New York City is at the epicenter of the COVID-19 pandemic, with area hospitals racing to keep up with the volume of cases as well as with inquiries from those who have been potentially infected with the virus.
Among the medical organizations dealing with the pandemic in NYC is Mount Sinai Health System, which spans eight hospital campuses in the New York metropolitan area, with more than 7,200 physicians. As part of its pandemic response, Mount Sinai was able to quickly develop a text-to-chat platform that allows patients to have live chats with clinicians for care guidance.
In one case, thanks to a live chat, Mount Sinai identified an elderly patient who needed immediate help and dispatched an ambulance in a matter of minutes. In another case, Mount Sinai was able to discover that an infected patient lived in a group home and immediately contacted the home, which isolated the patient, mitigating the spread of the virus in the facility.
"We found during our initial consumer research that healthcare relies too much on legacy communication modalities like phone and secure email, and those modalities are simply not as convenient and scalable as they should be," David Kerwar, chief product officer and head of Consumer Digital Innovations at Mount Sinai Health System, told ITPro Today. "Our chat platform enables real-time integration to Epic, our electronic medical record [EMR], for common transactions like scheduling and registration, and allows streamlined handoff to higher fidelity forms of communication like voice and video."
The text-to-chat platform has a chatbot at the front end to improve staff efficiency and ensure issues are routed to the most appropriate clinical or administrative resource, Kerwar said.
Mount Sinai's Tech Stack
The new text-to-chat app is part of Mount Sinai's larger efforts to make healthcare easier and more accessible from a digital perspective. It's an approach that is built on top of an agile technology stack.
"Our platform strategy requires a modular, component-based architecture powered by scalable microservices," Kerwar said. "Our build strategy started with a novel data model and relationship-based security system that bakes in HIPAA [Health Insurance Portability and Accountability Act] compliance from the ground up."
In terms of the tech stack, Mount Sinai Health System is using cloud infrastructure building blocks from Amazon Web Services (AWS) and custom-built frameworks on top of that. The open-source Go programming language is the primary back-end stack because of its robust characteristics around type safety, lightweight compiling, raw performance and a strong concurrency model, he said.
To help enable the text-to-chat functionality, Mount Sinai is making use of communications APIs from Twilio.
"Twilio provides us with HIPAA-eligible SMS and Programmable Video APIs that allow us to build and scale based on our needs to engage with consumers where, when and how they want to engage via an omnichannel platform," Kerwar said. "Twilio was a natural infrastructure partner with powerful APIs for SMS and video that let us control the user experience."
Integration Is the Key to Success
The path to building the text-to-chat app hasn't been an entirely straight line; rather, it's an evolving process.
"Every time we tackle one use case, we stumble across three other opportunities to improve the consumer experience," Kerwar said.
For instance, the initial idea for chat was just a HIPAA-compliant version of texting, he said. But Mount Sinai developers soon realized that the real value is to build a more graphical user experience that converts tapping on a keyboard to widget-based interactions.
"We also realized that not every chat is one and done—many require follow-up from the agent or provider," Kerwar said. "Therefore, organizing that follow-up leveraging workflow management and CRM capabilities that automate communication back to the consumer became essential too."
Kerwar's team spent a lot of time calibrating its DevOps and observability subsystems to ensure that releases could be made quickly and safely, he said. That led Mount Sinai to leverage many GitHub plugs including CircleCI for continuous integration. Developers are also funneling all data telemetry from Datadog, Bugsnag, AWS and custom monitors into Slack channels, bringing the alert system to where the dev team lives.
"Spending enough time on core foundational work is critical to building a new platform with so many interwoven features," he said.
With the technology in place, the larger challenge that Kerwar and his team had to deal with was the complexity of the healthcare system.
"Our biggest challenge has been rationalizing the myriad systems, processes and protocols that are a necessary evil of the American healthcare supply chain," Kerwar said. "That’s a big reason why we love working with Twilio. They make it incredibly easy to communicate with the patient across different customer engagement channels."