legacy-medical-device-security-frameworks

FDA and Regulatory Considerations for the Pattern C MFA Shim

This is not regulatory advice. This document describes the regulatory questions that any deployment of an inline MFA shim raises, and the considerations that must be addressed in consultation with regulatory affairs counsel and the relevant manufacturer.

The central question

When an HDO (or a manufacturer) inserts an in-line hardware element between a medical device’s service port and the technician’s tooling, has the medical device been modified in a way that requires re-clearance, change notification, or other regulatory engagement?

The answer is jurisdiction- and case-specific. This document outlines the framework for analysis.

United States — FDA

Relevant regulations and guidance

Analysis framework

For an HDO operator deploying the shim, the FDA’s primary jurisdictional concern is whether the deployment changes the device’s intended use or fundamental scientific technology. If the answer is no, the deployment is in the operator’s compliance domain (HIPAA, internal risk management) rather than triggering FDA premarket review.

Key questions:

  1. Does the shim modify the device’s clinical function? No — it gates access to the service port, which is not a clinical-function path.
  2. Does the shim change clinical data flow? No — clinical data flows through different interfaces (HL7, DICOM, vendor protocols on dedicated network ports), not through the service port.
  3. Does the shim affect the device’s labelled performance characteristics? No — performance specifications (rate accuracy, alarm latency, etc.) are not impacted.
  4. Does the shim affect cybersecurity properties of the device? Yes — favourably. Postmarket cybersecurity guidance (December 2016) explicitly recommends compensating controls for legacy devices.

The deployment most closely resembles the addition of a network compensating control (firewall, IPS, gateway), which is a routine HDO operational practice not requiring FDA engagement.

Caveats and risk areas

Production-deployment path

For production deployment beyond a research or quality-improvement pilot, recommended steps:

  1. Pre-submission (Q-submission) meeting with FDA CDRH if the deployment plans extend across a fleet of any meaningful size, particularly across multiple facilities.
  2. Manufacturer engagement for each affected device class. Best case: the manufacturer endorses the shim under their security advisory. Acceptable case: the manufacturer is informed and not opposed. Worst case: manufacturer asserts that the deployment voids service warranty — proceed only with explicit organisational acceptance of that risk.
  3. Risk management file update (ISO 14971 cl.10) treating the shim as a postmarket risk control measure with corresponding residual risk evaluation.
  4. Quality system process for shim deployment, monitoring, and decommissioning, integrated with existing clinical engineering work-order processes.

Pilot deployment criteria

A pilot deployment limited to a small number of devices in a controlled clinical environment (e.g., 10–20 pumps in a single ICU) for a defined evaluation period is the lowest-risk regulatory posture and is what cjr-templates/examples/cjr-leaked-service-pin.md documents.

Pilot deployments should:

European Union — MDR

Relevant regulations

Analysis

The EU MDR’s Article 16 prohibits any person from making changes to a medical device that “may modify its compliance with the requirements of this regulation” without satisfying the obligations of a manufacturer. This is more stringent than the FDA position.

A reasonable reading: the shim does not modify the device itself; it sits in the technician-tooling-to-device connection. By analogy with network compensating controls (firewalls, gateways), it should not constitute a device modification. However, this analogy is not formally established in MDR jurisprudence.

For EU deployment:

MDCG 2019-16 — Cybersecurity Guidance

The EU’s medical device cybersecurity guidance (MDCG 2019-16) explicitly endorses compensating controls for legacy device cybersecurity. The shim is consistent with this guidance in principle. The procedural pathway for deployment is the question that requires notified-body engagement.

Other jurisdictions

The pattern of analysis above generalises. Key questions in any jurisdiction:

  1. Does the regulation distinguish between modification of the device and operation of the device?
  2. Does the regulation accommodate compensating controls for legacy devices?
  3. Does the regulation require manufacturer involvement in inline accessory deployment?

For Health Canada, the TGA (Australia), the PMDA (Japan), and other regulators, similar consultations should occur before any production deployment.

What this artifact provides

This repository provides:

  1. The design and reference implementation of Pattern C, sufficient to support pilot evaluations.
  2. Documentation sufficient to inform regulatory analysis.
  3. An Apache-2.0 patent grant ensuring no party can monopolise the design.

This repository does not provide:

  1. FDA pre-submission documents or 510(k) content.
  2. Notified body assessments.
  3. Production-grade hardware (the prototype is software-only).
  4. Clearance, certification, or any form of regulatory approval.

Any party wishing to deploy the design in production should engage their own regulatory affairs function, the affected device manufacturer(s), and the relevant regulatory authorities before scaled deployment.

Disclaimer (re-stated)

The authors of this repository are not regulatory affairs professionals. This document is the authors’ best-effort summary of relevant considerations, intended as a starting point for organisational regulatory analysis, not as a substitute for it. Decisions about deployment must be made by the deploying organisation in consultation with their own regulatory, clinical, and security functions.