The global healthcare sector is grappling with unprecedented challenges. For decades, chronic underinvestment and recruitment hurdles have converged with the soaring demand for services from an aging population, creating significant strain.
This perfect storm has led to fragmented access to care and alarming rates of stress and burnout among staff. The World Health Organization (WHO) warns that the current shortfall of healthcare workers is projected to escalate to a staggering 11 million workers by 2030.
In this urgent quest for sustainable solutions, many healthcare providers are now placing their hopes on agentic AI. A recent report by KPMG reveals that a significant majority, 68%, have already begun integrating AI agents into their operational frameworks.
This innovative technology is being deployed across various functions, from automating complex back-office processes to collaborating with medical teams and even triaging patients. The ultimate goal is to alleviate the cognitive load on clinicians and elevate the quality of patient care, especially as the human healthcare workforce dwindles.
Beyond Traditional Digital Tools
The journey of digitalization in healthcare, until now, has often presented a mixed bag of results. Many staff members have found that slow or outdated technology inadvertently adds to their administrative burden rather than reducing it.
For example, the migration of U.S. patient data to electronic health records (EHRs) in the early 2000s, while a step forward, has left data fragmented and overly reliant on cumbersome manual inputs. This often hinders seamless information exchange and efficient workflows.
Even newer innovations like telehealth services and digital care tools, such as remote monitors, have had their limitations. Dr. Ashis Barad, MD, Chief Digital and Technology Officer at Hospital for Special Surgery (HSS), notes that while these technologies improved access by removing geographical barriers, they often struggled to replicate the quality of in-person care or fully earn patient trust.
However, Dr. Barad insists that agentic AI represents a fundamental shift from these earlier technologies. Unlike systems that depend on rigid frameworks or manual data entry, AI agents can adeptly navigate nuanced and complex scenarios.
They possess the unique ability to make autonomous decisions, retrieve crucial information from expert clinical sources, and continuously learn and improve over time. As Dr. Barad vividly puts it, “Agentic AI takes your workflow and collapses it, augments it, supercharges it, and makes it more performant,” empowering clinicians to dedicate their expertise to higher-level patient care.
Agentic AI in Action: The HSS Story
At the Hospital for Special Surgery (HSS), a leading academic medical center specializing in musculoskeletal health, agentic AI has already demonstrated transformative capabilities. One notable success lies in streamlining complex backend processes, such as insurance claims.
Previously, these claims could take several weeks to process, requiring extensive involvement from HSS staff and often a third-party contractor to manage the sheer volume. Now, AI agents handle an impressive 1,100 claims per month.
Since their implementation nine months ago, these agents have dramatically reduced the appeals stage from 45 minutes to five and boosted the success rate of those appeals from 65% to a remarkable 100%. This efficiency gain has allowed HSS to manage all claims entirely in-house.
Building on this success, HSS is now expanding its AI agent deployment into non-clinical, patient-facing settings through an AI scheduling and triage service. This initiative, developed in collaboration with enterprise agentic AI developer Ema Unlimited, is accessible to patients 24/7 via web, text, or phone.
Utilizing advanced conversational AI, the service asks clarifying questions about a patient’s condition and then intelligently books appointments with the most appropriate clinician. It meticulously factors in location, insurance coverage, and physician availability, effectively “completing the whole loop,” as Dr. Barad describes it. This system, trained on HSS’s comprehensive context, rules, and knowledge base, provides patients with streamlined access to world-leading specialist care.
Building Trust and an AI-Powered Future
Given the high-stakes decisions delegated to AI agents, HSS has meticulously embedded safeguards into its triage service. Sensitive, complex, or uncertain scenarios are always escalated to human specialists, ensuring patient safety and peace of mind.
Every decision made by the AI agent is fully auditable, and human staff retain the ability to intervene at any point. Patient data remains secure, and the system is rigorously trained on all HSS protocols, policies, and care pathways, striking a crucial balance between efficient automation and human-informed decision-making.
As agentic AI becomes more prevalent, healthcare providers must prioritize the integration of such robust guardrails. At HSS, all decisions regarding this technology are vetted through a dedicated AI subcommittee, co-chaired by Dr. Barad and a senior nursing executive.
This ensures that AI agents impacting patient care undergo far more rigorous scrutiny than those handling only backend processes. To further democratize access and understanding, Dr. Barad plans to establish a dedicated AI lab at the HSS main campus in New York City.
This lab will offer informative classes and one-on-one training, making agentic AI accessible to all staff looking to understand or even build these agents. This aligns with research from Deloitte, indicating that leading adopters redesign end-to-end workflows rather than focusing on isolated use cases.
The key, it appears, is to integrate AI agents across the entire enterprise, treating them not as narrow solutions but as a general-purpose technology, analogous to electricity. Achieving this value requires providers to lay a strong foundation, particularly a unified data strategy that integrates fragmented data sources into a single, comprehensive source of truth.
Fragmented data, often spread across multiple departments and legacy IT systems, impedes AI agents from seamlessly retrieving and assimilating tacit knowledge—a core differentiator of this technology. By enhancing data interoperability, patient-facing AI agents at HSS can draw from a patient’s full clinical history and current symptoms, making informed decisions and escalating situations to specialists when necessary.
Dr. Barad sees immense potential for AI agents to revolutionize healthcare, easing the pressures on resources, access, and patient care. He envisions a future where 90% of non-clinical health-care tasks could be administered by AI agents, freeing clinicians to focus on “white-glove work”—the most complex, specialized, and sensitive cases.
This optimism is widely shared across the industry, with KPMG’s research indicating that 84% of providers are already comfortable entrusting AI agents with decision-making for specific processes. As Dr. Barad concludes, “We’re spending so much time on keyboards and computers right now that we’re actually not doing what we should be doing. This is going to rehumanize health care.”
Source: MIT Tech Review – AI
