In-vitro diagnostic reagents are important auxiliary means for disease diagnosis and treatment. In addition to optimizing reaction patterns and systems, reasonable performance analysis and evaluation are mandatory procedures and the basis for quality assurance for a newly developed reagent to truly emerge. Generally, performance evaluation of IVD reagents includes detection limit, linear range, reportable range, accuracy, precision, interference experiment, stability, and reference range.
The detection limit refers to the lowest detectable concentration of the analyzed substance by the detection method, also known as analytical sensitivity.
Analytical sensitivity: Generally calculated using a 95% confidence limit, repeat blank sample measurement 20 times, calculate the mean (X) and standard deviation (SD) of the 20 reaction measurements, and this is the analytical sensitivity of the in-vitro diagnostic reagent, which is also a required content in the registration data.
Functional sensitivity: After diluting the low-value sample several times, repeat the measurement at least 10 times, calculate the mean, standard deviation, and coefficient of variation (CV) of the detection signal for each low-value sample, and choose the low-value sample average concentration corresponding to a CV greater than 20%. This is the functional sensitivity of the IVD reagents.
The linear range refers to the range where the final output value (concentration or activity) of the detection system is proportional to the analyzed substance's concentration.
To establish a quantitative detection method's linear range, 7-11 concentration levels need to be selected within the expected detection range. Prepare sufficient high-value (H) and low-value (L) samples, and mix them in different proportions to prepare intermediate concentration samples, with the mixing relationship according to equal intervals or other determined proportions. Some important concentration levels need to be considered:
The lowest analytical concentration or linear range lower limit;
Different medical decision level values;
The highest analytical concentration or linear range upper limit.
The reportable range refers to the concentration range of the measured substance that is significant for testing.
Lower limit of the reportable range: Use the CV value indicated by the method performance as an acceptable critical value, and select the lowest concentration level with a CV value equal to or less than the acceptable critical value from the data as the lower limit of the reportable range.
Upper limit of the reportable range: When the measured value exceeds the upper limit of the linear range, the measured value should be regarded as an inaccurate value, and the sample needs to be diluted. Due to matrix effects, any sample cannot be diluted indefinitely. That is to say, each experimental project has its maximum dilution factor. The product of the maximum dilution factor and the upper limit of the linear range is the upper limit of the reportable range.
The accuracy refers to the consistency between the detection results and the measured true value.
Recovery experiment: Essentially, it is the ratio of the concentration of the added standard substance obtained by experiment measurement to the concentration of the added standard substance calculated theoretically.
Methodological comparison: For comparative methods, use laboratory methods that meet the manufacturer's requirements or use worker reference methods.
The precision usually refers to the consistency between independent test results under specified conditions. The degree of precision is obtained by statistical methods and is represented by the digital form of inaccuracy, standard deviation, and coefficient of variation (CV).
Only evaluate within-batch precision: The determination of within-batch precision should use the same type and batch number of the reagent box calibration material (if possible, only perform one calibration) and perform at least 20 repeated determinations of the sample within a batch.
Simultaneously evaluate within and between-batch precision: Test two batches of samples each day, measure duplicates for the same sample in each batch, and work for 20 days. Evaluating the precision requires 40 pairs of data or 80 test results at the end. The within-batch precision can be obtained from the duplicate measurement results of the same sample, and the between-batch precision can be obtained from two batches of 80 data.