Welcome to Health Care POV | sign in | join
Stepwise Success

Reproducible Cutoffs

Published April 22, 2013 7:25 AM by Scott Warner

A million monkeys on a million microscopes probably couldn’t bang out an accurate WBC differential in a million years, but a lab tech could teach a gibbon to run a chemistry analyzer in about ten minutes.

I’m kidding. But the subjective variability of microscopic analysis versus objective reliability of engineered automation is known. WBC differential, abnormal cell identification, manual cell count, cell count estimate, crystal analysis, urinalysis, or Gram stain often have significant variation. It’s the nature of the primate.

Reliability combines accuracy and precision; let’s briefly consider both.

A reason for inaccuracy is confirmation bias, described by Science Daily as “a phenomenon wherein decision makers have been shown to actively seek out and assign more weight to evidence that confirms their hypothesis, and ignore or underweigh evidence that could disconfirm their hypothesis.” In other words, we tend to find what we’re looking for.

For example, if we glance at a report and see the patient MCV is 75 fL, our confirmation bias will seek microcytes on the peripheral smear. Techs who don’t know the instrument MCV may or may not see microcytosis.

A reason for imprecision is our tendency as scientists to seek a greater degree of precision than necessary. We think in procedural, stepwise fashion and not outcomes.

For example, in urinalysis we may report white cells as 0-4, 5-10, 11-15, 16-20, 21-30, etc. It’s difficult to imagine a laboratory can reproduce this suggested precision given the significant variation that exists in spinning the sample, preparing the slide, time under the scope, etc. I’m not sure it matters to a physician.

While we can’t always prevent confirmation bias, we can change our microscopic criteria to be reproducible. One approach is to base grading around significant cutoffs. In urinalysis, we might report squamous epithelial cells as less than or greater than 10 per LPF, for example, a decision limit to determine contamination.

Choosing reproducible cutoffs may not help a million monkeys read a differential, but it might keep us techs from going bananas at the scope.

NEXT: Corrective Action

0 comments

leave a comment



To prevent comment spam, please type the code you see below into the code field before submitting your comment. If you cannot read the numbers in the image, reload the page to generate a new one.

Captcha
Enter the security code below:
 

Search

About this Blog


    Scott Warner, MLT(ASCP)
    Occupation: Laboratory Manager
    Setting: Critical Access Hospital
  • About Blog and Author

Keep Me Updated