Summarize this content to 2000 words in 6 paragraphs in Arabic Unlock the Editor’s Digest for freeRoula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.The writer is executive chair and co-founder of the Open Data InstituteAs of 2023, fewer than 1 in 4 people were satisfied with the NHS — the lowest number ever recorded. New UK health secretary Wes Streeting is determined to address failings, linking the health of the nation to the health of the economy. Data and technology have the potential to tackle many NHS challenges, revolutionising care and research. As early as 2012, data analysis by the Open Data Institute (ODI) showed the NHS could save £200mn annually by changing the prescribing of statins. Yet the opportunities remain largely unrealised. In the race to adopt artificial intelligence, the government has overlooked data governance, as our recent paper highlights. But for the NHS to reap the benefits of AI, strong data infrastructure is essential.Good data governance could improve care quality, safety and cost-effectiveness in the NHS. AI technologies, according to IBM, could reduce treatment costs by 50 per cent and improve outcomes by 40 per cent. The NHS holds vast amounts of data, often in siloed IT systems, making it hard to use across the health service ecosystem. Before the boom in AI, data already played a crucial role in NHS research and healthcare. Initiatives such as OpenSafely and Health Data Research UK improve our understanding of diseases, enable personalised treatments and enhance efficiency. The UK leads globally in health data reuse. The NHS AI Lab uses machine learning to predict disease, identify high-risk patients and recommend treatments, and AI tools such as “C the Signs” have significantly increased cancer detection rates. Of course, using sensitive health data raises questions about privacy and security. US data analytics company Palantir’s involvement with the NHS Federated Data Platform has left people concerned about whether the benefits of data sharing outweigh the risks of their information being misused. Winning and maintaining public trust is therefore crucial. Following the 2021 GPDPR campaign, patient opt-outs from sharing health data nearly doubled, so organisations will have to reassure the public of their trustworthiness. UK data protection laws pre-date AI and large-scale data sharing, and must be updated.Data sharing is complicated, as 189 NHS trusts collect patient data through different electronic systems. AI models also use data from medical imaging, genomics and wearable devices. The ODI recommends that the NHS and the new government adopt Fair principles for governing and sharing this data. Well-curated data infrastructure is essential for delivering AI’s value effectively and responsibly. This includes interoperable open standards, secure data storage, efficient data processing and strict data protection regulations.While we await a national data infrastructure, machine learning developments allow researchers to access and analyse sensitive data without compromising privacy or security. Privacy-enhancing technologies have also emerged to give people direct access and control over their own health data.Federated learning (FL) allows algorithms to access data from multiple local data sets without exchanging the underlying data, and has shown promising results, such as Oxford university’s Curial-Federated platform, which developed a Covid-19 screening test by training data across four NHS Trusts, with hospitals retaining data custody. Secure data platforms like the Secure Data Environment (SDE) and OpenSafely give researchers access to anonymised NHS data.To improve the NHS, we need a collective commitment to ethical, transparent and innovative data practices, from ensuring access to high-quality, well-governed data to enforcing data protection and intellectual property rights. Without these measures, Streeting’s vision of Britain as a MedTech “powerhouse” and the NHS unlocking AI’s potential will be difficult to achieve.
