Joanna Engelhardt, vice president of product management at Health Gorilla
Photo: Anastasia Tveretinova
Good data is the lifeblood of good healthcare. Ensuring the right information is available to the right care team members at the right time is essential to delivering high-quality care.
Two recent trends in healthcare are underscoring this fact.
The first, unavoidably, is artificial intelligence. It is critical to have clean and well-governed to make AI trustworthy and useful. The second is accountable care, where rock-solid datasets are required to help provider organizations reach their value-based reimbursement goals.
We spoke recently with Joanna Engelhardt, vice president of product management at Health Gorilla, a vendor of infrastructure and APIs designed to access patient data securely and comply with CalHHS DxF and TEFCA – where the company also serves as a qualified health information network.
Engelhardt has more than 15 years of experience with electronic health records, ambient clinical documentation, coding and value-based care at companies, including athenahealth, Nuance/Microsoft and Agilon Health.
She offers a fresh perspective on the future of health information management, practical AI in care delivery and building product teams that deliver outcomes, not outputs. Here's what she had to say about the importance of data quality.
Q. Most health IT professionals know you need good, clean and relevant data when working with AI. Why do you say deduplication, plausibility checks and context are prerequisites for useful AI?
A. Before anything else, AI requires a complete and accurate picture of the patient. Finding and filling data blind spots is just as critical as cleaning and structuring the data itself. Too often, AI is limited by incomplete information like missing records, disconnected systems or outdated inputs.
Consider, for example, that in the U.S., patients are still expected to recall their own medical histories in detail: past surgeries, dates, anesthesia reactions, medications and dosages, and other information that should already be part of a shared, longitudinal record.
But patients may have difficulty remembering all the specifics of their medical histories, which means clinical records are often full of gaps. Because AI can only be as powerful as the data it's built on, even the most advanced models will be constrained when a patient's medical history is incomplete.
The challenge goes beyond digitization. Even if every piece of health data were captured and shared, the ecosystem still needs to make that information reliably accessible in real time. Patient data – such as addresses, insurance coverage, medications or new encounters – changes constantly, and health systems must be able to exchange updates seamlessly.
True AI utility in healthcare depends, not only on data quality, but also on responsiveness, completeness and contextual accuracy. Until the industry achieves that, AI's promise will remain limited by the gaps in the data itself.
Q. That was AI. Let's talk the same for value-based care. How must data show up for hospitals and health systems in order for them to succeed in VBC?
A. For value-based care to succeed, data must enable organizations to define their populations, determine appropriate interventions and make those interventions actionable at the point of care. Today, those who seek care aren't always the ones who need it most.
Effective value-based care depends on identifying high-risk, high-cost patients early and proactively addressing their needs. That requires continuous access to timely, accurate and contextually relevant data to identify where risk exists and support clinicians in focusing their time and resources on the patients who will benefit most.
The key is transforming data into actionable intelligence that drives both better outcomes and lower costs. Hospitals and health systems need insights that guide care prioritization, highlight care gaps and measure the impact of interventions. But it's not just about analytics – it's also about user experience.
The technology community should challenge itself to make these insights intuitive and delightful to use so clinicians and administrators can act on them easily and confidently. Value-based care success depends, not just on data availability, but on how seamlessly and meaningfully that data integrates into everyday decision making.
Q. You say when it comes to software and data, outcomes are greater than outputs. Please elaborate, and explain how to build teams that ship what clinicians will truly use.
A. A recent personal experience illustrates this perfectly. When my son injured his knee, he had an X-ray taken at urgent care, which the orthopedic office could not access electronically. The only way to share it was by burning it to a CD, which I had to physically pick up and bring to the appointment, only to find the file unreadable.
Eventually, urgent care emailed me the image, leaving the orthopedic clinician to squint at it on my phone screen. On paper, the system worked: the X-ray was digitized, shareable and technically interoperable. But the outcome was a confusing, inefficient and frustrating process for both patient and provider. This was far from acceptable.
This story highlights the difference between output and outcome. Output is about checking boxes, which revolves around completing tasks or meeting technical requirements. Outcome is about delivering real value and usability.
Teams that succeed focus on solving problems holistically and designing systems that make clinicians' lives easier and improve patient experiences. That requires collaboration between developers, clinicians and administrators to understand real-world workflows, test systems in practice and continually refine them.
Success in healthcare technology isn't measured by the number of features released. Instead, it's measured by how effectively those features improve care delivery and user experience.
Q. Please offer a prediction for 2026 on the subject of healthcare data.
A. Rather than a prediction, this is a necessity: In 2026, healthcare must remove the barriers that restrict data flow. To improve patient outcomes, patients themselves must have true ownership of their data. That means the ability to share it securely, through granular consent, with any provider or application they choose.
The pendulum must swing away from overcautious data hoarding toward responsible, privacy-preserving data mobility. Many lives could be improved simply by allowing information to move more freely and efficiently between systems and stakeholders.
This conversation also needs to expand globally. People don't live, fall ill or receive care within a single country's borders. Many patients have rich medical histories abroad, yet those records rarely follow them. As healthcare becomes increasingly global, the question of what constitutes a "complete chart" will become even more complex.
In 2026, progress should focus on unblocking domestic data flows. In 2027, perhaps, we'll move forward on addressing the international gaps that keep patients from having a truly unified health record.
Follow Bill's health IT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.
WATCH NOW: EHR transitions: How to ensure adoption and satisfaction
