This data quality survey is the third project conducted by the Northern Territory (NT) Department of Health over a 13 year period to measure the accuracy of demographic information collected by NT public hospitals about hospital inpatients. The NT component of the survey was conducted by the Health Gains Planning (HGP) Branch of the NT Department of Health. HGP Branch staff interviewed 892 inpatients of the five NT public hospitals during their hospital admission. Patients were asked about their age, sex, Indigenous status, country of birth, and place of residence. In addition, the NT survey included questions about the patient’s place of residence one year and five years ago. The survey results were compared with information recorded in the NT hospital information system, which is shared by all NT public hospitals.

A data quality audit was conducted in late 1997 in the five NT public hospitals. Patients were interviewed while they were in hospital by an interviewer independent of the hospital patient adminstration processes. The interviewer asked each patient their Indigenous status, sex, date of birth, marital status, place of residence and country of birth. Data obtained at interview were compared to the same data items in the hospital morbidity dataset.

The purpose of this paper is to report on conceptual, methodological and reporting issues involved in the various population data types. This includes both Australian Bureau of Statistics (ABS) and non-ABS sources. The appendices to this paper provide a comprehensive list of the most recent population data available to users.

Accurate recording and updating of hospital administrative data for inpatients is critical to the management of hospitals, statistical reporting and the planning of future policy and funding. In the Northern Territory (NT), the patient management system CareSys is used to record patient care details and key demographic information. CareSys is used across all five NT public hospitals. Central to statistical reporting is the quality of demographic information contained in hospital records: age, gender, Indigenous status and place of residence. Accurate collection of all key demographic items is important, however, Indigenous status and place of residence are particularly important. The accurate recording of Indigenous status is vital in the NT due to the proportionally high Indigenous population compared to other states and territories. Monitoring service use by the NT Indigenous population is essential to establishing the success of various health initiatives, identifying where gaps exist in health service provision, and determining funding of future programs and campaigns. Accurate collection of a patient’s usual place of residence allows identification of patients from interstate, allowing for cross-border charging, as all states and territories are required to meet the cost of their residents’ inpatient services regardless of where treatment is received in Australia. Place of usual residence also shows where the greatest need for hospital services exist in the NT, and highlights travel costs associated with the provision of hospital services. The study found high levels of agreement between the hospital dataset and interview dataset for all variables; sex (99%), Indigenous status (97%), country of birth (98%), year of birth (94%), date of birth (90%) and NT health district of residence (88%). A comparison between the 1997 and 2008 levels of observed agreement for the key demographic variables shows improvements for all variables, with the exception of sex which remained unchanged.