Health care organizations can use AI to solve practical business problems in transformational ways
November 12, 2020 at 12:00:00 AM
Artificial intelligence (AI) is already helping to make some aspects of health care more efficient. As data—including health care and social determinants of health (SDoH)—becomes more interoperable and secure, we expect AI will become a critical engine driving digital transformation and data analytics. For example, with many health systems reeling from the COVID-19 pandemic, AI-enabled solutions could help reduce costs by automating some types of manual work. AI could also be used by health plans to develop new products and improve consumer engagement.
Kumar Chebrolu, managing director, Deloitte Consulting, LLP
The pandemic overwhelmed many hospitals and health systems and exposed limitations in their ability to deliver care and reduce costs. Since March, many health systems have experienced a significant shift to virtual health, fueled by necessity and regulatory flexibility, according to results of Deloitte's 2020 survey of physicians. The pandemic also opened the aperture for AI and digital technologies to solve problems.
Our latest report, Smart use of artificial intelligence in health care, offers a thorough look at how health care organizations are using AI. We determined that health care organizations can scale up their AI investment by pairing it with a robust security and a data-governance strategy.
AI is a key component in the future of health
AI is already being used to automate processes in health care. In our vision of the future of health, we view radically interoperable data as central to the promise of more consumer-focused, prevention-oriented care. Data analytics will be critical for generating actionable insight from the vast data that will be generated by ubiquitous sources. AI is already embedded into data analytics and is likely to become even more so in the future.
AI uses algorithms and machine learning (ML) to analyze and provide insights based on data. It also can be used to automate some types of repetitive work and has the potential to augment decision-making among operational and clinical staff. By reducing the time spent on administrative tasks, humans can focus on more challenging, interesting, and impactful management and clinical work.
Today, health care organizations often experience pervasive problems across their value chains, which can span every process along the continuum. In the future, health care organizations that apply AI across every process—from care to cure—will likely be able to improve the health and well-being of consumers.
Five areas AI could improve
Many day-to-day, non-clinical operations (e.g., submitting and paying claims) are ripe for AI. For example, natural language processing can understand unstructured data from electronic health records. AI can automate tedious administrative work and generate insights for monitoring fraud and abuse or physician practice patterns. Here are five areas where we see AI having the biggest impact in health care:
1. Improving patient care: AI-based solutions can effectively streamline diagnostics and suggest personalized treatment options by tapping into vast amounts of structured and unstructured medical data across institutions. Real-time, data-driven insights can help clinicians and care teams make more-informed decisions, which can be altered and implemented based on their personal expertise.
2. Simplifying administrative processes: Operational issues (e.g., number of staff, staff availability, skills, and specific equipment required) have been a major challenge for many health systems since the onset of the pandemic. AI-powered solutions can assist in accurately scheduling and planning clinical staff rotation.
3. Catching errors and reducing burnout: AI can help minimize patient risk by identifying medication errors, which traditional rule-based clinical-decision support systems might miss. For example, an algorithm might determine the dosage for a written prescription is incorrect and send an alert to the physician’s system. At the same time, another algorithm could prioritize the alerts that the physician sees and remove those that don’t require confirmation or intervention. This can help reduce alert fatigue.
4. Increasing efficiency and improving the patient experience: The ability to quickly examine large amounts of information can help hospital and health plan administrators optimize performance, increase productivity, and improve resource utilization, resulting in time and cost efficiencies. Additionally, AI-enabled solutions can speed up and strengthen the insight-generation process by allowing the organization to gain the holistic picture it needs to make data-driven decisions. AI can also deliver personalized experiences by facilitating conversations with patients through virtual assistants.
5. Identifying fraud, waste, and abuse (FWA): AI-based analytics can help health plans detect and reduce improper billing practices, which can streamline reimbursement. Additionally, pharmacy benefit managers (PBMs) can use AI to detect potential problems with prescription claims data, reduce improper claims payment, and minimize the manual effort often required to review claims. AI can also help to effectively and proactively identify potential fraud and abuse.
Two key AI applications in health care
1. Enhancing customer experience through conversational AI: Growth in virtual health helped to lay the groundwork for a digital front door. Deloitte’s DocTA is an omni-channel experience powered by conversational AI integrated at the contact center to resolve administrative inquiries and steer consumers to appropriate care without human intervention.
2. Realizing efficiency through FWA detection and prevention: Deloitte Risk and Financial Advisory’s Program Integrity solution integrates and analyzes claims data using ML models to identify potential fraud and improper payments.
What are some of today’s major challenges?
Deloitte’s State of AI survey, which was released in late 2019, looked into how organizations are adopting, benefiting from, and managing AI technologies by industry. The survey found that about 75% of large organizations (e.g., annual revenue of over $10 billion) invested more than $50 million in AI projects/technologies, while approximately 95% of mid-sized organizations (e.g., annual revenue of $5 billion to $10 billion) invested less than $50 million. And 73% of all organizations said they expected to increase their funding in 2020.
The three most cited reasons for using AI were to make processes more efficient (34%), enhance existing products and services (27%), and lower costs (26%). Respondents from health care organizations reported that their main concerns about AI investments were the cost of the technologies (36%), integrating AI into the organization (30%), and implementation issues, including AI risks and data issues (28%).
Investing in AI while confronting risk
As investments in AI increase, and as AI-powered solutions become more widespread in health care settings, the industry should address a new set of challenges both from the data used (including cyber threats) and the potential for bias in the AI algorithms. The strategy should comply with regulations and patient-privacy rules.
AI algorithms can create risks such as variability in patient diagnoses and treatment, data bias, and traditional IT risks such as change management. Health care organizations should work to verify the integrity and accuracy of their AI algorithms by focusing on data strategy, testing, and monitoring. Best practices for health systems and health plans range from confirming stakeholder buy-in, creating a set of strong governance practices, safeguarding patient data privacy, and implementing protection from cyber threats. Providing transparency to consumers about how their data is used is a key component of AI governance.
Health care organizations should consider ramping up AI investments
Every health care stakeholder has opportunities to use AI effectively.
• Hospitals and health systems: To prepare for the future of health, hospitals and health systems should advance past the experimental stage by learning to leverage data and advanced AI as a core capability. Many health systems are under financial stress. Their short-term focus might be on AI approaches that can help reduce costs. Examples include provider profiling for equipment (e.g., supply chain), FWA detection and prevention, and automating health care operations. Over the longer term, health systems can invest in more transformative AI applications to improve their competitive positioning, achieve profitable growth, engage consumers, and deliver more personalized customer experiences. Health systems should actively cultivate their relationships with AI start-ups, technology and professional services firms, and academia, and consider taking a more active role in AI innovation. They should also encourage stakeholders—including physicians, clinical team members, and administrative staff—to champion AI and promote an AI-augmented workforce.
• Health plans: Some health plans are already investing heavily in data and AI-based analytics. AI could be used by health plans to develop more cost-efficient health coverage options, improve consumer engagement and satisfaction, create insights for medical and organizational management, and generate better outcomes for clients and members to complement the company’s overall positioning and strategy.
• Pharmacy benefit managers: PBMs could use AI to improve their communications and interactions with health plans. AI could help optimize distribution activities, as well as claims and clinical program management. PBMs should consider investing in AI to enable proactive and personalized engagement, which can result in effective savings programs and pharmacy programs designed to improve the patient journey.
AI is already beginning to deliver significant business benefits throughout the health care sector, and it has the potential to shape it more dramatically in the future. Health care organizations that remain in the experimental pilot phase too long could be left behind by both traditional and unconventional competitors.