Testing underpins the entire medical device development process. It provides the tangible evidence of device efficacy and safety. Testing generally begins with a reasonably ad hoc approach in Phase Zero for proof of concept and then shifts to a more formalized testing strategy to address safety and efficacy. At this point in time, product definition has matured to include the identification of specifications for the device to be tested. The following is a brief overview of a systematic approach to device testing, from laying the groundwork to writing the final report.
Establish the groundwork
It starts with the rationale for testing the device. What is the current objective of the device? Is it going to be used for an animal trial, clinical trial or regulatory submission? The end use of the device has the largest influence on the testing requirements. These may include standardized testing, custom testing protocols or simplified testing strategies of standardized tests (pre-compliance testing). A comprehensive review of applicable medical device standards should be performed to determine testing requirements and methods. Now is a good time to involve your Regulatory Affairs & Quality Assurance (RA/QA) representatives. They can help identify any specific regulatory and/or quality requirements for your device to ensure your design and testing plan addresses the requirements. Work with them to determine specific institutional review board requirements for clinical studies. For regulatory submissions that have unique testing requirements, consider the merits of pre-submission meetings to highlight gaps and increase confidence in your approach.
Once the ground work for your testing strategy has been established, a strategic plan can be generated including resourcing requirements, identification of testing equipment/materials and schedule creation. The strategic plan considers the following:
- Will you have a single monolithic protocol or multiple focused versions? A single protocol may reduce the number of samples required and potentially reduce the testing schedule and cost, but it has the downside of potentially adding documentation and test execution complexity. For more complex testing, protocols which define a single test are often more efficient to review, execute and repeat if required.
- Estimate how many samples are required for each protocol. Testing sample lead-times and costs are often non-trivial. The budget and schedule implications should be part of the project plan in order to avoid unexpected cost overruns or delays.
- Identify long lead time tests. Certain tests such as biocompatibility, accelerated aging and IEC 60601 can take weeks or months to finish. Make sure you initiate these tests as early as possible once design / manufacturing processes have been frozen.
- Identify if you need any specialized equipment to perform testing and if you need to outsource the testing to 3rd party test houses. Specialized equipment tends to have long lead times and direct costs. Specialized testing houses often have long lead times, set-up constraints and queues (if devices cannot be tested in parallel).
- Understand the possibility and impact of a design or manufacturing change between now (time of testing) and your objective. Changes made to the device after testing has been completed will, at best, require the creation of justifications to demonstrate equivalence. At worst, they will necessitate retesting. Prior to testing, it is always a good idea to perform a design review, check your Design History File (DHF) one last time, carry out a risk analysis review prior to testing and examine the critical components. A small amount of due diligence, even with the added effort, can save money and schedule in the end.
- Plan response strategies for failed tests. The associated mitigations and risk response strategies will then get added to your program risk registry and program plan as appropriate. This is a big part of reducing program risk.
Detailed Testing Protocols
With the strategic plan in place, individual tests can be planned at the detailed level. The methods of writing testing protocols and reports tend to be company-specific, but consider the following general guidelines:
- Provide a brief overview of the of the device and the objective of the testing (refer to the specification being tested)
- Determine the sample size and define the statistics that will be used. Look in regulatory standards to determine if the sample size is defined for the individual tests. Otherwise, perform statistical analysis to determine the appropriate samples size. Statistical tests that will be used to analyze the results should be defined in the protocol.
- List all samples, materials and equipment that will be used for testing (serial numbers and calibration dates should be included)
- Describe procedures that will be used to test the device. The procedures should allow the experiment to be independently reproduced.
- Define the acceptance criteria. Ensure that there is no ambiguity.
- Provide the results and any discussion. Any deviations from the testing protocol can be discussed and justified.
- State the conclusions. This should be a repetition of the acceptance criteria, stating if testing resulted in a pass or fail.
Summarize the Results
Finally, create a summary report of all the testing that has been performed. The report should facilitate both executive level and “deep-dive” reviews: program level implications of the testing for the former and also a detailed review of the methodology and test results for the latter.
In summary, approach testing in a systematic and planned manner. You will not only reduce or mitigate device related risks, but also reduce project risks such as schedule slip and cost overrun.
Image: StarFish Medical
Paul Hulme is a StarFish Medical Project Manager. His professional experience includes working with the Canadian Space Agency and 5 years in Switzerland working at Zimmer GmbH. Paul studied Mechanical Engineering at the University of Victoria, completed his Masters of Mechanical Engineering at the University of Calgary, and his PhD in Biomedical Engineering at the University of Bern.