At the recent White House health tech event, CMS and more than 60 organizations—including Amazon, Epic, Apple, Google, and OpenAI—committed to modernizing healthcare with better data infrastructure, interoperability, and patient-facing digital tools.
The goal:
- Patient-centered digital health
- FHIR-based real-time data exchange
- National provider directories and ID management
- Trusted app libraries for Medicare
- A definitive move away from paper workflows and fax machines
In a standout video, Amy Gleason, strategic advisor to HHS & CMS shared how uploading her daughter’s records into an AI assistant revealed a long-missed diagnostic discrepancy. It may open the door to a clinical trial after 15 years of waiting.
“Let’s stop accepting what’s broken and start building what works.”
That message resonates beyond AI and patient apps. It applies directly to how we manage testing and validation for the systems behind clinical care.
Are You Still Testing Like It’s 2005?
Too many teams still rely on:
- Excel test scripts
- Manual screenshots saved in folders
- Paper approvals or emailed sign-offs
- Last-minute evidence collection in Word or PDF
These methods weren’t built for today’s complexity, pace, or staffing realities. They're hard to scale, hard to manage, and often delay go-lives. And with growing regulatory expectations and leaner teams, we no longer have time to work around inefficient tools.
What Modern Validation Really Requires
Upgrading your platform is only part of the solution. Validation needs to evolve from the inside out. Here’s what that means:
1. Risk-Based Prioritization
Not every function or integration needs the same level of testing. Risk-based validation means asking:
- What happens if this fails?
- How likely is it to go wrong?
- How difficult is it to detect an issue?
- How directly does this affect patients or compliance?
These are the questions at the heart of Computer Software Assurance (CSA)- the FDA's shift away from documentation-heavy Computer System Validation (CSV) toward a more modern, risk-based approach. CSA encourages:
- Critical thinking over volume of test records
- Testing approaches based on intended use and impact
- More flexibility and less burden for low-risk systems
- The use of modern tools including automation and collaborative platforms
You can read the FDA's draft guidance on CSA here: 📄FDA CSA Draft Guidance
Although CSA is designed for manufacturing and quality systems, its principles apply equally to lab systems, middleware, EHR modules, and other regulated health IT environments.
2. Fit-for-Purpose Tooling
There’s no one-size-fits-all solution. The right mix often includes:
- Automation where test cases are repeatable and structured
- Manual validation where human judgment is critical
- Cloud-based tools that simplify evidence capture and approvals
- Dashboards that show real-time test progress, resource load, and open issues
Spreadsheets and folders are too fragile for this. Testers and approvers need a platform that helps them work efficiently and accurately, not just document effort.
3. Clinician-Friendly Approvals
Clinical users need to see the right information, at the right time, in the right format. That means:
- Clear test context and evidence side-by-side
- Web-based approval from anywhere
- The ability to flag issues without toggling between documents
- Clean audit trails without backtracking through emails
Making the process easy for clinicians improves both engagement and quality of review.
4. Continuous Readiness
Testing is no longer a quarterly or annual project. With frequent upgrades, integrations, and releases, validation needs to be ready when the next go-live or inspection arrives.
That includes:
- Clear test coverage and traceability
- Up-to-date documentation and audit trails
- Fast reporting without last-minute formatting
- A system where nothing falls through the cracks
It’s Time to Stop Tolerating What’s Broken
Tools like Cymetryc are helping health IT teams catch issues earlier, reduce test effort by up to 40 percent, and deliver the visibility and control needed to eliminate delays and uncertainty caused by disconnected, disparate tools.
But this shift isn’t just about technology. It’s about leadership.
Validation doesn’t have to be a bottleneck. It can be faster, more collaborative, and more meaningful. And it must be—especially as we step into the era of AI, where trust in the systems behind the data becomes more critical than ever.
