While most clinical assays have linear calibration responses, where the increase in a subsequent calibrator concentration produces a proportional increase in assay signal, this is not always the case.
Some calibration curves appear to "drop off" at the highest concentration calibrators. In this case, there is an increase in signal which is not proportional to the signal that would be expected if the assay was truly linear. If this is the design of the clinical assay, then a non-linear calibration curve may be acceptable.
