What a PER actually proves
A Performance Evaluation Report demonstrates that your IVD performs as claimed — that its sensitivity, specificity, and other metrics hold up against a credible comparator on a representative population. It is the clinical heart of an IVD submission.
Clinical
Design the study before you touch a sample
The biggest red flag for a reviewer is a statistical plan that looks written after the data arrived. Pre-specify everything:
- Primary endpoints (sensitivity, specificity) and secondary endpoints (PPV, NPV)
- Sample size powered against the lower confidence bound of your claim
- A credible comparator — gold standard or an accepted reference method
- How indeterminate results will be handled
Lock the statistical analysis plan before the first sample runs. Any deviation needs a written, dated rationale.
Common reasons a PER gets bounced
- Underpowered sample size relative to the performance claim
- A weak comparator (e.g. another rapid test) instead of a reference method
- Single-site studies with no geographic diversity
- Indeterminate handling decided after the fact
Anticipate the reviewer's questions during design, and your PER clears on first review instead of coming back with a stack of queries.