Overview

As software-based medical products continue to reshape diagnostics, monitoring, and therapeutics, the regulatory spotlight has shifted sharply toward clinical evidence. Nowhere is that more critical—or more misunderstood—than the requirement to define a robust Clinical State of the Art (SotA) during Software as a Medical Device (SaMD) development.

For companies submitting SaMD for FDA or EU MDR approval, a poorly executed SotA can derail your regulatory strategy. But it’s more than a compliance task—it’s foundational to clinical safety, performance benchmarking, and ultimately, user trust.

What Is Clinical "State of the Art"—and Why It Matters

Under EU MDR Annex XIV and increasingly in FDA interactions, companies must describe “the current knowledge/state of the art in the corresponding medical field.”

That means going beyond your device or algorithm to include:

• Peer-reviewed literature

• Clinical guidelines and consensus statements

• Real-world clinical practice

• Safety and performance benchmarks of alternative or predicate options

Why it matters: Regulatory reviewers use this as a reference frame to evaluate your product’s safety, performance, novelty, and risk-benefit balance. If your clinical evidence is unsupported by current practice—or worse, contradicts it—approval may stall or fail.

Four Pitfalls SaMD Developers Should Avoid

Focusing Too Narrowly on the Product

Many teams build their SotA around the innovation—what their tool does—without framing it against how clinicians currently solve the problem. For example, an AI triage tool should be benchmarked against manual triage protocols, not just past AI systems.

Over-Reliance on Peer-Reviewed Literature

Guidelines from professional societies, health technology assessments (HTAs), and even consensus white papers are equally valid. Ignoring grey literature can leave dangerous evidence gaps.

Weak Search Methodology

Using only one database or excluding older-but-relevant studies can be seen as cherry-picking. Regulators want transparency in how you searched, what you excluded, and why.

Failure to Address Alternative Pathways

Every device has competitors—even if they’re non-digital. If you fail to compare your SaMD to the standard of care, you risk inflating benefits and underplaying risks.

Best Practices: How to Get SotA Right

• Start Broad, Then Narrow

  Begin with a scoping review of the medical problem, current practice, and unmet need. Only then anchor to your SaMD.

• Use Multiple Sources

  Search PubMed, Embase, Cochrane, clinicaltrials.gov, and regulatory databases. Include guidelines, expert consensus, and HTA reports.

• Engage Experts Early

  Clinicians can surface practice patterns or guideline conflicts that literature alone may miss.

• Update Frequently

  The clinical landscape evolves rapidly—especially in fields like oncology, cardiology, and radiology. Re-review your SotA every 3–6 months during development.

Where Gesund.ai Comes In

At Gesund.ai, we’ve built a platform specifically designed to help companies develop and validate AI-enabled SaMD—with SotA rigor embedded.

Our platform enables:

• Curated Literature Integration: Support for embedding peer-reviewed and grey literature directly into validation workflows, annotated and searchable across model versions.

• Search Traceability & Documentation: Teams can log, tag, and version all sources used in SotA analysis—with review-ready exports for FDA or MDR documentation.

• SotA-Aware Validation Metrics: Benchmark your model’s performance not just against internal thresholds, but against clinical comparators defined in SotA.

• Clinician-in-the-Loop Collaboration: Bring subject matter experts into the validation workflow via secure, no-code interfaces—ensuring your model fits real-world care pathways.

• Audit-Ready Clinical Justifications: Auto-generate sections of the Clinical Evaluation Report (CER) and Summary of Safety and Clinical Performance (SSCP) using platform-logged evidence and annotations.

Why This Matters Now

With the rise of AI in healthcare, the line between software and clinical intervention is blurring. That’s why both the FDA and EU are placing greater weight on how well you contextualize your solution—not just how smart it is.

A strong Clinical State of the Art assessment shows regulators:

• You understand your product’s clinical environment

• You’ve benchmarked against reasonable alternatives

• Your risk-benefit claims are grounded in current practice

At Gesund.ai, we help ensure that’s more than a PDF—it’s a validated, living part of your development lifecycle.

📍In SaMD, it’s not enough to be innovative. You have to be clinically relevant, evidence-aligned, and regulator-ready.

→ Learn how Gesund.ai helps clinical and regulatory teams streamline validation, build compliant SotA documentation, and accelerate AI-based approvals: https://gesund.ai/get-in-touch-gesund

Bibliography

1. Scarlet. How to Avoid Common Pitfalls for SaMD Clinical State of the Art.

   https://www.scarlet.cc/post/how-to-avoid-common-pitfalls-for-samd-clinical-state-of-the-art

2. Topflight Apps. Mastering SaMD Clinical Evaluation: A Comprehensive Guide.

   https://topflightapps.com/ideas/samd-clinical-evaluation

3. Mantra Systems. Five Common Pitfalls in Writing a Clinical Evaluation Report (CER).

   https://www.mantrasystems.co.uk/articles/five-common-pitfalls-when-writing-a-clinical-evaluation-report-cer

4. FDA. Software as a Medical Device (SaMD): Clinical Evaluation Guidance.

   https://www.fda.gov/media/100714/download