Secrets to unlocking the value of clinical data from a team of data engineers, analytics aficionados, and digital architects
From evaluating physician performance to increasing marketing efficiency, there are few initiatives clinical data cannot support. The only problem: 97 percent of clinical data is never used. Diverse sources and structure across the equivalent of 25 trillion pages of information make clinical data an “untamed wilderness,” leaving clinical teams with little utility and control over the data they collect.
To solve this challenge, Q-Centrix product development senior director, Bryan Wojnowski, enlisted a team of experts. Abandoning old technologies and processes, they began to build a modern, centralized clinical data infrastructure to unlock the value of clinical data.
“Our goal was simple: centralize data in one place and make it analytic ready. Most of us cannot fully appreciate the complexity of that goal. I’m proud of our team’s dedication to doing what was right, not what was easy.” – Bryan.
When the project began in 2019, the Q-Centrix team identified two vital conceptual opportunities, the first of which was data structure. To offer easy data access across departments and facilities, data would need to be centralized in one place and organized efficiently. The second opportunity was clinical data reporting and analytics that would truly meet the requirements of hospital leaders across the enterprise.
Creating a clinical data repository – or data lake
Lead data architect Sujay More joined Q-Centrix in 2020 to support the journey to modern data infrastructure for the existing offerings and the challenges ahead. When he joined, Sujay was one of the few team members tasked with creating a Q-Centrix clinical data repository. This repository is designed to consolidate data from many disparate clinical sources in one place. The data can then be used to support a varied list of uses. Typically, this data’s highest use case is to facilitate the management of specific clinical departments.
In the two years since Sujay began working on this project, five additional engineers have joined his team. Together, they continue looking for ways to improve health information exchange.
“Think of data sources as having their unique personalities and languages. This makes it extremely difficult to get the data to work together. To do so, we need to understand the data source – and the processes by which it is collected and organized – as quickly as possible. From there, we can begin to create meaningful interactions that give way to transformational insights.” – Sujay
Why is a clinical data repository necessary?
When a clinical data repository holds data organized for analytics, it becomes what’s known as a clinical data warehouse: a digital tool used to accept, store, and translate information from a rapidly growing list of data formats. When used together, these tools provide several unique advantages to healthcare teams.
More value in less time
For the staff using the clinical data, the efficiency obtained from centralized data translates to new capabilities to track and trend critical information, such as patient flow or physician performance, in real-time.
“Typically, the process of getting usable information from registry data is lengthy. A data warehouse allows us to provide feedback in two hours, and we’re working on making that even faster. That’s the power of storing all your data in the same place.” – Sujay.
Flexibility to keep up with frequent industry changes
As healthcare processes and protocols change, combining different data can be tremendously cumbersome, resulting in updates that cause even greater burdens for the team. However, when data is stored in the same place and in the same language, changes become simple, even intuitive.
While striving to build analytics for all major use cases, the Q-Centrix team also prioritized the flexibility of ad-hoc reporting to ensure their tools continue to best serve end users even as the industry changes.
Lehigh Valley Health Network is one of the hospitals that has benefited from the ad-hoc reporting abilities made possible by the Q-Centrix data repository. Specifically, the health system creates reports for HBIPS and Severe Sepsis measures that point to specific improvement opportunities most important to their overall strategy.
Reduction in waste and inefficiencies
Armed with centralized and reliable data, hospital staff have a lot more time to do what they’re best at – adding value to their organization and improving patient care. For example, care coordinators using centralized data tools can reduce or remove the burden of abstraction from their daily tasks. They are then free to focus on monitoring care and implementing monthly improvements.
Understanding the clinical data analytics needs of healthcare professionals
A clinical data repository alone is not enough to solve healthcare teams’ complex challenges. With all the data in one place, these individuals need a way to interpret the data and understand actionable next steps.
Alicia Johnson, Q-Centrix business intelligence product manager, joined Q-Centrix in 2014. When Alicia began her career at Q-Centrix, her colleagues manually reviewed every detail of operational and clinical data in vast Microsoft Excel spreadsheets. While Excel is a familiar resource to many clinical data experts – and a testament to their patience – the manual analysis was not very efficient. Finding answers to simple questions often involved a lengthy, multi-team effort. Accordingly, Alicia regularly built massive workbooks of information to keep up with demand.
“Our users wanted every detail available. Building something new allowed us to determine which details were truly actionable – delivered answers – and which did not.” – Alicia.
To create tools that help clinical teams access the information they needed, Alicia and others collaborated with the Q-Centrix partner community and the Q-Centrix team of experts. They identified the most valued data and metrics, how and what they were being used for, and what questions remained unanswered.
While Alicia focused on operational data, the data and analytics product manager, Cody LaPapa, led the development of clinical measurement and outcome dashboards. While interviewing key stakeholders, Cody and the team understood that the primary goal was to offer users quick and easy access to analyze their registry data and improve care delivery.
“We focused on understanding established and nationally monitored metrics first, like the benchmarks associated with the National Cardiac Data Registry CathPCI registry. Then, we query hospital partners, executive leaders, and clinical experts from our business to gain the best perspective on what clinical teams need. We found that every facility is different. In some cases, the priority may be surgeons who are adamant about viewing their performance measures. In other cases, the hospital board members’ needs are a much higher priority. We obtain feedback from all stakeholders to produce something universally valuable. The best way to produce something universally valuable is to give users the flexibility to create exactly what they need.” – Cody.
Building better healthcare data analytics tools
With typical registry participation, clinical teams abstract, audit, curate, and submit clinical data directly to the professional organization using a capture and submission tool. They then receive performance summaries that may or may not include actionable insights. While some of these registries deliver updates once a week, others share aggregate findings a few times a year, if ever. The system Cody and the team developed offers partners updated analytics every two hours.
“Our reports support both our hospital partners and our internal team of experts. They validate the accuracy of the data before submission, identify measures that offer opportunities for celebration or improvement, and create filtered lists to answer specific questions. The goal was to create flexibility that enables every use case—and I think we have succeeded.” – Cody.
Connecting the data to a purpose
Teodoro (Teo) Gomez, a lead user experience architect at Q-Centrix, was responsible for connecting the data to purpose through insights-driven reports. To Teo, that undertaking was one of the most challenging projects of his life. Throughout his five and a half years at Q-Centrix, he studied the way clinicians employ clinical data.
“Our hypotheses of how clinical teams engaged with data proved false many times. While working with other teams, we quickly learned to distinguish between ‘not enough to use’ and ‘too much to make sense of.’ We went from one chart that could do multiple things to multiple charts for everything and discovered how important flexibility was in our user experience. Now, the tool can be used as intended, and that progress is exciting to see.” – Teo.
As Teo and his colleagues worked through the build, they focused on adding functionality, enhancing the experience, and improving performance. Today, the tool offers users access to their data through reports within seconds. Tomorrow, that could be just one small piece of what is possible.
What’s next for the clinical data pioneers?
The Q-Centrix team members empowered a population to make real-time, data-driven decisions from clinical data. What is next for these pioneers?
Make fast even faster and make flexible even more flexible to expand the use cases of clinical data. For Sujay’s team, this means incorporating more data sources into their data lake and expanding the clinical data warehouse, ensuring that all data is accessible. For Alicia and Cody, that means creating more reports and adding new metrics to help various teams more quickly understand their data. And for Teo and his team, it means working to create experiences where users easily connect purpose to their clinical data.
“We see the improvements, and we’re incredibly proud, but our team is open to anything that can help make a better user experience. Ultimately, this project is about helping more hospitals use the assets they already have to guide industry-changing development.” – Teo.