GROS mid year small area population estimates GROS death registrations (using year of death, not registration) Sample survey data - self-assessed health reported as good or fairly good from the Scottish Household Survey. Index of various data for SIMD. HLE is derived by combining estimates of life expectancy ( LE) with data on self-assessed health (from surveys). LE calculations are based on the Chiang (II) methodology; HLE based on the Sullivan method. Estimates of HLE are less robust than estimates of LE due to the use of survey data; the fact that health status is self-assessed brings in an element of potential bias to the estimates. HLE estimates have much wider confidence intervals than LE estimates. A small proportion of death records (0.7%) could not be assigned to a deprivation category and have therefore been imputed. The self-assessed health data only covers ages 16+. Therefore the results for the age group 16-19 are applied to all age groups below 16. This approach, the relatively small numbers surveyed within an age/sex category, and the variability of the SAH estimates over time, have led to the apparent rise in HLE for females in the 15% most deprived datazones in 2001-2002 compared to 1999-2000 and to 2005-2006. The methodology will be reviewed and different estimates may be published in future. |