Private Hospital Data Collection Review Final Report

3.1 Findings

Private Hospital Data Collection Review - Final Report

Page last updated: 14 June 2012

      Figure 2 and Figure 3 illustrate the data flows for private hospitals collectively and the data burden for an individual private hospital, with respect to the data collections within scope of this Review. They serve to highlight the reporting burden that these collections represent for the private hospital sector.

      Figure 2 Overview of data collection landscape for private hospitals.
      Figure 2 shows an overview of data collection landscape for private hospitals.

      Figure 3 Data outflows for a private hospital.
      Figure 3 shows the data outflows for a private hospital.

      On 8 June 2011 a workshop was held in Canberra to discuss private hospital and public hospital comparability (Work Stream 2). Workshop participants (see Appendix B) were sent a background information kit, which consisted of Section 3.2 and a summary of work to date. The workshop involved a discussion of the project, a background on work to date and also an in-depth discussion on identified areas for improved comparability between public and private hospital data.

      Participants were asked what outcomes they wanted to see from the workshop. Participants identified that in relation to public and private hospital data collections they wanted:

      • simplification;
      • streamlining;
      • wanted to make sure that the veracity of the data that is currently collected is not lost;
      • decrease duplication;
      • improve completeness of data;
      • increase accuracy;
      • decrease burden, and time taken to comply; and
      • demonstration that there is value in the data that is being collected, and in conducting the review.

      Many of these outcomes relate to streamlining (Work Stream 1 – such as simplification and streamlining) and not directly to the goal of improved comparability. Others are directly related to comparability (for example, completeness of data and accuracy). Within one workgroup at the workshop, participants were using “comparability” and “streamlining” interchangeably.

      Importantly, a significant number of participants noted the inherent difficulty of comparing the two sectors and whether comparing the two sectors should be a priority. In making these observations, the threads of the discussion were that current levels of comparability are adequate for the stakeholders’ purposes; the effort involved in delivering comparability in some areas is not warranted; and the fundamental operating models for the two sectors make pursuing some comparisons pointless or render it impossible to resolve differences in counting rules.

      3.1.1   Cancer data

      There was a general acceptance that the processes for collecting cancer data were largely independent of other processes and there were no opportunities for substantial streamlining in relation to those national collections.

      There is a standard, minimum data set defined by the ACD data items92. These items are provided for each cancer case (live case or cancer related death), by each State or Territory cancer registry. At the State or Territory level, case records are produced by collating and summarising data from multiple sources – pathologists, medical specialists, public hospitals, private hospitals and others. The range of data items required at this level is determined by legislation and supporting policy directions as is, to varying degrees, the process for submitting the data.

      92 National Cancer Statistics Clearing House Protocol, 2009. Australasian Association of Cancer Registries and Australian Institute of health and Welfare.

      As a result, the range of data items required by each State and Territory usually exceeds the minimum national requirement. Large private hospital providers with hospitals operating in multiple jurisdictions are confronted with developing separate data extraction procedures for each jurisdiction.

      However, the scope for standardising procedures and formats for submitting private hospitals’ data across States and Territories is limited, at best. This Review also notes those processes are not within the direct influence of the Department and the Commonwealth more generally.

      Top of page

      3.1.2   Perinatal data

      As with cancer data, there was a general acceptance that the processes for collecting perinatal data were largely independent of other processes and there were no opportunities for substantial streamlining in relation to the NPSC.

      This observation must be qualified with the observation that it applies to the national level and does not apply to the same degree at the State and Territory level. There is variation among States and Territories in the data items required by perinatal statistics units at that level. In addition, there is variation in the level of information technology used in each jurisdiction. For example, some jurisdictions still use a paper based form for data submission to the perinatal collection.

      There is a Perinatal National Minimum Data Set93. States and Territories collect additional data items beyond what is required for the Perinatal National Minimum Data Set. There is also variation in the forms used in different states for collecting perinatal data. Large private hospital providers with hospitals operating in multiple jurisdictions are confronted with developing separate data extraction procedures for each jurisdiction.

      93 Perinatal NMDS 2011-2012. Australian Institute of Health and Welfare.

      There may be a case for undertaking a comparative review of State and Territory perinatal reporting requirements and forms, with a view to identifying opportunities for reducing the variation among them. There may also be opportunities for States and Territories to share technological solutions developed locally, to make submission by large private hospital providers easier.

      However, this Review notes that these are State and Territory level issues and beyond the direct influence of the Department and the Commonwealth more generally.

      3.1.3   NHCDC

      The purpose of the NHCDC is to produce benchmark data for use by hospitals so that they can compare their costs to other similar hospitals. The NHCDC also produces national cost weights for AR-DRGs and other statistics relevant for hospital service costing and planning.

      The main concerns with respect to the NHCDC related to the issue of data quality, rather than to the burden associated with data preparation and submission. This partly reflects the voluntary nature of NHCDC participation for private hospitals.

      A link to reporting burden arose in this context when considering how data quality might be improved for the private hospital component of the NHCDC. The main mechanism for improving data quality was seen as being capture of patient level cost or resource use data, through the implementation of “feeder systems”. However, stakeholders saw no intrinsic business case for a private hospital to invest the cost and effort to adopt such systems and modify their business processes to apply them.

      There are a number of differences which prevent proper cost comparisons between private and public sectors. The key difference in the reporting of costs between public and private sector hospitals in the NHCDC:

      • predominance of ‘cost modelling’ to produce cost estimates in the private sector, in comparison to ‘patient costing’ for the majority of public hospitals;
      • treatment of teaching, training and research costs;
      • differing admission practices within and between sectors;
      • pharmacy costs for private hospitals;
      • prostheses costs; and
      • public and not-for-profit private hospitals are partially exempt from paying fringe-benefits tax (FBT) and are not required to pay payroll tax.94

      Top of page

      94 Productivity Commission 2009, Public and Private Hospitals, Research Report, Canberra.

      Not only are there the above mentioned differences but the process of supplying data to the NHCDC process differs for public and private hospitals. The primary difference being that public hospitals submit their data to a State/Territory Health department officer to assess the data before it is sent on to the Department. Whereas private hospitals are required to undertake this quality assessment process themselves.

      However this work focuses on obtaining private hospital establishment data for the Australian Hospital Statistics publication. It is about obtaining current hospital resources information. Currently that report uses PHEC data from the previous financial year for private hospitals, and uses NPHED data from the report’s financial year for public hospitals. That is, the 2009-10 AHS describes private hospitals in terms of PHEC data from 2008-09. Compounding this is the proximate timing of release of the AIHW’s report and the ABS report from PHEC. Australian Hospital Statistics 2009-10 was released on 29 April 2010 while the PHEC output was released 7 weeks later, on 17 June 2010.

      Stakeholders generally were comfortable with the notion of AIHW operating as the collection agency for PHEC or an equivalent collection. The private hospital sector felt that the AIHW's governing legislation offers the same level of protection of the data as does that of the ABS. They and other stakeholders felt that the AIHW has in place well established and effective protocols for providing different levels of access to the data, while protecting privacy and confidentiality.

      This support for the AIHW operating the PHEC or replacing it with a similar collection was qualified in several ways, as follow:

      • the range of data collected and definitions used would need to be maintained, so that the integrity of time series data was preserved;
      • the coverage of private hospitals contributing to PHEC would need to be maintained or improved over the current high level achieved by the ABS;
      • the ABS would need to have access to at least those aspects of the PHEC data required for national accounts and other reports relying on the PHEC data other than private hospitals Australia reports; and
      • private hospitals would need to have control over the level of access provided to their individual data in a similar way to the protection is currently afforded States and Territories in relation to the APC96. It may be possible to develop a standard agreement – analogous to the National Health Information Agreement – that could be used to effect such control.

      Top of page

      96 The APC data are provide to AIHW for specific national reporting purposes. From time to time, AIHW makes data available at a level that would allow the identification of States and Territories. Before allowing this to happen, the AIHW requires the consent of all States and Territories to the use of the data for the desired purpose. The process of obtaining such consent is well established and accepted by all parties concerned as offering both access to the APC data for legitimate research and policy analysis purposes and reliable protection of the privacy and confidentiality of the data. This arrangement is governed by the National Health Information Agreement.

      3.1.5   Safety and quality data

      Safety and quality reporting burden varies enormously across the private hospital sector in Australia. As discussed in Chapter 2 some States and Territories impose significant reporting requirements specific to safety and quality while others impose very few. In addition, health insurers impose their own requirements for reporting related to quality of care. This requirement is imposed through conditions inserted into contracts between private hospitals and health insurers.

      The requirements of insurers, in relation to data and reporting for quality and safety related purposes vary enormously. There is variation in the specific data and supporting information requirement, in specifications for data items sought (for example, one insurer may require address as a single, free text field, while another requires it to be split into several fields) and the frequency with which data are required (for example, some require data quarterly while others require it monthly). A number of insurers also require an annual summary report to be provided by a private hospital. The level of detail and range of coverage required in this report also varies significantly with some insurers requiring only a page or two and others requiring up to 40 pages of information.

      The new accreditation arrangements are likely to be implemented across all hospitals and day procedure services by January 2013. The ACSQHC is hopeful that this will lead to a greater sharing of common safety and quality data which will allow health insurers to reduce their demands for safety and quality reporting from private hospitals. At the least, the existence of national standards and indicators for safety and quality reporting will permit insurers to consider the standardisation of such reporting within their part of the industry.

      The details of national performance monitoring and reporting of safety and quality accreditation are, as yet, not determined. It is unlikely that there will be a single national regulator for accreditation as State and Territory Health Departments will regulate the public sector and it is yet to be determined who will regulate the private sector. Whatever the detail of resolution, the regulators will receive accreditation outcome data from accrediting bodies on health services. If there is to be a national collection, that may well fall to the NHPA.

      Top of page

      Representatives of the private hospital sector have been closely involved in the development of the accreditation reforms. Nonetheless, they are clear on the view that the private hospital sector has not been party to, nor is it a signatory to the national health care reforms and agreements that have been negotiated among Commonwealth, State and Territory governments.

      This Review found that there is significant uncertainty about future data collection from private hospitals related to safety and quality. The ACSQHC and AIHW have recently undertaken a project to populate safety and quality indicators developed by the ACSQHC. This project required access to hospital level data in the APC. In order to provide this level of access, the AIHW approached private hospitals, through States and Territories, for their consent.

      While hospitals generally were happy to provide their consent, States and Territories were prepared to provide their supporting consent only if they received copies of the results for each private hospital within a jurisdiction. Private hospitals were concerned that allowing this would lead to their individual data becoming subject to freedom of information legislation within States and Territories. They saw this as risking commercially sensitive information falling into the public arena. Consequently, they chose to withhold their consent.

      This experience highlights the following issues:

      • ownership and custodianship of private hospital admitted patient data, which is provided to the APC by the jurisdictions, but "owned" and generated at source by private hospitals with differing jurisdictional arrangements;
      • technical difficulties in identifying individual private hospitals within the APC, in order to generate nationally risk adjusted indicators through efficient central mechanisms; and
      • the lag in achieving centrally collated, national data (somewhere between 5 and 17 months) for efficient and consistent central generation of a series of measures.

      It also suggests that the national reporting of private hospitals performance in other operational areas may be similarly difficult to implement.

      In spite of these practical issues, the development of consistent and well-specified flows that support the generation of well designed measures will potentially provide insurers and other payers with a standard suite of well designed measures. In turn, this may create opportunities to reduce the burden for private hospitals of providing different reports – in terms of scope and presentation – to different insurers.

      Top of page

      3.1.6   HCP, PHDB and APC

      There is variety in the effort involved among hospitals in the process of extracting and providing data for the different collections in scope. The variation largely related to issues of information systems’ capability, economies of scale and access to suitable information management skills. Some stakeholders referred to “up to 4 days” of effort required to produce HCP and PHDB extractions while others stated that it required “the push of a button”.

      A consistent view from stakeholders was that there is a subset of day hospitals that simply lack the scale and information systems to collect and provide the data sought by collections such as HCP, PHDB and APC. These day hospitals are often, but not always, small scale providers, whose data collection is limited to what is required to provide good clinical care and to manage the business aspect of the hospital. In some cases, these are day procedure centres providing services for which no insurance claims are made. As such, they may not be declared hospitals under the Commonwealth Private Health Insurance Act 2007, thereby not being subject to HCP and PHDB requirements.

      Overview of HCP, PHDB and APC data flows

      The general view was that the data collections themselves are warranted and that the range of data items collected should not be reduced. However, all stakeholders expressed concern over the fact that the HCP, PHDB, and the APC collect a significant set of common data items with common definitions and code sets. They also expressed concern over the substantial effort required to generate monthly extracts for these collections.

      On a monthly basis a private hospital within any jurisdiction in Australia will submit a large amount of health data to insurers, State and Territory health authorities and the Department. Figure 4 illustrates the processes for the submission of data to the APC – via jurisdictional admitted patient collections – and the HCP and PHDB collections. The figure excludes the state based perinatal and cancer registry collections (see Section 3.1.1).

      Top of page

      Figure 4 also does not illustrate another issue with participating in these data collections, which is the validation of the monthly data submission. This is where the private hospital receives an “edit report” or similar about their submission. This edit report will specify whether submitted episodes are either rejected, flagged as a warning or accepted based on the business rules of the organisation to which they are submitted.

      For example, in Victoria there are 403 different edits or “business rules”. These edits are classified as rejection, fatal, warning and notifiable. Examples of these edits include where a hospital submits an invalid Medicare number (rejection) or a 14 year old is listed as being married (warning). A rejection edit requires the hospital to check, correct and re-transmit that particular episode97. This is not a simple task for the hospital.

      97 It is important to note that this describes what happens only in Victoria, should a record fail validation check and be rejected. That is, only the rejected episode record need be resubmitted once it has been corrected. In Queensland, if a single episode record is rejected then the entire batch of records must be resubmitted once that error has been corrected.

      In addition insurers may have similar or different business rules to each other and to State health authorities. This means an episode of care may be accepted by an insurer but rejected by the State health authority. Private hospitals also reported instances of insurers advising of rejected record is six months after the original data had been submitted. Understandably, it is much more difficult to go back and correct an error for record that is more than six months old. This was not an uncommon occurrence.

      These difficulties are clearly more burdensome for those corporate private hospital providers who operate private hospitals in multiple jurisdictions than they are for individual private hospitals that operate in single jurisdictions. Nonetheless, they are the sources of significant inefficiency in the national private hospital data collection processes.

      Figure 4 Overview of HCP, PHDB and APC data submission processes.
      Figure 4 shows an overview of HCP, PHDB and APC data submission processes.

      Data overlap

      The other main area of inefficiency in relation to these three data collections is the area of data overlap. There are two main differences between the PHDB and the HCP. They are the scope of the collections and four data fields. A PHDB submission contains all private hospital separations while an HCP submission contains data only for insured patients for whom a benefit is being claimed.

      Table 6 summarises the key differences between the HCP and PHDB data specifications. Aside from these key differences, the two datasets capture essentially the same data items, with some minor differences in code sets.

      Top of page

      Table 6 Differences between the HCP and PHDB.

      Field No or issue HCP PHDB
      1 Insurer Membership
      Identifier – valid value added
      Insurer Membership Identifier – blank filled
      2 Insurer identifier – the health fund registered three character code. Example:
      AHB - Defence Health
      AUF – Australian Unity
      Etc.
      Payer Identifier – indicator of the type of funder of the episode:
      IH – Insured with Agreement with Hospital
      IN – Insured with no Agreement with Hospital
      SI – Self Insured
      WC – Worker’s Compensation
      TP – Third Party
      CP – Contracted to Public Sector
      CV – Department of Veterans’ Affairs patient
      DE – Department of Defence patient
      SE - Seaman
      OT - Other
      3 Family Name Family Name – Blank filled, as not required for reporting to DoHA
      4 Given Name Given Name – Blank filled, as not required for reporting to DoHA

      Both the PHDB and the HCP require data to be submitted in a fixed field, ASCII file format. The specified format for the PHDB is different to that for the HCP, in spite of the fact that they have so many items in common.

      The use of a fixed, ASCII format also means that when changes are made to the HCP or PHDB requirements then the process of modifying data extraction procedures is more difficult than would be the case if a more flexible data interchange format was used.

      There is a significant amount of overlap between the PHDB and APC (see Table 7). Very few items included in the APC are not also included in the PHDB specification or able to be derived from PHDB items. For example, APC contains an Australian State or Territory identifier for the hospital, whereas PHDB contains a private hospital provider number that can be used to derive State or Territory for the hospital.

      There are a large number of items within the PHDB that are not available in the APC. By far the majority of these items relate to private hospitals’ charges and supporting information for different types of specialist care. For example, 3 of these items relating to Special Care Nursery – Special Care Nursery Charge, Special Care Nursery Days and Special Care Nursery Hours.

      Table 7 Relationship between PHDB and APC data items.

      Items common to APC and PHDB 18
      Items mappable from PHDB to APC 15
      Items in PHDB and not in APC 43
      Items in APC and not in PHDB 4

      Top of page

      Missing private hospital identifiers in APC

      The other key difference between PHDB and APC data items is the fact that some States and Territories do not provide private hospital identifiers when submitting data for private hospitals within their jurisdiction. These jurisdictions maintain that the basis on which they collect and hold the private hospitals’ data is as custodians and that the data remain the property of the private hospitals. As such, they are not willing to provide the information that would allow individual private hospitals to be identified within the admitted episode data submitted to the APC.

      These jurisdictions also are not willing to provide a data linkage field that would allow episodes from the same hospital to be identified as such, without explicitly identifying the hospital itself. This reflects the view that it would be possible, by comparing episode profiles for each hospital in the jurisdiction, to then identify individual hospitals.

      Relative coverage

      Comparing the most recent complete year of data for PHDB and APC with the corresponding year’s data for PHEC shows that both PHDB and APC under enumerate numbers of private hospitals and of total admitted private hospital activity in Australia (see Table 8).

      The PHDB applies only to private hospitals declared under the Private Health Insurance Act 2007, whereas APC and PHEC both aim to capture data from all licensed private hospitals in Australia. The differences between APC and PHEC coverage are due to private day hospitals in the ACT and NT not submitting data to the APC and to 2 other hospitals not providing data for and NT not submitting data to the APC and to 2 other hospitals not providing data for 2009-10.

      Table 8 Comparison of private hospitals and admitted patient separations covered for 2009-10: PHDB, APC and PHEC.

      Data collection 2009-10 separations 2009 -10 private hospitals
      PHDB98 2,599,163 Approximately 570
      APC99 3,461,715 573
      PHEC100 3,590,800 581

      98 PHDB Annual Report 2009-10. Australian Department of Health and Ageing.
      99 Australian Hospital Statistics, 2009–10. Australian Institute of Health and Welfare. April 2011.
      100 Private Hospitals, Australia 2009-10. Australian Bureau of Statistics. Catalogue no. 4390.0, released 17 June 2011.

      Top of page

      ECLIPSE

      Some stakeholders pointed out that an opportunity exists to submit data via the ECLIPSE claiming system. ECLIPSE stands for Electronic Claim Lodgement and Information Processing Service Environment and is an online claiming system developed by Medicare Australia101. It is used by private hospitals to lodge claims electronically with a health insurer and facilitates the checking of eligibility and payment of the claim by the insurer.

      101 ECLIPSE

      ECLIPSE contains within its file specification the HCP data specification. This was originally included in 2006, when ECLIPSE was being considered as a mechanism for collecting HCP data. It has not become a tool for submission of HCP data, and so this part of the record specification remains unused. In addition, the HCP specification within the ECLIPSE record has become outdated and no longer matches the current HCP specification.

      According to Medicare Australia all health insurers are using ECLIPSE for online claiming and eligibility checking. This does not mean that all insurers are using it for private hospital claiming, as ECLIPSE can also be used for online claiming of services delivered by other providers such as dentists, physiotherapists and medical specialists. It also does not mean that an insurer who is using ECLIPSE for private hospital claims is using it with all private hospitals with whom it deals. For example, one major health insurer informed this Review that around 50% of its private hospital claims are handled using ECLIPSE. The DVA also indicated that around 50% of claims from contracted hospitals and day procedure centres are handled using ECLIPSE.

      Between May 2010 and April 2011, Medicare Australia processed approximately 588,000 in-hospital claims. While some of these claims may have been for hospitals other than private hospitals, it is likely that the majority related to private hospital episodes. Though some recently available data on hospital separations in Australia, this suggests that around 19% of insured private hospital episodes are claimed using ECLIPSE102. Health insurers are actively pursuing further roll-out of ECLIPSE for managing claims lodged by private hospitals generally.

      Top of page

      102 Australian Hospital Statistics 2009-10. AIHW. In 2009-10. There were 2,767,947 insured episodes and 199,732 DVA episodes in Australian private hospitals. The publication also indicated that private hospital activity had been growing at around 3.2% per annum leading into 2009-10.

      Any consideration of ECLIPSE as a preferred mechanism for collecting HCP data must consider issues of intellectual property and the possibility that other online claiming solutions may be preferred by insurers or private hospitals. The intellectual property inherent in ECLIPSE belongs to the Commonwealth, through Medicare Australia103.

      103 From 1 July 2011, Medicare Australia will cease to exist as a separate agency and will become part of the Australian Department of Human Services. The intellectual property held by Medicare Australia will remain with the Commonwealth.

      The marketing of ECLIPSE as an online claiming solution has been the subject of a previous court case, brought by the distributor of a competing online claiming solution against Medicare Australia, alleging contravention of the Trade Practices Act104. A finding against Medicare Australia would have required a fee to be charged for use of ECLIPSE. That case was settled out of court in October 2009 and at present ECLIPSE remains freely available for use by insurers and hospitals. However, there remains a possibility that other competing solutions for online claiming may enter the marketplace in the future and this legal question to be revisited.

      104 Budget Strategy and Outlook: Budget Paper No. 1, 2009‑10. Commonwealth of Australia, 2009. Canberra:CanPrint Communications.

      HCP data for privately insured patient stays in public hospitals

      Workshop participants (see Appendix B) identified the need to obtain information on privately insured patient stays in public hospitals as top priority among the comparability issues. The clear consensus was that public hospitals should be providing this information; there has been continued growth in these separations from public hospitals (see Figure 5); and a plan needs to be devised to obtain this information.

      An option was suggested at the workshop, to link AIHW inpatient data with insurers’ claims’ data. However, this approach was unlikely to succeed, as jurisdictions would be unlikely to provide the blanket permission required to use the AIHW data in this way, and the AIHW data lack the necessary identifiers to effect reliable linkage.

      The preferred approach suggested by participants was for the Department to work closely with a jurisdictional health department and an insurer to undertake a data linkage exercise using episode and patient unit record numbers rather than just date of birth, gender, hospital, admission and separation dates.

      Top of page

      Figure 5 Privately insured patient stays in public hospitals 2005-06 to 2009-10.
      Figure 5 shows the privately insured patient stays in public hospitals 2005-06 to 2009-10.

      The following quote from the reporting requirements for HCP and PHDB highlights the issue with the lack of data for privately insured patient stays in public hospitals.

      “It is also expected that hospitals will work toward providing health insurers with data that complies with the HCP. No timeframe has been set for public hospitals to provide health insurers with a complete HCP dataset as per the specifications. This recognises the needs of both parties and allows additional time for hospitals to build a complete data provision capability, whilst continuing to provide existing base levels of data. DoHA expects public hospitals and health insurers to continue working towards providing the full HCP dataset.”

      Table 9 highlights the number of privately insured patient stays in public hospitals. The Private Hospital Insurance Act 2007 requires HCP data be submitted for any and all admitted patient episodes where an insurance claim is lodged with an insurer. This requirement is imposed on all declared hospitals105. Under this Act both private and public hospitals may be declared.

      105 Section 121-5 of the Private Hospital Act 2007 states “,em>A hospital is a facility for which a declaration under subsection (6) is in force.” and Section 121-6 states “The Minister may…declare that a facility is a hospital”. As such, a “declared hospital” is one that has been declared by the Minister under this Act.

      However, the HCP admitted and rehabilitation data requirements are usually less onerous for public hospitals than for private hospitals. Historically, public hospitals usually do not have service contracts with insurance funds and so are able to claim only the default benefits for insured patients. As a result, public hospitals have only provided that subset of HCP data required for payment of default benefits, meaning that much of the detail of the patient episode is not provided. When the Private Hospital Insurance Act 2007 was drafted, States and Territories were not prepared to agree to provide full HCP data for public hospitals. Consequently, public hospitals continue to provide a less complete HCP data set to health insurance funds, although the long-term aim is for public hospitals to move to full HCP provision106 as noted in the quote above.

      Top of page

      106 Reporting Requirements for Hospital Casemix Protocol (HCP, HCP1 & HCP2), GT-Dental And Private Hospitals Data Bureau (PHDB).

      Table 9 Separation by funding source107.

      Separation Funding Source: Private health insurance Percentage
      Public Hospital 501,819 15.3%
      Private Hospital 2,767,947 84.7%
      Total 3,269,766 100.0%

      107 Australian Institute of Health and Welfare 2011. Australian hospital statistics 2009–10. Health services series no.40. Cat. no. HSE 107. Canberra: AIHW.

      3.1.7   Governance of data development and implementation

      There have been issues with disconnection between data development and implementation of APC NMDS items, and operating constraints for the private hospital sector. This Review found the issues most likely have arisen because the data development and implementation processes largely have focused on the public sector and not adequately dealt with factors unique to the private sector.

      There are 2 key factors that differentiate the private sector from the public, in terms of implementing changes to data standards and systems. The first is that costs of implementation for private hospitals and insurers need to be recovered through fees and charges and the second is that contractual arrangements dictate the data that private hospitals need to collect and report to insurers.

      Recently, an issue arose with disconnection between implementation of changes in the code set for ICD-10 AM under the APC NMDS and the private hospital sector operating environment. Private hospitals are bound to provide insurers with diagnostic and procedural information according to the AR-DRG versions, ICD versions or MBS codes specific in their contracts. When the most recent changes to ICD-10 AM were implemented, these led to AR-DRG changes for some patients. In turn, this would have resulted in reduced revenue for those patients under existing private hospital contracts.

      As a result, significant work had to be carried out by private hospitals and by jurisdictional staff to develop a “work around” that allowed the hospitals to submit different data to different recipients. Finding ways to avoid the need for this sort of effort represents an opportunity for streamlining data collection from private hospitals.

      Risks of similar situations arising in the future would be reduced if there was greater involvement of private hospitals and insurers in national data development work and implementation planning. Such involvement, particularly of insurers, would also place the sector on a more informed footing, which could allow greater consideration of data related issues when insurers and private hospitals are negotiating contract terms and conditions, further reducing risk of future problems.

      Top of page

      3.1.8   An authoritative list of private hospitals

      There are considered to be at least five lists of private hospitals in use within the Department and federal agencies:

      • private hospitals declared under the Private Health Insurance Act 2007;
      • private hospitals that submit data to the HCP collection108;
      • the AIHW list of private hospitals that contribute to the APC;
      • the list of hospitals within local hospital networks109; and
      • a list of private hospitals that contributes to the myHospitals website.

      108 Note that this is a list that also includes public hospitals which submit data to the HCP collection.
      109 Note that this is a list that includes both public hospitals and private hospitals within the hospital networks.

      The need for an authoritative list is re-inforced by the fact that the national health-care reforms herald greater transparency and visibility of hospital performance. As a consequence, there is an expectation that there will be monitoring and reporting of performance of hospitals in both public and private sectors110. A fundamental prerequisite for such transparency is a common an agreed understanding of what private hospitals are delivering services across Australia.

      110 National Health Reform Amendment (National Health Performance Authority) Bill 2011. Parliament of Australia.

      Departmental stakeholders agreed that any authoritative list should comprise those hospitals that are declared under the Private Health Insurance Act 2007.

      When the private hospital is declared under that Act, the Department obtains a set of information about that hospital (see Section 2.3.1). The concern for the Department is that the information may change over time but the Department has no formal mechanism for updating hospitals' details when this happens. Consequently, changes in hospital name, profile of services or other key details are often out of date within its list of declared hospitals.

      As discussed in Section 2.3, the Department does obtain updated hospital information from States and Territories on an informal basis. One possibility is that States and Territories may be willing to establish a formal arrangement for providing the Department with updated details for private hospitals. Indeed, Western Australia indicated that it would be happy to enter into such a formal arrangement. However, the variability of licensing regimes across jurisdictions and differences in things such as classifications of clinical service profiles means that this does not present a reliable means for keeping information in the Department's list of declared hospitals up-to-date.

      This state of affairs is exacerbated by the fact that some jurisdictions do not license all declared private hospitals. For example, South Australia does not license day hospitals and so can provide no information on those hospitals.

      This Review also found that there is no obvious mechanism provided in the Private Health Insurance Act 2007 that would allow the department to demand private hospitals informed the Department when key details change.

      Top of page

      3.1.9   Other areas identified for improved comparability

      Other areas identified in the workshop as potential avenues to improved comparability were the governance arrangements for the HCPWG and PHWG, as well as establishing a clearing house for data collection. The workshop’s participants felt these avenues, in the long term, would lead to improved comparability between the public and private sectors.

      The issue of governance of the HCPWG and PHWG working groups primarily related to the fact that the two groups do not meet to consider matters of common interest. Participants suggested that at least once a year the groups should meet.

      A clearing house for data would act as a recipient of data from public and private sectors. However there was not a great deal of detail suggested for this notion, nor for how it would lead to improved comparability between public and private sectors.