I think the best way to start this task would be to try and relate the product in the project to another product already on the market and then see what are the risks within that product. You can do this for multiple products to cover all the bases of risks within products in different fields and uses. Then you can categorize your product based on its components after some development, like based on whether or not is has electronics within, it would have different risks than a product that doesnt, and so on until you cover all your bases.

The method that eh76 described is a very sound way of identifying and assessing the risks for a medical device that has a similar medical device already on the market. Additionally, if the category and classification of a medical device are ambiguous since it has characteristics of multiple different medical devices, it can also prove difficult to assess risks as the device is novel. 

If the device is a mix and has characteristics that are of 2 or more categories, then it could prove beneficial to conduct tests to assess the risks from all the devices it is similar to. This would allow one to garner relevant data, but it begs the question of how to quantify the data as the "mixed" device might respond differently to the same stimuli as other devices it is being compared to.

It is a good idea to compare the product at hand to another one on the market and assess the risks of the similar product, however, I would like to add that it is important to take into account the customer opinions and customer reviews. As Dr. Simon mentioned in this week's lecture, it is important to have a customer service representative on the team, as they can give insight into what customers complain the most about, or in terms of risk, what hazards customers have faced regarding the product. Furthermore, a very thorough research of the components of the product should be conducted to assess risk. 

Similar to es446's post mentioning DFMEA, is Failure Mode Effects Criticality and Analysis (FMECA), which is a method to analyze each risk listed. I believe the best way to assess and evaluate risks of a medical device is to adopt very thorough and commonly used methods, such as FMECA, in addition to performing thorough research on customer complaints and/or hazards previously involved with the product. 

In addition to fault tree analysis and FMECA which are widely used, revealing severe hazard potential as early as possible is very important. Preliminary hazard analysis conducted as early as in the conceptual design stage could help save time and effort. 

HAZOP (Hazard and Operational Study) is a suitable method for new complex products. Identifying deviations from design intent makes this suitable for unexplored new products. A team with members having different expertise is normally tasked with the HAZOP. 

Furthermore, performing additional analysis on procedures might be required for a new product. A what-if approach is used in these to identify possible hazards associated with the procedure. Undesirable consequences for each process step are identified in this approach. 

To create a best output it is always better spend a while on assessing the risks. Failure Mode Effects and Criticality Analysis (FMECA) is the most widely used method to analyze and evaluate the risks of a Medical Devices. It identifies where additional preventive measures are necessary to minimize the risks. Risk assessments begin with the evaluation of hazardous situations that have potential to cause harm. FMECA assess the risk associated with the failures and prioritize issues for current actions. Most serious concerns are addressed first to reduce the affect on the end users or patients.

Having worked in the complaints department, we work with risk documents very frequently. Risk management data used in the risk management documents are compiled from historical complaints data as well as anticipated failure rates that come from the tests done before releasing a product. In my opinion, the best way of assessing and evaluating the risks of a medical device is from the validation and verification tests done prior to releasing of the products. Additionally, tests done on cadavers or animals, ethically of course, would help to ensure that once released, the occurrence rates of the failures on the field are not higher than anticipated rates on the risk management documents or else action will need to be taken such as a recall, or action on the manufacturing floor. 

So there's always the choice of comparison. Your product can be compared to that of other companies that have created a similar medical device. Depending on the type of medical devices, there can never be too much testing. The risks of a medical device can easily be assessed with testing the product before release or even documentation of risk assessment.

I would definitely try to compare the product to what is already on the market to get an idea of possible risks that could potentially be detrimental to my product. To assess any risks, research is key step. If you are well informed about the potential risks, there can be fail safes put in place to reduce the amount of risk something can produce or completely eliminate a risk. At the same time, there are some risks that you will not find out until the product is on the market and has undergone extensive consumer use. If you were able to identify most of the severe risks, the product should still be allowed on the market. The issue is that with the historical data of complaints and more in-depth research studies are performed, the product might have to undergo major changes.