In ISO 14971, the process has ALWAYS required the manufacturer to first determine if risk can be reduced by designing out the hazard, or reducing the impact by design. Second if that is NOT POSSIBLE, then a manufacturer can use protective measures, and finally, the manufacturer, when confronted with the situation were protective measures CANNOT be implemented after finding design changes cannot be performed, may as a LAST RESORT use the LEAST EFFECT?IVE risk control and that is information for safety, including user training. Currently the requirements are expressed in ISO 14971:2019 7.1 a), b), c). Explanations of these requirements can be found in the Guidance, ISO TR 24971:2020 7.1
You do have to identify significant Overall Residual Risks in the Information for Safety, as required in Clause 8 of ISO 14971:2019.
------------------------------
Edwin Bills MEd, CQA, RAC, BSc, CQE, ASQ
Principal Consultant
Overland Park KS
United States
elb@edwinbillsconsultant.com------------------------------
Original Message:
Sent: 03-Feb-2021 10:35
From: Scott Bishop
Subject: Statistical Significance - Cautionary Statements
Thank you for reaching out -
The state of the art standard currently is ISO 14971:2019.
Critically, in the application of risk impact the frequency of occurrence is considered with the potential severity of the event. An issue that occurs frequently (perhaps up to 1% of devices or cases) that results in the potential for slight procedure delay of up to 5m is rather different than an issue that occurs at an expected rate of 1 / 10,000 but could lead to the death or permanent impairment or damage to the patient (or user) and must be addressed differently.
As well, the approach of listing or warning via labeling, while still important, is not the preferred solution. Rather, modification to reduce or remove risks through design is the approach favored.
You might benefit from connecting with a consulting resource to address risk management for your product - as it can be challenging to understand the implementation if you haven't exercised it in your QMS / business. A functioning risk process and documentation is incredibly valuable for your business (and required by FDA / ISO / etc) and is utilized in consideration of almost all QMS elements.
------------------------------
Scott Bishop
Houston TX
United States
Original Message:
Sent: 02-Feb-2021 13:00
From: Anonymous Member
Subject: Statistical Significance - Cautionary Statements
This message was posted by a user wishing to remain anonymous
Hi everyone, I have a question I'm hoping some of you could weigh in on. Is there any guidance document that gives a threshold for statistical significance when it comes to adding a cautionary statement to your literature? Or even a contraindication for that matter. For example, say there is 1 report of event X occurring, the doctor won't offer an opinion on if there was a connection, but we have 100k devices in circulation. Possible 1 out of 100k, probably no rationale for adding a cautionary statement. But say there is 10 reported event X's, doctors offer no opinion, but there are only 50 devices in circulation. Possible 10 out of 50, probably rationale for making some type of cautionary statement (at least).
What I'm curious about is if anyone is familiar with any type guidance document or standard that lists a percentage threshold for this situation.
Thanks in advance.