How Manipal Hospitals is driving tech innovations in healthcare

At the peak of the pandemic in India, patients who fell sick but were unsure if they should visit the emergency department turned to Manipal Hospitals’ video consultation app on their smartphones to get medical advice from a doctor.

By enabling doctors to triage a patient’s symptoms, the app had not only cleared up any confusion for patients about their symptoms, but also helped to take some pressure off the emergency departments in Manipal’s hospital network.

Today, the app has expanded its capabilities beyond telehealth and now lets patients view lab reports, pay bills, book medical appointments, and even request for home healthcare services, underscoring Manipal’s ongoing efforts to harness technology to improve patient care.

The hospital group, which operates 33 hospitals across 17 cities in India, also offers home delivery of medications to reduce waiting time at pharmacies, keeping patients who would otherwise drop out of care within its healthcare network.

Patients who have undergone heart stenting, angioplasties and joint replacement surgeries can also be put on a home monitoring platform powered by Singapore-based ConnectedLife, which combines the use of Fitbit fitness trackers with its virtual platform to monitor patient progress.

Karthik Rajagopal, chief operating officer of Manipal Hospitals, told Computer Weekly on the sidelines of Google Cloud Next ’24 that the home monitoring platform serves as a “gatekeeper” that directs patients to the right specialist for post-surgery care, enabling them to get back on their feet sooner than later.

Rajagopal acknowledged that remote monitoring, however, is still a work in progress. For now, patients can only use Fitbit trackers, but if other types of monitoring devices are added to the mix, then adoption rates should improve, he added.

Besides patient care, Manipal is tapping technology to improve its ward operations. Partnering with Google Cloud, it built a mobile app that runs on the hyperscaler’s Compute Engine infrastructure to automate nurse rostering, reducing personnel requirements and freeing up senior nurses for more valuable tasks.

Before the app was launched, scheduling was done manually by senior nurses and could take up to 45 minutes a day. Sometimes, too few nurses were rostered on certain shifts, drawing complaints from patients and doctors. The app has helped to reduce stress on nurses, enabling them to focus on patient care.

Rajagopal said plans are afoot to work with Google Cloud to leverage artificial intelligence (AI) to further support nurses and improve their productivity. “For example, AI can help to decipher discharge summaries and facilitate shift handovers using large language models [LLMs],” he added. “Our intent is to look at two or three very large initiatives that will enable caregivers to spend more time with patients.”

Madhur Gopal, associate vice-president of marketing at Manipal Hospitals, said the healthcare service provider plans to take existing LLMs, finetune them using data from its digital platforms, and focus on use cases rather than the underlying technology.

Rajagopal added that with multiple languages supported in LLMs like Google’s Gemini, Manipal could translate documents such as discharge summaries and radiology reports, which are typically in English, into Indic languages for patients. “There is a certain amount of maturity in Google’s ability to look at not transliterating, but translating information into languages that patients prefer,” he said.

But whether global LLMs, which are primarily trained in English, can suit the needs of India’s diverse multilingual population remains to be seen. In recent years, Indian firms and organisations have been building indigenous LLMs which are touted to perform better than global models in understanding cultural and linguistic nuances.

“We will only see the number of models grow – there will be first-party models and many more special-purpose models,” said Vikas Arora, director of sales and consulting at Google Cloud India. “Which is why Google is taking the approach of allowing customers to choose the models that suit them through our Vertex AI platform, which has a model garden and tools to help them build on those models or create their own models”.