From reactive to predictive: equipping general departments with AI
Sahyadri Hospitals, a private healthcare provider in India, recently equipped its non-ICU units with AI.
Founded in 1996, the group claims to be the largest hospital chain in Maharashtra in western India, with thirteen hospitals with more than a thousand beds.
THE PROBLEM
For years it relied on traditional patient monitoring systems and regular manual checks by experienced nursing and clinical teams.
“However, this approach had its limitations, including reliance on periodic monitoring and the potential for delays in identifying early signs of deterioration, especially in non-ICU settings,” admitted Dr. Kapil Borawake, Director of Critical Care at Sahyadri Hospitals .
In an interview with Healthcare IT newsDr. Borawake noted these key patient safety challenges: early detection of patient deterioration, improving response times and reducing physician workload.
SOLUTION
“The lack of continuous monitoring in non-ICU units meant a reactive rather than proactive approach to patient safety,” said Dr Borawake.
Recognizing this gap, Sahyadri considered setting up an AI-powered center in partnership with medical device company Dozee.
“The AI-powered command center was launched to bridge these gaps, enabling continuous, real-time monitoring and providing early warnings to healthcare teams. This initiative aims to transition from a reactive care model to a predictive and preventive model, improving patient safety and clinical outcomes.”
MEETING THE CHALLENGE
A proof of concept with Dozee showed positive results, Dr. Borawake said.
“The AI-powered early warning system showed a remarkable ability to do that predict the patient’s deterioration up to 16 hours in advancegiving doctors critical time to intervene. This led to a significant reduction in adverse events and an optimized workload for physicians by streamlining monitoring processes.”
Dr. Kapil Borawake, Director of Critical Care, Sahyadri Hospitals
“In addition, operational efficiencies improved, resulting in cost savings while maintaining high standards of patient care.”
RESULTS
For now, the non-ICU wards of Sahyadri’s Hadapsar and Deccan branches are fully AI-enabled. It includes AI-based ballistocardiography for non-contact monitoring of vital parameters such as heart rate, respiratory rate, blood pressure, SPO2 levels, temperature and ECG, and an EWS that tracks vital trends and provides alerts on early clinical deterioration.
On the clinical side, Sahyadri pursues the following goals: reducing mortality rates, ensuring timely interventions and improving patients’ recovery outcomes.
“We also aim to reduce the number of code blue incidents in non-ICU settings,” Dr Borawake added.
On the operations side, the hospital seeks to improve resource utilization, reduce physician workload and promote seamless communication among care teams.
Sahyadri is expected to expand the AI-powered command center to all thirteen branches “in the coming months.”
“Patient safety is at the core of our activities. With this initiative, we are moving from a reactive model to a predictive and preventive model… In today’s evolving healthcare landscape, integrating AI into clinical practice becomes essential to meet the high standards of care that patients now expect,” emphasizes Dr. Borawake.