Infrastructure to Support Healthcare Analytics and Decision Support
Cloud infrastructure supports clinical collaboration in many ways. The combination of redundancy, security, geographic distribution and proximity to end users helps minimize downtime, Saul says. That’s especially valuable for genomics, imaging and other workloads that require a lot of computing resources. An added benefit here is the concept of the clean room, which lets organizations grant access to a data store while retaining ownership of the data.
Meanwhile, unified and normalized data can give organizations a centralized location for conducting real-time analytics, Saul says. “When you have low-latency data processing, and you can integrate with third-party services, then you can build complex yet responsive workflows around your data pipelines.”
Ritu Mukherjee, vice president of product management at Zoom, highlights cloud infrastructure’s potential to support documentation and other automated workflows. Tools such as Zoom Workplace for Clinicians are equipped to do this for both virtual and in-person encounters and can apply different templates so clinicians spend less time formatting and reviewing notes.
In addition, clinicians can see AI-generated patient histories and visit summaries prior to an appointment. They can also set reminders to bring up certain topics during the visit, based on insights surfaced in the summary. “The more that doctors have that rich context at their fingertips, the better the outcomes,” Mukherjee says.
READ MORE: Embrace AI and cloud solutions for optimized collaboration.
Enable Monitoring in the Hospital and at Home
The combination of cloud computing and AI has also significantly improved remote patient monitoring, both in the hospital and at home.
In its infancy, RPM was little more than a dashboard, Saul says. “You need someone to monitor the data — and even if things were flagged, it was difficult to process them and understand their significance.” Plus, given the sheer volume of data generated in inpatient care — blood pressure, heart rate, temperature, oxygen saturation and so on — little has historically made its way into the patient’s chart beyond infrequent anomalies.
“AI can discern subtle patterns that may not be noticed, but you need the fidelity of a full data stream. That requires infrastructure that few hospitals can support on-premises,” Saul continues. “Now, we can uplevel our care with data processing what wasn’t possible before.”
Saul points to the AWS partnership with Validic, which aggregates data from medical and consumer health devices. An AWS analytics layer atop Validic’s data streams can not only detect anomalies but also send alerts to physicians. “The volume and speed of the cloud can process data in real time and generate insights that drive clinical decision-making.”
As capabilities evolve, providers may find themselves with fewer reasons to see patients in person, Saul says. Video streaming, real-time data processing and asynchronous communication are poised to make clinical-grade physician therapy assessments, heart rate monitoring and other services possible from home.
