Palliative Chemotherapy: what are the demographic, clinical and temporal factors associated with receiving chemotherapy in people who go on to die from cancer

Problem

The decision whether or not to prescribe chemotherapy for someone who is possibly or likely going to die from their cancer is complex and challenging. While it may extend quantity of life, chemotherapy often has a negative effect on quality of life. In this cohort of cancer decedents (people who went on to die from cancer) we examine what demographic, clinical and temporal factors were associated with whether or not patients received chemotherapy.

Approach

Retrospective cohort study of all 2,443 people who died from cancer in NHS Tayside, in Scotland, between 01/01/2012-30/06/2015. Clinical population datasets including the Cancer Registry (SMR06) and GRO Death Data, were linked to routinely collected clinical data using the Community Health Index (CHI) number. The CHI is a unique patient identification number used in all clinical encounters in NHS Scotland. Anonymised CHI-linked data were analysed in SafeHaven, with descriptive analysis, using binary logistic regression for adjusted associations.

Findings

Cancer decedents who were under 65 years old were substantially more likely to have receive chemotherapy than those who were aged >85 (AOR 0.03 95CI(0.017 to 0.051). Cancer decedents who lived in the least deprived areas - SIMD4 (AOR 1.58(1.07 to 2.33) and SIND5 1.80(1.20 to 2.71) - were nearly twice as likely to receive chemotherapy as those who lived in the most deprived areas.Compared to people with lung cancer, those with upper GI malignancies (AOR 0.52(95CI 0.37 to 0.73)) and bowel cancer (AOR 0.57(95CI 0.38 to 0.85)) were much less likely to receive chemotherapy. People who died from prostate cancer were 20 times less likely to receive chemotherapy than those who died from lung cancer (AOR 0.05(95CI 0.02 to 0.138)).Timing of diagnosis was strongly associated with whether or not people who died from cancer received chemotherapy. Compared to cancer decedents diagnosed 0-12 weeks before death, those who were diagnosed 13-25 weeks before death were seven times (AOR 7.64(4.73 to 12.31)) more likely to receive chemotherapy, decedents diagnosed 26-38 weeks before death were twenty-three times (AOR 23.61(14.66 to 38.03)) more likely to receive chemotherapy, decedents diagnosed 39-51 weeks before death were twenty-six times more likely to get chemotherapy and cancer decedents diagnosed ≥52 weeks before death were nearly thirty times (AOR 29.52(19.28 to 45.19)) more likely to receive chemotherapy, compared to cancer decedents who were diagnosed 0-12 weeks before death.

Consequences

There are multiple factors to consider when offering chemotherapy to patients with advanced cancer. While some of the associations observed in this study have direct correlations with side-effects, outcomes and anticipated benefits of chemotherapy, including age and anticipated prognosis, the variation in the provision of chemotherapy based on deprivation was unexpected and requires further analysis to explore a causal relationship.