Silent theatre 2 – Tuesday 5 November

10:30 am-11:00 am

Room: Hall 4


Programme session type(s): Silent theatre

How representative are patients responding to the Cancer Patient Experience Survey (2010-2014) of the wider cancer population in England? Analyses for breast, prostate, lung and colorectal cancer
Speaker: Saleh Alessy
Affiliation: King’s College London

Abstract:

The National Cancer Patient Experience Survey (CPES) assesses English cancer patients’ experiences of NHS care each year. Studies to date have used these data to show systematic sociodemographic variations in cancer patients’ experiences. However, the extent to which the survey responders represent the wider cancer patient population is not known. We examined how representative CPES responders are of the wider cancer population in England with respect to age, sex, socioeconomic status, ethnicity, and median survival for the four most common cancers.

Method

From the cancer registry, we identified 103,186 colorectal, lung, breast and prostate cancer patients who responded to at least one of the surveys carried out between 2010 and 2014. In addition, for each CPES responder we randomly selected one patient who was not a CPES responder based on their cancer site and time of diagnosis (same yearly quarter). The distribution of the patient characteristics (age, sex, socioeconomic status, ethnicity) and tumour characteristics (stage) were compared between CPES responders and the cancer registry population using chi-square test. We also compared median patients’ survival (in years) between the two groups.

Results

Survey responders were younger, more likely to be white, less deprived, and diagnosed with earlier disease stage than the cancer registry population across all cancer types. Median survival (in years) for the survey responders was higher than the cancer registry population group across all cancers (colorectal: 4.8 vs 3.2; lung: 2.0 vs 0.3; breast: 5.7 vs 5.5; and prostate: 5.7 vs 5.2).

Conclusion

This study showed that the representativeness of CPES results are limited in terms of patient and tumour characteristics between different cancers. These limitations need to be acknowledged when interpreting findings and future surveys should consider adopting sampling methods to include the missing care experiences of patients with lower survival, older age, and those from different ethnic minorities.

Geographical variation in access to ovarian cancer treatment in England.
Speaker: Ewa Zotow
Affiliation: Cancer Research UK

Abstract:

Data from the International Cancer Benchmarking Partnership has demonstrated poor one-year survival for ovarian cancer in the UK compared to other high-income countriesdriven mainly by survival differences in advanced disease, and older age. There is also substantial regional variation in survival within the UK. This study aimed to investigate geographical variation in treatment access rates and determine the extent to which it can be explained by patient factors. 

Method

We used linked cancer registry, COSD, SACT and HES data from patients with stage 2-4 ovarian cancer diagnosed between 20152016 in England (N=9,188)A multivariable multinomial regression model was used to explore the extent to which case-mix variables account for differences in treatment. We calculated case-mix adjusted treatment access rates for each Cancer Alliance to determine whether these rates are significantly different from the England average. All analyses were repeated in a sub-sample of patients aged 75 and over 

Results

53% of patients received surgery with chemotherapy (the NICE-recommended treatment), 12% surgery only, 13% chemotherapy only, and 22% had no recorded treatment. Older age, emergency presentation, and poorer performance status were the strongest predictors of having no recorded treatment. There was substantial variation in access to treatment by Cancer Alliance, ranging from 43%-70% for surgery with chemotherapy, 9%-23% for surgery, 8%-33% for chemotherapy, and 6%-25% for no recorded treatment. Older patients showed similar geographical variation but were significantly more likely to have no recorded treatment (43%) 

Conclusion

This study provides evidence that there is geographical variation in access to treatment that cannot be fully explained by the differences in case-mix. The identification of patient factors that are associated with both higher and lower access to treatment should feed in to the development of strategies to reduce variation. Further analyses could investigate variation in ovarian cancer survival and the underlying cause of these differencesThese results should inform policy makers and treatment providers of existing inequalities and guide future initiatives aimed at reducing the international gap in ovarian cancer outcomes. 

A Systematic Review of the Economic Costs of Gastrointestinal Consequences of Treatment for Cancer Patients and Healthcare Providers
Speaker: Mala Mann
Affiliation: Cardiff University, Specialist Unit for Review Evidence (SURE), Cardiff

Abstract:

One in two patients treated with cancer is expected to survive 10 years after diagnosis. Research so far has focused on toxicities directly interfering with the delivery of treatment. Gastrointestinal (GI) toxicity can seriously impact patients’ ability to live a normal life, nearly 1 in 2 patients has chronic side effects (eg diarrhoea, faecal incontinence, abdominal pain) after pelvic radiotherapy. These toxicities also reflect an economic burden to patients, healthcare providers and society. We carried out a systematic review of the economic costs of GI toxicities following cancer treatment (CRD42017067800).

Method

Ten databases were searched from 2006. In addition, relevant websites searched, and citation tracking of included studies were carried out. Published and unpublished studies were included. Two independent reviewers undertook screening, data extraction and quality assessment. Quality assessment was based on an adapted CHEERS checklist. Costs not expressed in Sterling Pounds were converted using the Purchasing Power Currency (PPP) exchange rate. All costs were inflated using the Bank of England calculator.

Results

Nineteen relevant studies were identified. Most studies were decision models and based in the USA. The study sample size ranged from 204 to 68,462 and the follow-up period varied from 6 weeks to 10 years following the completion of cancer treatment. Cost per episode of GI toxicity event ranged from £345 to £11,548 whereas cost per patient varied from £1,606 to £120,440 (following non-fatal intestinal perforation). Costs were mainly limited to hospitalization and secondary care data. Only one study calculated lost income.

Conclusion

Evidence is limited, nonetheless it offers a good sense of the burden on the NHS/third party payer perspective. The burden and the impact of chronic toxicities on patients’ quality of life should be incorporated when comparing the cost-effectiveness of alternative cancer treatment pathways.