There is no question that organizations worldwide are increasing their investment in AI.
There is also little doubt that AI is starting to impact many different sectors. The health care and life sciences sectors are no exception, with many organizations investing in new technology. The real issue is how to maximize the impact of those investments.
A recent study on AI and data in the Asia–Pacific region provides an opportunity to understand more.
The state of play
There is a wide range of levels of AI maturity in health care across the Asia–Pacific region. Around five percent of organizations are still evaluating the technology, and 17% are still in the planning stage.
However, around 60% of organizations already use AI, albeit at different stages of competence ranging from functional through short-term focus to integrated. Around 18% of organizations successfully use AI to transform their business models and make a real difference in health care.
A key focus for most health care organizations in the region is optimizing the return on their AI investments.
Around a third of organizations expect a return of at least two times the investment, and 47% expect over three times the return.
There are three important areas of investment:
- Using AI to enhance operational workflows.
- Improving patient outcomes.
- Reducing fraud in claims processes.
AI-driven personalized care and patient management improvements, including safety, are expected to generate business value. This approach is expected to combine better financial performance, improved data security and more substantial regulatory compliance.
Challenges and benefits
The top benefits reflect this combination of priorities.
- Health care fraud cost containment: At the top of the list are health care fraud and cost containment. Fraudulent claims, billing inaccuracies, and medically unnecessary procedures contribute to the escalating costs of health care, impacting insurers, providers, and patients.
- Using real-world and synthetic data to accelerate regulatory approvals: Patient data is often limited due to privacy concerns, ethical restrictions, and lengthy data collection processes, which slow down drug approvals.
- Optimizing patient and drug safety, medical resources and clinical trial design: Timely identification of adverse drug events and safety signals is critical to protect patient safety. Additionally, poor resource allocation in hospitals and inefficient trial designs increase costs and delay patient access to new innovative treatments.
Organizations also expect to achieve a range of benefits from AI. The most common expectation was better end-to-end data management. This was followed by faster time-to-market and synthetic data generation, which was more a matter for pharmaceutical or device companies than health care providers. Some organizations also cited improvements in internal processes, including better collaboration across teams and more agile decision-making processes.
However, these benefits are not without their challenges.
- Lack of specialized skilled personnel.
- Concerns about data or IP loss due to improper use of AI.
- Data foundation lacks sufficient governance processes.
Approximately a third of the region's health care and life science companies stated that they had particular data challenges. They cited the problems of eliminating unused or unnecessary copies of data sets across their data infrastructure. They noted that excluding old or expired data from AI-based modeling was challenging.
The way ahead
What should health care and life sciences organizations in the Asia–Pacific region do to deliver on their AI investments? The study's findings suggest that the most important action is shifting from tactical AI deployments to a more strategic approach.
Organizations must integrate AI across the care continuum by aligning AI initiatives with their core objectives, investing in scalable infrastructure, and promoting cross-functional collaboration. This will enable them to build and create trustworthy AI-based systems and provide a foundation for models supporting data security and fair and unbiased outcomes for customers and patients.
Health and Life Sciences organizations must develop capabilities that will optimize their tech investments. These include developing a robust AI strategy, adopting a data management and governance life cycle approach to AI, implementing a responsible AI policy, and improving the monitoring of the deployed AI model's performance. These will help them achieve better ROI from use cases like health care fraud and cost containment, resulting in improved performance and better patient outcomes.
Data and AI Pulse: Asia Pacific, 2024
One of the most critical areas is improving data governance management. Problems in this area have been the significant blocker to earlier AI investments. The focus should be on data integrity and interoperability across systems.
Many organizations bring external expertise from consulting companies to help with AI projects and invest in software platforms or the cloud. They should also consider improving the skills of existing employees, especially health care professionals, to ensure they can leverage AI tools effectively.
There is a trend in the implementation of health data platforms and data interoperability. It is estimated that more and more health care organizations will adopt industry cloud platform technology.
The platforms will help eliminate data silos, structure data from various data sources and streamline workflows around ingestion, mapping, cleansing, deidentification and data quality.
All these are aspects of developing the capabilities to enable the organization to maximize its tech investments. Generating the best possible return on investment in use cases like fraud and cost containment will improve performance and provide options to invest in better patient outcomes.
To achieve a society with more accessible, sustainable and cost-effective health care services, we need a more accurate and timely understanding of how patients interact within the health care ecosystem.
AI is only as good as the data that trains it. Improving data interoperability and governance during the AI deployment could benefit earlier returns on AI investments.