Hello Anon,
Unfortunately the answer to your question is probably not answerable based on the information provided because it depends on many factors including type of assay, results being obtained, intended use, type of disease, disease state, monitoring, diagnosis, subject population, etc. If your regulatory pathway is a 510(k) often times the predicate device will have the type of study and number of samples performed in the 510(k) Summary which you should consider at least mimicking or doing similar to those performance studies already performed. If you are really unsure on the sample sizes for the different performance studies you can also have a Q-Submission meeting with the FDA to outline and describe your performance testing studies to gather feedback on the sample size (in addition on acceptance of having sites outside the US.) Often performance studies can be performed outside the U.S. but again depends on the disease, patient population, ethnicity concerns, interference concerns, etc.
------------------------------
Richard Vincins RAC
Vice President Global Regulatory Affairs
------------------------------
Original Message:
Sent: 20-Jul-2020 10:24
From: Jeffrey Freedman
Subject: Performance evaluation IVD
While you didn't specify what samples you are referring to, I'm assuming you mean the samples to calculate estimates of clinical sensitivity/specificity. The number of samples required to demonstrate that your product's clinical performance is substantially equivalent to the predicate requires statistical justification. You will need to calculate point estimates of your clinical sensitivity and specificity as well as the 95% confidence intervals; the number of samples you test will determine how broad the 95% CIs are. I'm sure the FDA would prefer that at least one of your clinical sites is located in the U.S., but you should consider submitting a pre-submission to obtain their feedback on your study designs.
------------------------------
Jeffrey Freedman
Regulatory Affairs Specialist II
Lowell MA
United States
Original Message:
Sent: 20-Jul-2020 08:13
From: Anonymous Member
Subject: Performance evaluation IVD
This message was posted by a user wishing to remain anonymous
Hi to all,
We are developing a new IVD and we are planning to apply for a 510(k) clearance.
We would like to initiate performance evaluation studies and we are trying to calculate the amount of samples we have to analyze in order to demonstrate that our device is at least as safe and effective as the predicate device.
Furthermore, it is necessary to conduct those studies in USA?
Thanks in advance