The National Health Information System in New Zealand[a]

• Abstract
• Success and failure in information systems
• Health informatics and information management
• Gaps in the jigsaw puzzle
• Existing sources of data
• Contract-based data
• Data processed from manual sources
• Claim-based data
• Making use of existing data
• Multi-dimensional approach – a template for success
• Barriers and enablers
• Conclusion
• References

Abstract
This paper examines some of the difficulties faced in creating successful information systems in general, and national health information systems in particular. The value and scarcity of the health informaticist is discussed. New Zealand’s existing national data sources are described, gaps identified and ways in which these data are being used to develop outcome measures are discussed. The paper concludes with a recommendation to make the best use of existing data sources before seeking to create new ones, and highlights some difficulties that may be encountered in attempting this.

Success and failure in information systems
Worldwide, it has proved difficult to identify the value realised from the development of information systems,^[1] and the health sector is no exception. In fact, the health sector has been referred to as “still a rather immature user of information technology compared to other parts of society”.^[2] Others have observed that innovation tends to proceed relatively slowly in health care generally and, since the implementation of information systems is often associated with innovation, this factor may be one of the reasons for the difficulty in realising the value of information systems.

A scan of the history of health information systems in New Zealand is no more encouraging. The Wave Report, published by the Ministry of Health (MoH) in 2001, outlined significant structural issues for New Zealand health information and made 79 recommendations along eight separate work streams. Such a broad range of issues cannot concentrate on any single area in detail and the Wave project did not greatly explore the causes of failures to implement information systems strategy, which largely stemmed from progressive under-funding. The writers seemed to be aware of potential problems with strategy implementation when they indicated that, were recommendations not acted upon rapidly, “the sector [would] be writing another large-scale sector information strategy in five years time”.^[3]

Unfortunately, it is precisely this desire to make significant change in a short period that often creates problems, and, of course, it is easier to devise strategy than to implement it. It has been observed that the Wave report reiterated issues delineated in earlier strategy documents, and the Wave Report itself notes the health sector is “awash in strategy”.^[4] Almost five years have elapsed since the publication of the Wave Report and, although progress has been made according to the Office of the Auditor General it has been slower than anticipated and much work still remains to be done to fully implement the Wave recommendations.^[5]

It has been acknowledged that successful implementation of health information systems is not impossible but does require considerable expertise as well as patience and tenacity over time.^{[6, 7]} The difficulty of creating a new information system is often so great that the ability to use it once completed is impaired, almost as if the system’s original purpose becomes forgotten in the process of building it. People and organisational problems of this nature are not new nor are they technological.^[8] Even with good intentions, failure to understand the complexity of the health sector and the different needs of groups within it can lead to imperfect outcomes in the development of information systems.^[9]

Health informatics and information management
Lorenzi has defined the cornerstones of health informatics as:^[10]

• structures to represent data and knowledge
• acquisition and presentation of data
• management of change
• integration of information.

Lorenzi notes that these cornerstones extend well beyond the skills associated with traditional data processing and information systems and acknowledges that human factors, not technical considerations, constitute the greatest obstacles to informatics success. Gardner reinforces this view: “The success of a project is perhaps 80% dependent upon the development of the social and political interaction skills of the developer, and 20% or less on the implementation of the hardware and software technology”.^[11]

Leaders in the development of health informatics need to have, in addition to credentials in the technology arena, the ability to develop metrics to measure the value of enterprise processes. This view is reinforced by Ball’s statement that the value of informatics “resides in the relationship between cost containment, customer service and satisfaction, and superior clinical results or outcomes”.^{[12, 13]}

Skilled health information managers are in short supply. This is partly economic: the relatively small New Zealand population will not support large numbers of highly specialised knowledge workers in any field. It is also an issue of perceived value: District Health Boards (DHBs) striving to remain fiscally responsible make hard choices about what will be and what cannot be supported. The value created by health information managers is realised over the long term, and it can be difficult for a small number of specialists to communicate and justify their needs. Information managers tend to be more oriented toward informatics than technology and executive management tends to have a greater understanding of, and sympathy for, the latter. Consequently, the role of the health information manager has been eroded.

The true health informaticist has an effective understanding of how a health care system works and knows how information technology can be deployed to enhance that system. This has equal application in the delivery of clinical care and the management of public health in the widest sense.

Gaps in the jigsaw puzzle
The New Zealand health care system has for years been attempting to bridge gaps in information availability. There is no consistent national data collection for primary care contacts or outpatient events, although projects to establish these are now underway. Gaps also exist in other areas, eg, there is no electronic information regarding the state of wellness of the population, all the data that do exist are by-products of traditional treatment-of-illness models. Data on prescriptions and laboratory test ordering and information on other diagnostics do exist, but they are held in different systems that do not interface well and are either coded inconsistently across the sector or not coded at all. This problem is exacerbated by the considerable variety of systems in use in public hospitals.^[3]

Nevertheless, a significant range of information is available. The following sections discuss the existing national information systems and data available in the New Zealand environment.

Existing sources of data
Nationally consistent collections of health care data and the information systems to collect and maintain these data are regarded as fundamental to the management of the health care sector. Health statistics are required to inform key policy initiatives and to provide useful material for researchers to create sample frames and direct data on outcomes.

The MoH has invested considerably in the collection and dissemination of public health-related data. Since 1992, the chief vehicle for this work has been The New Zealand Health Information Service (NZHIS). NZHIS is responsible for the collection and dissemination of health-related data and operates the National Health Index (NHI), the Medical Warning System (MWS) and a number of other data collections. NZHIS has a stewardship, rather than an ownership, role: the data are held for the public good and are available, within the constraints of ethics and privacy considerations, to researchers and policymakers as well as the public at large.

Understanding the nature of existing data collections is fundamental to any discussion of the creation of new data sets. Data collections currently operated by NZHIS fall approximately into four categories:

1. Demographic register (National Health Index)
2. Contract-based data
3. Data processed from manual sources
4. Claim-based data.

More details, including full specifications for these data sets, can be found online at www.nzhis.govt.nz.

In general, data within the NZHIS collections are classified using the WHO International Statistical Classification of Diseases and Related Health Problems (ICD) and the WHO International Classification of Diseases for Oncology (ICD-O).

Contract-based data
Contract-based data sets are derived directly from hospital information systems in both public and private hospitals. Hospitals are contractually required to provide these data. Table 1 lists data sets that are contract based.

Table 1: Contract-based data sets

Data processed from manual sources
These data sets are derived from coding and data entry of data provided on paper forms. Teams of specialised coders working to very specific protocols examine cases in detail and maintain create highly accurate data sets suitable for use in epidemiological research. These data sets are described in Table 2.

Table 2: Data sets processed from manual sources

Claim-based data
Claims for payment for government subsidies of primary care, in particular general practice consultations, community laboratory tests and fulfillment of prescriptions in retail pharmacies, must, contractually, include basic information about the claim to support payment. HealthPAC, the payments and processing agency, receives and processes the claims. Once the claims are processed, associated data are passed to NZHIS for storage in the data warehouses. These data sets are described in Table 3.

Table 3: Claim-based data sets

Making use of existing data
The investment of scarce resources in developing new data sets may not be the most effective way of improving measurement of outcomes. Information can be improved using currently available data. New Zealand already holds a vast amount of data in the national systems alone.

Practice management software is now ubiquitous throughout the country for patient and practice administration. Approximately 80 percent of GPs are now using software for clinical purposes, such as the electronic generation of prescriptions and recording details of patient health encounters.^[14] Electronic claims from primary care providers number approximately 66 million per year, excluding Accident Compensation Corporation claims.

It is acknowledged that there are a number of problems and issues with all of this data. However, new developments will not necessarily solve these problems, as such developments take a great deal of time to deliver results. As noted earlier, creating new information systems is difficult, and the cost involved and the risk of failure are high. All too often, existing sets of data are derided and bold attempts to create better data embarked upon, only to find that the new data has all the same problems as the old.

NZHIS is frequently used by researchers familiar with it, and there are numerous examples of national data sets being used by researchers to contribute to the general body of knowledge.^{[15, 16]}

Clinical benchmarking data has been provided to all public hospitals by NZHIS to inform the quality and cost-effectiveness of health care services. The indicators representing the main aspects of health care are calculated at a Diagnostic Related Group (DRG) level from the hospital discharge data in the NMDS. These data are distributed annually to all DHBs, which are encouraged to monitor their performance against benchmarks determined for the same groupings (eg, tertiary hospital in relation to all tertiary hospitals).

Multi-dimensional approach – a template for success
Successful outcomes in health informatics require an adequate synthesis of health-care knowledge and information systems knowledge (see Figure 1). Consequent on accepting the merit of the model shown in Figure 1 is the establishment of an adequate synthesis of health care knowledge and information systems knowledge.

Figure 1: Dimensions of successful systems outcomes


An example of what can be achieved using this approach is illustrated by the Elective Services Performance Indicators web-site (www.electiveservices.govt.nz/), which uses data from the NBRS and the NMDS, among other sources, and delivers a series of key performance indicators using business intelligence technology tools. The data provided by NZHIS and other MoH units, combined with the knowledge of the health care domain provided by the MoH Clinical Services Directorate and the technical expertise of the website provider, have resulted in an online key performance indicator system.

Barriers and enablers
Both dimensions of the model are of critical importance. In the absence of clinical interpretation, the data are without context, which is a requirement for good information. Without the technical ability to deliver a sound and useable technology platform, the information cannot be processed and disseminated and its value is constrained.

Performance indicators in health care must be developed within the constraints of consumer privacy and information security. New Zealand has comprehensive privacy legislation and regulation with regard to health information; despite this, the level of consumer concern about privacy is considerable, as noted by Professor Jocelyn Chamberlain in her review of Breastscreen Aotearoa:

. . . the offer of preventative service is regarded with extreme suspicion, fearing that a paternalistic medical profession is taking away people’s freedom of choice . . . I found the level of concern about protecting privacy extraordinary . . . If the popular view remains “Privacy at all costs”, then it must be recognised that one of those costs is ineffective and inefficient Public Health systems.^[17]

The gulf between the consumers’ desire for total privacy and the desire of clinicians and researchers to make use of individual data for the public good is considerable. The result is that the latter are reluctant to engage in constructive discussion on the issues. This is understandable, given the general lack of public awareness of basic scientific methods and issues. However, unless such debate takes place, some consumers will remain uninformed, suspicious and hostile towards initiatives designed to improve health outcomes for all.

Conclusion
Significant improvements in the measurement of health outcomes do not necessarily require extensive monetary investment. History has demonstrated that large investments in new information systems do not necessarily pay off as expected or that they show benefits only in the long term.

Researchers, clinicians, administrators and educationalists must work together with skilled health information managers to achieve the best outcomes. Experience indicates that where this multi-dimensional approach is used, the benefits are considerable. Many existing data sets are very useful sources of data that are currently underutilised. At the national level, New Zealand holds five years of mental health contact data, nearly eight years of laboratory tests, 14 years of pharmaceutical dispensing data, 20 years of hospital discharge information, 30+ years of mortality data and over 50 years of cancer diagnoses. There are robust technology and governance mechanisms for protecting individual privacy while allowing analysis of these data to the most sophisticated degree. Few other countries in the developed world can boast as much. It is a national treasure and an epidemiologist’s dream.

References

1. Strassman PA. Crash: The squandered computer. Evaluating the business alignment of information technologies. New Canaan, Conn: The Information Economics Press; 1997.
2. Klein ;. Standardization of health informatics: results and challenges. In: Haux R, Kulikowski C, eds. Yearbook of Medical Informatics. Stuttgart: Schattauer. 2002.
3. Ministry of Health. From strategy to reality: the WAVE project. Report of the WAVE Advisory Board to the Director-General of Health. Wellington, New Zealand: Ministry of Health; 2001.
4. Robin Gauld. The Troubled History and Complex Landscape Of Information Management and Technology in The New Zealand Health Sector. Health Care and Informatics Review OnlineTM. February 2006. /journal/index.cfm?fuseaction=articledisplay&FeatureID=020306
5. Office of the Auditor General. Progress with priorities for health information management and Information technology. 2006. Available at: http://www.oag.govt.nz/2006/wave/docs/wave.pdf (accessed 25 May 2006).
6. Gamm LD, Barsukiewicz CK, Dansky KH, Vasey JJ. Investigating changes in end-user satisfaction with installation of an electronic medical record in ambulatory care settings. J Healthc Inf Manag 1998;12(4):33–65.
7. Gordon D, Geiger G. Strategic management of an electronic patient record. J Healthc Inf Manag 1999;13(3):113–123.
8. Lorenzi NM, Riley RT, Blyth AJC, et al. Antecedents of the people and organisational aspects of medical informatics: review of the literature. J Am Med Inform Assoc 1997;4(2):79.
9. Myers MD, Young LW. Hidden agendas, power and managerial assumptions in information systems development: an ethnographic study. Information Technology and People 1997;10(3):224–240.
10. Gardner R. Davies Keynote Lecture. Proceedings of the Computer-based Patient Record Institute Conference. Washington DC: CPRI, 1998. In: Lorenzi NM, Riley RT. Managing change: an overview. J Am Med Inform Assoc 2000; 7:116–124.
11. Ball MJ. Better health through informatics: managing information to deliver value. In: O’Carroll P, Yasnoff W, Ward E, Ripp L, Martin E, eds. Public Health Informatics and Information Systems. Gaithersburg, Md: Springer-Verlag; 2002.
12. Smaltz DH. The elevation of CIO roles: organisational barriers and organisational enablers. J Healthc Inf Manag 2000;14(1):81–91.
13. Didham R., Martin I., Wood R., Harrison K. Information Technology systems in general practice medicine in New Zealand. Journal of the New Zealand Medical Association, 23-July-2004, Vol 117 No 1198.
14. Cox B, Sneyd MJ, Paul C, et al. Vasectomy and risk of prostate cancer. J Am Med Assoc 2002 Jun 19;287(23)
15. Blakely T, Woodward A, Pearce N, et al. Socioeconomic factors and mortality among 25-64 year olds followed from 1991 to 1994: the New Zealand census-mortality study. NZ Med J 2002;115(1149):93–97.
16. Chamberlain J. Breastscreen Aotearoa – an independent review. 2002. Available via the Internet (www.moh.govt.nz/chamberlainreview) Accessed 3 Sep 2004.

Footnotes

(a). Abridged and updated from its original form which appeared in Health Information Management: Delany R. Evolution, not revolution: measurement and management of health outcomes in New Zealand through efficient use of national information systems. Health Information Management 2004; 32(3&4): 118-125.
(b). Ray Delany is a former Group Manager of the New Zealand Health Information Service. All material contained in this article is the opinion of the author.