From Joe Squire, Director of Analytics and Data Science at UPMC
Feat. Jen Eberhard, Sr. Consultant of Enterprise Technology at Q-Centrix
- A centralized enterprise data model at Q-Centrix partner, The University of Pittsburgh Medical Center (UPMC), has improved patient access, competitive market analysis, research capabilities, day-to-day quality initiatives, and more.
- Joe Squire, director of analytics and data science at UPMC, stresses the importance of iterative collaboration – balancing technical skills and data expertise with effective communication – to achieve a well-executed data strategy.
- As hospitals transition to a data-driven future, considerations for data governance, security, and automation become crucial for sustained success.
As a clinical data abstractor, Joe Squire spent many of his early days in healthcare thinking about where the data he curated was going. What actions and changes were the data driving? How could the data be made more useful?
Eventually, this curiosity led him to advance his technical skills, become Director of Analytics and Data Science, and develop a dedicated analytics team at UPMC. Today, Joe’s team uses data to make a positive impact across the entire heart and vascular program:
- Subject matter experts partner with individuals across the institute to guide quality initiatives and data literacy within important heart and vascular registries.
- Data scientists and analysts work with UPMC’s research teams to connect and combine clinical registry, EHR, and claims data to produce research data sets that guide improved outcomes.
- Data engineers build the infrastructure to answer critical questions about processes and operations that contribute to improved care and growth across the system in a sustainable manner.
Over the last decade, Joe and his team have connected the heart and vascular institute through clinical data, which has empowered both his team and the institute to grow substantially in their capabilities. Their progress and achievements show the value of a well-executed data strategy and the need for this enterprise, centralized approach to data management. In conversations with Joe, he shared several key considerations for ultimate enterprise data management and utilization. Each builds upon the others for a comprehensive and tactile data strategy.
Best practices for building a data-driven decision-making foundation.
“Prioritize clinical expertise.”
Investing in clinical data expertise is the linchpin for effective data-driven decision-making. Without a clinical perspective supporting data teams, models won’t be as accurate, and the information they contain won’t be as potent. In Joe’s words,
“Data expertise helps you choose the suitable data sources and tools for the job. Clinical expertise is what gives meaning to the data so a story can be told. Each build upon the other, and that clinical perspective has the be the foundation or you won’t get very far.” – Joe
Jen Eberhard, Sr. Consultant of Enterprise Technology at Q-Centrix, also underscored the importance of clinical expertise supporting data analysis decisions.
“The focus of clinical data analytics has shifted towards understanding the impact on patient safety first and foremost. Quality directors, CQOs, and other clinical leadership roles must now look beyond their traditional roles and dive into the analytics space to elevate the effectiveness of data–driven decisions.” – Jen
Jen, a Q-Centrix advisory solutions team member, helps facilities embark on the process improvement journey through technology and data. Her insights brought forth a critical realization – data analytics in cardiovascular spaces demands support, especially for professionals transitioning from roles primarily focused on patient safety.
Speaking with Jen, she emphasized the intricate combination of expertise, continuous learning, and collaboration that lays the foundation for a future where data becomes a catalyst for elevating healthcare facilities and careers.
“For those accustomed to focusing on patient safety, modern clinical data analytics can feel complex and removed from the role. But this isn’t the case anymore. These leaders must constantly learn new information to articulate their projects thoughtfully and precisely, especially if they regularly work with physicians. They will want to be able to prove they know what they’re talking about so their team can engage with them and the data at a high level.
It’s not easy – especially with all the requirements on heart and vascular teams already – but it is crucial. That’s why much of my job is dedicated to helping our partners understand their data and processes for including it in key decision-making.” – Jen
Where to start:
- Consider which data identifiers you collect. What would be the best identifiers to help any data source work well with your other available sources?
- Elevate your data literacy. Consider how you will help people understand the metrics you create, the national bodies’ metrics, and which changes will make an impact on which metrics.
- Train, or hire clinical data expertise. It can take years to train abstractors to a full, comprehensive understanding of data and its impact. For “hair on fire” situations, clinical teams such as Q-Centrix’s help leaders quickly understand their data, trust in its accuracy, and use it for process improvement.
“Then hire and train data professionals.”
Joe stressed the importance of data professionals as another key foundational piece of data-driven healthcare. Whether it’s skilled data engineers or data analysts with extensive warehousing experience, these individuals provide the foundation that every other data project is built upon.
“Support these team members, give them the places and the tools they need to store your data and build a strong foundation – or work with a partner to help you do this. This approach is the best way to ensure your data scales with you and will be useful for a long time to come.” – Joe
Joe also cautioned against prioritizing data scientists over engineers.
“A data scientist should never be your first hire. They can deliver a lot of insight, but they need a solid infrastructure to be potent. If you hire a data scientist before you have your infrastructure in place, they won’t be able to do their job, and they will probably leave before you’ve made any progress.” – Joe
Where to start:
Joe’s perspective speaks to the strategic sequence of building a data-driven team. While hiring a data scientist first may seem like a move to expedite insights, it’s akin to hiring a professional chef without having a kitchen.
Starting with data engineers and analysts ensures the necessary groundwork is laid for a data science team to flourish. In essence, prioritize essentials first to enhance the longevity of your data strategy for a more impactful data-driven decision-making journey.
Building communication pathways will be your most important skill
If there’s one insight from this blog that supersedes all others, it’s this: Communication is everything in the intricate world of clinical data work. In Joe’s words,
“Communication is the most essential skill in data work. The work we do is transforming data into information. To do that, you need to cross the divide between understanding how to work with data at a tabular level and distilling that into a story that people can understand, take learnings from, and use.” – Joe
Most healthcare professionals have been trained in the scientific method. Any data project is going to look similar. Develop a strong hypothesis and then be willing to test it through multiple iterations.
“It’s a back-and-forth process. A data analyst will develop a proof of concept, but it’s not the final product. It needs to be returned to the stakeholders and clinical teams that developed the original hypothesis. Their input is critical in refining and evolving the model for the best answers.” – Joe
Where to start:
Data-driven care becomes reality through united expertise. As the source of the data and questions that drive improvement, the end user (often clinical teams) must assume a pivotal role in the decision-making process. They provide critical feedback, filling in the details that aren’t immediately available or apparent to the data specialists. This dynamic interplay between data experts and end users is where the magic of turning raw data into actionable insights crystallizes.
“Think and work for the long-term.”
Sustainable clinical data practices demand more than short-term triumphs. A strategic approach to data-driven decision-making transcends any individual ownership and thrives on the dissemination of knowledge.
“Ten years from now, are your infrastructure and tools still going to exist? If any individual on your team leaves, will there be processes that fail and fall apart? Don’t have functions that one individual owns; transfer knowledge around.” – Joe
Joe’s team started centralizing their clinical data warehouse in 2015, meaning that by the time they have fully integrated with UPMC’s latest acquisitions, the system-wide project will have taken roughly ten years to complete. Just moving their clinical registry data into one central bucket for their analytics group took several years.
While these timelines depend on the scale of the project, the existing infrastructure, and the internal politics any change initiative faces, the lesson remains the same. Centralizing clinical data management activities is a long-term process.
Where to start:
Building a durable clinical data management system doesn’t happen overnight. It requires a solid plan that is good enough to garner and maintain executive leadership buy-in.
“For any data strategy to deliver on its promise, you must have people willing to support the concept of ‘How can we use data better? How can we build the infrastructure to give us that ability?’
Just like any change initiative, you need executive-level support driving it. If you don’t have that, you’ll be fighting uphill the whole way. Fortunately, these activities often make sense to leadership.” – Joe
“Be willing to have constructive data governance and security conversations.”
How are the goals of data democratization balanced against the necessity of maintaining controlled access? The importance of this question has only increased in recent years. Unfortunately, there’s no single solution. As an advocate of democratizing data, Joe spoke to the nuances of these conversations.
“Giving anybody willy-nilly access to the data is not a path you want to start going down. There’s a point where you must be able to delineate who needs access to what and what access isn’t necessary. That isn’t about saying no, but being willing to discuss what will work.
I’ve been on both sides of this conversation. I’ve had people say no to me when asking for access to their data multiple times. It’s frustrating because you know you wouldn’t abuse the access; you’re just trying to do your job, but the person on the other end doesn’t know that. You have to be sensitive to that confusion.”
Where to start:
The disconnect between intention and perception can often be bridged with transparent communication and a shared understanding of responsible data practices. Joe’s insights remind us that this dialogue is not a straightforward path. Knowing who needs what and having a willingness to discuss what truly works is what shapes a balanced approach.
The future of care: Data-driven, patient-centric
Robust infrastructure, strategic partnerships, and transparent communication: these are some of the most critical forces driving the future of healthcare. Amidst these considerations, Joe’s excitement for the future resonates and echoes Q-Centrix’s.
As UPMC transitions to a single EMR platform, Joe’s team works on new automation strategies in data collection for clinical registries, streamlining datasets across inpatient and outpatient settings. Outside of UPMC, the anticipation of generative AI adds another layer of excitement. The evolution of these tools, particularly within the data and analytics space, presents even more opportunities.
In the context of the broader healthcare landscape, Joe’s team and their progress is a rarity. Q-Centrix exists to help bridge this gap and empower more providers to reap the benefits of this intersection of technical expertise and human dynamics. At the core of these advancements is one goal – enhancing the quality of care delivered to patients.