Article Text
PDF
Abstract
Introduction/Background Complete cytoreductive surgery and platinum-based chemotherapy is standard of care in the United Kingdom (NICE, 2011), yet studies indicate substantial variation in the utilization of both (Kumar et al, 2016, Hall et al). Recent work from the Netherlands shows variations in treatment for ovarian cancer across regions; however, contribution to survival was unclear (Timmermans et al, 2019). Care that is not compliant with guidelines is also seen in other countries, including the USA (Warren et al, 2017).
As part of the Ovarian Cancer Audit Feasibility Pilot, geographic variation in treatment was investigated with the objective of informing improvements in treatment and outcomes for all women diagnosed with ovarian cancer in England.
Methodology Ovary, fallopian tube and primary peritoneal carcinomas (‘ovarian cancers’) diagnosed between January 2016 and December 2018 were audited using data extracted from the national cancer registry (Henson et al,). Borderline tumours were excluded. Data is routinely collected for every patient with cancer in England through a national dataset; Cancer outcomes and Services Dataset (COSD). This information was supplemented with relevant data from the Systemic Anti-Cancer Therapy (SACT) dataset for patients receiving chemotherapy and Hospital Episode Statistics (HES) for admitted patients. Linear probability models were constructed adjusting for tumour morphology, stage at diagnosis, patient age at diagnosis; Charlson comorbidity index, area income deprivation. Tumours with stage 1 disease at diagnosis were excluded from analysis of variation in treatment. Treatment variations across the 19 cancer alliances (units of geography) were evaluated. Survival analyses were extracted from a previous cohort diagnosed 2013–2017.
Results Treatment received in 13,889 ovarian cancers was analysed. The weighted average probability (range for cancer alliances) of a stage 2–4 ovarian cancer receiving any treatment, any surgery, and any chemotherapy across England was 73.8% (70.4% - 79.3%), 51% (37.2% - 58.9%) and 66.5% (61.8%-73.6%) respectively (figure 1). One-year net survival for the 19 Cancer Alliances in England varied between 62.9% and 75.2%, 5-year net survival varied between 28.6% and 49.6%. Cancer Alliances that were statistically less likely to undertake surgery generally had lower than average survival (figure 2).
Variation in treatment for women with advanced OCData from 13,889 women diagnosed between 2016–2018, figures from Treatment variation report of the National Ovarian Cancer Feasibility pilot auditFig 1A. Variation in receiving any anticancer treatmentFig 18. Variation in receiving surgeryFig 1C. Variation in receiving chemotherapyOuter Dashed lines indicate 3 standard deviations
A) Variation across 19 Cancer Alliances in England for patients with advanced OC receiving surgery: women diagnosed 2016–2018B) Survival analysis from 19 Cancer alliances in England women diagnosed 2013–2017Interpretation - Stressing differences in the time coverage & cohort definitions, cross-referencing treatment variation & survival analyses suggests Cancer Alliances less likely to undertake surgery had generally lower than average five year survival
Conclusion Significant variation in treatment and survival across England are demonstrated in this audit. The population-based nature of this robust audit indicates that our findings are likely to be relevant to international settings. Efforts to understand and reduce variation in treatment decision making and reducing the proportion of women not receiving treatment are critical to improving survival in ovarian cancer. Ongoing audit of treatment will be key to driving and monitoring progress.