NOTE: Stabilized control materials must be at two different analytic levels (i.e., "normal" and
"high"). Three levels of control is a conceptual carryover from clinical chemistry, and does not
apply to hematology particle counting. Dilute, "low-level" (e.g. leukopenic and thrombocytopenic)
"oncology" controls are less informative indicators of calibration status, and are neither required
nor recommended. For example, a 10% calibration bias will be numerically most apparent in a
high-level control, less apparent in a normal-level control, and perhaps inapparent in a low-level control; it would be quite extraordinary for a low-level control to indicate a calibration problem that
is not revealed by the other controls. There should be some relationship between the frequency
of control runs and the numbers of patient specimens processed. If the frequency of commercial
control use is less than two control specimens per 24 hours, one or more of the additional
approaches to QC must be employed to produce a total of at least two different data points per
24 hours.