Private Hospital Data Collection Review Final Report

1. Purpose

Private Hospital Data Collection Review - Final Report

Page last updated: 14 June 2012

    This document forms the final report for the Private Hospitals Data Collection Review (the Review), which was commissioned by the Australian Department of Health and Ageing (the Department). The objectives of the Review are:

    1.   Increase the collection, management, and handling efficiency of private hospital data, so as to reduce data management burdens where possible;
    2.   To support increased comparability between the public and private sectors; and
    3.   Recommend a mechanism for creating and maintaining an authoritative list of private hospitals.

    In commissioning the Review, the Department prescribed two streams of work to be undertaken by KPMG:

    • current and proposed (where known) private hospital data collections and data supply chains from point of collection through to state and Commonwealth authorities and groups, andrecommendations of options for streamlining (Work Stream 1); and
    • public and private hospital service data and development of a feasible data convergence plan for public and private hospitals (Work Stream 2).

    Work Stream 1 focused on objectives 1 and 3 listed above, while Work Stream 2 focused on objective 2. This report presents the findings and recommendations for both streams. It also includes a draft implementation plan for delivering those recommendations.

    Project Background

    The 2008 and 2010 health reforms will create a far more data reliant environment than currently exists. Further changes to those reforms, agreed to by the Council of Australian Governments (COAG) in 2011, have not altered this fact. Nationally, data will become essential for system design, financing and accountability. Thus far, implementation of data standardisation and expansion of patient level collections have focused on public hospitals, as they are the most critical to the reforms.

    However, the reforms have implications also for the private sector. In particular, those reforms centred on accountability and transparency, make reference to the private sector and are likely to drive further development of private hospital data. In this context, it is noted that the 2008 National Healthcare Agreement1 includes indicators that apply equally to the private and public hospital sectors, as do those of the Australian Commission on Safety and Quality in Health Care (ACSQHC) indicators of patient and hospital level safety and quality. In addition, the National Health and Hospitals Network Agreement2 envisages that private hospitals will participate in the accountability and transparency reforms, including the hospital performance reports that will be prepared by the National Health Performance Authority (NHPA), at the hospital and hospital network levels.

    1 Council of Australia Governments (2008). National Healthcare Agreement (Intergovernmental Agreement on Federal Financial Relations).
    2 Council of Australian Governments (2010). National Health and Hospitals Network Agreement.

    Private hospitals themselves have expressed interest in participating in hospital reporting and some larger providers actively contributed to the recently launched myHospitals web site. Notwithstanding this interest, private hospital managers, their peak bodies and other interested parties have observed that such reporting incurs a cost to the hospitals. They have pointed to issues of the diverse range of reporting required of them by the Commonwealth, States and Territories, representative groups and others; duplication of information among mandatory collections; and unnecessary complexity of data supply chains.

    If reporting was rationalised, data duplications removed and data supply chains streamlined, the reporting burden on private hospitals would be reduced and more active involvement in accountability and transparency activities would follow.

    1.2   Focus of the Review

    1.2.1   Reducing Data Collection Burden

    The focus for Work Stream 1 was on those collections where the data collected ultimately ends up in the hands of the Commonwealth or in the hands of a Commonwealth agency or instrumentality. Thus the focus was on data collections that are national in scope. In addition, this work stream focused on reduction in data collection burden but not in the range of data items collected, for those collections deemed to be within scope.

    In considering the data collection burden, the Department asked KPMG to identify and review data collection processes both for those collections within scope and other collections operating at a local or state level. The purpose of this component of the Review was to gather a fuller picture of the overall data reporting burden borne by private hospitals. The Department recognised that the Review would not be able to make recommendations influencing the operations of those collections as lie outside the direct control and influence of the Commonwealth.

    Nonetheless, it was expected that the Review might comment on relationships between those collections and reporting processes and those within scope of the Review.

    1.2.2   Increased comparability of public and private sectors

    The focus for Work Stream 2 was on the same collections as Work Stream 1 – where the data collected ultimately end up in the hands of the Commonwealth or in the hands of a Commonwealth agency or instrumentality. In considering these collections, this work stream focused on identifying where data are collected in both sectors and there are issues with the comparability of the data between sectors, or where (national) data are collected for the public sector and no comparable data are collected for the private sector.

    In considering the comparability of data for the two sectors, the Department asked KPMG to identify how private sector data collections might be modified to move the private sector data closer (in terms of comparability) to public sector data.

    This work stream focused on the comparability of data collected (nationally) from the two sectors and the ability to directly compare measures based on such data. In doing so, there is a question of the relevance of other aspects of comparability – those that relate to the rationale for wanting to compare public and private hospitals for a given measure and those that relate to confounders.

    Confounders are factors that affect how measures should be interpreted or adjusted, such as the impact of the lower proportion of complex medical patients with multiple presenting comorbidities in the private system, relative to the public system.

    The other issues relate to fundamental issues of comparability between the two sectors. For example, there are fundamental differences in the operating models for public and private hospitals that affect whether meaningful comparisons of medical workforce can be made, using hospital level data.

    Confounders are more strongly related to the interpretation of comparisons between the sectors, rather than to questions of (technical) comparability. As such, this review has focused on the more fundamental issues of comparability that affect the rationale for wanting to compare the two sectors, which have greater relevance to decisions to change existing data collection arrangements.

    1.2.3   An authoritative list of private hospitals

    Currently, the Department makes use of different lists of private hospitals for different purposes. In addition, there are other lists being used for purposes outside of the Department as well. With the advent of the national health reforms and the increased emphasis on performance measurement and reporting, it is important that private hospitals are well defined and that there is a common understanding of what health services should be included and which should be excluded when considering the performance of the private hospital sector.

    Consequently, the Department had identified a need to rationalise the existing multiple lists and develop a mechanism for determining a single, authoritative list of private hospitals in Australia. That mechanism also should allow for ongoing maintenance of the list, at the least including processes to update information for existing hospitals, add information for new hospitals and to remove information for hospitals that cease to operate.

    Therefore, this Review needed to consider the purposes to which such a list would be put, the existing lists of private hospitals - both at national and State or Territory levels – and available mechanisms for obtaining and maintaining the necessary data on private hospitals.

    1.3   Methodology

    Figure 1 summarises the approach to this Review. The process for deciding which private hospital data collections, data supply chains and reporting processes are within scope was agreed in project initiation. Broadly, the process consisted of asking the following questions in relation to identified data collections:

    • Does this collection comprise data generated by and used by private hospitals and which ultimately ends up in the hands of the Commonwealth?
    • Do the data from this collection end up in the hands of the Commonwealth agency, department or instrumentality? Such agencies include the Australian Institute of Health and Welfare (AIHW), Australian Bureau of Statistics (ABS) and Federal Government departments.
    • Is the collection one that the Department of Health and Ageing is likely to be able to influence?

    Based on this clarification, the following collections were identified as warranting consideration for inclusion within scope of the Review:

    • the National Hospital Cost Data Collection (NHCDC);
    • the Hospital Casemix Protocol (HCP);
    • the Private Hospital Data Bureau (PHDB);
    • the ABS Private Health Establishments Collection (PHEC); and
    • the AIHW Admitted Patient Collection National Minimum Data Set (APC).

    Relevant collections implied by the work of the Australian Commission on Safety and Quality in Health Care (ACSQHC or the Commission) also were deemed to be within scope.

    With respect to the NHCDC, the processes and systems for collecting the NHCDC data are not within the scope of this Review. The scope will be limited to identifying overlaps with other collections and collection processes and opportunities for streamlining other collection processes to reduce double handling and incompatible data.

    Databases of claims for the medical component of private hospital care, held by the Department (Medicare Benefits Schedule claims) and by the Department of Veterans’ Affairs are out of scope.

    Throughout the stakeholder consultation process in Work Stream 1 two other national collections were often mentioned by stakeholders. These collections were:

    • Perinatal National Minimum Data Set; and
    • Cancer registry data, collated by the National Cancer Statistics Clearing House (NCSCH) from state cancer registries.

    Subsequently, these two collections were considered by the Review.

    Figure1 Overview of Review stages.
    Figure one shows a flow chart of the Review Stages including an information interchange between Work stream 1 and Work stream 2.

    1.3.1   Work Stream 1

    Following project initiation, there was a step to obtain and review documents relevant to the objectives for Work Stream 1. The types of documents reviewed include:

    • Commonwealth, State and Territory legislation (Acts, Regulations and other legislative instruments);
    • user manuals, guides, data dictionaries and other operational documents related to the data collections within scope (and those considered to be out of scope) for Work Stream 1;
    • working documents provided by stakeholders, including some not publicly available and others that had been published;
    • reports of previous reviews and of Government and non Government agencies; and
    • submissions to previous reviews and Government sponsored inquiries.

    In analysing this body of information for Work Stream 1, particular lines of enquiry were followed. These lines of enquiry and their relationship to each of the Review’s objectives relevant to Work Stream 1 are summarised in Table 1. This enquiry framework was applied with the focus on the data collections identified as within scope for this Review (see Chapter 2).

    Table 1 Enquiry framework used in analysing and synthesising information gathered for Work Stream 1.

    Objective Focus of enquiry
    Data collection streamlining Identify the major concerns and issues with respect to reporting and data submission burden in the private hospital sector.

    Understand the processes associated with reporting and data submission by private hospitals.

    Identify areas of duplication, overlap or inefficiency.

    Identify potential opportunities for improvement.
    An authoritative list of private hospitals Identify the reasons for such a list.
    Identify existing lists and their use, within the Commonwealth.
    Develop a definition of “private hospital” suited to those purposes.
    Identify potential sources of information for populating such a list.

    These documents were used to confirm the inclusion of the above data collections and also to identify other potential data collections for inclusion. In addition, the documents were reviewed for relevant information on those collections determined to be within scope of the Review.

    Subsequently, consultations were undertaken with a wide range of stakeholders. These stakeholders included private hospitals and their representative bodies, health insurers and their representative bodies, State and Territory governments, representatives of national agencies and Departmental officers. The full list of stakeholders consulted is provided at 4.4.

    Stakeholders were asked to provide their views and information relating to the data collections that should be within scope of the Review, other data collection and reporting burdens imposed on private hospitals, existing processes for data collection and opportunities for streamlining.

    Stakeholders were also asked about legislated and licensing requirements to which private hospitals are subject within different jurisdictions and their implications for both objective 1 and objective 3 of this Review.

    Throughout the consultation process additional documentation was identified, obtained and reviewed to complement and supplement information obtained through the consultations themselves.

    At the completion of the consultations and review of documentation, the full range of information found and reviewed was subject to analysis, synthesis and interpretation. Initial findings were tested selectively with stakeholders and with the Advisory Group formed by the Department for this Review. Potential recommendations were also tested in this way.

    The last step in the process for Work Stream 1 was the drafting of the draft work stream report, which summarised the findings and presents the recommendations relating to the data collection streamlining objective and to the authoritative list objective for the Review.

    1.3.2   Work Stream 2

    Following project initiation the first step involved developing a comparison framework that stepped out how the collections within scope were going to be compared. The details of the comparison framework are discussed further below (see Comparison framework, page 8).

    Next, there was a detailed, item by item comparison between data items listed in the APC and five of the private hospital collections within scope of the Review (HCP, PHDB, PHEC, Perinatal and Cancer collections).

    The comparison was done using an Excel© spreadsheet. For each item by item comparison seven attributes were considered. Those attributes were:
    1.   Existence of same item in each collection (yes/no).
    2.   Naming congruence (yes/no).
    3.   Definition congruence (yes/no).
    4.   Domain congruence (yes/no).
    5.   Ability to map data items (yes/no).
    6.   Required for reporting (mandatory/optional).
    7.   Commentary on particular issues with the item.

    The spreadsheet containing the results of this comparison has been provided to the Department separately from this report.

    The next step involved obtaining and reviewing data collection manuals, survey forms and data set specifications relevant to the objectives for Work Stream 2. In this step the main reports where public and private hospital sectors are compared were reviewed. This included AIHW’s Australian Hospital Statistics and the Productivity Commission’s Public and Private Hospitals reports. The reason for this was that these reports had already highlighted a number of issues with comparing the public and private sectors.

    Subsequently a workshop was run on 8 June 2011 with a number of stakeholders. These stakeholders included private hospitals and their representative bodies, health insurers and their representative bodies, representatives of AIHW and ABS and departmental officers. The full list of stakeholders consulted is provided in Appendix B.

    Stakeholders were asked to provide their views and information relating to the data collections within scope of the review in relation to improving comparability between the public and private sectors. They were provided with three areas identified to improve comparability and to prioritise these options and also suggest possible convergence plans in these areas. Stakeholders were also asked to suggest other mechanisms to improve comparability between public and private sectors.

    Discussions at the workshop often focussed on streamlining of data collections rather than improved comparability. Of the three areas identified to improve comparability stakeholders from private hospitals and insurers identified that obtaining data on private patient stays in public hospitals was the number one priority. The other areas of improving comparability were in terms of cost and reporting of private hospital establishment data in Australian Hospital Statistics publication.

    For improving comparability in the domain of cost, stakeholders did not identify this as a priority. They stated that comparability within the sectors was more important than trying to achieve comparability across the public and private sectors. For improving the reporting of private hospital establishment data all stakeholders did not view the issue as a priority but they agreed that AIHW and ABS should continue to work together to resolve the issue.

    At the completion of the workshop the full range of information found and reviewed was subject to analysis, synthesis and interpretation. The last step in the process for Work Stream 2 has been the drafting of this report, which summarises the findings and presents a feasible data convergence plan for data collections from public and private hospitals.

    Different stakeholders see different potential benefits from increased capability to directly compare public and private hospitals on a range of performance measures. Governments, insurers and other purchasers of public and private hospital services3 have a strong interest in being able to compare treatment outcomes and treatment costs when devising policy or making purchasing choices. The hospitals themselves and their operators – both public and private – are interested in benchmarking their own performance against peers, to better understand the quality of their services and where they might improve that quality.

    3 For example, self insured individuals purchase treatment as a private patient in public and private hospitals; the Department of Veterans’ Affairs purchases hospital services in both sectors; and State and Territory governments contract private hospitals to provide treatment for public patients.

    Comparison framework

    In carrying out the detailed comparisons of collections, 6 dimensions for comparison were used:

    • scope;
    • definitions;
    • counting rules;
    • units of measurement;
    • costs; and
    • practicability of change.

    Each of these dimensions is described further in the following sections.

    Scope of the Collection

    This dimension required consideration of the following attributes of the collection:

    • defining the hospitals that contribute or are required to contribute data;
    • the frequency of collection or data submission;
    • the basis for participation, such as whether or not participation in the collection was required under legislation or under licensing conditions; and
    • the area of hospital operations covered by the collection, such as admitted patient treatment or financial management or workforce.

    Data Item Definitions

    This dimension involves sourcing and detailing the various data dictionaries and collection manuals for each of the private hospital data collections. The collections within scope collect a large amount of personal, episode, establishment and financial information. These dictionaries and manuals were used for comparing all the items collected in the AIHW Admitted Patient Care NMDS to each of the private hospital collections (apart from the NHCDC).

    This dimension involved sourcing the data dictionaries and collection manuals. These dictionaries were then used in the detailed item by item comparison between the AIHW Admitted Patient Care NMDS and the private hospital collections within scope.

    Counting rules affecting data items

    To compare the information collected it is important to understand the counting rules. This involves identifying what is the lowest level of information collected or the “statistical unit.” For example, the statistical unit for the AIHW Admitted Patient Care NMDS has a statistical unit defined as “Episodes of care for admitted patients” while the Cancer registry has a base statistical unit of a “tumour.”

    This dimension identified the “statistical unit” of each collection.

    Units of measurement, including code sets

    There are a number of classifications that are quite particular to health collections. These include classifications such as ICD-10-AM, which refers to the Australian modification of the WHO ICD-10 base classification system. ICD-10-AM is a classification of diseases and health problems. Another classification is the Australian Classification of Health Interventions (ACHI) 5th edition. There is also the Australian Refined Diagnosis Related Groups (AR-DRGs), which is a patient classification system that provides a clinically meaningful way of relating the types of patients treated in a hospital to the resources required by the hospital.

    Classifications as those listed above are subject to regular revisions and jurisdictions may not always be using the same version of these classifications. Occasionally jurisdictions may make slight modifications to these collections.

    This dimension identified the main health classifications being used by each collection and which version of the classification the collection used (or if the collection accepted many versions).

    Costs of collection and changes to collection

    The data collections utilise a variety of collection methods. The ABS PHEC only collects information from private hospitals. This is done via both a survey that a private hospital responds to and a submission of data from the health department in the jurisdiction in which the private hospital is located. However the AIHW APC NMDS collects information from both public and private hospitals via the health department in each jurisdiction.

    This dimension identified some of the main costs of the collection. The costs were not detailed in dollar amounts but by broadly identifying the main costs in terms of participating in the collection.

    Practicability of changing the collection

    Many of these data collections have been running for over a decade. For example, the ABS has been collecting data from private hospitals and jurisdictions for the PHEC since 1996-97. Not only do jurisdictional departments have established systems in place to supply private hospital data to these organisations they also have well established working relationships with ABS.

    This dimension identified some of the barriers and enablers to changing the collection.