Key Takeaways:
- SaMD uses software to perform medical functions without relying on proprietary medical hardware.
- Regulators such as the FDA and IMDRF assess SaMD based on intended use, risk, clinical evidence, and lifecycle controls.
- Development teams need cybersecurity, interoperability, quality management, and post-market monitoring built in from the start.
Software as a Medical Device (SaMD) is changing how clinicians diagnose disease, monitor patients, and deliver care at scale. Because SaMD operates independently from hardware, it follows its own regulatory and technical requirements that product teams need to understand before development begins.
If you’re planning a regulated digital health product, you need to understand how software is classified, how it differs from adjacent software categories, and what evidence, security controls, and lifecycle processes regulators expect.
How to Determine Whether Your Software Qualifies as SaMD
You can usually determine whether software qualifies as SaMD by reviewing four factors in order.
- Intended use and indications for use. Start with what your software is intended to do, for whom, and in what clinical context. If the software is intended to diagnose, treat, mitigate, cure, prevent, or drive a clinical decision, it may meet the definition of a medical device.
- Medical purpose. Regulators look at the specific device software function, not just the platform. A cloud service, smartphone app, or desktop product can qualify if it performs a medical purpose for a patient, clinician, or care team.
- Standalone operation. SaMD performs its medical function without being part of a hardware medical device. If the software is necessary for a specific device to operate, control, or power that device, it is more likely Software in a Medical Device (SiMD).
- Data handling only. Software that only stores, transfers, converts formats, or displays data may fall outside the device definition under section 520(o). An EHR system is a useful boundary case: software that only organizes or displays clinical records is not automatically SaMD, but an EHR feature that analyzes patient-specific data and recommends diagnosis or treatment may create a regulated device software function.
| Software Category | Medical Purpose | Hardware Dependence | Example | Likely Regulatory Implication |
| SaMD | Yes | Operates independently from a specific hardware medical device | A skin lesion analysis app that supports diagnosis | Usually regulated as a medical device, with class and submission path based on risk |
| Software in a Medical Device (SiMD) | Yes | Depends on the parent medical device | Pacemaker control software | Regulated with the parent device and its system-level controls |
| accessory software | Supports a regulated device use | Often linked to a device ecosystem or intended accessory role | Software that configures an infusion pump or supports device-specific workflows | May be regulated as an accessory or as its own device software function |
| general purpose software | No specific medical purpose | Independent | A spreadsheet, messaging platform, or video meeting app used in a clinic | Usually not regulated as a medical device unless a medical function is added |
What Is Software as a Medical Device?
Defining SaMD in the Modern Healthcare Ecosystem
Software as a Medical Device represents software that performs a medical function on its own. According to the International Medical Device Regulators Forum (IMDRF), SaMD is “software intended to be used for one or more medical purposes that perform these purposes without being part of a hardware medical device.” This distinction matters because software that analyzes, interprets, or acts on clinical data can be regulated even when it runs on a phone, tablet, laptop, or cloud platform.
The FDA has aligned its framework with this international definition. In practice, your intended use and indications for use determine whether software is treated as a medical device, while platform choice does not. A smartphone app and a cloud service can both qualify as SaMD if they perform the same regulated function.
The Critical Distinction Between SaMD and SiMD
SaMD functions independently of the hardware device to perform a medical purpose, while software in a medical device (SiMD) is embedded in or required for the device itself. This distinction affects architecture, documentation, testing scope, and regulatory planning.
Real-World Applications Revolutionizing Patient Care
The practical applications of SaMD span the healthcare continuum, from prevention to diagnosis to treatment management. In radiology departments, AI-powered SaMD solutions detect abnormalities in medical images and help clinicians prioritize urgent findings. These systems can support earlier cancer detection, stroke triage, and faster review of critical cases.
For chronic disease management, SaMD has become a standard part of many care models. Diabetes patients now rely on apps that track glucose levels, predict trends, recommend insulin dosages, and alert caregivers to dangerous fluctuations. These healthcare mobile apps can turn smartphones into tools for monitoring and intervention.
Mental health represents another area where SaMD is expanding access to care. Cognitive behavioral therapy apps, mood tracking platforms, and AI-assisted mental health assessments extend support beyond traditional clinical settings. These tools are especially useful where provider capacity is limited and ongoing monitoring matters.
SaMD Regulation: FDA, IMDRF, and Global Standards
FDA Classification and IMDRF Risk Categories
The FDA’s approach to SaMD regulation focuses on the risk of the device software function (DSF), the significance of the information it provides, and the condition it addresses. SaMD is still classified under the FDA’s standard device classes: Class I for low risk, Class II for moderate risk, and Class III for high risk. Many SaMD products fall into Class II because they inform diagnosis, monitoring, or treatment without acting as fully autonomous life-sustaining systems.
| FDA class | Example app type | Likely regulatory implication |
| Class I | Wellness application paired with a low-risk device, such as a wearable that displays non-diagnostic hydration or activity data. | May fall under general controls, with a lower documentation burden if claims remain non-diagnostic. |
| Class II | RPM application connected to a Bluetooth blood pressure cuff, ECG recorder, or glucose monitor. | Often requires 510(k) clearance, design controls, documented risk management, and validation evidence. |
| Class III | Life-supporting or implantable device software, such as clinician-facing controls for a cardiac support system. | Requires the highest level of evidence, extensive clinical validation, and premarket approval. |
Not every healthcare-related software feature is regulated as a device. Certain functions that only store, transfer, display, or organize information may fall outside the device definition under section 520(o). That boundary matters for products such as EHR systems, workflow tools, and communication platforms that may operate in clinical environments without performing a medical purpose.
The FDA does not publish a single running count of cleared SaMD, since no database query isolates the category cleanly from connected hardware. Independent tracking by medtech consultancy Orthogonal, built from public FDA submission data, put the number of identifiable FDA-cleared SaMD products at over 600 as of late 2024, with radiology and cardiology applications making up the largest share.
IMDRF adds another layer by categorizing SaMD based on the healthcare situation and the role the software plays in clinical decision-making. This framework helps you assess global risk posture early, especially if you plan to launch across multiple markets.
Premarket Submission and Documentation Requirements
Once you classify the software, you need to map the likely submission path. Lower-risk products may qualify for enforcement discretion in limited cases. Moderate-risk products often proceed through a 510(k) premarket notification) if a suitable predicate exists, while novel products may require a De Novo request. High-risk products generally require Premarket Approval (PMA) with more extensive evidence.
Submission depth depends on risk and product complexity. Higher-risk SaMD may require enhanced documentation covering software architecture, hazard analysis, cybersecurity, verification, validation, human factors, and labeling. Clinical evaluation is also part of the evidence story. You need to show that the product performs as intended, supports safe use, and fits the target population and workflow.
The FDA also evaluates app-based products through related policies for mobile medical applications. Software can still be regulated when it runs on a smartphone or in the cloud if the intended use is medical.
Global expansion adds parallel requirements. In Europe, SaMD is typically assessed under the EU Medical Device Regulation (MDR), while software intended for in vitro diagnostic purposes may fall under the EU IVDR. These frameworks require technical documentation, clinical evaluation or performance evaluation, post-market surveillance, and alignment with privacy requirements such as GDPR.
Post-Market Surveillance and Change Control
Regulatory obligations do not end at launch. SaMD manufacturers need complaint handling, vulnerability management, version control, corrective and preventive action, and documented post-market surveillance processes. Secure updates, audit trails, and field action procedures are especially important because software changes can be deployed quickly and at scale.
Cybersecurity is now part of that lifecycle expectation. The FDA’s cybersecurity guidance expects manufacturers to address secure design, risk management, SBOM planning, and update mechanisms as part of ongoing compliance. Quality controls also need to align with the FDA’s Quality Management System Regulation (QMSR) and the organization’s broader design control process.
FDA’s Regulatory Framework for AI/ML-Powered SaMD
AI/ML-powered SaMD adds continuous-learning questions that traditional hardware-centric rules did not fully address. The FDA’s Total Product Life Cycle approach and Predetermined Change Control Plan (PCCP) framework allow some anticipated model updates to be reviewed up front, rather than handled as entirely new submissions each time.
In practice, that means you need governance for model-drift monitoring, rollback criteria, release approval, training-data controls, and audit trails. Good Machine Learning Practices also matter. Regulators expect representative datasets, traceable performance testing, transparent change management, and ongoing monitoring after release.
If you’re planning regulated AI features, review our compliance and risk mitigation for AI Healthcare apps.
SaMD Market Size and Growth Drivers
Market Adoption and Investment Trends
The SaMD market continues to expand as healthcare organizations adopt software-first tools for diagnosis, monitoring, workflow support, and remote care. Growth is tied to chronic disease burden, broader use of connected devices, and the operational advantage of updating software faster than hardware. According to Data Bridge Market Research, the global SaMD market was valued at $1.58 billion in 2024 and is projected to reach $6.87 billion by 2032, a compound annual growth rate of just over 20 percent. Market sizing varies by analyst firm depending on how broadly SaMD is defined, but the direction is consistent across sources: steady, multi-year expansion.
North America remains a major market because of regulatory maturity, capital access, and strong digital health infrastructure. Asia-Pacific is also growing quickly as healthcare systems invest in mobile access, hospital digitization, and scalable care delivery models, with several forecasts pointing to Asia-Pacific as the fastest-growing region through the early 2030s as smartphone adoption and government digital health initiatives expand access in countries like China, India, and Japan.
Key Growth Drivers Reshaping Healthcare Delivery
Several forces are accelerating SaMD adoption across healthcare systems. Value-based care models reward earlier intervention, better monitoring, and more consistent follow-up. SaMD can support all three when teams integrate it into clinical workflows and evidence plans from the start.
The COVID-19 pandemic also accelerated remote monitoring and virtual care. Many organizations that were once cautious about software-led monitoring now treat it as part of normal service delivery.
Artificial intelligence and machine learning are another major driver. Modern SaMD doesn’t just collect data. It can identify patterns, support risk scoring, and prioritize action when clinicians need faster insight.
Investment Trends and Market Dynamics
Venture capital investment in SaMD has increased alongside broader healthcare technology funding. Major technology companies are entering the space with cloud, AI, and user experience expertise. Traditional medical device manufacturers are also expanding their software capabilities because software-based products can support faster iteration and recurring service models.
The competitive landscape continues to shift as established players compete against startups and platform companies. That pressure is driving innovation, acquisitions, and more specialized products targeted at specific clinical use cases.
Critical Technologies Powering SaMD Innovation
How AI and Machine Learning Power SaMD
AI and ML have changed SaMD from basic data tools into clinical decision support systems. Modern AI-powered SaMD can analyze medical images, predict disease progression, and personalize treatment recommendations with greater speed and consistency.
Deep learning algorithms, particularly convolutional neural networks, have changed medical imaging interpretation. SaMD solutions can detect diabetic retinopathy in retinal scans, identify suspicious lung nodules in CT images, and support cardiovascular risk assessment from echocardiograms. These systems do not replace clinicians, but they can improve case review and prioritization.
Natural language processing enables SaMD to extract insights from unstructured clinical notes, research papers, and patient communications. This capability is useful for identifying adverse drug reactions, predicting readmission risk, and supporting clinical research. The ability to process large volumes of text helps clinicians make decisions based on broader information sets.
Cloud Computing: Enabling Scalability and Accessibility
Cloud-based deployment has emerged as a common architecture for SaMD because it supports scale, centralized updates, and shared access across distributed care teams.
Healthcare app development teams are increasingly adopting cloud-native architectures that support real-time data processing, advanced analytics, and global accessibility. Cloud platforms provide the computational power necessary for complex AI algorithms while maintaining the flexibility to scale resources based on demand.
Security concerns, once a barrier to cloud adoption in healthcare, are being addressed through advanced encryption, secure APIs, and compliance-focused cloud platforms. Major cloud providers now offer HIPAA-compliant infrastructure designed for healthcare applications, providing the security and reliability required for medical-grade software.
For a deeper look at infrastructure planning, review these scalable backend solutions.
Internet of Medical Things: Creating Connected Care Ecosystems
The Internet of Medical Things represents the convergence of connected medical devices, wearable sensors, and SaMD platforms. This ecosystem enables continuous patient monitoring, automated data collection, and real-time clinical interventions.
Wearable devices equipped with sophisticated sensors feed data directly to SaMD platforms for analysis. Heart rate variability, sleep patterns, activity levels, and blood oxygen saturation can be continuously monitored and analyzed for early warning signs of health deterioration. This continuous monitoring model is especially useful for chronic disease management and post-surgical recovery.
Integration with Electronic Health Records (EHRs) turns SaMD from an isolated tool into part of the clinical workflow. When SaMD solutions can populate records, trigger alerts, and support care coordination, they become more useful to providers and easier to operationalize.
If your roadmap includes connected sensors and companion apps, review these IoT development services.
Overcoming Implementation Challenges for Security, Privacy, and Interoperability
Cybersecurity in the Age of Connected Healthcare
The digitization of healthcare has created significant cybersecurity challenges. SaMD systems handle sensitive patient data and often connect to multiple healthcare networks, creating a broad attack surface. The consequences extend beyond data theft because compromised medical software can also affect patient safety.
The FDA’s cybersecurity guidance calls for security measures throughout the SaMD lifecycle. Manufacturers need secure design principles, regular vulnerability assessments, update mechanisms, and SBOM planning. These controls improve transparency and support faster response when vulnerabilities are discovered.
Encryption has become non-negotiable, with industry standards requiring protection both in transit and at rest. Modern SaMD solutions use TLS for secure communications, AES-256 for data storage, and increasingly, architectures that verify every request regardless of network location. Zero-trust models are becoming standard for products that move clinical data across multiple systems.
Data Privacy and Regulatory Compliance
Privacy concerns remain a major barrier to widespread SaMD adoption. Patients want to know how their medical data is collected, stored, and used, especially when AI systems depend on large datasets for training and operation. Addressing these concerns requires technical controls and clear communication.
HIPAA compliance in the United States, GDPR in Europe, and emerging privacy regulations worldwide create a complex compliance landscape for SaMD developers. These rules affect not only storage and transmission, but also algorithm design, user interface decisions, consent models, and cross-border deployment strategies.
Federated learning and differential privacy techniques offer practical options for some AI use cases. These methods allow models to learn from distributed datasets without centralizing sensitive information. Teams still need to validate that privacy-preserving methods maintain acceptable clinical performance.
Interoperability Breaking Down Healthcare’s Data Silos
Healthcare’s fragmented data landscape presents ongoing challenges for SaMD implementation. Different systems use incompatible data formats, communication protocols, and terminology standards. This lack of interoperability limits data exchange and prevents a unified view of the patient.
Standards like HL7 FHIR (Fast Healthcare Interoperability Resources) are gaining traction because they provide a more consistent framework for data exchange. FHIR’s API-first approach fits well with modern software development practices, enabling SaMD solutions to integrate with diverse healthcare systems. Many organizations still rely on legacy systems, so adoption remains uneven.
The Office of the National Coordinator for Health Information Technology’s (ONC) Interoperability Roadmap outlines a connected-care vision, but implementation still requires significant integration work. SaMD developers often support multiple standards and custom interfaces across provider environments. That complexity increases development effort and slows deployment.
Successful SaMD implementations often rely on middleware and integration platforms that translate between systems. These architectures add complexity, but they also make it possible to deploy across real clinical environments without rebuilding every workflow from scratch.
Success Stories for SaMD in Patient Outcomes
Diagnostic Tools in Clinical Practice
The real-world impact of SaMD is best illustrated through concrete examples that show better patient outcomes and more efficient workflows. LumineticsCore (previously known asIDx-DR) is an FDA-authorized autonomous AI diagnostic system for diabetic retinopathy screening. Deployed in primary care settings, it supports earlier detection where access to eye specialists may be limited.
In oncology, Paige’s AI-powered pathology platform supports cancer diagnosis by helping pathologists review tissue images more consistently and prioritize suspicious findings. This kind of software is useful in workflows where accuracy, speed, and case volume all matter.
Microsoft’s recent innovations with MedImageInsight and MedImageParse show the potential of broader imaging analysis. These AI models can identify abnormalities while also supporting segmentation and classification, helping radiologists localize issues and plan next steps more precisely.
Chronic Disease Management Breakthroughs
SaMD has changed chronic disease management by shifting many workflows from reactive care to continuous monitoring. Prenosis Inc.’s Sepsis ImmunoScore received FDA authorization for sepsis risk assessment, supporting earlier intervention in patients who may deteriorate quickly. Integration with electronic health records shows how effective SaMD implementation goes beyond model development.
For cardiovascular care, Kestra Medical Technologies’ ASSURE Wearable Cardioverter Defibrillator uses adaptive algorithms to filter motion artifacts and reduce false alarms while autonomously detecting and treating life-threatening arrhythmias in at-risk cardiac patients.
Edwards Lifesciences’ Acumen Assisted Fluid Management software shows how SaMD can support acute care workflows. By analyzing real-time hemodynamic data during surgery, the system helps clinicians make more informed fluid management decisions.
Population Health and Predictive Analytics
Beyond individual patient care, SaMD is changing population health management through predictive analytics and risk stratification. Healthcare systems use AI-powered platforms to identify high-risk patients before acute events occur, enabling preventive interventions that can reduce hospitalizations and improve quality of life.
One important use case is the prediction of mental health crises. Platforms that analyze patterns in patient communications, appointment histories, and clinical notes can identify individuals who may need earlier intervention. These systems can help connect patients with appropriate resources before a crisis escalates.
In hospital operations, SaMD solutions are being used to predict patient flow, optimize resource allocation, and reduce readmission risk. These systems analyze clinical, operational, and demographic information to support more efficient healthcare delivery.
Best Practices for SaMD Development and Implementation
Adopting a User-Centric Design Philosophy
Successful SaMD development begins with a clear understanding of end-user needs and clinical workflows. Unlike consumer apps, poorly designed medical software can affect patient safety and clinical outcomes. That risk makes user-centered design a core development activity rather than an optional refinement.
Healthcare app developers must engage clinicians, patients, and other stakeholders throughout the development process. This is not just about gathering requirements. It is about understanding how emergency physicians make time-sensitive decisions, what nurses need at medication administration, and how interfaces perform for users with varying technical proficiency.
Usability testing in realistic clinical environments reveals issues that laboratory testing might miss. Screen glare in operating rooms, gloved-hand interaction, alarm burden, and cognitive load during critical situations can all affect SaMD performance. Iterative testing and refinement, guided by user feedback, produce safer and more usable products.
For related design considerations, see our article on balancing aesthetics and functionality in healthcare app design.
Implementing Robust Quality Management Systems
Quality management for SaMD extends beyond traditional software development practices. ISO 13485, IEC 62304, and ISO 14971 provide frameworks for medical device software that address safety, reliability, traceability, and risk control across the lifecycle.
These standards map to specific activities. ISO 13485 supports the quality system and aligns closely with the FDA’s QMSR. IEC 62304 structures software development, maintenance, and problem resolution. ISO 14971 governs risk management across design, verification, validation, release, and post-market surveillance. Clinical evaluation connects those processes to real-world evidence by showing that the product performs as intended for the target users and use cases.
Documentation requirements for SaMD can seem extensive, but they serve critical purposes. Design history files, software development plans, verification and validation records, and traceability artifacts support audits, submissions, and root cause analysis when issues arise. Strong documentation also makes change control more reliable as products evolve.
Building for Scalability and Maintainability
SaMD success requires planning for long-term sustainability. Unlike traditional medical devices that may remain unchanged for years, software products evolve to incorporate new evidence, address security threats, and adapt to changing clinical practice.
Cloud-native architectures provide the foundation for scalable SaMD solutions. Microservices architectures enable independent scaling of different components, helping teams maintain performance under varying loads. Container orchestration platforms such as Kubernetes support deployment across environments while maintaining consistency and reliability.
Post-market surveillance adds a software-specific dimension to maintenance. Ongoing monitoring of algorithm performance, user behavior patterns, and clinical outcomes creates the feedback loops needed for improvement. Teams that can update products quickly while maintaining regulatory compliance are better positioned to scale safely.
SaMD Trends: AI/ML, Digital Therapeutics, and Global Access
AI Evolution From Assistance to Autonomy
The trajectory of AI in SaMD points toward increasing autonomy and sophistication. Current systems often serve as clinical decision support tools, giving recommendations that clinicians evaluate and act on. Future systems may take on more autonomous functions in narrow scenarios where rapid response is critical, or expertise is not immediately available.
Generative AI and large language models are opening new possibilities for SaMD. These technologies can synthesize medical information, generate patient education materials, and support preliminary patient intake. Their use in regulated settings still requires careful attention to accuracy, bias, and the risk of plausible but incorrect outputs.
Federated learning and swarm intelligence approaches may also enable SaMD to learn from broader patient populations while preserving privacy. This model could improve diagnostic performance without requiring centralized storage of all patient data.
Digital Therapeutics as Software-Based Treatment
Digital therapeutics (DTx) represents SaMD’s expansion from diagnosis and monitoring into treatment. FDA-authorized DTx solutions are already being used for conditions such as substance use disorder and ADHD, delivering evidence-based therapeutic interventions through software interfaces.
The integration of virtual reality and augmented reality into DTx opens new treatment options. VR-based exposure therapy for PTSD, AR-guided rehabilitation exercises for stroke patients, and immersive pain management tools show how software can deliver interventions that are difficult to provide through conventional workflows alone.
Prescription digital therapeutics, where software is prescribed like medication, mark a meaningful shift in care delivery. As regulatory frameworks mature and clinical evidence grows, software may be prescribed alongside or instead of some conventional therapies.
Global Health Impact and Healthcare Access
SaMD may have its greatest long-term impact on global health access. In regions with limited medical infrastructure, smartphone-based diagnostic tools can extend specialist support to remote communities. With the right training and oversight, community health workers can use these tools to support earlier screening and more consistent follow-up.
The COVID-19 pandemic showed how software can support large-scale public health responses through symptom checkers, remote monitoring, and triage systems. Future outbreaks will likely involve even more sophisticated SaMD deployments for monitoring, prediction, and treatment coordination.
Partnerships among technology companies, healthcare organizations, and global health initiatives are accelerating SaMD deployment in low-resource settings. These collaborations are addressing connectivity limits, device access, and cultural adaptation, leading to software built for specific public health needs.
Strategic Considerations for Businesses Entering the SaMD Market
Market Entry Strategies and Business Models
Success in the SaMD market requires strategies that balance innovation with regulatory requirements, and clinical needs with business sustainability. Unlike many software markets, SaMD development requires deliberate planning around safety, evidence, and lifecycle control.
Direct-to-consumer models are gaining traction for lower-risk SaMD focused on wellness and chronic disease management. These approaches can reach patients directly through app stores and digital channels, but they still require careful handling of medical claims, oversight, and trust.
Enterprise sales to healthcare systems are common for higher-risk SaMD that requires workflow integration. These B2B models involve longer sales cycles, pilot programs, and evidence review. Success depends on technology quality, clinical validation, procurement alignment, and change management.
Partnership strategies are increasingly important as the SaMD ecosystem matures. Collaborations between technology companies and established medical device manufacturers combine software expertise with regulatory experience and market access. Strategic relationships with healthcare systems also provide real-world testing environments and clinical validation opportunities.
Building Sustainable Competitive Advantages
In the rapidly evolving SaMD market, sustainable competitive advantages require more than good algorithms or polished interfaces. Long-term success depends on clinical evidence, regulatory expertise, and trusted healthcare relationships.
Clinical validation remains a strong differentiator. SaMD backed by peer-reviewed studies, real-world evidence, and demonstrated workflow fit is more likely to earn clinician trust and support adoption.
Regulatory expertise also becomes a competitive moat as requirements grow more complex. Companies that understand FDA pathways, international certifications, and lifecycle compliance can reduce delays and manage change more effectively.
Network effects and data advantages can create durable value in SaMD. Platforms that aggregate data from multiple sources, support integrations, and coordinate care become more useful as adoption grows, especially when switching costs are high.
Getting Started With SaMD Development
SaMD development requires clear classification, evidence planning, cybersecurity controls, and lifecycle governance from the start. If your product may qualify as a medical device, you’ll need to align intended use, software architecture, clinical validation, and post-market processes before launch.
If you need support from an experienced healthcare app developer, the next step is to validate your product category, regulatory path, and implementation plan early.
Request a free consultation to discuss your mobile healthcare app project.
SaMD FAQs
What makes software SaMD instead of SiMD?
SaMD performs a medical purpose independently of a hardware medical device. Software in a Medical Device (SiMD) is embedded in, or required for, the device to function. If the software can deliver its regulated medical function on its own, it is more likely SaMD. If it operates as part of a specific device, it is more likely SiMD.
How do I know if my software is a medical device?
Start with intended use and indications for use. Then assess whether the software performs a medical purpose, whether it operates independently, and whether it only stores, transfers, or displays data. Functions that fall under section 520(o) may be excluded from device regulation, but diagnostic, treatment, or patient-specific analytical functions often are not.
What FDA class is most SaMD?
Many SaMD products fall into Class II because they support diagnosis, monitoring, or treatment decisions without acting as high-risk autonomous therapy systems. Final classification depends on the device software function, the seriousness of the condition, and the significance of the information the software provides.
Does SaMD need ISO 13485, IEC 62304, or ISO 14971?
Those standards are not interchangeable, but they are commonly part of a mature SaMD program. ISO 13485 supports your quality system, IEC 62304 structures software lifecycle activities, and ISO 14971 governs risk management. You may also need clinical evaluation, cybersecurity controls, usability evidence, and post-market surveillance processes to support compliance.
Can cloud-based or smartphone-based software still be SaMD?
Yes. Cloud platforms, desktop applications, and smartphone apps can all qualify as SaMD if they perform a regulated medical function. The platform does not remove device obligations. In some cases, the FDA may also evaluate app-based products under its mobile medical applications guidance.
