E-poster Presentation 2014 World Cancer Congress

Identifying variation in cancer mortality in the elderly; not as easy as it sounds (#893)

Elizabeth Jane Maher 1 , Jenny Ritchie-Campbell 1 , Rachel White 1
  1. Macmillan Cancer Support, London, United Kingdom


Age adjusted cancer mortality rates have fallen by 11% over the last decade. This improvement is focused on those under 75; 15% reduction in those under 75, 5% reduction for those 75 and over. It is widely observed that this trend is not uniform across the UK.


To uncover the extent to which mortality rates vary in those 75 and over and how to identify which areas warrant further investigation.


UK registry data was extracted from the UK Cancer Information Service. We extracted age-adjusted cancer (ICD-10 C00-C97) mortality rates and numbers of deaths, cross tabulated with higher tier local authorities or equivalents (LAs) for two decades up to 2011. We explored the data using a range of methods.


There is variation in the age-adjusted mortality rate of those aged 75 and above; the area with the highest rate has more than double the rate of the area with the lowest rate. These areas are significantly different to the UK average. However, much of the variation in other geographies was found to be variation due to chance.


It is important to be able to accurately identify areas with genuinely higher mortality rates so that we can focus rationed efforts to improve outcomes.

The next stage is to unpick why variation may be present; this can be partially achieved through studying how mortality rates vary over time. If the mortality rate needs improvement in a stable system a systematic change of the whole system is probably needed. Alternatively the variation over time may indicate that something unpredictable is happening. These unpredictable changes are likely to be caused by an altering external factor that we need to understand.