60601 edition 3 is possibly the biggest change in the regulatory side of medical device development since the introduction of the FDA’s QSR. Until now, electrical safety testing was a black box test wherein a sample device was submitted and it passed or failed depending on a set of specific physical tests. There was no consideration of how the device was developed, just, “Is the device electrically safe?”
If we go back to where all this started in the last century, electrical testing was needed because consumers were getting shocked and fires were getting started by electrical device failures. The pattern of failures resulted in some basic design mitigations involving insulation, grounding, and overload protection (fuses). Industry specific standards were introduced to address the issues unique to that industry. As knowledge was gained about how devices fail, the standards were updated to address these failures. A point has now been reached where most devices are very safe electrically. The current generation of designers have grown up in a world where the requirements in the standards are common practice and the off-the-shelf components all comply with the general standards or some device specific standard. The main cause of injury to users and patients now is use error. To address this, the new electrical safety standard includes usability engineering and risk management.
Let’s ignore the obvious question of how we ensure non-electrical devices are safe. Surely, they must have some of the same use error issues.
The test houses, which until now, have tested the device against a prescriptive standard, must now look into the processes used to develop the device. In other words, people trained to perform objective tests in a lab must now also audit the design process. This is a big change.
The approach taken by most test houses has been to create checklists of all the clauses referencing the risk management file (RMF) and asking for where in the RMF the clause has been addressed. In other words, the same prescriptive pass/fail testing approach is being applied to risk management and usability engineering as was applied to dielectric strength. If the intent of edition 3 was to give developers more flexibility, the ironic result is a prescriptive and more burdensome approach. The attempt to allow alternative solutions justified by risk management has made the standard less well defined and so the pass/fail approach becomes a much more subjective and controversial test. The ultimate irony would be for this to result in devices that are less safe from electrical failure and the return to users and patients getting shocked and fires being started.
Surely the ISO 13485 registrars, FDA reviewers, and the notified bodies are better equipped to determine design control compliance. One telling result is that despite 60601 referencing IEC 14971 and 62366 as normative standards, most EU notified bodies do not accept 60601 certification as evidence of compliance for either one. This brings up a question. If the notified body requires a change in the risk management file during the review after certification, does the manufacturer submit that change to the 60601 test house? Who has the last word?
The responsibilities overlap between electrical safety test house, regulatory reviewer, 13485 registrar, and notified body is going to make life very difficult for medical device developers. Perhaps electrical safety certification should be just that—certification that a device won’t shock the user/patient and won’t set fire to its surroundings.