This page contains information relevant to both the Online Services Report (OSR) and national Key Performance Indicators (nKPI) collections. Further information specific to each collection is provided in Interpreting OSR data and Interpreting nKPI data.
Where to go for more information
This page contains general information to aid interpretation of the OSR and nKPI collections. This should be used in conjunction with additional information contained in the:
While some organisations constitute an individual health care clinic, others have multiple clinics, and provide combined data for all their clinics. Other organisations are intermediaries (for example, Primary Health Networks), which might also combine the data for the clinics where they subcontract services.
In both collections:
- A client may attend more than one organisation. The extent to which this occurs is not known and is not adjusted for.
- Aboriginal Community Controlled Health Organisations (ACCHOs) report all activity (regardless of funding source). Non-ACCHOs should only report activity funded under the Indigenous Australians Health Programme (IAHP) but some report all activity. The extent to which this occurs is not known and may vary by period.
Data collection and submission
Each organisation reporting to the OSR and nKPI collections records service provision in their Clinical Information Systems (CIS). While the CISs contain many variables related to individual clients, only those specified as required for the OSR and nKPI collections, aggregated for each organisation, are extracted for use.
Data are aggregated using cohort definitions and specialised software and then submitted to the Australian Institute of Health and Welfare (AIHW) via the Health Data Portal (the HDP). The HDP is the Department of Health’s secure web-based data submission platform.
Most CIS used by First Nations-specific primary health care organisations are able to extract, aggregate and upload the de-individualised data directly to the HDP through a direct-load process. Organisations with systems unable to complete the direct-load process can manually enter aggregate data directly into a web-based form through the HDP.
The HDP applies a series of pre-defined validation rules (originally developed by the AIHW) to the data to identify any data quality issues, for example, that the numerator is less than the denominator, that the numerator sums to the denominator and that related indicators or questions are consistent. In addition, the data are compared with data from the previous period and:
- for the OSR collection, differences of more than 20% are flagged
- for the nKPI collection, differences of 25% to 100% are flagged, depending on the size of the numerator/denominator.
If any validation rules are triggered, the organisation is asked to review their data and either amend it, or confirm that it is correct and provide a comment.
The data are then submitted to the AIHW for review. Where the AIHW identifies possible errors or inconsistencies in the data, a comment is added to the HDP containing a description of the issue and the organisation is invited to resubmit their data. This process is repeated until no data quality issues are identified, the relevant organisation indicates it is unable or unwilling to resupply corrected data or the collection period is closed. At this point, the data are considered to be finalised for that organisation.
Where unresolved data quality issues are identified, the AIHW excludes these data from national reporting (this may be a single data item for an organisation or all data for an organisation).
Comparisons over time
Trends over time are presented where possible, noting that the organisations reporting to the OSR and nKPI collections, and data quality, can vary over time. For example:
- While for the most part, it is the same organisations contributing to the collections, as a result of changes in funding, auspicing or reporting arrangements at the local level:
- the organisations that are funded to provide services vary between periods
- the funded organisations that report data each year vary (for example, an organisation may be given an exemption from reporting for specific periods).
- Some organisations may be unable to report accurate data in particular periods or for particular data items (for example, because of changes in their clinical information systems or record-keeping practices) and these data are excluded from national reporting. As such, the number of organisations submitting valid data, on the whole or for a particular data item, vary between periods.
Also, unlike in previous years where reporting to the OSR and nKPI collections was a mandatory condition of receiving funding, for 2019–20, 2020–21 and 2021–22 reporting was made voluntary in acknowledgement of the additional pressures on organisations because of the COVID-19 pandemic. COVID-19 also affected the organisations reporting to the collections in a variety of other ways. See Summary and Australia’s Health 2022 for more information.
Estimated resident population (ERP)
The estimated resident population (ERP) used as comparison for the OSR and nKPI client cohorts varies. The ERP chosen, or calculated, for each period is determined based on what is considered to be the most representative population for that period.
As the nKPI collection is based on census dates at 30 June and 31 December each year, the most appropriate ERP to use for comparison with First Nations regular clients is the ERP at the same date as the census date.
As the OSR collection covers a financial year, and estimates as at 30 June not considered appropriate for use when calculating rates based on financial year data, estimates for 31 December (that is, the midpoint of the financial year) are needed.
While the Australian Bureau of Statistics (ABS) produces estimates for the overall Australian population for 2 time points each year – namely, as at 30 June and 31 December – they only produce estimates for the First Nations population as at 30 June. As such, ERP at 31 December are calculated by averaging the June population estimates before and after the relevant December. For example, 31 December 2020 estimate = (30 June 2020 estimate + 30 June 2021 estimate)/2.