Treatment variation in early breast cancer in the UK
BMJ 2020; 371 doi: https://doi.org/10.1136/bmj.m4237 (Published 01 December 2020) Cite this as: BMJ 2020;371:m4237- David Dodwell, senior clinical research fellow1,
- Yasmin Jauhari, clinical research fellow2,
- Toral Gathani, senior clinical research fellow3 4,
- David Cromwell, professor5 ,
- Melissa Gannon, research fellow5,
- Karen Clements, project manager6 ,
- Kieran Horgan, consultant breast surgeon7
- 1Nuffield Department of Population Health, University of Oxford, Oxford, UK
- 2Clinical Effectiveness Unit, Royal College of Surgeons of England, London, UK
- 3Cancer Epidemiology Unit, Nuffield Department of Population Health, University of Oxford
- 4Department of Surgery, Oxford University Hospitals NHS Foundation Trust, Oxford, UK
- 5Department of Health Services Research and Policy, London School of Hygiene and Tropical Medicine, UK
- 6Public Health England, Birmingham UK
- 7Department of Breast Surgery, St James’s University Hospital, Leeds, UK
- Correspondence to: D Dodwell david.dodwell{at}ndph.ox.ac.uk
One in seven women will develop breast cancer. Although the number of people diagnosed with this common disease is increasing, mortality continues to fall. In the UK, almost 80% of women diagnosed with breast cancer are alive 10 years after their initial diagnosis.1 The survival gains seen over recent decades have largely been achieved through earlier diagnosis and more effective treatments. Nonetheless, there is still room for improvement. Population based mortality for breast cancer among those aged <75 years in the UK ranges from 13 to 32 per 100 000,2 and UK survival rates from breast cancer are still reported to be below those in countries such as the United States and Australia.3
The treatment of breast cancer is multimodal, comprising local treatment (surgery and radiotherapy) and systemic treatments (endocrine therapy, chemotherapy, and biological therapy). Breast conserving surgery is, in most cases preferred to mastectomy (removal of all the breast tissue) providing equivalent oncological outcomes, but with quality of life benefits. The choice of systemic therapies is driven by risk of recurrence, determined by stage, grade, and molecular pathological characteristics. For example, patients with ER (oestrogen receptor) positive cancer are treated with endocrine therapy and those with HER2 (human epidermal growth factor receptor 2) positive disease commonly receive trastuzumab and chemotherapy. Although randomised controlled trials have provided good evidence for optimal treatment, variation still occurs in the UK and may be one factor in reported lower survival rates.
Treatment variation
Surgery is the most important treatment for early breast cancer. The 2018 National Institute for Health and Care Excellence (NICE) guidelines maintain that patients should be treated “irrespective of age, with surgery and appropriate systemic therapy, rather than endocrine therapy alone, unless significant comorbidity precludes surgery.”4
However, the National Audit of Breast Cancer in Older Patients has shown important differences among hospitals in the surgical care of older patients in England and Wales. For patients diagnosed between 2014 and 2016, rates of surgery varied from 85% to 100% (mean 96%) in those aged 50-69 and from 46% to 97% (mean 76%) in those aged ≥70.5 Among the same cohort, rates of breast conserving surgery were 46-85% in patients aged 50-69 and 22-77% in those aged ≥70 (table 1).
These contemporary examples mirror the variation in patterns of surgical practice seen around 20 years ago. The Breast Cancer Clinical Outcomes Measure Audit of symptomatically presenting breast cancer reported mastectomy rates that varied from 36% to 53%, and 19% to 92% at regional and individual surgeon level, respectively.6
Older patients with ER positive early breast cancer may receive primary endocrine therapy (PET) rather than initial surgery for valid reasons—for example, serious comorbidity or short life expectancy. Figure 1 shows the percentage of women in England and Wales aged ≥70 years receiving initial surgical treatment by NHS organisation and ER status. Substantial organisational variation occurred for women with ER positive cancer but not those with ER negative disease.10 This variation remained after adjustment for comorbidity and frailty, suggesting that different perceptions of appropriate care for women with ER positive disease may be responsible.14
Randomised controlled trials have found that primary endocrine therapy has inferior outcomes for local control and disease-free survival compared with initial surgery,15 and overuse of primary endocrine therapy may be associated with the poorer relative survival seen in the UK compared with other European countries.1617
Substantial variation in clinical management extends to the use of radiotherapy,111218 adjuvant systemic therapy,1013 and immediate breast reconstruction.789 (table 1).
Causes of treatment variation
Many reports have highlighted inequalities in breast cancer care based on age, geography, ethnicity, and disability,2381017 but variation in the delivery of treatment persists even after trying to account for these factors. Furthermore, given the magnitude of variation across treatment methods, its persistence over time, and the large numbers of patients involved, the influence of factors such as patient preference or poor data quality is likely to be small.
The combination of rising breast cancer incidence and an increasing number of complex treatments has placed substantial demand on resources, and this may have contributed to variations in care.1920 In England, the 2012 Health and Social Care Act led to the introduction of clinical commissioning groups. Although these groups were intended to improve the delivery and rationalisation of health services for diverse local populations, they also created an opportunity for differential introduction of new treatments because of disparities in local commissioning priorities and processes. This was seen with the introduction of adjuvant bisphosphonates after the identification of significant benefit in post-menopausal women.21 Bisphosphonates are inexpensive and yet there was prolonged geographical variation in their availability. It took three to four years from the publication of a definitive meta-analysis confirming benefit before prescribing in the UK became uniform.2223
Although multidisciplinary teams should mitigate the effect of individual clinician preferences, clinicians may still be influenced by local practice, their involvement in clinical research, and the influence of their seniors. Several prominent trials of primary endocrine therapy occurred during the 1990s and 2000s, and these may have unduly influenced subsequent UK practice.242526
How to reduce variation in care
Unexplained variation in care is not a new problem and has a reported history of over 80 years, encompassing a huge range of medical interventions.27 There are, however, many reasons to expect greater uniformity in the care of patients with early breast cancer in the UK. These include the NHS principle of equitable healthcare, the volume and accessibility of high quality published trial evidence, the emphasis on multidisciplinary approaches to care, the quality assurance processes for screen detected disease, the long history of national audits (see supplementary data on bmj.com), and, until recently, national cancer peer review.
NICE offers advice on the principles and practice of shared decision making, but its guidance that a treatment be “offered” (strong recommendation) or “considered” (less certain benefit) is arguably open to variable interpretation. Other countries differ in the strength of their advice. In Denmark treatments are “recommended.” Radiotherapy in uniform dose and fractionation is routinely provided after mastectomy for node positive invasive disease or breast conserving surgery. Deviations from national recommendations require explicit explanation in the patient’s medical records and noting in the Danish Breast Cancer Group database. Jensen and colleagues reported that compliance with new or changed Danish breast cancer guidelines was over 90% within an average of one to two years.28
Other recent initiatives to reduce variation in care include a strengthened requirement for national clinical audits, supported and commissioned by the Healthcare Quality Improvement Partnership (HQIP), to identify outlying trusts (those with values >2 standard deviations (SDs from the mean for agreed performance indicators for care delivery and outcomes)). Trusts will be encouraged to review and understand outlying treatment rates that are ≥2 SDs from the mean and investigate when ≥3 SDs from the mean as part of their internal quality monitoring. This process will also support Care Quality Commission inspections using national clinical audit data.29
The Getting it Right First Time (GIRFT) initiative has established an ambitious programme of clinically led assessments of breast surgical services, including visits to trusts to understand and minimise undue variations in care.30 GIRFT’s initial priorities are to reduce, where possible, length of stay after breast surgery in line with British Association of Day Surgery guidelines, to encourage open access follow-up to allow patients to determine when and how they interact with breast teams, and reduce unnecessary hospital attendances, operations, and tests. The GIRFT national report on breast surgery is awaiting publication.
The National Audit of Breast Cancer in Older Patients has published information on care pathways and treatment rates at trust level.11 The ready availability of data concerning care provision, that can be individualised to a trust, team, or clinician is a priority for national breast cancer audits and Public Health England.11313233 Transparency and accountability to promote debate, discussion, and audit are powerful tools in minimising the risk of substandard practice. The formal involvement of patients in national audits and in the governance activities of Public Health England is now routine and is key to ensuring that outputs reflect patient priorities as well as those of clinicians and policy makers.34
Variations in the care provided to patients with early breast cancer have been described over decades, and it has been easy to attribute these to poor data quality, case-mix differences, and patient choice. Lack of accord concerning optimum levels of intervention, defining unacceptable variation, and an inability to confirm causality from observational data are also serious obstacles in trying to confront this problem.
The management of breast cancer is closely studied in terms of national audit and related initiatives, but the effect of variation in care for other cancers is equally important to investigate and should be a focus for cancer charities and professional groups. HQIP already supports key national audits in gastrointestinal, lung, and prostate cancer.35
Key messages
Substantial variations occur in the care of patients with early breast cancer in the UK and may be leading to poorer outcomes
Variation is of particular concern in management of older patients
Patient factors cannot fully explain differences in treatment
Data transparency and support for outliers may help reduce the gap between evidence and its uniform implementation
Acknowledgments
DD is funded by Cancer Research UK (grant C8225/A21133) and NABCOP. We thank Birgitte Offerson, Martine Bomb, and members of Independent Cancer Patients’ Voice. NABCOP is commissioned by the Healthcare Quality Improvement Partnership and funded by NHS England and Improvement and the Welsh Government.
Footnotes
Competing interests: We have read and understood BMJ policy on declaration of interests and have no relevant interests to declare.
Provenance and peer review: Not commissioned; externally peer reviewed.
References
Log in
Log in using your username and password
Log in through your institution
Subscribe from £173 *
Subscribe and get access to all BMJ articles, and much more.
* For online subscription
Access this article for 1 day for:
£38 / $45 / €42 (excludes VAT)
You can download a PDF version for your personal record.