5 tips helping evaluate RISK using ChatGPT

Today RISK assessment results

There are several reasons that medical devices may be recalled. While the implications are severe and may result in loss of life or financial depths, this article is targeted to the engineering phase in trying to see how the new technologies and discoveries can help our industries to perform better. Looking at some of the reasons for recalls from the past 2022 events, we have highlighted the following:

• Design defects: This can include issues such as poor materials selection, inadequate testing, or subpar manufacturing processes that result in defects in the device.

• Performance issues: This can include problems such as device failure, inaccurate readings, or a lack of efficacy.

• Adverse events: This can include incidents such as patient injury or death, as well as problems such as infections or allergic reactions associated with the use of the device.
• Non-compliance with regulatory standards: This can include issues such as failing to meet requirements set by the FDA or other regulatory bodies, such as the European Union.
• Labeling or packaging errors: This can include incorrect labeling, inadequate instructions, or misbranded packaging that may need to be clarified or lead to incorrect usage of the device.

It is important to note that medical device recalls can have severe consequences for patients, healthcare providers, and the medical device industry. Therefore, medical device companies must evaluate thoroughly, in every step of the product lifecycle, the risk associated with their devices and take steps to mitigate those risks to ensure patient safety.

What is ChatGPT

ChatGPT is a language model developed by OpenAI that is trained to generate human-like responses to natural language text inputs. It is a form of artificial intelligence (AI) that uses deep learning algorithms to generate text based on large amounts of data it was trained on.

In the context of medical device design, ChatGPT can help identify risk by analyzing large amounts of data, such as design specifications, user scenarios, and regulatory standards, to identify potential hazards and assess their severity. This can include evaluating the likelihood of a hazard occurring and the potential harm it may cause to users or patients.

By using natural language processing (NLP) techniques, ChatGPT can understand and process the language used in these sources of data and provide relevant information to engineers and designers. This information can help inform risk mitigation strategies and assist in ensuring that the medical device design is compliant with relevant regulatory standards.

Overall, the use of ChatGPT in the design of medical devices can help streamline the risk evaluation process, provide engineers with accurate and up-to-date information, and improve the overall safety and efficacy of these devices.

Here are five advantages to using ChatGPT to help engineers perform risk evaluation during the design phase of medical devices:

1. Speed and efficiency: ChatGPT can assist engineers in performing risk evaluations quickly and efficiently. The model can process large amounts of data and provide relevant information in a matter of seconds, allowing engineers to make informed decisions in a timely manner.
2. Improved accuracy: ChatGPT can provide engineers with accurate and up-to-date information, reducing the likelihood of errors and ensuring that risk evaluations are based on the latest data and regulatory standards.
3. Enhanced collaboration: ChatGPT can facilitate collaboration between engineers, as the model can provide a common platform for sharing information and data. This can help ensure that all relevant parties are on the same page and that risk evaluations are based on the best available information.
4. Increased consistency: ChatGPT can provide engineers with a consistent and systematic approach to risk evaluation, reducing the likelihood of subjective interpretations and ensuring that all evaluations are based on the same criteria.
5. Cost savings: By streamlining the risk evaluation process, ChatGPT can help reduce the time and resources required for this process, leading to cost savings for medical device companies.

Overall, using ChatGPT for risk evaluation during the design phase of medical devices can help ensure that these devices are safe and effective and can help companies bring their products to market more quickly and cost-effectively.

1. Identify potential hazards: ChatGPT can assist in identifying potential hazards associated with the medical device during the design phase. This includes analyzing user scenarios and using cases to identify potential risks, such as electrical shock, mechanical failure, or user error.
2. Analyze risk severity: Once potential hazards have been identified, ChatGPT can assist in evaluating the severity of each risk by analyzing the likelihood of occurrence and the potential harm to users or patients.
3. Consider mitigation strategies: ChatGPT can help evaluate various methods for mitigating risk, such as design changes, warning labels, or user training. The model can also assist in determining the effectiveness of these strategies and assessing their impact on the overall risk profile of the device.
4. Evaluate compliance with regulations: ChatGPT can assist in evaluating the medical device design against relevant regulatory standards, such as those set by the FDA or the European Union. This includes analyzing the device’s design, materials, and manufacturing processes to ensure compliance with these standards.
5. Monitor for post-market risks: ChatGPT can assist in monitoring post-market risks associated with the medical device, such as adverse events or product recalls. The model can help evaluate the causes of these incidents and determine what changes, if any, need to be made to the design to prevent similar incidents in the future.

Example Case Study

The following example comes from a use case that happened just recently during the year 2022 on a recall made to HeartWare Ventricular Assist Device (HVAD) made by Medronics that was exposed to power failure. The question that was asked to ChatGPT:

What possible risk causing recall could be in HeartWare Ventricular Assist Device?

The answer coming back from ChatGPT

“Several potential risks could cause a recall of the HeartWare Ventricular Assist Device (HVAD). Some of these risks include:

• Device malfunctions: The HVAD is a complex device, and there is always a risk of malfunctioning that could cause the device to stop working correctly.
• Infections: Infections can occur at the site where the HVAD is implanted and can cause serious complications.
• Thromboembolic events: Thromboembolic events, or blood clots, can occur as a result of HVAD and can cause serious health problems.
• Hemolysis: Hemolysis, or the destruction of red blood cells, can occur due to HVAD and cause serious health problems.
• Electrical malfunctions: Electrical malfunctions can cause the HVAD to stop working properly and could result in a recall.

These are some of the possible risks associated with HVAD that could lead to a recall. It is important to note that the HVAD is a lifesaving device for many people, and the benefits of using the device often outweigh the risks. However, it is always important to be aware of potential risks and to talk to your healthcare provider if you have any concerns.”

Diving deeper into each section could explore more valuable information; here is an example of how to pill the risk one level more.

What type of electrical malfunctions can cause the HVAD to stop working properly and could result in a recall.

The ChatGPT responded…

“Electrical malfunctions in the HeartWare Ventricular Assist Device (HVAD) can result in several issues that could cause the device to stop working correctly and could lead to a recall. Some of the types of electrical malfunctions that could occur include:

• Power failure: The HVAD is powered by a battery, and if the battery fails, the device will stop working.
• Leaks in the electrical circuit: If there is a leak in the electrical circuit, the device may not function properly and could stop working altogether.
• Controller failure: The HVAD is controlled by a computerized controller, and if this controller fails, the device may stop working.
• Software malfunctions: The HVAD relies on software to control its functions, and if the software malfunctions, the device may stop working.
• Electrical interference: The HVAD can be affected by electrical interference from other medical devices, and if this interference is significant enough, it could cause the HVAD to stop working.

These are some of the types of electrical malfunctions that could cause the HVAD to stop working properly and could result in a recall. It is essential to be aware of these potential risks and to talk to your healthcare provider if you have any concerns about the functioning of your HVAD.”

Although the answers are not specific, they can expand the engineer’s imagination on what attention needs to be given to which functions of the device; AI ensures you don’t miss anything obvious. Mitigation of such an event could be of several natures, but I will leave it to our fantastic engineers, who will know how to address it.

Here are some sample prompts for ChatGPT

What risk medical device batteries may cause when experiencing electrical faults in [MedicalDeviceTypeName]what type of failure in medical device manufacturing could cause a recall?What type of RISK could be caused when using Bluetooth keyboards in medical devices?What are the top common errors in the packaging of a medical device?

Conclusion

Orcanos recommend constantly reviewing the output coming out from the results of the API tools you intend to use and not taking it as it is directly into your product. We are beginning a new era and researching how to incorporate AI technologies into information management systems such as Orcanos ALM & QMS may introduce game-changing value. New information will be available to all stakeholders in the product lifecycle, helping put the best effort to prevent the following recall.