Decision Guide to Help Determine Digital Health Regulatory Pathways

I’ve spent years on the payer side, building GTM strategies and pricing models for digital health products. But here’s the truth: the FDA approval pathway is still murky, even for people who live in this world.

The FDA website has guidance documents—lots of them. But if you’re a founder trying to figure out whether your product needs approval, it’s overwhelming. The guidances are dense, full of nuance, and frankly, hard to parse without a regulatory background.

This matters because getting your regulatory strategy wrong is expensive. I’ve seen startups delay commercial launch by 2+ years or burn millions in unnecessary capital because they didn’t think through the FDA pathway early enough. The question of “Do I need FDA approval?” isn’t just a legal checkbox—it fundamentally shapes your business model, target market, and how you’ll get paid.

So how do you figure this out? I’ve created a decision tree to help you think through it. Quick disclaimer: I’m not a regulatory expert, and this isn’t regulatory advice. But if you need help with your regulatory strategy, reach out—I can connect you with the right regulatory experts.

The Most Important Question in Digital Health

Understanding whether your digital health product requires FDA approval can save you millions of dollars and years of development time. Getting this wrong can mean:

  • FDA warning letters and shutdown orders
  • Wasted development on products that need approval
  • Unnecessary delays pursuing approval you don’t need
  • Legal liability and patient safety issues
The Core Principle:
It’s NOT about how sophisticated your technology is, what disease you address, or whether you use AI.

It’s about WHAT YOUR SOFTWARE DOES and WHAT CLAIMS YOU MAKE.
START: Does your digital health product need FDA approval?

Follow this decision tree step by step. Answer each question honestly based on what your product actually does (not what you wish it could do).

1
Is your product SOFTWARE or does it have a software component?
NO – It’s hardware only
Physical medical devices without software (surgical instruments, implants, diagnostic equipment, etc.) → Different regulatory framework, not covered by this decision tree.
YES – It’s software or has software features
Continue to Question 2 →
Examples: Mobile apps, web platforms, AI algorithms, SaaS products, embedded software, predictive analytics tools
2
Does your software perform ANY of these ADMINISTRATIVE/IT functions?
YES – It’s primarily administrative
These functions are EXCLUDED from medical device definition by statute:
Excluded functions:
• Electronic health records (EHR)
• Appointment scheduling
• Insurance claims/billing
• Telemedicine platforms (video calls)
• Patient portals
• Hospital bed management

⚠️ If it ONLY does these → NO FDA APPROVAL NEEDED
If it does these PLUS medical functions → Continue to Question 3
NO – It does more than administrative functions
Continue to Question 3 →
3
Does your software make MEDICAL CLAIMS or is it intended for MEDICAL PURPOSES?
NO – General wellness, fitness, or lifestyle only
If your software is ONLY for general health/wellness WITHOUT medical claims:
General wellness examples (NO FDA):
• Fitness tracking (steps, calories)
• Meditation apps
• Nutrition logging
• Sleep tracking (not diagnosing disorders)
• Lifestyle coaching

✓ RESULT: NO FDA APPROVAL NEEDED
YES – Makes medical claims or intended for medical use
Continue to Question 4 →
Medical claims include:
• Diagnose, treat, cure, prevent disease
• Manage disease symptoms
• Replace/augment clinical decisions
• Analyze medical data for clinical insights
5
🎯 CRITICAL: What does your software’s OUTPUT do?
A) Provides information/education ONLY
→ Likely NO FDA
Examples: Medical reference apps, health education, symptom checkers with disclaimers
B) DIAGNOSES disease or conditions
→ YES – FDA required
Examples: AI analyzing images, ECG interpretation, diabetic retinopathy detection
C) TREATS, manages, or prevents disease
→ YES – FDA required
Examples: Digital therapeutics, disease management platforms, dosing calculators
D) PREDICTS disease risk or health outcomes
→ USUALLY requires FDA
Examples: AI predicting cardiovascular risk, multi-modal health trajectories, genetic risk assessment
F) CONNECTS patients to healthcare providers
→ NO FDA approval needed
Examples: Telemedicine platforms, patient-provider messaging
⚠️ The PROVIDER makes all medical decisions, not the software

📊 Outcome Summary

✅ NO FDA APPROVAL NEEDED
Your product qualifies if it:
→ Performs administrative/IT functions only
→ Is general wellness/fitness without disease claims
→ Provides health information/education only
→ Connects patients to providers (telemedicine)
→ Displays user-entered data without interpretation
What you CAN do:
→ Launch immediately without FDA submission
→ Iterate and update freely
→ Make general wellness claims
What you CANNOT do:
→ Make claims to diagnose, treat, cure, or prevent disease
→ Provide patient-specific clinical recommendations
→ Replace healthcare provider decision-making
🔴 YES – FDA APPROVAL REQUIRED
Your product requires FDA if it:
→ Diagnoses disease or medical conditions
→ Treats, manages, or prevents disease
→ Predicts specific disease risk or acute events
→ Analyzes medical device data or images
→ Uses multi-modal medical data for clinical purposes
Most Likely Pathways:
510(k) Clearance: 3-6 months, $200K-$1M
De Novo: 6-12 months, $500K-$2M
PMA: 1-3 years, $1M-$10M+ (rare for software)
Next Steps:
→ Pre-Submission Meeting with FDA (highly recommended)
→ Identify predicate device (for 510(k) route)
→ Plan clinical validation studies
→ Budget 6-18 months, $500K-$2M+

📱 Real-World Examples

Calm / Headspace
What it does: Meditation, mindfulness, sleep sounds

Why NO FDA: General wellness, no disease claims
✅ NO FDA APPROVAL NEEDED
Teladoc / MDLive
What it does: Video/text with physicians

Why NO FDA: Communication platform, doctors make decisions
✅ NO FDA APPROVAL NEEDED
Propeller Health
What it does: Smart inhaler + app for COPD/asthma management

Why YES FDA: Disease management, analyzes device data
🔴 510(K) CLEARANCE
Click Therapeutics
What it does: Digital therapeutics paired with pharmaceutical products (“software-enhanced drug”)

Why Significant: Early adopter of PDURS model, positioning software as drug companion
🔴 PDURS PIONEER

💡 Pro Tips for Digital Health Founders

1. When in doubt, ask FDA: Pre-Submission meetings are free and can save you millions.

2. Your marketing determines regulation: Claims matter more than technology.

3. Document your decision: Write a regulatory strategy memo. Get legal review.

4. Don’t try to outsmart FDA: If you’re providing medical value, embrace regulation.

5. Plan early: Regulatory strategy should be part of product development from day one.

Important Disclaimer:

This decision tree provides general guidance based on FDA regulations as of January 2025. It is NOT legal or regulatory advice. Always consult with qualified regulatory professionals.

Resources: FDA Digital Health Center of Excellence • Clinical Decision Support Guidance (2022) • PDURS Draft Guidance (Sept 2023)

Leave a comment

I’m Anita

Welcome to my blog. This is where I share insights at the intersection of healthcare, digital transformation, and AI—grounded in real-world experience.

I write for leaders navigating complex change in healthcare, sharing practical lessons and clear, distilled insights from my doctoral studies at Johns Hopkins—shaped by real-world experience across payer and pharma. I focus on what actually drives impact, beyond the buzzwords.

If you’re interested in thoughtful, practical takes on leading change and building meaningful digital solutions, I’m glad you’re here.

Let’s Connect!