HMH moved all its clinical, imaging, and administrative files to a data lake built with Cloud Storage. By doing so, it replaced siloed data storage strategies with a consistent data platform that can accommodate any kind—and amount—of data.

To improve its ability to serve up the right insights to individual stakeholders, HMH used BigQuery to build a central data warehouse that contains 30,000 tables and more than 100,000 column views of data from Cloud Storage. “The hyper-indexed column store structures we use in BigQuery make it much faster and easier to analyze our data,” Stegman says. “By increasing data integrations in BigQuery, we also simplify data access for our network providers so we can improve continuity of care for our patients.” 

In 2022, HMH adopted Looker to create a semantic data layer that developers named Ekam, which is a Sanskrit word meaning one source of truth. The Ekam semantic data layer gives teams a uniform, governed data platform that can be used for different analytics purposes, as well as BI development tools that require minimal learning curves. Over the course of 12 weeks, HMH worked with Google partners to use Looker to architect and implement the semantic data layer that specifies what data developers and BI consumers can see in BigQuery, as well as standard calculations for measuring KPIs. Going forward, developers can use Looker to add data sources to its Ekam data platform and also modify governance controls and KPI calculations rather than having to implement these types of changes in individual systems.

“Our hospital GPS navigation tool has proven to be a game-changer in the realm of healthcare, revolutionizing how we identify and address critical measures for better patient care and external reporting,” says Dr. Benjamin Schleich, Principal Analytics Translator at HMH. “The secret to our success lies in the seamless integration of Google Cloud and Looker, which facilitates real-time benchmarking, providing a clear roadmap to areas that demand immediate attention. By offering drill-down capabilities and measure-level stratification, our application enables healthcare providers to pinpoint exact areas for improvement.” As a result of these insights, HMH was able to update administrative processes, including how it codes and documents medical procedures. Schleich continues, “The changes we made resulted in better patient outcomes, particularly in discharging patients to their homes.”

By building its modern data platform in Google Cloud, HMH can deliver faster, more comprehensive insights into more data. A query that used to take 2,500 seconds on HMH's legacy system now takes just 2 seconds on BigQuery. “Developers are always conscious of queries' time and cost,” Stegman says. “By using Google Cloud, we remove performance as a constraint because we know it's going to easily rip through billion-row tables, so we can write more complex queries.” Developers can also easily add hardware resources to support specific projects or workloads, paying only for what they use rather than having to provision hardware that could be underutilized. 

With its enhanced abilities to analyze more data, HMH is now using Looker to rewrite hundreds of existing reports to ensure consistency and to better match stakeholders' requirements. HMH is also using its semantic data layer as a foundation for a Google AI solution that will give clinicians rapid, on-demand insights into their patients' clinical data. In a recent test, developers used an early version of their AI tool to immediately summarize 100 pages of data about a patient, including clinician notes. “With our transformative approach and lightning-fast dashboards that we build with BigQuery and Looker, we're harnessing the power of technology to drive innovation, improve patient care, and elevate the entire healthcare experience,” Schleich says.