repeat rates on IHC
There was some interest in out PI project and how we are going about it.
And thanks already to those who have responded with their repeat rates.
So, for those who are interested, here is a brief overview of what I did
and the parameters that I set up to review our repeat rates.
1) All slides are counted if they failed to stain correctly, even if the
pathologist choose not to repeat that specific antibody.
2) Slides that were repeated multiple times (due to adherence problems,
processing, etc.) were only counted once.
3) the last 100 slides/antibodies per tech were selected to form the
4) information is based on what the pathologist documented on the IHC QC
sheets (given to them per case.)
Our first task was to establish a baseline repeat rate, which we did and
discovered that , we averaged a 7% repeat rate:
1.5% = tissue adherence problems (high pH HIER combined with breast tissue
not optimally processed)
1.0% = non specific background in unknown (the bulk of these directly
related to Kappa & Lambda)
1.0 % = weak staining
1.0% = control is fine, unknown has some questionable staining or
1.5% = unknown, no comment by pathologist on QC chart
1) Staff will establish a goal repeat rate to aim for based on our existing
rate, the overall rate received from outside labs with pathologist imput.
2) Each staff member will set their own individual goal for repeat rates.
3) Repeat rates will be monitored again for the next 100 slides per tech.
By the by, we are also doing this with repeat/recut rates on routine H &
E's and special stain repeats. We may place this process into our
quarterly QA monitors.
About our lab:
We are a clinical , diagnostic, Hospital lab with no research, but a small
amount of outside cases
We have 500 beds, 6 pathologists and no residents
We use a DAKO stainer
We have a stable of 65 antibodies
We have 4 techs rotating through the IHC bench, one week at a time
We perform about 250 slides/antibodies per month.
Control and unknown are on same slide
Colorado Springs, CO
<< Previous Message | Next Message >>