Trusting AI developers with healthcare data

Healthcare's dirty data problem will take a united effort - and AI - to solve

Developing AI for healthcare - who should be trusted?

Recently we ran a Linkedin poll asking the healthcare community who should be entrusted to develop the AI tools necessary for solving healthcare’s dirty data problem - only specialist AI healthtech companies, or open the floor to everyone? The response was overwhelmingly in favour of leaving it to the experts, with many concerned by the possibility for unethical misuse of data, hallucinations, misdiagnosis, and breaches of privacy.

You can’t beat the experts

It is certainly true that AI healthtech companies have the greatest insights into complex factors like clinician workflows, patient behaviour or the clinician-patient dynamic, using them to inform the design or deployment of their products. When healthcare IT professionals and C-suite leaders are staking their careers on a software recommendation, and the success and uptake - or not - of that software comes down to factors like the amount of clicks it takes for a clinician to complete a task, they need to be supremely confident that they’ve chosen the right - future proofed - product, with the right partner.

The risks are real

These risks are very real - plus the more familiar issues of purchasing technology that later becomes redundant, choosing partners that don’t go the distance, or, per the World Health Organization - that don’t consistently exercise the necessary caution required for any new technology such as Large Language Models, and adherence to vital values of transparency, inclusion, public engagement, expert supervision, and rigorous evaluation.

But the benefits of collaboration can’t be denied

But it’s also true that the greatest innovations come from the coming together of diverse perspectives. The potential benefits of involving a broader community are clear - greater inclusivity, creativity, pooling of resources and efficiencies of scale, to name a few. But how to ensure that non-health experts conduct themselves in a responsible, ethical way and successfully navigate the unique pitfalls and responsibilities of the healthcare ecosystem?

Build strong partnerships

We believe in a collaborative approach that balances expertise with inclusivity. Our goal is to ensure the responsible development of AI by observing the WHO’s key values, while leveraging diverse insights from our partners like Pieces Technology , and our many customers. One of our key objectives as an organization is understanding the strategic objectives of health-system C-suite leaders, and designing and deploying solutions to enhance those outcomes.

Remain open to new ideas

But we also believe that the best ideas can come from anywhere, which is why we open many source tools in our Orchestral Health Intelligence Platform . Over the next few months we’ll be exploring how to foster collaboration between health-tech experts, C-suite leaders and the wider AI and healthcare communities, and looking for examples of successful innovations and improved data management.

Ultimately we’re excited by the idea of a united effort to address healthcare’s dirty data problem, provided it’s guided by health tech experts to ensure sustainable, secure progress.

Leverage your data or get left behind.

The AI horse has well and truly bolted, and businesses of all kinds are scrambling to keep up. We believe that the healthcare organizations that are best able to fully unify and harness their data, using reliable AI products developed by the most innovative, forward-thinking teams, are the ones that will become leaders in their field - much like Amazon, Uber and AirBnb before them.

This post originally posted on Orionhealth.com