Matching Dental Practice Board data to the health authority population register

Objective: To evaluate the process of matching Dental Practice Board (DPB) registration data to the health authority (HA) register for a defined locality. This included testing for the presence of socio-economic bias in the postcodes held in the DPB data sets and to determine if patients travelling to obtain primary dental care would have an effect on the matching process at health authority level. Basic research design: All 21 general dental practitioners in Ellesmere Port, South Cheshire gave written consent for the DPB to provide files of their registered patients to the health authority. The registered patients were electronically matched against residents in the HA register in a four stage process. Testing for the presence of a socio-economic bias in the postcodes included in the DPB data set was undertaken by constructing frequency distributions of the proportions of postcoded, and non or partially postcoded DPB records by Super Profiles Lifestyle group. A chi square test was used to test for significant differences. The extent of patients travelling to access primary dental care was assessed by comparing the proportions of registered patients of under 6 years, 25-60 years and those aged 65+ who lived outside South Cheshire Health Authority yet obtained their GDS care within Ellesmere Port. Main outcome measures: Out of a total of 39,474 records, 458 (1.2%) were duplicates. Some 35,447 (90.9%) of DPB records could be matched to those contained in the HA register. No significant (χ2 = 10.85, df=9, P=0.27) socio-economic difference was found in the proportions of DPB records originally with, and without postcodes by Super Profiles Lifestyle group. Some 1,688 (47.3%) of the unmatched subjects were resident in health authorities other than South Cheshire and therefore could not be expected to be found in the HA register. Nearly 90% of children under 6 years registered with an Ellesmere Port GDP were South Cheshire residents, compared to approximately 85% of adults of working age and older adults. Conclusions: In this locality DPB data could be matched to the HA register with a reasonable degree of success. However, because of the time taken to clean and make the data sets compatible, the methodology used is inappropriate for routine use. In this locality no socio-economic bias could be found in the records with missing postcodes in the DPB data set. Patients travelling to access dental care would have little influence on the matching process at health authority level in this locality, irrespective of the age group under study.