- What is the National Health Index?
- History of the NHI
- In Defence of Public Good
- The NHI upgrade project
- Assessment of the NHI
The National Health Index (NHI) is regarded as the cornerstone of health information in New Zealand. Within the health care community in that country, the index has been so well known for so long that the acronym NHI has achieved the status of a descriptive word for many health care professionals, and is scarcely ever spelled out. However, outside the health care community, there is much less awareness of the index, and its purpose and role in linking together the information transactions that inform the delivery and management of health care in New Zealand is poorly understood. Further, certain myths have grown up around the NHI, which â€“ while they may have once had some elements of truth â€“ continue to be widely believed long after the facts have changed.
What is the National Health Index?
The purpose of the public health management vision for the NHI has never been precisely defined in public. The purposes to which the index have been put have altered over the considerable number of years the index has been in place. Much of the material publicly available seems to concentrate on defining what the NHI is not, presumably to mitigate concerns over the privacy of health-related information. However, it is easy to infer that the NHI has the following purposes:
- Identification: Different people with the same or similar names will have different NHI numbers, ensuring that the information pertaining to one does not get confused with that pertaining to the other.
- Linking: Data held about the same individual in many different databases can be brought together for the purposes of obtaining maximum information regarding a particular patient, as well as analysis and research.
- Consistency: When people change their names for any reason, the NHI number remains the same, thus ensuring longitudinal consistency with respect to information held about that person.
The NHI is widely perceived as a significant national asset. International commentators frequently find it remarkable that this index has been in place for many years, while other countries have yet to establish such an index.[1, 2]
In the late 1980s and early 1990s, a convergence of public policy focusing on the corporatisation of public services and the technology trend to move away from centralised computing services, because of the emergence of distributed computing and open systems standards, led to the privatisation and subsequent sale of government-run organisations responsible for the management of the centralised mainframe system. Later, the introduction of a funderâ€“provider split in health care and the break-up of Area Health Boards into competition- and profit-oriented Crown Health Enterprises triggered the establishment of independent and different hospital information systems throughout the public health system. In order to prevent fragmentation of health care information, the New Zealand Health Information Service (NZHIS) was established in 1992 to act as a centralised repository for key national standards and minimum data sets built according to those standards. The first priority for NHZIS was the establishment of the NHI, and the systems established at that time are largely the same as those in use today.
In 1994, New Zealandâ€™s Privacy Commissioner produced the first version of the Health Information Privacy Code. Since the legislation underlying the code was primarily concerned with unique identifiers in all sectors of society, one of the 12 rules in the code is devoted to the regulation of the NHI. For the next 10 years there was considerable tension between the office of the Privacy Commissioner and the NZHIS, the one being concerned with ensuring the safety of the population from privacy violations and the other being concerned with maximising the distribution of health care information. One consequence of this tension was a disinclination to promote the societal benefits of the NHI, with the result that public awareness of the NHI was, and continues to be, low.
In the mid-1990s, the corporatised model of health care shifted the focus of health information systems away from obtaining high-quality statistical information and towards providing transactional information to assist with managing the multitude of service contracts associated with the new approach to management of the sector. One consequence of this was a drive to introduce the use of the NHI number into primary health care transactions. As a result, the protocols for management of the NHI, which had been designed around a relatively limited volume of transactions associated with people who made use of public hospital services, now needed to incorporate both a change in the nature of its transactions and a huge increase in the volume of transactions. The measures taken to achieve this were, in hindsight, quite inadequate. Techniques such pre-issuing a tranche of NHI numbers and mailing these to general practices on CDs resulted in a huge number of duplicate index entries being created. Payment of subsidies for general practice consultations were conditional on NHI numbers being embedded in the subsidy claim, but there was no mechanism to ensure that number provided matched the patient that was seen. Consequently, the level of confidence in the accuracy of the NHI, and its utility for primary care, plummeted.
In 2001, the Ministry of Health published the Wave Report. By this time, the issues with the NHI were so well canvassed in the NZ health sector that the Wave Report did not even explain in detail what they were, but produced the simple recommendation to â€œFix up the NHI . . .â€
In Defence of Public Good
The Wave Report refers to the â€œpublic goodâ€ aspects of the NHI. This phrase seems to encapsulate the public health management vision and purpose of having such an index. The NHI itself is a pivot point through which can be shared both individualsâ€™ information (such as the national medical warning system which provides the facility to record patient allergies) and other key information that can be available to clinicians anywhere in the country and can be used independently of name changes over the entire period of a person’s lifetime. The NHI number is also a standard index number that is routinely used by analysts and researchers to undertake unit-level analysis that would not be otherwise possible. The ability to link and correctly assign data in both practical care and research makes it possible to address the two major challenges facing health care services worldwide: (1) minimising the risk of adverse events and (2) maintaining high quality outcome information on public health interventions. Despite these rather obvious potential benefits, the NHI has frequently been the target of criticism.
The most common criticism of the NHI has been that the accuracy of the data held in the index is questionable. In the mid-1990s, the burgeoning of information technology and systems across all areas of society prompted a high level of interest in privacy issues related to these systems. Privacy Commission staff singled out the NHI in public forums as an example of a risk to privacy stemming from poor data quality. Criticism of the NHI typically pointed out that the NHI was â€œobviouslyâ€ full of duplicate entries, since there were at the time over six million entries on the index for a population of less than four million New Zealanders. This type of commentary was based on a misunderstanding of the difference between â€œactiveâ€ NHI numbers and the total number of entries in the index. NHI numbers are never re-issued once used. When an individual dies, the NHI is updated with the date of death, but the NHI index entry remains, becoming inactive. Active NHI numbers in theory identify people who are alive and living in New Zealand. This is quite different from the total number of entries in the index. Deceased individuals are identified with great accuracy from the register of deaths. Keeping accurate records of whether or not people remain resident in New Zealand is more difficult, nevertheless the accuracy of the NHI is generally higher than it is given credit for. Duplicate entries are a constant difficulty, and the legacy of earlier poor processes remained in the NHI system for many years, although the problem was never as great as external commentary alleged.
Another example of the kinds of criticisms levelled at the NHI in past years can be found on p 35 of the Wave Report. One of the anecdotal case studies found throughout the document is used to illustrate problems in the NHI. In this story, PHARMAC, the government agency responsible for managing and funding government subsidies for pharmaceuticals, advises a researcher that â€œproblems with the NHI systemâ€ meant that ethnicity data was unreliable and, thus, would not be suitable for a comparative study of the effectiveness of antipsychotic drugs in Maori populations. The same story goes on to say that even if it had been logistically possible to conduct such a study based on the evaluation of hard copy prescriptions, this would not have yielded the required ethnicity information.
There are two observations to be made about this story. The first is that the information regarding the alleged problems with the NHI ethnicity data were provided to the researcher not by the agency responsible for the NHI but by a third party. The second is that the key data required would apparently not have been available even from a detailed search of the relevant paperwork, yet its absence in the national register was seen as a failure of that system.
While there is no doubt that the accuracy of ethnicity data in the NHI has been questionable in the past, which is a function of the processes used to determine ethnicity at the health care provider level, researchers have shown that this problem is not insurmountable and can be corrected by triangulating the NHI data with census data from Statistics New Zealand.
The fact that the NHI does not provide an instant answer for every question at the push of a button does not detract meaningfully from the fitness of the index for the purposes outlined earlier.
International experience has echoed anecdotal findings in New Zealand that manual transcription of unique identifiers erodes their value.Maintaining the credibility of a national standard electronic system is difficult, and has proved a challenge over many years.
A key component of the NHI upgrade programme was the upgrading of the APIs. New APIs were seen as key to the provision of real-time access to the NHI by primary care systems, such as GP practice management systems. De-scoping of the NHI upgrade project in 2005 to fit within budget constraints meant that the only way primary care could access the NHI was via the new NOAH[b] system. This is not suitable for the needs of GPs using integrated practice management systems, and so the uptake of the NOAH system was low. The New Zealand Auditor Generalâ€™s review of the progress since the publication of the Wave report recommended that upgrading the APIs should be a priority, and NZHIS has plans to implement this as a separate project, together with a number of other projects stemming from the Health Information Strategy for New Zealand, published in 2005.
Assessment of the NHI
In 1998, the US National Committee on Vital and Health Statistics (NCVHS), a public advisory committee to the US Secretary of Health and Human Services, produced a white paper called â€œUnique Health Identifier for Individualsâ€. This paper was produced in response to the Health Insurance Portability and Accountability Act of 1996 (HIPAA) which mandated the use of unique identifiers for individuals, amongst other standards. The paper discussed a number of options for a unique identifier for healthcare transactions to support HIPAA compliance. The paper cites the Standard Guide for Properties of a Universal Healthcare Identifier, published by the American Society for Testing and Materials (ASTM; a standards development organisation accredited by the American National Standards Institute) as the appropriate tool for evaluation of a unique identifier. This standard identifies 30 separate criteria for a unique identifier. These criteria have been used to evaluate the New Zealand NHI and the resulting assessment is set out in Table 1.
Table 1: Criteria for a unique identifier (Standard Guide for Properties of a Universal Healthcare Identifier)
|1. Accessible||Available when required.||The NHI has been available to all public hospitals in New Zealand since 1992. Development since the Wave report has increased availability to primary care.|
|2. Assignable||Can be assigned when needed by trusted authority after properly authenticated request.||The NHI is specifically designed to meet this criteria.|
|3. Atomic||Single data item; no sub-elements having meaning.||The NHI number is an essentially random series of letters and numbers.|
|4. Concise||As short as possible.||The NHI number is seven characters in length, three alpha and three numeric characters plus a check-digit.|
|5. Content-free||No dependence on possibly changing or unknown information.||The NHI number is routinely used in isolation from the main index.|
|6. Controllable||Only trusted authorities have access to linkages between encrypted and non-encrypted identifiers.||One-way encryption algorithms have been developed. It is possible that this encryption could be cracked using the power of modern computers.|
|7. Cost-effective||Maximum functionality with minimum investment to create and maintain.||The cost of operating the NHI has become progressively lower over time.|
|8. Deployable||Implementable using a variety of technologies.||A wide range of different systems are currently using the NHI.|
|9. Disidentifiable||Possible to create a number of encrypted identifiers with same properties.||One-way encryption algorithms have been developed for the NHI.|
|10. Focused||Created and maintained solely for supporting health care - form usage, and policies not influenced by other activities.||The NHI is specifically reserved for health care purposes only, and this is regulated by the Health Information Privacy Code 1994.|
|11. Governed||Has entity responsible for overseeing system - determines policies, manages trusted authorities, and ensures proper and effective support for health care.||The NHI has been governed by NZHIS, a department of the Ministry of Health since its inception.|
|12. Identifiable||Possible to identify the person with such properties as name, birth date, sex, etc, by associating these with the identifier.||These demographics are contained in the NHI database.|
|13. Incremental||Capable of being phased in.||The NHI has been phased in, starting with public hospitals, although it is now ubiquitous throughout the health sector in New Zealand.|
|14. Linkable||Can link health records together in both automated and manual systems.||The NHI is embedded in national collections of health data.|
|15. Longevity||Designed to function for foreseeable future with no known limitations.||The NHI system is a remarkably long-lived system. It has functioned in its current form without major changes since 1992.|
|16. Mappable||Able to create bi-directional linkages between new and existing identifiers during incremental implementation of a new identifier.||NHI numbers can be linked one to another to eliminate duplicates, but there is no specific mapping facility.|
|17. Mergeable||Can merge duplicate identifiers to apply to the same individual.||The NHI has always had this functionality since its inception, and it has recently been improved.|
|18. Networked||Supported by a network that makes services available universally.||The NHI network currently covers all public hospitals and is available to primary care via a number of different paths.|
|19. Permanent||Never to be reassigned, even after a holderâ€™s death.||NHI numbers are never reassigned.|
|20. Public||Meant to be an open data item â€“ the owner can reveal it.||Unencrypted NHI numbers are routinely used on hospital medical records, lab test requests, prescription forms, etc.|
|21. Repository-based||Secure, permanent repository exists to support functions.||This repository exists and is operated by NZHIS.|
|22. Retroactive||Can assign identifiers to all existing individuals when system is implemented.||NHI numbers are estimated to have been issued for 95% of the New Zealand population.|
|23. Secure||Can encrypt and decrypt securely.||Encryption facilities exist.|
|24. Splittable||Able to assign new identifier to one or both people if the same identifier is assigned to two people.||These facilities have recently been improved in the NHI.|
|25. Standard||Compatible if possible with existing or emerging standards.||The NHI was one of the first standards set in New Zealand for health care data.|
|26. Unambiguous||Minimises risk of misinterpretation such as confusing number zero with letter O.||The rules for issuing an NHI number are designed to mitigate this problem by not using certain ambiguous letters and combinations.|
|27. Unique||Identifies one and only one individual.||Although one person can have more than one identifier, each identifier can only be for one individual.|
|28. Universal||Able to support every living person for the foreseeable future.||The number of NHI identifiers that can be generated is finite. There are sufficient unused numbers to serve the needs of the New Zealand population for at least another decade or two.|
|29. Usable||Processable by both manual and automated means.||The NHI is routinely used in many different modes in different systems throughout the NZ health sector.|
|30. Verifiable||Can determine validity without additional information.||The NHI incorporates a check-digit to assist with validation.|
New Zealand has a significant national asset in the current NHI. Many of the developed nations have yet to achieve a single identifier with the level of coverage that has been obtained here or have yet to even develop an identifier system. New Zealandâ€™s success with its NHI is partly due to the small population base. It should be remembered the NHI is not free from difficulties. However, this should not blind us to the fact that, overall, the system is conceptually sound and the practical difficulties that arise from time to time can be resolved. The New Zealand experience has useful lessons that any country attempting to implement similar systems can benefit from.
- For a discussion of the US progress with unique identifiers see, for example, Yasnoff et al, J Am Med Inform Assoc 2004 Julâ€“Aug; 11(4): 332â€“338. doi: 10.1197/jamia.M1616.
- For a discussion of unique identifiers in electronic health records see Ceusters W, Smith B. Tracking referents in electronic health records. Connecting medical informatics and bio-Informatics: Proceedings of MIE2005 - The XIXth International Congress of the European Federation for Medical Informatics. Studies in Health Technology and Informatics 2005; 116: 71-76.
- Robin Gauld. The troubled history and complex landscape of information management and technology in the New Zealand health sector. Health Care and Informatics Review Online. February 2006. Available at: /journal/index.cfm?fuseaction=articledisplay&FeatureID=020306.
- Ministry of Health. The WAVE Advisory Board to the Director-General of Health. From Strategy to Reality: The WAVE Project. Ministry of Health Wellington New Zealand, 2001.
- Ajwani S, Blakely T, Robson B, Atkinson J, Fawcett J, Kiro K.. Unlocking the numerator-denominator bias for the 1980s and 1990s. NZCMS Technical Report No. 4. ISBN 0-473-09111-9. Available at http://www.wnmeds.ac.nz/nzcms-info.html. Wellington: Department of Public Health, Wellington School of Medicine and Health Sciences, University of Otago: 2000. p 130.
- Dalrymple AJ, Lahti LS, Hutchinson LJ, O’Doherty JJ. Record linkage in a regional mental health planning study: accuracy of unique identifiers, reliability of sociodemographics, and estimating identification error. J Ment Health Adm. 1994 Spring; 21(2):185-92.
- Office of the Auditor General. Progress with priorities for health information management and Information technology, 2006. Available from http://www.oag.govt.nz/2006/wave/docs/wave.pdf (accessed 25 May 2006) .
- (a). 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.
(b). NOAH is an acronym for NHI Online Access for Health.