On the news this morning, I heard a story that got me thinking about risk management. The headline was “Eight N.H. patients possibly exposed to fatal brain disease”. Further research about the situation reveals just how complex this case is.
It all started when neurosurgeons in a New Hampshire hospital operated on a patient with Creutzfeldt-Jakob disease – a rare but fatal brain disease. However, at the time of the operation the surgeons were not aware of his disease. After the surgery, the instruments were sterilized per normal protocols, and re-used on additional patients. The disease is transmitted by small proteins called prions. These prions are not susceptible to normal sterilization methods, and can only really be killed under extreme conditions with harsh chemicals or environmental conditions, or a combination of the like.
Typical safety risk management logic is such that hazards set off by a sequence of events create a hazardous situation. Exposure to a hazardous situation leads to harm. There are many different ways to describe and document a hazard, but the main point of risk management is to implement risk reduction measures to ‘break’ the chain of events, and ultimately reduce the risk level.
Let’s look at the safety hazard of the instruments in this situation in detail. I’m going to define the hazard as ‘disease carrying proteins’ or ‘prions’.
The sequence of events that give rise to the hazardous situation is:
- Instruments are used on a patient with Creutzfeldt-Jakob disease
- Sterilization does not kill disease carrying proteins
The hazardous situation is:
- Prion contaminated instruments
Exposure to the hazardous situation can lead to the harm:
- Contraction of fatal brain disease
Above I mentioned that in order to reduce risk, you aim to break the chain of events by first looking at ways to modify the design, then protective measures, and then information for safety.
To reduce the possibility that instruments are used on a patient with Cretuzfeldt-Jackob disease, information for safety is the only possibility (i.e. contraindications or warnings could be provided in the provider manual). However, the effectiveness of this control is limited by the fact that the disease has an incubation period before symptoms develop. In the case here, it took a few months before the hospital realized the potential impact to other patients.
To reduce the possibility that sterilization does not kill disease carrying proteins the manufacturer could alter their recommended sterilization protocols for the tools and validate the effectiveness of the sterilization. A few issues are posed with this control – the first being that the tools may have to be re-designed to withstand extreme sterilization conditions. Additionally, the hospital’s sterilization equipment may not be able to handle these conditions. The other possibility is for the manufacturer to recommend disposal of the instruments, but since it is routine practice to re-use instruments of this type, this option may not be viable either.
The severity of this harm could directly relate in death of the patient, putting it at the worst severity level. However, the probability of this situation may be low based on the prevalence of the disease state. Regardless there is some level of residual risk. In this situation, device manufacturers are stuck between a rock and a hard place – the risk level of this situation can only be reduced so far. In this scenario, it becomes important for manufacturers to disclose the residual risk of their device.
You can see how complex this case is, and what a tough job risk management is. ISO 14971 does provide guidance, but ultimately manufacturers have a responsibility to develop a risk management system and determine their comfort with the overall risk level of the devices they make. I thought this was a practical example of where risk management fits into the device lifecycle.
Image Credit: karmaOWL @ Flikr