5.1 Introduction

In the Climate Change Bill the UK government is committed to ‘cut carbon dioxide emissions by at least 60% by 2050, based on 1990 levels’. To ensure the NHS helps government achieve these targets the NHS’ Sustainable Development Unit (SDU) has produced the NHS Carbon Reduction Strategy which aims to reduce carbon emissions in 3 key areas of NHS activity; energy consumption, transport and procurement. The NHS has no national data on which to base its 1990 carbon footprint and approached the Stockholm Environment Institute to calculate its carbon footprint in 2004.
  
NHS expenditure on energy, transport and consumption in 2004 forms the basis for calculating CO₂ emissions. It is estimated that the NHS produced the equivalent of 18 million tonnes CO₂ in 2004, of which 3.41 million tonnes is attributed to travel, equivalent to 10.5 billion kilometres.

Figure 5.1: Stockholm Institute estimates of NHS CO₂ emissions in 2004

Stockholm Institute estimates of NHS CO2 emissions in 2004

Transport related CO₂ emissions for the NHS were calculated on the basis of the NHS Estates Returns Information Collection (ERIC) system which records expenditure on transport services provided by NHS organisations and their staff business travel claims. Additionally the Stockholm Environment Institute predicted that patient and visitor travel would contribute a further 10% of all travel CO₂ emissions but had no means to measure this element.

Cancer Registries have excellent records for patients diagnosed with cancer since 1981, which include patient addresses and hospitals of treatment. The work reported here develops a method for assessing the carbon footprint of patient travel for one small area of NHS activity, radiotherapy treatment for breast cancer, in the West Midlands region in both the government’s baseline year 1990 and in 2004. The work described will use GIS and patient registers to address two key information gaps:

• Patient travel to NHS services.
• Comparable data for patient travel in 1990 and 2004.

5.2 Data Sources

An extract for all breast cancers recorded as treated with radiotherapy or having a ‘planned radiotherapy’ treatment in each of our target years, 1990 & 2004 was extracted from the cancer registry database. In instances where the radiotherapy centre was not recorded in the database, if the hospital of diagnosis was a radiotherapy centre at time of treatment the hospital of diagnosis is used as a proxy for radiotherapy treatment centre. There were, more cases in 1990 with no radiotherapy treatment centre recorded than in 2004 but by adopting hospital of diagnosis for these cases we were able to include over 97% of all cases identified. The number of patients receiving radiotherapy for breast cancer in 2004 almost doubled from 1990 levels.

Table 5.1: Estimated emergency department attendance counts for assault in the resident populations of PCTs in the West Midlands, 2005/2006

The West Midlands Cancer Registration database does not currently record evidence of the radiotherapy regime delivered to each patient so we have relied on the results of two Royal College of Radiologists audits1, 2 of patient treatment regimen in 1993 and in 2003 and applied the relevant percentages to our patient cohorts. Table 5.2 summarises the percentage allocation of cases to each regime in the two years. The number of fractions is taken to be equivalent to the number of days attendance at radiotherapy centre for each patient.

Table 5.2: Allocation of treatment regime to patients based on published research

The diagnosis postcode of each patient is allocated to its Lower Super Output Area and provides a start location for each journey and the radiotherapy centre of treatment ultimately determines how far they travel. One return journey to treatment is recorded for every fraction of radiotherapy administered regardless of whether this is given as an inpatient or day patient, which allows for some account of visitor travel. No account is made of consultations and other appointments attended prior to, nor following a course of radiotherapy. No account is made of additional ‘booster’ radiotherapy treatments added to a standard regime for some patients. The percentage of patients allocated to each regime assumes a standard distribution of disease stage geographically across the region.

The Ordnance Survey flagship road product, the Integrated Transport Network (ITN) is used with the ArcGIS Network Analyst extension to determine the routes to radiotherapy centres based on road hierarchy. An origin-destination matrix from the population weighted centroid of every LSOA in the region to every NHS Acute Trust (n=20) is produced which can be reused in future analyses with minimum effort. Travel routes from each LSOA to radiotherapy centre are identified with priority given to main road routes over minor roads. The distance along these routes is calculated for every patient to give a total annual single return distance travelled. Total distance is allocated in proportion to each radiotherapy regime and factored up to provide total distance travelled by all patients following all regimes.

Total distances travelled are converted to CO₂ emissions based on DEFRA’s greenhouse gas conversion tables (2008). The conversion factor takes account of a mixed age range and fuel type of cars typical in 2008 (0.2042 kg/km was used). Similar conversion factors for cars licensed in 1990 and 2004 do not appear to be available. No account of travel by public transport is included in the model as we felt that although the region has some good train and bus routes most women would aim to travel by private vehicle or taxi for at least one leg of their daily journey. Even if they did not drive themselves their driver would generate a minimum of one return car journey. Some double counting may have been made for those patients that used NHS patient transport services.

5.3 Results

Routes are assigned from the source LSOA to each radiotherapy centre. Map 5.1 provides an illustration of these routes for patients treated at University Hospital Coventry and Warwickshire. The longest distances travelled in 1990 were from North Staffordshire and whilst these are not repeated in 2004 more patients now travel from Mid Worcestershire and from Walsall and Dudley reflecting cancer network commissioning patterns.

Map 5.1: Routes to University Hospital Coventry and Warwickshire in 1990

Routes to University Hospital Coventry and Warwickshire in 1990

From these origin destination matrices the cumulative distances travelled by all patients to access one dose of radiotherapy treatment were calculated in each year. Although patient numbers had doubled and some referral patterns changed during the period total return journeys still doubled from 1990 to 2004.

Map 5.2 Single Return Journey Distance generated per LSOA by all patients receiving radiotherapy for breast cancer in 1990

Single Return Journey Distance generated per LSOA by all patients receiving radiotherapy for breast cancer in 1990

Map 5.3: Single Return Journey Distance generated per LSOA by all patients receiving radiotherapy for breast cancer in 2004

Single Return Journey Distance generated per LSOA by all patients receiving radiotherapy for breast cancer in 2004

In 1990 a patient would travel an average of 34.5km (21.4 miles) to access each fraction of radiotherapy treatment for breast cancer but by 2004 this had risen to 38.5km (23.9 miles) a 12% increase in distance.

Table 5.3: Cumulative Distances travelled by patients once treatment regime taken into account

Year No. Fractions % Patients No. Patients allocated Total Distance Travelled CO2 emissions generated Target CO2 emissions for 2050(tonnes) Km Miles Kg Tonnes 60% reduction 26% reduction 1990 15 24 403 208,275 130,172 42,530 42.53       20 17 285 196,388 122,743 40,102 40.10       25 35 587 505,614 316,008 103,246 103.25       Other 24 402 13,851 8,657 2,828 2.83       TOTAL   1677 924,127 577,579 188,707 188.71 75.49 139.64 2004 15 45 1503 867,232 542,020 177,089 177.09       20 16 535 411,594 257,246 84,047 84.05       25 21 702 675,091 421,932 137,854 137.85       Other 18 601 23,119 14,449 4,721 4.72       TOTAL   3341 1,977,036 1,235,647 403,711 403.71 161.48 298.75

In 1990 1,677 patients received radiotherapy for breast cancer and generated over half million miles of road travel and almost 190 tonnes of CO₂ emissions. By 2004 there had been a doubling in the number of patients treated with radiotherapy which contributed over 1.2 million miles travel and over 400 tonnes of CO₂ emissions.

5.4 Conclusions

The numbers of patients treated with radiotherapy for breast cancer almost doubled from 1990 to 2004 and despite a reduction in the number of radiotherapy fractions administered and the introduction of new Cancer Network commissioning patterns there was no curtailment in the volume of travel generated by treatment for breast cancer. If we consider our findings in terms of the measures produced by the Stockholm Environment Institute for the NHS, patient travel to radiotherapy for breast cancer in West Midlands would account for 0.012% CO₂ emissions currently identified in NHS estimates.  The West Midlands population has a fairly typical age structure and accounts for just short of one tenth of the English population so assuming it to be of typical geography one might expect radiotherapy for breast cancer in England to account for 0.12% of all NHS CO₂ emissions for travel.

We know that radiotherapy treatment is only one aspect of treatment for breast cancer if we add the carbon footprint of diagnostic tests and other patient consultations it seems likely that the whole carbon footprint of breast cancer may be nearer 1% of the SDU allocation for the carbon footprint of travel in the NHS. Add to these figures consultations and treatment for all other cancers, visitor journeys to inpatients and the SDU estimates for all NHS travel and transport activity may well become an under-estimate of patient and visitor travel associated with the NHS.

The linkage of standard geographical data to Cancer Registry data provides a sustainable method for assessing the burden NHS cancer services place on the NHS Carbon Footprint. By producing an origin-destination matrix at LSOA level new patient postcodes can easily be included in the analyses and the method applied to any NHS hospital led activity. It is sustainable because the road network data and hospital locations are unlikely to change in the short term.

Despite the assumptions implicit in our model it does provide a novel yet efficient way for the NHS to gain a real understanding of the carbon footprint of patient and visitor travel. It is easily reproducible across England as it uses standard geographical datasets. Cancer registry databases provide good historic records of patient treatment patterns, records for other diseases may not be so readily available.

5.5 Implications of NHS Carbon Reduction Strategy on the Delivery of Cancer Services

In order to contribute at a ‘pro rata’ rate to government targets radiotherapy for breast cancer would need to reduce its carbon footprint to 161 tonnes p.a.  If based on 2004 data and to 76.5 tonnes p.a. when based on the Government’s 1990 baseline (equivalent to an 81% reduction in CO₂ emissions). When Department of Transport anticipates a 35% increase in transport CO₂ emissions for UK by 2035 one has to ask whether the NHS can contribute to reaching these targets without radically rethinking its service configuration.

If radiotherapy treatment for breast cancer is in anyway representative of the delivery of health services in the region it would appear a radical change in either clinical practice or the pricing policy of health services will be required in order to achieve a 60% reduction in CO₂ emissions. If a ‘carbon footprint cost for travel and procurement’ is not included in commissioning toolkits we may be forced to make patients use public transport when attending health service appointments and out patient treatment. Alternatively current high cost drugs may suddenly become more affordable or will centralised services become a thing of the past to ensure we make our contribution to the NHS carbon reduction strategy? The Government has to be clear strategies currently seen as cost savings may be in conflict with its environmental strategies and manage the service planning for long-term future as well as short term cost savings.

We cannot deny patients treatment for cancer but can we really support the government’s targets to reduce CO₂ emissions by 60% when so little is known of our true carbon footprint. Attaining a carbon reduction target must depend on conceptualising the scale of the task. By calculating 1990 and 2004 CO₂ emissions based on patient activity records and carbon conversion factors typical of car ownership patterns in both 1990 and 2004 we may be able to better understand the travel element of the NHS carbon footprint.

References

1. ‘Non-surgical Management of Early Breast Cancer in the United Kingdom: Radiotherapy Fractionation Practices’ JR Yarnold, P Price and GG Steel, on behalf of the Clinical Audit Sub-Committee of the Faculty of Clinical Oncology, Royal College of Radiologists, and the Joint Council for Clinical Oncology. Clinical Oncology (1995) 7:223-226.


2. ‘National Survey of Radiotherapy Fractionation Practice in 2003’  MV Williams, ND James, ET Summers, A Barrett, DV Ash On behalf of the Audit Sub-Committee, Faculty of Clinical Oncology, Royal College of Radiologists, London, UK. Clinical Oncology (2006) 18: 3–14.


3. ‘The UK Standardisation of breast radiotherapy (START) Trial B of radiotherapy hypofractionation for treatment of early breast cancer: a randomised trial START Trialists' Group’. SM Bentzen, RK Agrawal, EG Aird, JM Barrett, PJ Barrett-Lee, JM Bliss, J Brown, JA Dewar, HJ Dobbs, JS Haviland, PJ Hoskin, P Hopwood, PA Lawton, BJ Magee, J Mills, DA Morgan, JR Owen, S Simmons, G Sumo, MA Sydenham, K Venables, JR Yarnold. Lancet (2008) 371:1098-1107.


4. ‘Saving Carbon, Improving Health: a draft carbon reduction strategy for the NHS in England’ NHS Sustainable Development Unit. May 2008.


5. ‘NHS England Carbon Emissions: Carbon Footprinting Report’ Sustainable Development Commission 2008.

Acknowledgements

• Catherine Lagord and Christopher Lawrence for their assistance in extracting and interpreting data from West Midlands Cancer Registration Database.

• Samuel Jones for his perseverance in setting up the ITN road network files and running the origin destination matrix.