatment within a year and 60% finished within three months. Only 3% of skin cancer patients had treatment exceeding a year. This pattern is likely due to a high proportion of early diagnoses, with 283 out of 319 patients (88%) diagnosed early.

In contrast, patients with CRC often undergo longer treatment cycles. 73% of colorectal cancer patients were treated within a year, with 34% completing treatment within three months. However, 27% of colorectal cancer patients were treatment exceeding a year. This is primarily linked to a significant proportion of late diagnoses, with 672 out of 853 patients (78%) diagnosed late.

The detailed analysis of skin and colorectal cancer treatment timings can be found within Supplementary Material (Appendix 1).

Cancer 75% target (to achieve 75% detection at stage 1 or 2 by 2028) [17]

Over four years, a total of £80.7 million was expended on treating 4,596 patients. The early-detection cohort (2,277 patients) incurred treatment costs of £25.5 million, while the late-stage cancer cohort (2,329 patients) required significantly higher expenditure at £52.2 million. Although this varies considerably across cancer types overall approximately 50% of cancers from our study population are detected early (Table 2).

Table 2.

Potential financial impact of the hypothetical shift towards a 75% early detection ate within our study population

	Early Stage	Late Stage	Total
Current 4-year average detection rate	≈50%	≈50%
Number or patients	2,277	2,319	4,596
Avg Cost of total care pathway/per patient in £	11,200	23,814
Current total cost	£25,502,400	£55,224,666	£80,727,066
If 75% of patients from our study population are diagnosed as stage 1 and 2	75%	25%
Number or patients	3,447	1,149	4,596
Avg cost of total care pathway/per patient in £	11,200	23,814
Total cost	£38,606,400	£27,362,286	£65,968,686
Saving of cost (4 years)			£14,758,380
Per year			£3,689,595
Or potential release of capacity to treat further patients			1,028
Increase of treatment capacity by			22%

Legend: Avg- average; ≈- Approximately

In the scenario of hypothetical shift towards an NHS Long Term Plan 75% early detection rate, with the same sample size, could potentially yield treatment cost savings of £14.7 million over four years (£3.68 million annually) (Table 2). Alternatively, this increased early detection rate could potentially enhance treatment capacity by 22% or potential release of capacity to treat further 1028 patients.

Socio-economic value of early cancer detection [29–32]

Using the described method, we were able to make an approximate adjustment for the effect of cancer survival upon life expectancy. Our approach assumes that patients who remain cancer-free for five years are at no greater risk than the general population beyond that point. The cancer survival statistics measure from date of diagnosis, while life expectancy is based on age at start of treatment. It is assumed that the interval between diagnosis and treatment start is minimal and that no adjustment is needed for this period. We observed that including survival rates by stage of diagnosis had a significant effect on the economic value of treatment, largely due to the significantly poorer prospects of patients diagnosed at late stages (Table 3).

Table 3.

Potential financial impact of the hypothetical shift towards a 75% early detection ate within our study population

Cancer stage	Avg treatment length in days	Avg Cost of total care pathway/per patient in £	Life Expectancy	Avg HLE in years	Economic Value in £ using £20k per NICE	Present Value in £ discounted at 1.5%	Economic Value in £ using £70k per Green Book	Present Value in £ discounted at 1.5%
Adjusted for Cancer Survival Rates
Early Stage	190	11,200.74	18.88	10.3	205,984	170,866	720,944	598,029
Late Stage	204	23,814.04	8.04	4.31	86,207	75,540	301,726	264,390
Overall Average	197	17,565.03	13.41	7.28	145,548	122,767	509,419	429,685
	−14	−12613	10.83	5.99	119,777	95,326	419,218	333,639
The difference between Early Stage and Late Stage
Unadjusted for Cancer Survival Rates
Overall Average	197	17,565.03	19.91	10.74	214,706	189,544	751,472	663,404
The difference that cancer makes			−6.50	−3.46	− 69158	− 66777	− 242,053	− 233,719

Legend: Avg- average; HLE - Health Life expectancy

Without adjusting for Cancer Survival Rates obtained from the NDRS, our calculations, using National Institute for Health & Care Excellence (NICE) £20k valuation, showed that successful cancer treatment, on average, adds treatment adds 10.74 years of healthy life expectancy per patient. This translates to a value of £189.5K when discounted at 1.5% and results in a favourable cost-benefit ratio of 10.8:1. Notably, earlier cancer detection would yield even more significant benefits, adding an average of 2.43 healthy life years, and £40,275K value per patient. For comparison, using the HM Treasury Green Book value of £70,000 per Quality-Adjusted Life Year (QALY), the average economic value of life added by cancer treatment is £663K when discounted at 1.5%.

With the adjustments for Cancer Survival Rates, the average economic value of life added by cancer treatment is £122.8K (or £145.5K discounted to present value at 1.5%), when using NICE £20K valuation. This represents a cost-benefit ratio of 8.3:1 (or at present value 7:1). The inclusion of these survival rates reduced the life expectancy of early-stage diagnosed patients by 3 years (from 21.9 to 18.88 years), but for late-stage patients, life expectancy was more than halved, dropping from 17.94 years to 8.04 years when survival rates were taken into account. This highlights the significantly increased potential gain in additional life years and healthy life years that would be achieved if cancer were detected early. Based on our findings, investment in screening and other measures to detect cancers early would be cost-effective if it could be done for less than £12,613 per cancer detected early, which represents the difference in the cost of treatment between early and late-stage cases.

The economic impact of the diagnosis of cancer in situ

To illustrate the financial impact of early detection, we analysed the treatment costs for 174 patients with stage 0 breast cancer identified in the cancer registry. Treating breast cancer at the earliest stage (stage 0) incurs an average price of £10.3K. In contrast, treatment of stage 3 breast cancer averaged £23.8K, and Stage 4 breast cancer treatment costs rose to £35.1K (Appendix 1). These findings underscore the substantial financial benefits of early detection and treatment of breast cancer.

Discussion

This study extends the current research by analysing the costs of care incurred by cancer patients diagnosed at a large DGH serving an ethnically diverse and deprived population in East London. Our approach, which includes data from multiple NHS databases (SCR, HES, PLIC, ONS and NDRS) to create a novel resource for analysing cancer treatment costs, mirrors the well-established SEER-Medicare Linked Data Resource in the USA [33]. This methodology is unique compared to most existing reported data from NHS hospitals, which are typically episodic, because it allows for the reconstruction of complete patient care pathways for each cancer type.

Our analysis, conducted at a DGH, strongly advocates for radical interventions by the NHS and public health services to enhance the early diagnosis of cancer within the UK. While a small increase in early cancer diagnosis rates has been observed since the COVID-19 pandemic, the current rate of 55% represents a significant gap from the planned NHS Long term objectives of 75% by 2028 [17, 34]. Despite this, we are encouraged to report that our early stage diagnosis rate has increased over the last four years, from 37% in 2020 to 53% in 2024.

The study demonstrates a significant economic burden associated with late-stage cancer diagnosis. At £23,814, the average cost of a late-stage diagnosis is nearly double that of an early cancer diagnosis (£11,274). This phenomenon can be attributed to prolonged treatment length, increased hospital resource utilisation including hospital admissions and intensive care admission, and a higher number of required outpatient appointments for cancer care. Our data also highlights a clear correlation between patients’ deprivation index, late-stage (3 and 4) diagnosis, and increased cancer care pathway costs. This finding is reinforced by research from Cancer Research UK, which notes that cancer death rates are almost 60% higher in UKs most deprived areas, indicating that individuals from socioeconomically disadvantaged areas are more likely to be diagnosed at an advanced stage [35]. Our study found that late-stage diagnosis was 12% more common in highly deprived areas. Achieving the recommended detection rate of 75% on the sample size would generate a saving of £14.7 million over four years or £3.6 million per year. This can be transformed into release capacity to treat a further 1028 patients.

Informed by these findings, our Trust is prioritising a strategic and collaborative approach to cancer care. While transitioning to a unified electronic healthcare records system which is already implemented across other Nort East London NHS Trust, BHRUT aims to leverage data to gain new insights into cancer prevention, diagnosis and treatment, thereby improving care and resource allocation. The team is now sharing this methodology with regional and national cancer alliances to guide policy and ensure public funding achieves the greatest impact especially considering reports of underfunding in cancer care in Northeast London compared to other regions. The findings of this work encouraged local stakeholders to a local GI endoscopy strategy that incorporates innovative technologies like robotic colonoscopy and trans nasal endoscopy. As a part of this new strategy, we become a pilot hospital for lowering the faecal immunochemical test (FIT) threshold in bowel cancer screening from 120ug/gm to 80ug/gm, which successfully diagnosed asymptomatic cancers which would have otherwise been missed. Additionally, the GI strategy has promoted the establishment of a community diagnostic centre which offers a broad spectrum of services including cytosponge for upper GI suspected cancer cases and colon capsule for lower GI. Our Trust successfully embarked on NHS Targeted Lung Health Check Programme, providing free lung checks and low dose CT scans to at risk individuals to facilitate early diagnosis. These local management efforts informed by the study findings, show a clear commitment to improve patients’ outcomes and healthcare efficiency through collaborative and innovative approach.

We consider that a part of the success behind achieving a 75% early cancer detection rate goal nationally is the importance of leveraging scientific innovation, including artificial intelligence tools and multi-cancer detection blood tests. For example, the BARCODE-1 study demonstrated the efficacy of saliva tests compared with blood tests in prostate cancer diagnosis [36]. Although these early trial results are encouraging, their successful translation into clinically implementable diagnostic tools remains unclear. The NHS England update on the Galleri multi-cancer detection blood test trial highlighted the need for more compelling evidence before proceeding with a large-scale pilot program [37]. Furthermore, timely access to appropriate treatment is crucial for successful early cancer diagnosis. In 2023, England experienced the worst year on record for cancer waiting times, with only 64.1% of patients commencing treatment within 62 days of suspected cancer and nearly 100,000 individuals experiencing longer-than-acceptable waiting times [38]. We recognise that achieving national goals for early cancer diagnosis can only be accomplished through enhancements in populations where cancer rates are the highest and screening or early diagnosis is the lowest, particularly in disadvantaged areas. The absence of a comprehensive National Cancer Control Plan (NCCP) in England poses a significant challenge. A robust NCCP is essential to provide a framework for practical cancer control efforts, including early diagnosis, timely treatment, and equitable access to care.

When analysing the literature on expenses related to cancer care we note that the quality of cost data can vary across hospitals and over time. While other hospital cost data exists, they are typically aggregated at an episodic level, preventing a comprehensive assessment of costs across the entire cancer care pathway. Methodologically, reconstructing complete patient care pathways is a novel and resource-intensive undertaking that requires specialised expertise and capacity. This level of detailed cost analysis is currently not mandated or regulated by the NHSE, limiting the availability of comparable benchmarks. To reduce variability, the analysis relied exclusively on cost data from the study, which may restrict the generalisability of the results. It is also crucial to note that the current national cost collection method primarily emphasises episodic data, which obstructs a comprehensive understanding of resource use. Furthermore, the study did not include patients with complex or recurrent cancers, which potentially narrows the applicability of the findings to this particular patient group. Although methods such as a solid board assurance framework, Costing Assurance Assessments, and thorough reconciliation of patient activities and costs have been utilised to guarantee data quality and trustworthiness, these limitations should be considered when interpreting the findings.

The linked dataset approach has potential to be reproduced within other UK institutions and to facilitate the creation of new research questions in health economics. We sought to provide evidence regarding the allocation of cancer care resources in a DGH setting. Being one of the first within the NHS to analyse the data in this way, we would like to contribute to the efficient and most important effective utilisation of current and future healthcare resources to increase the quality of care that we provide as well as improve patients’ quality of life.

This study has several limitations, primarily stemming from the retrospective design and nature of the data sources used. One of the main limitations is that the final sample size is relatively small, as the data was sourced from two hospitals in East London and as a result from our strict exclusion criteria the findings may not be widespread. We understand that the patient population served by BHRUT are characterised by a high level of deprivation, which as our study demonstrates is associated with a later cancer diagnosis and higher treatment costs for the NHS. This context means our findings may not be representative of other regions with different socioeconomic profiles, infrastructure or demographics.

Furthermore, our reliance on secondary data sources (SCR, HES, and PLIC) presents another limitation. While the data quality within these sources is anticipated to improve over time, current issues include incomplete cancer data, specifically, a lack of information on “cancer in situ,” which hinders the optimisation of early detection strategies. Furthermore, a substantial portion of patients in the SCR, approximately 20% of the sample, lacked cancer staging information. This missing data precluded an analysis of the impact of cancer stage on treatment costs, and imputation was not feasible. The application of our exclusion criteria, while necessary to achieve a clean sample size for a like-to-like comparison led to a final sample of only 4596 patients, representing around 22% of total sample of 20,535. This reduction in the sample size may introduce an exclusion bias. Additionally, due to data unavailability, the study excluded costs associated with crucial supportive care services such as primary care, social care, and pre/rehabilitation. Financial data limitations, including incomplete data for patients treated at tertiary centres and potential biases within the data sources, may have influenced the findings.

Within the socio-economic value of early cancer detection analysis, we would have preferred to use net survival rates, but gaps in that data prevented this, representing another limitation of the chosen method. The use of overall survival rates means some double-counting of deaths by other causes (present in both the overall rate & ONS LE & HLE). This means LE & HLE are likely underestimated, erring on the side of prudence. Survival by stage of diagnosis information was not available for all cancer sites, though at least one was available for each of our 10 cancer groups; this limits the applicability of the survival rates, particularly for Upper GI and Head & Neck, where less than half the sites were available.

Future research should address these limitations by employing more robust data collection methods, including multivariate analysis, considering the broader context of cancer care within Northeast London or London, addressing the impact of socioeconomic disparities, and investigating their impact on early cancer diagnoses.

Conclusions

Our study presents a significant cost discrepancy between early and late-stage cancer diagnosis and treatment. The NHS faces higher expenses for patients diagnosed with advanced disease, which is attributed to the extended treatment length resulting from resource use and more invasive care needs. Presenting a DGH data we show “real world” evidence of potential cost savings associated with early diagnosis in a deprived area in east London and the eventual possibility of increasing the life expectancy for our patients. A hypothetical 75% early detection rate has the potential to save £14.7 million over four years within our Trust. If this strategy is extrapolated across the entire NHS, financial savings could be substantial, highlighting the significant impact of early detection on healthcare efficiency and resource allocation.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Conceptualisation: VB, PS, WC and SB. Data curation: PS, TG, DP and WC. Formal analysis: VB, PS, TG, DP and WC. Funding acquisition: N/A. Investigation: All authors. Methodology: VB, PS, TG, DP, AM, WC, NR, MB, MC and SB. Project administration: VB, PS, AM, NR, MB, MC and SB. Visualization: VB, PS, TG, DP, AM, WC, NR, MB, MC and SB. Writing–original draft: VB, PS, TG, DP, AM, WC, NR, MB, MC and SB. Writing–review and editing: VB, PS, TG, DP, AM, WC, NR, MB, MC and SB. All authors read and approved the final manuscript.

Funding

None.

Data availability

The datasets used and analysed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Due to the retrospective study design and anonymized documentation of underlying cost data, no ethical vote was needed. Institutional Approval was granted before the collection of anonymized data. This study was approved by the Divisional Governance Committee and adhered to all the local information and research governance guidelines. The analysis used anonymized and de-identified patient-level data to ensure compliance with NHS data management regulations and the Health Insurance Portability and Accountability Act (HIPAA). This study was not considered research by the NHS Health Research Authority decision tool hence no ethical approval needed [22].

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.