Hematocrit comparison acceptability criteria

Hi. I had this issue with Chem-8 cartridges last year and ended up doing months of testing just to find out I needed to switch my K2EDTA to K3EDTA in my DE customization settings.

This may very well not fix your problem, but I found out that which KEDTA setting you chose was based on what aligned better with your primary instrument.

Abbott has a technical bulletin about it you can find at this address called: K2EDTA and K3EDTA Customization for Hematocrit on the i-STAT System

We experienced this as well.  We could hit 4% for the Instrument Correlations but not Method.  I had searched and read on this list serv that the 4% was acceptability criteria for proficiency testing.  I created an SBAR for our Medical Director and referred back to the previous allowable limit of 7%.  Our "Assessment" was:  A 7% allowable difference is achievable for Method Correlations, and a 4% limit is achievable for Instrument Correlations.  Adopting the proficiency testing criteria for consistency is common, however, adjusting them internally based on practical considerations for different methods and platforms and approval by our Medical Director is desired to tailor our quality control processes to better fit our operational realities while still maintaining high standards of accuracy and reliability.  He approved and signed off on it so we now use 7% for Method Correlations and 4% for Instrument Correlations.   

Hi, we have this statement in our Analytical Correlation workbook and procedure, approved by our Senior Clinical Laboratory Director. " Results must agree within the Total Allowable Error limits for like devices. *** For i-STAT to Sysmex H&H correlations: use the TAE of 1.5 g/dL or 14% for Hemoglobin and use the TAE  5% (absolute difference) or 10% (calculated as %bias) for Hematocrit. For non-like devices concordance is applicable (when the value leads to the same clinical decision). We had issues in past with these results correlating and narrower AE criteria.

Hi All. Correlating the iSTAT to other methods is challenging because it measures the Hct and calculates the Hgb which is different from every other analyzer. Abbott suggests correlating the Hct and NOT Hgb.   
Per Abbott:
 EXPECTED LEVEL OF METHOD AGREEMENT Average i-STAT hematocrit results over a group of samples should normally agree with those from the comparative method within ± 2 %PCV at 29 %PCV and below, ± 3 %PCV from 30 to 50 %PCV, and within 10% above 50 %PCV when the following conditions are met: • i-STAT analyzers are customized correctly. • Comparative analyzer is calibrated correctly. • Sample handling is optimal for both i-STAT and comparative methods. • Samples are unaffected by factors listed in the i-STAT Cartridge and Test Information sheet for Hematocrit or in the user documentation for the comparative method 

We do not use the proficiency testing criteria necessarily. Our medical director approved 6% or 2.0 for the acceptable difference for hematocrit.

Thank you all for your responses! We did adjust our KEDTA setting to match the main lab's and it didn't work. How are you all testing the samples? Are you taking samples from the purple top before analyzing it on the Sysmex or some other way?

We use the correct tube types for the analyzers, so purple for Sysmex and green top or lithium green syringe for ISTAT.  We usually have the different specimens from patients drawn at the same time.

That's exactly what I was doing and had a hard time getting at or below 4%. I will discuss with my medical director about the possibility of increasing the acceptability criteria.

We also use the same acceptability criteria as Mindy Cotter.  Also, I have had no issues with passing proficiency testing using 6% or 2.0.