USE OF THE CENSUS SAMPLES OF ANONYMISED RECORDS (SARS) AND SURVEY DATA IN COMBINATION TO OBTAIN ESTIMATES AT LOCAL-AUTHORITY LEVEL

Synthetic estimation techniques are methods for obtaining improved loc al estimates by combining data that are available for local areas with other data that are not. The Samples of Anonymised Records (SARs) pro vide large representative samples of 278 small areas of Britain and ar e thus of great value to planners. This paper describes an approach wh ich takes advantage of the fact that the SARs comprise individual reco rds. Estimates of the proportions of local authority populations suffe ring serious illness were produced by use of data from the 4th Nationa l General Practitioner Morbidity survey and the 2% anonymised sample o f individual 1991 Census records. These estimates were compared with e xternal validation criterion, all-cause mortality. The correlation was high, providing some evidence of the validity of the approach. The me thod could be adapted to produce a variety of different estimates.