Published in:
Open Access 01-12-2018 | Research article
The usefulness of small-area-based socioeconomic characteristics in assessing the treatment outcomes of type 2 diabetes patients: a register-based mixed-effect study
Published in: BMC Public Health | Issue 1/2018
Login to get accessAbstract
Background
Assessment of the differences in the outcomes of care by socioeconomic status (SES) is beneficial for both the efficient targeting of health care services and to decrease health inequalities. This study compares the effects of three patient-based SES predictors (earned income, educational attainment, employment status) with three small-area-based SES predictors (median income, educational attainment, proportion of the unemployed) on the treatment outcomes of type 2 diabetes patients.
Methods
Mixed-effect modeling was applied to analyse how SES factors affect the treatment outcomes of type 2 diabetes patients. The treatment outcomes were assessed by the patients’ latest available glycated hemoglobin A1C (HbA1c) value. We used electronic health records of type 2 diabetes patients from the regional electronic patient database, the patients’ individual register-based SES information from Statistics Finland, and the SES information about the population of the postal code area of the patients from Statistics Finland.
Results
The effects of attained education on the treatment outcomes, both at the patient-level and the small-area-level are quite similar. Age and male gender were associated with higher HbA1c values and lower education indicated higher HbA1c values. Unemployment was not associated with HbA1c values at either the patient-level or the area-level. Income gave divergent results: high values of HbA1c were associated with low patient incomes but the median income of the postal code area did not predict the treatment outcomes of patients.
Conclusions
Our comparative study of three SES factors shows that the effects of attained education on the treatment outcomes are rather similar, regardless of whether patient-based or small-area-based predictors are used. Small-area-based SES variables can be a good way to overcome the absence of individual SES information, but further research is needed to find the valid small-area factors by disease. This possibility of using more small-area-based data would be valuable in health service research and first-hand planning of health care services.