Something I find interesting is how different companies assign responsibility for quality. In some places, there are entire teams dedicated to QA and QC, while in others, engineers are expected to handle quality as part of their design work.
Having dedicated QA/QC teams means there are experts who really understand compliance and testing, which can catch problems more effectively. But having engineers own quality might make them more careful from the start, since they know they’ll also be the ones held accountable. I think both approaches have pros and cons, and maybe the best system is a mix of the two.
So my question is: do you think it’s better for companies to separate QA/QC from engineering to keep quality independent, or should every engineer be directly responsible for quality in their designs? Which approach do you think leads to better products in the end?
I think it's important first to define what quality assurance and quality control's main roles are within the field of medical device development. Quality assurance (QA) is a proactive, process-focused system used to ensure that medical devices are consistently safe, reliable, and compliant with regulatory standards. In design controls, for example, QA ensures that every stage of research, prototyping, and engineering is documented and verified to meet safety requirements. Unlike QA, quality control (QC) is a reactive process focused on inspecting, testing, and verifying that manufactured components and final products meet predefined safety and performance criteria before distribution. Understanding this distinction matters because the question is not really whether QA/QC should exist. It is who performs these functions and how independently they operate from engineers who create the design.
Therefore, going back to the initial question on whether companies should separate QA/QC from engineering, I believe it depends on company size, risk classification, and regulatory requirements.
Overall, I do not think engineers should be excluded from thinking about quality. Engineers who understand and care about quality from the earliest design stages tend to catch problems sooner, when they are cheaper and easier to fix, rather than later during formal verification or, worse, after the product reaches the market. In the end, I believe the strongest system is one where engineers take ownership of quality as part of their design mindset, while an independent QA/QC team serves as a required checkpoint that cannot be bypassed.
I agree with dcapera that this isn't really an either/or question‚ but I'd push a little further on why independence matters so much in this industry specifically․ Engineers‚ no matter how much they care‚ will always have a bias towards their own solutions․ After spending a month solving a problem one way‚ it's hard to be as critical of that way of solving it as one might be if they weren't involved․ It's not a fault‚ but human nature to not be as critical of your own work․ A QA/QC function exists in part to provide a more rational perspective‚ without the emotional involvement․ There is also a regulatory aspect to this․ Under 21 CFR 820 and ISO 13485‚ device manufacturers have a legal requirement to implement a quality management system that has some degree of independence‚ e․g․‚ including people in design reviews who have no responsibility in the design stage under review․ So even in a small company that wants engineers to "own" quality as well‚ there's no escaping the structural separation for purposes of compliance․ It seems that the industry has settled the question that self-policing by the original designers should not be sufficient‚ no matter how good the engineers․Where I think there's real space for debate is at the smaller company or startup level‚ where hiring a dedicated full-time QA/QC team isn't feasible‚ and teams have had to rely on cross-functional reviews‚ external consultants‚ or a single quality lead who wears multiple hats in the name of efficiency․ That's a reasonable compromise‚ but it does raise the question of how much independence is "enough" independence․