Appendix A: Data quality information

This appendix includes a data quality summary and additional information relevant to interpreting the National Public Hospital Establishments Database (NPHED) and other data presented in this report, including variations in reporting and in the categorisation of hospitals as public or private

Information relevant to interpretation of the ABS’s Private hospitals Australia report is available on the
ABS website 

National Public Hospital Establishments Database data quality statement summary

Data source information and key data quality issues

For 2017–18, the NPHED is based on data reported by state and territory health authorities for the Local Hospital Networks/Public hospital establishments National minimum data set (LHN/PHE NMDS).

The AIHW has undertaken the collection and reporting of the data in this report under the auspices of the Australian Health Ministers’ Advisory Council, through the National Health Information Agreement.

The LHN/PHE NMDS is defined in the AIHW’s Metadata Online Registry (METeOR) (METeOR identifier 642698). 

The scope of the LHN/PHE NMDS includes 3 levels of heirarchical reporting:

  • public hospital establishments, including public acute and psychiatric hospitals, and alcohol and drug treatment centres. It also includes public hospitals that provide subacute and non-acute care (for example, rehabilitation and palliative care hospitals).
  • Local Hospital Networks (LHN)
  • at the jurisdictional level, all public hospital services that are managed by a state or territory health authority and are included in the General list of In-scope Public Hospital Services, which has been developed under the National Health Reform Agreement (2011) and excluding data which are already reported in the PHE or LHN levels (above).

The LHN/PHE NMDS allows the collection of recurrent expenditure, revenue, admitted contracted care and staffing information whether delivered and/or managed by hospitals or other administrative units (LHNs and state/territory health authorities).

Similar information at the public hospital establishments-level has been reported in the Australian hospital statistics reports since the first report on the 1993–94 and 1994–95 collection periods. Information at the LHN-level and at the jurisdiction-level has been reported since 2014–15. 

The LHN/PHE NMDS also includes data elements to allow the reporting of recurrent expenditure on contracted care and the number of beds available for contracted care—this information is not presented, as not all states and territories reported it, and the information did not appear to be comparable among them.

Where possible, information is reported at the lowest level of reporting possible (for example, by hospital establishment), and is not duplicated at higher levels of reporting. For example, expenditure data reported at the state/territory health authority level does not include any data reported at the LHN level or at hospital level.

At the establishment-level, the NPHED holds data for each public hospital in Australia, including public acute hospitals, psychiatric hospitals, drug and alcohol hospitals and dental hospitals in all states and territories. Hence, public hospitals not administered by the state and territory health authorities (hospitals operated by correctional authorities for example, and hospitals in offshore territories) are not included. The collection does not include data for private hospitals.

Local hospital networks are defined as those entities recognised as such by the relevant state or territory health authority.

Summary of key data quality issues

States and territories are primarily responsible for the quality of the data they report. However, the AIHW undertakes extensive validations on receipt of data, checking for valid values, logical consistency and historical consistency. Where possible, data in individual data sets are checked with data from other data sets. Potential errors are queried with the state/territory health authorities, and corrections and resubmissions may be made in response to these queries. Except as noted, the AIHW does not adjust data to account for possible data errors or missing or incorrect values.

Where possible, variations in reporting have been noted in the text. Comparisons between states and territories and between reporting years should be made with reference to the accompanying notes in the chapters and in the appendixes. The AIHW takes active steps to improve the consistency of these data over time.

  • In 2017–18, the NPHED included all public hospitals. It also included LHN-level and/or state/territory health authority-level reporting for all states and territories.
  • The number of hospitals reported can be affected by administrative and/or reporting arrangements and is not necessarily a measure of the number of physical hospital buildings or campuses. Changes in the numbers of hospitals over time can also reflect the opening of new hospitals, the closure of hospitals, the reclassification of hospitals as non-hospital facilities (or vice-versa) and the amalgamation of existing hospitals. For example:
    • Between 2013–14 and 2014–15, 46 very small reporting hospitals in Queensland and 3 establishments in South Australia that were previously classified as hospitals were reclassified as non-hospital facilities, accounting for most of the decrease in the national number of public hospitals. In addition, the Mater Children’s Hospital and Royal Children’s Hospital (both in Queensland) closed. A hospice in New South Wales and an aged care/rehabilitation facility in Victoria ceased reporting as separate campuses to the NPHED.
    • For 2014–15, the Lady Cilento Children’s Hospital (Queensland), the Fiona Stanley Hospital (Western Australia) and the Ursula Frayne Centre (Victoria) opened. Rankin Park Hospital (New South Wales) commenced reporting as a separate campus, whereas its data were previously amalgamated with another hospital.
    • For 2015–16, Byron Central Hospital (New South Wales) opened, and Byron Bay Hospital closed—both hospitals were reported. Nolan House, Albury (New South Wales) commenced reporting as a separate campus, whereas its data were previously amalgamated with Albury Hospital. The St John of God Midland Public Hospital (Western Australia) opened and Swan District Hospital closed—both hospitals were reported.
    • For 2016–17, the Sunshine Coast University Hospital (Queensland) commenced reporting. Reporting ceased for Byron Bay Hospital, St Vincent’s Lismore, Cudal War Memorial Hospital, Rankin Park Hospital (New South Wales), Next Step Drug and Alcohol Services, and the Royal Perth Hospital Shenton Park Campus (Western Australia).
    • For 2017–18, the Perth Children’s Hospital opened and the Princess Margaret Hospital closed.
  • In 2017–18, there was variation among states and territories in the administrative levels at which revenue, recurrent expenditure and staffing information were reported, including:
    • New South Wales reported this information for all 3 administrative levels.
    • Victoria reported information at the LHN and state health authority levels, and none at the public hospital level. Before 2014–15, Victoria reported this information at the network level for hospitals within networks that consisted of more than one hospital, and at the hospital level for LHNs that consisted of individual hospitals. LHN-level reporting in Victoria is therefore likely to be equivalent to the combination of hospital level and LHN-level reporting for other jurisdictions.
    • Queensland reported this information for all 3 administrative levels.
    • Western Australia reported this information for all 3 administrative levels.
    • South Australia reported this information at the hospital level only. Data attributable to the LHN level and state health authority level were included in the data provided at the hospital level.
    • Tasmania reported this information at the hospital level and at the LHN level.
    • the Australian Capital Territory reported this information at the hospital and LHN levels. Data reported at the LHN level include information for The Canberra Hospital. Data attributable to the territory health authority level were included in the data provided at the hospital and LHN levels.
    • the Northern Territory reported this information at the hospital level and data attributable to the LHN level and territory health authority level were included in the data reported at the hospital level.
  • Revenue data are not presented in this report because the data provided by states and territories for the category National Health Funding Pool differed from the 2017–18 funding reported by the National Health Funding Body (NHFB). These differences may be because:
    • the NHFB figures represent payments into the pool, not payments to service providers
    • there are differences in the timing of the reported data.
  • Available beds for admitted contracted care and Recurrent expenditure on contracted care are not reported in this publication. For 2014–15 to 2017–18, not all jurisdictions were able to report these data, and the comparability of the data was not adequate for reporting.
  • Information on hospital accreditation reported for the NPHED does not appear to be consistent with data reported by the Australian Commission on Safety on Quality in Health Care, and is not comparable across jurisdictions.
  • Differences in accounting, counting and classification practices across jurisdictions and over time may affect the comparability of these data. There was apparent variation between states and territories in the reporting of revenue, recurrent expenditure, depreciation, available beds and staffing categories. In particular, for 2013–14, data were not available at the LHN- and State/territory health authority-level, and were also not available for:
    • recurrent expenditure on different types of care, such as admitted patient care, non-admitted patient care, emergency care services and teaching, training and research
    • the type of salaried medical officers—whether a Specialised salaried medical officer or Other salaried medical officer
    • the non-salary recurrent expenditure categories for Administrative expenses-insurance, Administrative expenses-other, Depreciation-building, Depreciation-other, Lease costs and Other on-costs.
    • sources of funding (revenue), including appropriation from government sources
  • the range and types of patients treated by a hospital (casemix) can affect the comparability of bed numbers with, for example, different proportions of beds being available for special and more general purposes. In addition:
    • the average number of available beds presented in this report may differ from the counts published elsewhere. For example, counts based on a specified date, such as 30 June, may differ from the average available beds for the reporting period.
    • at the time of publication, the Northern Territory were unable to provide average available bed numbers for the Top End Health Service. Therefore, average available beds are underestimated for the Northern Territory, and overall.
    • Between 2013–14 and 2014–15, 46 very small reporting hospitals in Queensland were reclassified as non-hospital health services. The 46 hospitals combined reported 20 average available beds (in total) in 2013–14.
    • In 2014–15, Tasmania reclassified a number of mental health, aged care and same-day beds in hospitals, resulting in an apparent increase of 103 beds between 2013–14 and 2014–15. After adjusting for this change, Tasmania estimates that average available beds increased by about 0.8% between 2013–14 and 2014–15.
  • The collection of data by staffing category is not consistent among states and territories.
  • The outsourcing of services with a large labour-related component (such as food services and domestic services) can have a substantial impact on estimates of costs, and this can vary among jurisdictions.

Hospital funding and expenditure information

In this report, data presented on the funding of hospitals are sourced from the AIHW’s Health Expenditure Database (HED).

Financial data reported from the HED are not directly comparable with data reported for public hospital services from the NPHED. Hospital expenditure reported for the purpose of the HED collection may cover activity that is not covered by the NPHED. The HED data include trust fund expenditure, whereas the NPHED does not.

The 2017–18 data from the HED will be available in the second half of 2019.

Hospital funding is reported here as the money provided for the overall public and private hospital systems within each jurisdiction and nationally.

The original (or indirect) sources of funds are reported here rather than the immediate (or direct) sources. As such, the Australian Government is regarded as the source of funds for the contributions that it made for public hospitals via intergovernmental agreements, and for the contributions it made to private hospitals via the private health insurance premium rebates. For the purpose of this report, the sources of funding are disaggregated as:

  • Australian Government (including funding via intergovernmental agreements, Department of Veterans’ Affairs and private health insurance premium rebates)
  • state and territory governments          
  • non-government sources (including private health insurance, injury compensation insurers, self-funded patients and other sources of private revenue).

The information in this section was sourced from the AIHW’s Health Expenditure Database (HED), which draws data from a wide variety of government and non-government sources. Hospital funding estimates can differ from hospital recurrent expenditure reported to the NPHED—for example, depending on the administrative structures and reporting practices in the jurisdiction.

Financial data reported for public hospital services from the HED are not directly comparable with the expenditure data reported from the NPHED for the same period. The HED financial data included trust fund expenditure and central office costs, whereas the NPHED did not. The HED data for public hospital services reflect only that part of public hospitals’ expenses that were used in providing hospital services. That is, they exclude expenses incurred in providing community and public health services, dental care, patient transport services and health research undertaken by public hospitals.

Private hospital information

The most recent data available for private hospitals and private free-standing day hospital facilities is for 2016–17, based on the Australian Bureau of Statistics (ABS) in the Private Health Establishments Collection (PHEC).

Establishment information on private hospitals and private free-standing day hospital facilities were previously collected by the Australian Bureau of Statistics (ABS) in the Private Health Establishments Collection (PHEC). These data were reported in the ABS’s Private Hospitals Australia reports (ABS 2018, and earlier).

Counts of private hospitals can also vary, depending on the source of the information. Therefore, there may be discrepancies between counts of private hospitals from the ABS’s PHEC and the numbers of private hospitals contributing to the AIHW’s National Hospital Morbidity Database (NHMD). The states and territories reported the latter information, which may not correspond with the way in which private hospitals report to the ABS’s PHEC.

The PHEC data were discontinued after the 2016–17 reference period and therefore data for 2017–18 are not available.

For private hospitals, average annual changes are presented between 2012–13 and 2016–17, and between 2015–16 and 2016–17.

Annual change rates are not adjusted for any changes in data coverage, changes in metadata and/or re-categorisation of the hospital as public or private, except where noted in the text.

Contracted care

There is some variation between jurisdictions as to whether hospitals that predominantly report public hospital services, but are privately owned and/or operated, are reported as public or private hospitals. A list of such hospitals with information on how each is reported is in Table A2 available to download in the Data section of this report. The categorisations listed are those used for this report; reports produced by other agencies may categorise these hospitals differently.

For example, Peel and Joondalup hospitals are private hospitals that predominantly treat public patients under contract to the Western Australian Department of Health. The public health services provided by these two hospitals are reported separately from the private hospital activity.

The Hawkesbury District Health Service was categorised as a private hospital until 2002–03 and has been categorised as a public hospital in AIHW reports since 2003–04. From 2017–18, public hospital activity for the Hawkesbury District Health Service will be reported separately from the private hospital activity.

A list of all public and private hospitals contributing to this report is in table A.S1 available to download in the Data section of this report.

Data reported for the public hospital administrative levels

The collection of public hospital data at LHN level or at state/territory health authority level, in conjunction with the data reported at the individual hospital level, allows data to be reported by states and territories at the level relevant to service management and/or provision.

In sections of this report that present public hospital information on recurrent expenditure and full-time equivalent (FTE) staff, detailed information is presented for the total of all administrative levels. Summary data are presented for the three administrative levels:

  • Public hospitals—presents information reported for individual public hospitals.
  • Local hospital network—presents information reported at the LHN level.
  • state/territory health authority—presents information reported at the state/territory health authority level.

For 2017–18, there was variation among states and territories in the administrative levels at which revenue, recurrent expenditure and staffing information were reported. Table 1.1 available to download in the Data section of this report, summarises the comparability of the data reported by administrative level for each state and territory. For example, the data are comparable at:

  • the hospital level for New South Wales, Queensland and Western Australia
  • the LHN level for New South Wales and Western Australia
  • the combined hospital and LHN levels for New South Wales, Victoria, Queensland and Western Australia
  • at the state/territory health authority level for New South Wales, Victoria, Queensland and Western Australia
  • the total of all 3 levels for all jurisdictions.

For public hospitals, average annual changes are presented between 2013–14 and 2017–18, and between 2016–17 and 2017–18, unless otherwise stated.

Annual change rates are not adjusted for any changes in data coverage, changes in metadata and/or re-categorisation of the hospital as public or private, except where noted in the text.

The ‘major public hospital’ in each LHN was identified as the hospital with the greatest amount of admitted patient activity among the included hospitals.

Limitations of the data on staffing

Staffing information for public hospitals for 2013–14 was largely staff employed by individual hospitals, and did not include all staff employed by state or territory governments for the provision of public hospital services.

Between 2014–15 and 2017–18, staffing information reported to the NPHED includes FTE staff reported for public hospitals, for LHNs and for state/territory health authorities. For more information, see Table 1.1 available to download from the Data section of this report.

In addition, for 2017–18:

  • For Western Australia and the Northern Territory, Salaried medical officers were not disaggregated into Specialist medical officers and Other salaried medical officers as these sub-categories were not comparable with the data for other jurisdictions.
  • Western Australia reported estimated staffing information and associated salaries for 3 private hospitals delivering public hospital services.
  • For South Australia, all public hospital salaries for administrative, clerical, domestic and other personal care staff were estimated.
  • The collection of data by staffing category for public hospitals was not consistent among states and territories. In particular, there was variation in the reporting of Diagnostic and allied health professionals, Administrative and clerical staff and Domestic and other personal care staff.

Staffing numbers can include staff on contract (for example, nurses and medical officers), but exclude staff contracted to provide products (for example, contractors employed to refurbish an area).

Different reporting practices and the use of outsourcing services with a large labour-related component (such as food services, domestic services and information technology) can have a substantial impact on staffing figures and may also explain some of the variation in average salaries reported between jurisdictions. The degree of outsourcing of higher paid versus lower paid staffing functions affects the comparison of averages. For example, outsourcing the provision of domestic services but retaining domestic service managers to oversee the activities of the contractors tends to result in higher average salaries for the domestic service staff. Information was not available on numbers of visiting medical officers who were contracted by public hospitals to provide services to public patients and paid on a sessional or fee-for-service basis in public hospitals.

Limitations of the data on expenditure on public hospital services

For 2013–14, recurrent expenditure information on public hospitals reported to the NPHED was largely expenditure by hospitals and did not necessarily include all expenditure on hospital services by each state or territory government. For example, recurrent expenditure on the purchase of public hospital services at the state/territory or at the LHN level from privately owned and/or operated hospitals may not have been included.

Between 2014–15 and 2017–18, recurrent expenditure reported to the NPHED includes expenditure on public hospital services by public hospitals, by LHNs and by state/territory health authorities and includes expenditure on the provision of contracted care by private hospitals. For more information, see the ‘Data reported for the public hospital administrative levels’ section above, and Table 1.1 available to download from the Data section of this report. In addition:

  • between 2014–15 and 2017–18, for the purpose of reporting recurrent expenditure on public hospital services by public hospital peer group in this report, the AIHW assigned the recurrent expenditure reported by Victoria at LHN level to the ‘major hospital’ in the LHN—identified as the hospital with the greatest amount of admitted patient activity in the LHN.
  • between 2013–14 and 2014–15, Queensland reclassified 46 very small reporting hospitals as non-hospital services that accounted for about $89 million of recurrent expenditure in 2013–14. In addition, expenditure on pathology services for Queensland was not reported as these were purchased from a state-wide pathology service rather than being provided by hospital employees.
  • between 2014–15 and 2016–17, Tasmania reported estimated recurrent expenditure for all public hospitals.

Variation in expenditure on visiting medical officers may reflect differences in outsourcing arrangements. Variations in the outsourcing arrangements may also be reflected in variations in other recurrent expenditure categories reported in Table 2.7, available for download in the Data section of this report.

Estimated data indicators

For 2017–18, estimated data indicators were included for each category in Salary and wage expenditure, Non-salary expenditure and Revenue. The estimated data indicators specify whether the information reported reflected actual data, or estimated data.

At the public hospital level, Queensland provided estimated salary expenditure for 3 hospitals. All jurisdictions provided estimates for Revenue.

More information on estimated data is available in Table A1 available to download in the  Data section of this report.

Hospital beds

Differences in administrative practices and in the measures of beds used between public and private hospitals should be considered when interpreting the information presented in this section.

For public hospitals, counts of available beds are averaged over the reporting period and include:

  • Average available beds for same-day patients—beds, chairs or trolleys exclusively or predominantly available to provide accommodation for same-day patients 
  • Average available beds for overnight-stay patients—beds exclusively or predominantly available to provide overnight accommodation for patients (other than neonatal cots and beds occupied by hospital-in-the-home patients).

For private hospitals, the numbers of beds reported are licensed beds—the maximum number of beds specified in the hospital’s registration process. For private free-standing day hospital facilities, they include chairs, trolleys, recliners and cots.

Private hospital beds are not directly comparable to public hospital beds.