AIM: To assess the completeness of primary care data collected by the Royal New Zealand College of General Practitioners' (RNZCGP) Dunedin Research Unit, and assess the feasibility of creating a New Zealand national minimum data set for primary care. METHODS: Patients from 42 practices contributing data to the Dunedin Research Unit made up the study population. A six-month sample of data was evaluated for completeness, and compliance to a minimum data set structure. Rates of recording patient identifiers, sex, ethnicity, community services card status, consultation identifier and date, prescriptions and Read codes were calculated for each practice and registered patient. RESULTS: Patient demography, NHI and community services card status were all well recorded (date of birth 99.3%, sex 98.9%, NHI 94.8%, CSC 100.0%). Read codes and ethnicity were still poorly recorded, although there was wide variation between practices. CONCLUSIONS: The completeness of data collected by the Dunedin Research Unit appears to be improving, although there is wide variation between contributing practices. The capability to create a primary care national minimum data set exists, but this will not become a reality until suitable education programmes and support are supplied for general practitioners and other staff members who record patient information.
AIM: To assess the completeness of primary care data collected by the Royal New Zealand College of General Practitioners' (RNZCGP) Dunedin Research Unit, and assess the feasibility of creating a New Zealand national minimum data set for primary care. METHODS:Patients from 42 practices contributing data to the Dunedin Research Unit made up the study population. A six-month sample of data was evaluated for completeness, and compliance to a minimum data set structure. Rates of recording patient identifiers, sex, ethnicity, community services card status, consultation identifier and date, prescriptions and Read codes were calculated for each practice and registered patient. RESULTS:Patient demography, NHI and community services card status were all well recorded (date of birth 99.3%, sex 98.9%, NHI 94.8%, CSC 100.0%). Read codes and ethnicity were still poorly recorded, although there was wide variation between practices. CONCLUSIONS: The completeness of data collected by the Dunedin Research Unit appears to be improving, although there is wide variation between contributing practices. The capability to create a primary care national minimum data set exists, but this will not become a reality until suitable education programmes and support are supplied for general practitioners and other staff members who record patient information.