Sharks with laser beams
(With apologies to Dr. Evil!)
One of our clients has a saying, ‘Sharks with Laser Beams’ to describe the situation when you have been so convoluted in assessing the possible impact of use error that you end up with such a string of extremely unlikely events that all have to align to produce the harm. The product is disposed of incorrectly, which eventually leads to contamination of the sea, and sharks get mutated to develop laser beams.
Of course, it is ludicrous to go to such lengths with risk assessments. There has to be a balance, otherwise every single use error could potentially cause death. And no-one would be able to develop a safe product. There’s a risk-benefit balance that needs to be considered sensibly.
Regulators want you to show that you have understood and reduced the use-related risk. And to do so you need to think carefully about the impact of use error. But when should you stop? We all know that FDA do not like probability included in your uFMEA. But the sharks are not going to develop laser beam. We all accept that.
We often see HUGE uFMEAs, with every conceivable possibility described. And this has the effect of producing long lists of critical tasks, all of which put pressure on the validation study, because of course we don’t want to see use errors with critical tasks when validating.
So what should be the rules of thumb? How should you approach the analysis of use-related risk?
In our experience, there are some common mistakes that we see with uFMEAs, and by avoiding these you can significantly reduce the burden for validation.
Clinical practice: If you are including tasks such as ‘locate a suitable vein’, or ‘locate the opening to the urethra’, or ‘remove air from the line’ you are effectively taking on the responsibility for ensuring that a HCP practices safely. There are 3 million nurses in the United States; even the world’s largest medical device companies cannot ensure that they all practice safely. It is not the manufacturer’s responsibility to accept the risk posed by poor clinical practice. So be very careful that you are not straying over into accepting risk for poor clinical practice.
Clinical opinion: I remember a validation study we did which involved priming a saline line of air. Failure to purge the line before connecting to the patient’s central line was deemed to be a FAIL. However, we had a number of consultant anaesthetists who insisted that in their clinical opinion small amounts of air did not pose a risk of harm. So they did not fully purge the line. We marked it as a FAIL. But if the Consultant views the air bubble, judges that it is safe and goes ahead and connects the line, who’s risk is it?
Deliberate misuse: Of course you need to test use scenarios associated with an increased risk of harm. For example if your device is used at night, or by an untrained user. But if a user knowingly and deliberately uses your device for an unintended purpose, then you are not expected to anticipate the risks posed by that ‘abnormal use’; provided that you can show that you made it very clear to users how it should correctly be used.
Someone else’s product: Many devices are used in a system, and therefore need to be used in conjunction with other devices. You do not have access to the other manufacturer’s risk assessment, so you cannot be sure of the risk profile. And you did not design it, so you are not responsible for the risk it poses. If your device relies on the safe use of a different manufacturer’s device, you must make sure that you are not accepting the risk of the other device being used incorrectly. However, if your device needs to be connected to another device, the connection does fall into the scope of your risk assessment.
Iterative usability testing: The uFMEA is intended to be a living document. There are times where manufacturers only realise they need to conduct usability testing when their product is in the final stages of development. If usability testing is conducted iteratively throughout the cycle of design, then the study outputs can be used to continuously update the uFMEA, and provide justification as to why some risks can be removed; e.g. “formative testing has shown the device mitigates risk XX adequately, therefore we have decided to remove the hazard from the validation ”.
Sharks with laser beams: And finally, be realistic about the extent of the hazards. When you have completed your risk assessment, ask yourself ‘does this truly reflect the risk of using our product?’