Manual Differential agreement with Auto Diffs

[LG53 10]

I'm in the process of updating our Hematology manual. We have what I feel is a rather strange requirement. If we perform a manual differential, we are supposed to compare it to the auto diff and it has to agree within certain criteria:

Percentages of Neutrophils and Lymphs have to agree within 10 cells and all "other" cells have to agree within 5 cells. Furthermore, the manual diff has to be recorded on the instrument printout in case there is a question about the results.

I understand the logic behind double checking your diff results - avoids errors when performing the diff (wrong key type errors), but I have a problem with writing everything down on the printout. Currently, we are drowning in paper. We save normal CBC printouts for a month, and those with manual slides for 5 months.

I want to get away from the recording of the results on the printout to save time, but am encountering resistance because no one really knows exactly why we started. It may be some obscure Joint Commission rule, but I can't find it anywhere. I've asked friends who work at other hospitals if they do it and have always gotten "NO" for an answer. Any thoughts or recommendations on this would be greatly appreciated!

[SMILLER 708]

I don't think many labs go to this extent to "verify" manual diffs. For uncomplicated auto CBCs, the auto-diff is going to be more accurate anyway, so I would wonder why you would want to even do an auto diff.

We do perform manual diffs when the auto diff doesnt work for some reason, and in certian clinical cases, such as for leukemias and whatnot. But we do not document diff comparisons. Our diff procedure does direct the tech to correlate with analyzer results in general.

I would say that if you cannot find a reg to support the practice, stop doing it.

Scott

[PAWHITTECAR 89]

I agree with SMILLER, If your techs do an anual compentency why are you "checking" them every time they do a diff? We only do manual diffs when something is wacky on the auto diff and then I wouldn't think that they should match.

[nziegler 36]

we run one patient check per day to "verify" the auto diff vs manual diff. all other differentials are just put into the computer. we used to write the results on the instrument printouts as well, but no longer do that. there is no regulation stating you need to. i have a feeling that may have started "back in the day" when differentials could not be done with keystroke directly into the computer. the printout would have served as an "intermediate worksheet" and clerical errors would have to be checked.

as far as saving the printouts, we only save ours for a week. the only exception is when the auto diff is released, in NYS those histograms need to be saved for 6 months.