Evaluation of methods for calculating census health indicators for GP practices

Area of Science:

• Health Services Research
• Geographic Information Systems
• Demography

Background:

• General practices serve diverse patient populations across multiple census areas.
• Accurate demographic data is essential for practice planning and resource allocation.
• Existing methods for linking patient demographics to census data are often inadequate.

Purpose of the Study:

• To evaluate and compare different methods for aggregating census data to describe general practice patient characteristics.
• To determine the most accurate census-based method for estimating the proportion of elderly patients (aged 75 and over) within general practices.

Main Methods:

• Four distinct methods were employed to estimate the percentage of patients aged 75 years and over using census data.
• These census-derived estimates were compared against actual patient register data from 81 general practices in Suffolk, England.
• Methods included using census values for practice areas, surgery locations, and patient-weighted averages of ward and enumeration district data.

Main Results:

• Patient-weighted averages of ward and enumeration district data yielded the most accurate estimates.
• Ward-level data provided accuracy comparable to finer enumeration district data.
• Both superior methods achieved estimates within 2% of the patient register values, allowing for broad practice comparisons.

Conclusions:

• Patient-weighted aggregation of census data, particularly at the ward level, offers a validated approach for estimating practice populations.
• While census and patient register data have inherent discrepancies, these methods enable meaningful comparisons between general practices.
• The findings support the use of census data for understanding the demographic profiles of general practice patients.