Socioeconomic risk factors and population-based regional allocation of healthcare funds

Using the notion of professional uncertainty a population-based proxy need measure for hospital services was developed. Its relationship with socioeconomic variables and Standardized Mortality Ratios (SMR) was investigated in an attempt to develop an adjustment factor for socioeconomic risk factors beyond age-sex adjustment to be used for a population-based healthcare funding formula for Alberta. The data used are 1990, 1991, 1992 vital statistics and hospital separation abstracts, 1991 census data and Refined Diagnosis Related Group (RDRG) case weights. Geographic units studied were the 26 federal electoral districts in Alberta using postal codes as a linkage geo-code between census and hospital utilization and death data. SMRs, age-sex standardized per capita hospital utilization and proxy need rates were derived and correlated with socioeconomic variables derived from the census files. It appears that the poor, the less educated and aboriginals need more hospital services than the affluent, employed and educated, confirming previous findings. The unemployed tend to need more but use fewer services while immigrants and non-white ethnics tend to need and use fewer services. The unemployed, less educated and non-white ethnics are associated with positive correlation with premature mortality (SMR based on deaths under age 75 years), while the employed, highly educated tend to live longer. In general SMRs have positive but very low correlations with utilization and need rates suggesting that SMRs should not be used for resource allocation. Stepwise multiple regression analyses showed that the percentages of unemployed, immigrants, non-whites, aboriginals and those with education less than grade 9 explain about 90% of the variation in age-sex standardized hospital utilization rates. Percentages of unemployed, non-white ethnics, residents with education less than grade 9 and aboriginals explained 71% of variations in age-sex standardized per capita proxy hospital service need measures. Based on the results of regression analyses, a SEAM (Socio-Economic Adjustment Multiplier) scale was developed for utilization (SEAM-U) and proxy needs (SEAM-N). In essence a SEAM is a set of relative value (RV) multipliers applicable to a provincial common per age-sex adjusted capita allocation value to account for the impact of socioeconomic risk factors on hospital service needs or utilization. Finally, the resulting regression equations derived from the 26 Federal electoral district data were applied to Alberta's health regions, regional SEAMs were derived, and the impact of such adjustment was assessed.