Research Article

Unemployment rates: an alternative to the Jarman index?

BMJ 1991; 303 doi: (Published 28 September 1991) Cite this as: BMJ 1991;303:750
  1. D A Campbell,
  2. J M Radford,
  3. P Burton
  1. Department of Public Health Medicine, Central Nottinghamshire Health Authority, Ransom Hospital, Mansfield.


    OBJECTIVE--To compare the Jarman index with alternative deprivation measures with regard to its usefulness to district health authorities as an indicator of need at small area level. DESIGN--The Jarman index (UPA (8)), Townsend's index of material deprivation, the Scottish Development Department's index, the Department of the Environment's basic index, and unemployment rates were compared in respect of their correlation with measures of morbidity by electoral ward in a typical English district health authority. Measures of morbidity comprised standardised mortality ratios, admission rates (standardised and non-standardised), and permanent sickness rates. Spearman rank correlation coefficients were calculated for each combination of measures and were then ranked for each of the deprivation indices. SETTING--The 59 electoral wards of the Central Nottinghamshire Health Authority. RESULTS--The Jarman index consistently ranked lower in respect of its correlation with measures of morbidity than did the other deprivation measures. Current unemployment rates correlated well with morbidity measures, in particular with hospital admission rates, with correlations ranging from 0.669 to 0.830 for average and standardised all age admission rates. CONCLUSIONS--The Jarman index seems to be the least appropriate of these indices for health authority use. Unemployment rate merits further consideration as a simple, up to date marker for deprivation and consequent need for health service provision.