Software development for a medical device is organized around regulatory evidence, not only code quality. FDA guidance describes recommended documentation for premarket submissions covering device software functions, with two documentation levels that depend on risk and device function [1][2]. A separate guidance applies risk-based computer software assurance to production and quality management system software, distinguishing that work from software inside the device itself [3]. The development workflow therefore combines lifecycle processes, verification and validation records, traceability, and cybersecurity documentation. This article explains those distinctions using primary FDA documents and the IEC 62304 standard, then walks through a hypothetical evaluation example. It does not provide patient advice or clinical outcome claims.
Device Software Functions and Documentation Levels
FDA defines device software functions as software functions that meet the device definition under section 201(h) of the Federal Food, Drug, and Cosmetic Act [1][2]. The premarket submission guidance recommends documentation for FDA's evaluation of safety and effectiveness for those functions [1]. A manufacturer first states the documentation level and explains the rationale for that level [1]. The guidance outlines basic and enhanced documentation levels, with enhanced documentation adding complete configuration management and maintenance plan documents [1]. This structure means the same software category can require different evidence depending on the device context.
The documentation level is not a measure of programming sophistication. It reflects the information FDA needs to understand technical design details, how the software functions, and the associated risk controls [1]. FDA may request additional information beyond the listed elements when needed to evaluate a submission [1]. The guidance is nonbinding and describes current agency thinking; an alternative approach can satisfy applicable statutes and regulations [1][2]. Manufacturers should therefore treat the documentation outline as a starting framework rather than a fixed checklist.
Verification, Validation, and the Software Lifecycle
Verification and validation are distinct activities that together support a conclusion that software is validated [1][2]. FDA's General Principles of Software Validation describes validation as typically supported by verifications of outputs from each stage of the software development life cycle and checking proper operation of the finished software in the intended use environment [5]. Planning, requirements, traceability, testing, risk assessment, design reviews, and change management are identified as important activities supporting validation [2]. The device manufacturer has latitude in defining how validation will be accomplished, but validation should be a key consideration in deciding how and by whom software is developed [5].
IEC 62304:2006 defines life cycle requirements for medical device software and establishes a common framework for medical device software life cycle processes [7]. The standard was confirmed in 2023 and has one amendment from 2015 [7]. FDA guidance references software-related voluntary consensus standards and notes that documentation created only after actual development, verification, and validation efforts could raise concerns about design process control [2]. This means the lifecycle evidence should be generated during development, not reconstructed at submission time.
Documentation Level Selection in Practice
A practical first step is to state the documentation level and rationale before writing the full submission package [1]. Basic documentation includes the level statement and rationale, while enhanced documentation adds complete configuration management and maintenance plan documents [1]. Teams can use the Medical Device Software Guidance Navigator to identify potentially relevant guidances across the development life cycle [4]. The navigator links to guidance on software changes, interoperable devices, cybersecurity, and AI-enabled device software functions [4]. Reviewing these sections can highlight considerations that support premarket submission preparation [4].
Production and Quality System Software Assurance
Computer software assurance applies to computers and automated data processing systems used as part of medical device production or the quality management system [3]. This guidance describes a risk-based approach to establish confidence in automation and identifies where additional rigor may be appropriate [3]. It supplements the General Principles of Software Validation and applies a least-burdensome approach, where validation burden is no more than necessary to address risk [3]. The scope explicitly excludes a complete description of all software validation principles [3].
The distinction matters for development teams. Software inside a device may require premarket documentation under the device software functions guidance, while a spreadsheet or reporting tool used in manufacturing may instead follow risk-based computer software assurance [3]. FDA's example describes a spreadsheet field where incorrect text input is immediately visible and does not impact intended use, leading to a conclusion that the spreadsheet is acceptable for its intended use after a validation rule is added [3]. This example illustrates how intended use and risk determine assurance effort rather than a uniform testing requirement.
Cybersecurity, Traceability, and Third-Party Components
Cybersecurity documentation is part of premarket submissions for medical devices, including 510(k)s, De Novo requests, PMAs, and other submission types [8]. FDA recommends that manufacturers consider cybersecurity risks and controls associated with interoperability capabilities [8]. Third-party software components require attention, and FDA recommends establishing and maintaining custodial control of device source code throughout the product lifecycle [8]. A software bill of materials can aid in managing cybersecurity risks across the software stack [8].
Traceability links designs, requirements, and tests, and is described as a focal point of FDA review [6]. CDRH requires verification of device functionality but does not dictate how to verify it [6]. Early establishment of software quality system elements, including design requirements tracking, is recommended even during feasibility testing [6]. For off-the-shelf software, the device manufacturer with regulatory responsibility needs to assess the adequacy of that software [5]. This shifts the development burden toward documented assessment rather than assuming vendor compliance.
A Hypothetical Evaluation Workflow
Consider a hypothetical manufacturer developing a software function that calculates an insulin bolus suggestion from glucose readings. The team would first determine whether the function meets the device definition and select a documentation level with a written rationale [1]. They would then define user requirements, including intended use and the expected operating environment, hardware and software configurations, and safety-related functions [5]. Traceability records would link each requirement to design outputs and test cases [6].
For verification, the team would document outputs from each lifecycle stage and test proper operation in the intended use environment [5]. For validation, they would prepare a validation protocol, acceptance criteria, test cases and results, and a validation summary [5]. Because the function involves a dosing calculation, the team would also document cybersecurity risks, third-party components, and an SBOM [8]. This hypothetical is a suggested workflow only; it does not describe actual testing or establish that any specific approach satisfies FDA requirements.
Sources & references
- Content of Premarket Submissions for Device Software Functions · www.fda.gov
- Content of Premarket Submissions for Device Software Functions — Guidance for Industry and FDA Staff (June 14, 2023) · www.fda.gov
- Computer Software Assurance for Production and Quality Management System Software · www.fda.gov
- Medical Device Software Guidance Navigator | FDA · www.fda.gov
- General Principles of Software Validation - Final Guidance for Industry and FDA Staff · www.fda.gov
- NIH OER Regulatory Knowledge Guide Digital Health · seed.nih.gov
- IEC 62304:2006 - Medical device software — Software life cycle processes · www.iso.org
- Cybersecurity in Medical Devices: Quality System Considerations and Content of Premarket Submissions · www.hhs.gov