Considerations for Health Data Sharing Across Borders

Sharing my thoughts on a Saturday evening as I was excited to see the discussions Clive Tan shared about on data-sharing across borders. For context please read his original post below:

and a related white paper titled 'Responsible Data Sharing in Health and Healthcare'. Link to which can be found below.

Disclaimer: Minimal edits. Written on a weekend within 45min while dealing with 2 wonderful but highly demanding children. Please let me know if something doesn't make sense or needs an edit.

------------------------------------------------------------------------------

Just to add to the conversation my perspective of a start-up CEO in this data-sharing space. To be clear, Oncoshot is specifically enabling secure data sharing between healthcare systems within the space of oncology clinical trials with one difference being inclusivity of industry - which we feel is critical in building a sustainable long-term model and a data-driven ecosystem overall.

Data sharing is always thought of as the next step in the digital transformation process for healthcare systems, coming after the first step of data digitalisation. It is not wrong for hospitals or healthcare systems to think of this as a sequential process. From an impact and scale perspective, however, I believe we should actively consider opportunities in which both Data Digitalisation and Data Sharing can be tackled concurrently.

A few requirements need to be met for this and I’ve also attempted to share my thoughts on how these requirements can be shaped within a use case that has recently raised much discussion in Singapore

1.??????Strict and clearly defined data requirements

While it is more holistic to discuss EMR sharing, the reality is that discussions on data sharing will need to be very tightly defined at the onset. Instead of ‘EMR data’ being made accessible, framing discussions around use-cases where say ‘only 15 data fields centred around diagnosis, treatments, biomarkers, live/death outcomes etc’ is required is certainly a lot less daunting (particularly for clinicians/researchers/non-technical healthcare stakeholders) and for which the risk of data sharing and the value proposition to be gained can be better evaluated. These data fields will be defined, of course, from an acknowledged big enough problem or use case that needs a solution.

2.?????A problem where aggregated data sharing could be the solution

Identification of use cases where aggregated data sharing across healthcare systems or national ecosystems as a whole has value. Let's face it, sharing patient-level data is just downright challenging or impossible in most cases.

3. Addressing an impactful and high-value use case

It follows then that the requirement for data into an important use case where there are public health implications for the participating stakeholders.

Let’s stretch this discussion a little bit and look at the recent issues about funding for cancer drugs which the likes of Jeremy Lim , Salma Khalik and Hwee-Lin Wee have contributed to in both print and social media.

In deciding drug funding, funding agencies typically look at existing data on clinical outcomes (typically from publications which form the basis of drug approvals in the US/EU) and attempt to rationalise these outcomes with the cost of treatment.

To add to the complexity of drug reimbursement in the regional context (ASEAN/Asia), there are already existing issues where for example despite published evidence which included fairly diverse cohorts, many standard of care drugs may not be in the drug reimbursement lists of some countries due to the inability of these countries to rationalise costs and outcomes for their own local population.

In the context of our discussion on data, an issue that has been side-stepped, in the desire for quick-fix solutions, is that most countries are unable to analyse critical datasets in near real-time and inform drug funding decisions for both standards therapies and non-standard treatments/off-label drugs with precision. The recent discussion about rationalising drug costs and reimbursement adequacy is fundamentally an issue arising from our inability to gather, analyse and make decisions in a timely manner. Theoretically, it is a potential use case for which aggregated health data sharing could be a solution.

Singapore’s experience with this issue is not just unique to us and is an issue that every healthcare system in ASEAN/Asia faces. In relation to the example above, it could mean that insights on unique patient populations and outcomes of therapies are shared across countries in ASEAN to guide their own internal decision-making with regard to approvals and funding.

4.??????Enable value exchange amongst stakeholders

An important component that we also need to consider is the movement of financial incentives across stakeholders in addition to the movement of data/insights. In most cases, the inability to bring in financial incentives is a key reason why data sharing models never get established and are always perceived as a costly/net-negative exercise. If a particular use case is able to bring in appropriate funding and scale in adoption or usage, it could provide an important foundation for the sustainability of data sharing. Taking the example of a solution for the provision of aggregated analytics for drug-reimbursement decisions, this could translate into cancer pharma and biotech financing highly specific drug-impact analytics which need to support the application for funding/reimbursement review by the payor. The introduction of appropriate incentives would allow for healthcare systems/ministries to sustainably maintain both their digitalisation and data sharing systems over the long term as well.

Overall, while several other important requirements need to be in place for multi-national data-sharing; including appropriate safety, privacy, and data ownership frameworks and stakeholder alignment to act on the results derived from such systems or platforms, the points above will help stakeholders establish a clear vision with the intended outcomes that makes data-sharing one-to-one, one-to-many or even many-to-many plausible.

Links:

White Paper 'Responsible Data Sharing in Health and Healthcare: