Public-private collaboration matrix for accelerating use cases F I GUR E 3 Note: Connectors defined as NGOs, advocacy and patient groups, distributors, group purchasing organizations, investors and other organizations in healthcare. Relatively more supportive role Relatively more central role Why it matters: Research shows that traditional medical care accounts for only 10-20% of overall health outcomes. Genetics and social and behavioural drivers account for the rest. One way to make people healthier is to identify those at risk of becoming sick and intervene before falling ill. In both the US and Europe, however, governments allocate only about 3% of healthcare spending for preventive care, yet a ZS analysis places the ideal share at well over 10%. In a recent survey of more than 9,500 healthcare consumers and doctors from the US, Japan, China, the UK, Germany and Sweden, more than half said they only see a doctor when they are “sick”.7 Making progress: Given this dynamic, early disease diagnosis – whether chronic or acute – can be a matter of life and death.8 For example, the deadliest form of pancreatic cancer has a five-year survival rate of less than 10%, but early detection can increase the survival rate by up to 50%. Researchers at Cedars-Sinai, a non-profit hospital in the US, have developed an AI-based tool that can identify early signs of pancreatic cancer from CT scans with 86% accuracy up to three years earlier than doctors.9 Meanwhile, the German medical devices company Siemens Healthineers uses AIpowered image-reconstruction technology – based on convolutional neural networks – to accelerate brain MRI scans by 70% compared to traditional methods.10 This improvement means earlier diagnosis and better outcomes on a population scale. Yet not all diseases are binary. For many diseases, doctors hope to prevent an event, such as a stroke, asthma attack or sepsis. For these, AI is being used to stratify a patient’s risk and suggest appropriate interventions to prevent it from occurring. Nearly three out of five physicians predict AI technology will be most useful for fighting chronic disease.11 One US-based health tech organization, Ellipsis, is using deep learning to analyse voice samples recorded during clinical visits to detect semantic and acoustic patterns that can diagnose depression more effectively than traditional clinician-guided questionnaires. Some authorities warn that doing so at scale will prove difficult given the complexities of depression and multivariate factors that must be taken into consideration.12 Nevertheless, algorithms are now being trained to flag early signs of diabetes, asthma and other chronic diseases that benefit from prevention measures only possible with earlier signals. Scientists are even modelling environmental factors, such as climate change, to help strengthen earlier detection of diseases such as asthma and those transmitted by mosquitoes. Use case/ stakeholders Government Connectors Technology Life sciences/ medtech Providers Payers AI-driven diagnosis and risk stratification Infectious disease intelligence Clinical trials optimization 1.1 Use case for acceleration #1: diagnosis and risk stratification Scaling Smart Solutions with AI in Health: Unlocking Impact on High-Potential Use Cases 11
RkJQdWJsaXNoZXIy NzQwMjQ=