John Mathews: Benefits of sharing

FROM at least the time of Hippocrates, the medical profession has advanced its knowledge of disease, prognosis and treatment by pooling information about patients with similar histories.

Doctors built their reputations as they acquired new knowledge and skills, helping them to attract more patients.

That “small business” model, with practitioners competing for patients, was nevertheless leavened by a cooperative commitment to share expertise through teaching and publication. Until well into the 20th century, that model was sustained — patients who could afford it would pay the doctor directly, while those who couldn’t pay would depend on the charity of a doctor or hospital, with clinical records the property of the doctor and/or the hospital.

Much has happened since then. The rise of scientific medicine and the public funding of health services have seen the charity hospitals of the past transformed into the overloaded state hospitals of today, although the small business model of private practice, transformed by fee-for-service funding through Medicare, still allows the doctor–patient relationship to be maintained.

Where has this left the worlds of record-keeping and research?

With the advent of modern computing, there are now many publicly funded databases with individual records which can be accessed or linked, most easily within a single agency or jurisdiction, while still preserving privacy.

WA has been a pioneer both in using state-based data, and in linking WA data to Commonwealth data. For example, the linkage of WA health records to Pharmaceutical Benefits Scheme records has provided important insights into birth defects and other adverse effects of medication.  

Medicare records, although primarily established for administrative purposes, can also be of great value for research. For example, by linking de-identified Medicare records of CT scan exposures to de-identified records of cancer registrations, researchers have been able to show an average proportional increase of 16% in subsequent cancer risk for each CT scan before the age of 20.

Despite the value of such research, and the robust measures that protect privacy, government agencies and data custodians have often been reluctant to release data, and many researchers report long delays with approvals. Indeed, some ethically approved projects of particular relevance to health have had to be abandoned.

The difficulties with data release may reflect resource constraints, mistaken concerns about privacy, or concerns about whether research carried out by an external party might reflect upon the interests of the government agency or jurisdiction involved. Some of these matters are considered in more detail in a recent Croakey blog.

What about patient records in community or hospital practice? Because services are now supported by public funds, there is a strong rationale for being able to access the records for public benefit research, provided privacy is protected.

Furthermore, because of the uptake of computer systems by most practitioners and hospitals, it is now feasible to capture and link such records. For example, MedicineInsight, in a pilot study, is collecting anonymous patient and clinical care information from general practices and linking diagnoses to medications, other services provided and clinical outcomes.

The personally controlled electronic health record (PCEHR), with patient consent, will eventually allow identified health records to follow each patient through the health system, while still preserving privacy.

In the longer term, as more people sign up, and as the quantity and quality of captured data improve, the linkage and aggregation of anonymous data from the PCEHR will provide a much richer vein of clinical information which can be mined to identify adverse drug events and other risks, to evaluate diagnostic criteria and treatments, and to improve the efficiency and effectiveness of the health system as a whole.

So today, despite the great public benefits that can flow from data sharing and linkage, progress is being delayed, in some cases because of residual privacy concerns, public inertia, resource limitations or the potential conflicts of interest between government agencies, different jurisdictions and the research community.

As part of a wider public discussion of issues around data release and data linkage, and to canvass cooperative principles that could guide our nation forward, the Menzies Foundation will host a data linkage workshop in Melbourne on 16 August. It will be attended by community, medical and media representatives, leading researchers, senior public servants and members of parliament.

Let’s hope it helps to overcome some concerns and open opportunities for even more data links.

 

Professor John Mathews is director of the Menzies Foundation and a professorial fellow at the University of Melbourne.