Apollo Hospitals, Microsoft Roll Out AI Agent for Patient Safety Reviews

Apollo Hospitals & Microsoft Deploy AI Agent to Strengthen Patient Safety Reviews

Apollo Hospitals & Microsoft Deploy AI Agent to Strengthen Patient Safety Reviews

The initiative leverages AI to support the review of patient safety incidents and facilitate a continuous learning approach across clinical teams.

Apollo Hospitals, in collaboration with Microsoft, has deployed an AI-powered agent to support patient safety reviews, enabling clinicians to analyse clinical incidents, identify potential risks, and improve hospital safety processes.

The initiative leverages AI to support the review of patient safety incidents and facilitate a continuous learning approach across clinical teams.

It is aimed at helping clinicians examine cases, identify recurring patterns, and derive insights that can inform improvements in patient care and hospital operations.

Commenting on the new system, Dr Rohini Sridhar, Apollo’s Chief of Medical Services said, "The FDEs made that process much easier for us. This ability to listen to the idea, then devise a tool to draw in the data from wherever it is sitting on Fabric, bring it all together in a manner that works for the user. They have done it much faster than I thought would be possible."

Key Features

The system, called the Daily Tiered Huddle (DTH) agent, was developed by Apollo Hospitals with Microsoft’s forward-deployed engineering team using Microsoft Foundry.

Designed as a continuous learning system, it draws on hospital data stored in Microsoft Fabric to compile patient information, summarise reported incidents, prompt clinicians for additional details and suggest case assessments.

The AI agent supports reviews of incidents such as patient falls, medication errors and readmissions. It brings together information from electronic health records, including treatment timelines, medication details, and laboratory results, which clinicians can use during daily clinical governance meetings.

The platform also supports root cause analysis (RCA) and clinical outcomes and quality reviews (COQR), helping clinical teams identify cases requiring further investigation.

Alongside individual case review, the system also incorporates its cluster-analysis capabilities to identify recurring safety patterns, including infections and other adverse event trends across hospital units.

In one instance, it flagged a cluster of central line-associated bloodstream infections in a bone marrow transplant unit, prompting the clinical team to take corrective action.

By identifying recurring incidents and emerging patterns, the system is intended to facilitate early detection of potential safety risks and support timely clinical intervention.

Sharing her thoughts, Juliet Joji Varghese, Apollo’s Director of Nursing in Chennai, said,"At the unit level, we’re ready to answer questions from the leadership because the AI has generated questions to help us prepare. And we’re also learning from what’s happened at other Apollo hospitals. So, we go beyond looking at why or how something happened, to how we can prevent it in future.”

The AI tool is intended to serve as a decision-support tool, with clinicians retaining oversight of the review process.

According to Apollo’s reported usage metrics, patient falls and medication errors declined during July-September 2026 compared with the corresponding period in 2025.

Further, the hospital plans to explore the application of similar AI capabilities in other clinical workflows, including tumour board reviews and doctor handovers.

“There has been an almost 30 percent improvement in staff satisfaction in the way that we are handling errors. All the data is available to them automatically. They can read the data and draw their conclusions. We hope that using AI-enabled tools will allow us to deliver better care,” said Dr Rohini Sridhar, Apollo’s Chief of Medical Services.

Stay tuned for more such updates on Digital Health News

Follow us

More Articles By This Author


Show All

Welcome back.

Pick up where you left off on Digital Health News.

Sign in to DHN

For existing DHN and DHN Plus members.

or
Continue with Google
New to DHN?