Hey all,

Currently in our lab we are performing 20 (yes 20!) body fluid correlations per year. I am wondering if this is really necessary.

Here's the backstory:

We have 2 LH 780s, validated for body fluids with body fluid mode, WBC low end linearity >2.0x10^3, RBC low end linearity >0.01x10^6.  WBC below linearity defaults to manual (hemacytometer) count.

Does anyone know of a specific CAP requirement for this correlation?

My rational (for not performing these correlations) is this: both methods (manual and automated) have been independently validated, QC is performed daily/shiftly, and all scientists are subjected to yearly competenices and, of course, up-to-date certification. So, what are these correlations proving? I feel as though it is a waste of time/energy/money/etc because we only ever perform manual counts if the auto count is below linearity (i.e. we would never perform a manual count of a body fluid with a WBC of 40,000), so what is the use in correlating that?

Insights to what your lab does, supportive arguments I can use to change this practice, or justifcation for this practice are much appreciated!

1. It is not necessary to repeat your correlation/ validation constantly- this is a waste of your time and not required

2. Why are your linearity cut offs so high for WBC?? LH are basically linear to 0. We go down to 0.2 for our WBC counts.

3. Automated counts are great, but you must also check your cytospin slide to validate a relative WBC:RBC ratio because:

 4. I recently discovered that the RBC count on LH is not very valid anytime the WBC count is high- The WBC channel counts all nucleated cells, but unfortunately, the RBC channel counts all cells- nucleated or not. Therefore, because body fluids frequently have a high WBC relative to RBC, the RBC count becomes less and less accurate. For example, we had a 65,000 WBC count and guess what? our RBC count was .065 x106 uL = 65,000. Well, these were all the WBC being counted in that channel!!!! This becomes significant for us for any RBC reported as >50,000.

5. I would suggest that anytime the WBC count is over 10,000, that you subtract that from the RBC channel (would be .01). 

6. Accuracy in RBC counts of body fluids (except for CSF) not especially important except perhaps relative to the WBC count. This is why our policy and validations on RBC are reported (except for CSF that are clear and so probably below our linearity or any sample requiring a manual count- those that for whatever reason cannot go on LH) as either "<50,000", or a number if greater.