We do a lot of counting in the laboratory: white blood cells, abnormal red cells, urine formed elements, and microbiology colony counts. I’ve worked in labs where these are precise, for example, reporting urine microscopic red cells as rare, few, 0-1, 5-10, 11-15, 11-20, etc. Theoretically, if techs count so many fields to achieve an average there is reliability and less variability.
On the bench we know that isn’t true, of course. There is huge variation between techs, from day to day with the same tech, and from specimen to specimen. It seems nutty to me to attempt precision when counting something prepared from a wedge smear, wet prep, or streaked agar plate. These preparation techniques are designed to separate populations and not necessarily provide a quantitative measure (urine culture loops are an exception).
If the goal is to make sure Tech A counts the same one day as the next and the same as Tech B, there are better ways than vigorous procedural controls. The analytical precision of these semi-qualitative measures should be considered. Similar to a chemistry test, there is a cutoff below which results are too imprecise to be reported.
One approach is to develop significant cutoffs and reduce the number of reported values between them. This eliminates an illusion of precision but simplifies the report for physicians. Examples:
- Microbiology colony counts - count <10, 10-100, >100
- Urine white cells - count <2, 2-20, >20 per HPF
- Urine red cells - count <5, 5-50, >50 per HPF
- Peripheral smear red cells - count Present, >5 per HPF
Limits such as these are somewhat arbitrary. Is it significant if there are intact white blood cells in a urine sediment between 20 and 50 vs. greater than 50? I’m not sure. I think it’s more important for bench techs to concentrate on looking at the big picture and identifying abnormalities such as granular casts and other formed elements that are important. We can easily be lost in details if a procedure itself is too detailed.
But that’s just an idea. What does your lab do?
NEXT: A Better Marker for DKA