Nationally linked data to improve health services and policy

A well designed National Integrated Health Services Information Analysis Asset will improve services and policy

New Zealand continues to set best practice standards internationally for cardiovascular disease (CVD) risk prediction and management. A 2018 study, using data from the PREDICT general practice cohort linked with demographic, prior medical history and drug‐dispensing data highlighted that the risk of CVD is best estimated from longitudinal follow‐up of a contemporary nationally representative cohort initially free of disease.1 PREDICT is a risk tool that incorporates new predictors of socio‐economic deprivation and ethnicity and better reflects the population whose risk is being assessed. In comparison, the application of earlier international risk equations to Australasian populations, such as the Pooled Cohort Risk Equations (PCEs)2 in the United States and QRISK in the United Kingdom,3 are likely to substantially underestimate or overestimate risk, leading to either undertreatment or overtreatment. Where underestimates of risk occur, the individual is falsely reassured and no indication to commence preventive treatment is apparent. The reverse is true for overestimates of risk, where an indication to start preventive treatment is unnecessary. This is particularly true among socially disadvantaged and ethnically diverse populations, where both PCEs and QRISK underperform. New Zealand's ability to integrate its administrative health datasets with other data sources — in this case, primary care — has enabled the conduct of this policy changing research. The New Zealand's Ministry of Health has adopted and supported the roll‐out of the updated CVD risk management guidelines recommending that general practitioners use the new PREDICT‐derived CVD risk equation. It is thus important that Australia has a national repository that enables the combination of routine health datasets with other data sources, existing and emerging, to permit evaluation of health care and inform policy decisions.

The Australian Health Ministers' Advisory Council has approved the development of a National Integrated Health Services Information (NIHSI) Analysis Asset — owned by the Australian Institute of Health and Welfare together with the Commonwealth Department of Health and state and territory health authorities — containing anonymised data from participating states and territories from 2010–11 onwards.4 This initiative follows a successful national data linkage demonstration project, the findings of which will be published by the end of 2019. The project linked data from public hospital admissions, emergency services, the Medicare Benefits Schedule, the Pharmaceutical Benefits Scheme and the National Death Index. The anonymised data were used to document the patient journey from admission to and discharge from hospital, subsequent care in the community, and survival status over time. The available data are grouped by major health conditions and do not identify individual service providers or sites. Typical information is limited to health service utilisation, including visitations to primary health care providers, medical specialties and admissions to hospital. Such demonstration projects mirror what is routinely possible in New Zealand and provide future hope for advancing health services and policy if developed for public good in Australia.

The provisional datasets include those mentioned in the demonstration project and extend to cover admissions to private hospitals, together with outpatient and residential aged‐care data. There will be early limitations to NIHSI, such as who will have access and how this will be determined, the availability and type of data from each jurisdiction, and the research questions able to be asked from the available data. The Australian Institute of Health and Welfare Ethics Committee has approved the NIHSI Analysis Asset for the priorities of defining the patient journey, patterns of use, effective health services, and developing policies and programs.5

The existence and success of the NIHSI Analysis Asset will be contingent upon establishing explicit robust policies and standard operating procedures for access to the data from the outset. Australians will expect the highest level of governance, management and stewardship of the data, in addition to clarity and accountability as to the purpose of the collection and its use. More importantly, individuals will want assurance that all data made available for analysis are anonymised and that rigorous standards to protect personal privacy and data confidentiality are established.

For the NIHSI Analysis Asset to yield optimal return on investment through improving health services and policy, it will require the coverage of both public and private health sectors, processes for data linkage to additional datasets such as clinical quality registries and clinical trials, a streamlined and cost‐effective mechanism for access to and use of the data, and the development of “trusted user” status based on the “five safes” principles6 supported by training and accreditation.

In summary, a well designed NIHSI Analysis Asset paves the way for improving health services and policy in Australia. Better access to existing health data sources, together with established quality clinical registries and major cohorts, is central to improving population health. For high cost, high risk and high volume health care activities in cardiovascular and other chronic conditions, access to these data would provide essential information on comparative effectiveness of treatment strategies. Both funders and commissioners of health care would be able to target limited resources to those areas where the greatest long term benefit might be obtained. The advent of NIHSI Analysis Asset for public good will require exceptional leadership and comprehensive public discourse.