Population-based measurement of quality of diabetes care using HbA1c values in the state of Thuringia/Germany

Abstract

Introduction/hypothesis: There remains insufficient knowledge about the quality of diabetes therapy regarding structure, process and outcome given by primary care in Thuringia, and elsewhere in Germany. The project assesses a means of determining the quality of diabetes therapy in primary care by analysis of HbA1c values within a broad territory.

Methods: HbA1c tests of Thuringian patients were analysed by 20 participating medical laboratories between January 1 and March 31, 2002; each HbA1c test was identified by the postal code of the Thuringian General Practitioner (GP) who ordered the test and adjusted by a standardisation procedure (mean normal of healthy subjects 5 %). The German state of Thuringia (population of 2,421,871) consists of 23 urban and rural districts. This format was also used in the present study.

Results: Twenty participating laboratories contributed 59,702 HbA1c tests from diabetic patients in Thuringia. Adjusted mean HbA1c of entire Thuringia: 6.75 %, 36.4 % of all HbA1c tests were above 7.0 % and 3,0 % of all HbA1c values above 10.0 %. Regional percentage of HbA1c tests above 7.0 % ranged between 28.2 % and 47.3 %. Regional percentage of HbA1c tests above 10.0 % ranged between 1.8 % and 4.8 %. Subgroup analyses comparing urban vs. rural districts showed significantly lower mean HbA1c (6.72 % vs. 6.75 %, CI 0.01 - 0.06, p < 0.01) and percentage of HbA1c tests above 7 % (8.19 % vs. 8.24 %, CI 0.01 - 0.08, p < 0.01) in urban areas.

Conclusions: Mean HbA1c as an indicator of glycaemic control of diabetes patients in primary care in Thuringia was good. The percentage of patients who need immediate medical intervention to prevent acute complications varied between districts, which is indicative for differences in quality of diabetes care in Thuringia. However, the proposed method is still experimental and has not yet been evaluated. Consequently, considerable bias might influence the results.