Back to Blog

How low can we get?

When validating a new method, one of the key questions to be answered is how low analyte concentrations can be reliably detected. Here’s a short summary about relevant concepts and how they are studied with the help of Validation Manager. More details can be found from CLSI guidelines EP17 and EP06.

LoB – Limit of Blank

Many test methods give positive results for negative samples. To know whether a result really indicates a positive analyte concentration, Limit of Blank should be determined.

LoB means the measurement value under which nearly all (at least 95%) of results from negative samples remain. Samples giving results below LoB should be interpreted as negative samples.

In Validation Manager LoB can be established for quantitative methods using the brand new Classical Analytical Sensitivity study. A set of replicated negative samples is measured and the results are interpreted using statistical analysis (Classical approach) to find a value for LoB.

LoD – Limit of Detection

Similarly, as negative samples may give positive measurement values, samples with low analyte concentrations sometimes give such small measurement values that they are interpreted as negative samples. To know how high analyte concentration is needed that the sample is interpreted as positive, we need to determine Limit of Detection.

LoD means the lowest analyte concentration in a sample that can be consistently detected (in at least 95% of tested samples).

Validation Manager offers two options for establishing LoD. In both cases a set of replicated low positive samples is measured and the results are interpreted using statistical analysis to find a value for LoD.

  • Classical Analytical Sensitivity study offers a tool for determining LoD of a quantitative method. The study uses Classical approach. To establish LoD you also need to measure LoB.
  • For qualitative methods there is Analytical Sensitivity study that uses Probit approach for determining LoD. It is designed for PCR- based molecular diagnostic methods where LoB is effectively 0.

LoQ – Limit of Quantitation

When we are interested in the actual analyte concentration, it is not enough that we can reliably detect whether the sample contains the analyte or not. The quantitative values given by the test must also be consistent enough. Limit of Quantitation means the lowest amount of a measurand that can be quantitatively determined with stated accuracy. Measurement values between LoB and LoQ can be stated as positive, but reliable quantitative values can only be given for results exceeding LoQ.

Validation Manager does not yet support establishing LoQ. We will introduce a new study for this purpose in future.

Lower limit of linear range

The accuracy of a measurement is not enough to give you the correct quantitative value. The instrument needs to know how to convert measured signal to analyte concentrations. This is usually done with a linear function. As the measured signal is not strictly linear, it is important to know what is the linear range where the signal is close enough to linear to give relevant results. So, to get a quantitative value from the measurement, the measured value must exceed LoQ and fall between the linear range.

In Validation Manager the linear range is established using Linearity study. Linearity is discussed in our blog post The Linearity Story.

How to measure LoB and LoD using Classical approach?

Validation Manager helps you plan measurements to get a statistically adequate data set for determining LoB and LoD. The minimum number of samples and total number of replicates depend on whether you are establishing LoB and LoD of a new method or verifying the claims of a test manufacturer. The measurements should be divided to at least three days to get reliable results. If needed, you can determine LoB before planning LoD measurements.

After conducting the planned measurements, simply import your data to Validation Manager and tell Validation Manager which samples are negative and which ones are low positive. The report gives you the statistical results for LoB and LoD and access to sample specific results including results of each individual measurement.