Sloggett A. Socio-economic and socio-demographic inequalities in cancer incidence and survival in the older population of England and Wales. End of award report.. 2004
Data from the Office for National Statistics Longitudinal Study (ONS LS) was used to analyse variations in cancer incidence and survival in the older population by indicators of socio-economic and socio-demographic position. Cancer free life expectancies by region and socio-economic group were also estimated; and we examined whether cancer incidence led to changes in household circumstances, investigated the living arrangements and characteristics of co-residents of people with cancer and analysed whether these were associated with whether people dying from cancer were more or less likely to die at home.
We used relative survival methods in our analysis of differentials and survival and a special programme for analysis of longitudinal data to produce estimates of cancer free life expectancy. Neither of these methods had been applied to ONS LS data before. We made comparisons between incidence and survival results from the ONS LS and those from cancer registry data on the whole population and found a very good match, confirming the status of the ONS LS as a valuable source for study of cancer as, unlike registry data, it includes a wealth of individual level socio-economic and socio-demographic information.
We modelled relative excess mortality for all cancers combined and for eleven site specific cancers. For all cancers combined we found differences in survival by four indicators of socio-economic status: social class; housing tenure; access to a car, and Carstairs' index - an ecological indicator derived from characteristics of the ward of residence - in all cases socio-economic disadvantage was associated with poorer survival. A combination of car access and housing tenure was more sensitive than the ecological Carstairs measure at detecting socioeconomic effects on survival - confirming Carstairs effects where they occur but additionally identifying effects for other cancers. For ovarian cancer particularly, reported in larger studies as having only a small socioeconomic gradient when measured by ecological measures, car access identifies a strong socioeconomic effect on survival, confirmed by tenure.
We also found differences in survival by marital status and by household type. Widowed and divorced women had poorer survival for all types of cancer combined, even after adjusting for socio-economic factors. However this finding did not persist in the analysis for female breast cancer and so suggests that the ex-married groups of women do not necessarily have poorer survival across the board, but that they may have higher incidence of more rapidly fatal cancers. This could be investigated at a later date using the cancer incidence dataset. Particularly among men, those living with a spouse had better survival from cancer than those in other types of household.
Analysis of regional differentials in cancer free life expectancy was undertaken using a relatively new specialist program, application of which required some development work. Results showed that for men both overall life expectancy at age 45 and cancer free life expectancy at age 45 was highest on London and lowest in the North East. For women the East region had the best, and Mersey the worst, overall life expectancy and cancer free life expectancy. For both sexes the London Region shows the shortest number of expected cancer years. This is surprising because London has high overall life expectancy, the highest of all regions for men, and one would expect a pro rata proportion of time spent with cancer. Results therefore suggest some regional variations in needs for cancer related services, and differences in proportions of time spent with cancer by different social groups that will bear some further investigation.
Trends in the living arrangements of older people were found to be very similar for those with cancer and those without. As a consequence the proportion of cancer patients who lived alone was higher, and the proportion living with relatives other than a spouse lower, in 1991 than it had been ten years earlier By 1991 over half of female cancer patients aged 80 and over and a quarter of equivalent men lived alone. We found no evidence that household type transitions were associated with cancer diagnosis, for example among those living alone in 1981 a subsequent cancer registration was not associated with increased odds of no longer living alone by 1991. Of those cancer sufferers who lived with someone else, their primary co-resident was in most cases a spouse with much smaller proportions living with a child, a sibling or another person. 30% of spouse co-residents and 23% of other co residents had a limiting long term illness and 21% of the former and 17% of the latter were themselves aged 75 or over. Presence, type and health status of primary co resident was associated with differentials in the proportions of people dying of cancer who died at home, but not to the same extent with home death among those dying of other causes. Lower socio-economic status, both individual and area, was also associated with reduced chance of a home death, which is consistent with other findings. These results suggest increasing challenges to the delivery of home based palliative care and that planning of cancer services needs to take account of changing living arrangements of the older population.