Increasingly there is a greater emphasis on whole of government approach to service provision. Accordingly the Aged Care Unit and the Community Services Integration and Linkage Unit of the AIHW are carrying out research designed to follow the flow of care recipients through the aged care system. This research requires the use of data linkage methodology and must take into consideration issues of privacy and confidentiality. Ethical approval from the AIHW Ethics Committee is required and, for some research, clearance is required from the ethics committees of other organisations. Note that the Institute does not allow linkage of data for administrative or regulatory purposes.
About privacy and linkage protocols
The Australian Institute of Health and Welfare operates under the the Australian Institute of Health and Welfare Act 1987 which contains provisions to protect the confidentiality and security of data, as well as being bound by the provisions of the Privacy Act 1988. The AIHW has documented procedures, approved by its Board, covering these topics and a strong culture of ensuring data security. All staff sign a confidentiality agreement and the AIHW Act's confidentiality provisions contain penalties (which include imprisonment) for breaches of confidentiality ( Australian Institute of Health and Welfare Act 1987, Sect 29).
Statistical linkage
While linkage for patient or client management requires that individuals are unambiguously identified, for statistical purposes 100% accuracy is not required. A great deal of useful information can be derived from data linkage using a statistical linkage key (SLK).
Most linkage studies carried out in the Ageing and Aged Care Unit of the AIHW use a statistical linkage key initially developed for the HACC minimum data set. This linkage key uses the second, third and fifth letters of the persons family name, the second and third letters of the given name (note that all punctuation and spaces are removed first), the date of birth and sex. This SLK has been included in the HACC MDS, the CACP/EACH census data collections, the ACAP MDS, the CSTDA MDS and is under consideration by other data collections.
In some linkage research, for instance the feasibility study which looked at linking hospital data and residential care data, where letters of name are not available linkage protocols have linked on other information has been used to make a statistical linkage between data sets.