Is the Jarman score better than social class at assessing the need for prevention and primary care?

This paper reports an analysis by small areas of various measures of disease and the use of cervical smear services in the city of Sheffield. The correlation of these with social class and the Jarman underprivileged area score were compared. Wide variations in mortality rates between electoral wards in Sheffield were demonstrated, particularly for deaths from diseases with a large preventable component. Social class correlated more strongly with all-cause mortality (r = 0.69) and preventable mortality (r = 0.91) than did the Jarman score. There was no significant correlation between routine cervical smear rate and either social class or the Jarman score among women under the age of 35 years. Among older women, however, there was a high degree of correlation with fewest smears being taken in the most deprived wards. Social class was more strongly correlated with the invasive cervical cancer rate in electoral wards than was the Jarman score, and was thus a better indicator of the need for cervical screening. However, the Jarman score showed a greater degree of (negative) correlation with the uptake of cervical screening than did social class with disproportionately fewer smears being taken by general practitioners in areas of highest need. Social class may be better than the Jarman score as an indicator of both ill-health and the need for preventive health services in Sheffield. Information is routinely collected decenially on social class and needs little further computation, unlike the Jarman score. Furthermore, much is already known about the relationships between social class and both ill-health and the need for preventive services.