Dear colleagues!

Could you clarify for me please where you measure hemoglobin and hematocrit for transfusion?

Central or peripheral lines (venous, arterial) or in capillary bed by heelstick? 

I failed to find much info about that. For example this paper  https://doi.org/10.1053/j.semperi.2008.10.006 (pretty old one 2009) states that "central measurements are preferred".

If so is the any difference between UAC and UVC hemoglobin and hematocrit?

Many thanks!

Hi @Andrej Vitushka you can use both central or capillary. For cutoffs please check the PINT study. We use table 1 low threshold cutoffs for transfusion.

https://www.ncbi.nlm.nih.gov/pubmed/16939737

PINT trial.pdf

Hamed, thank you! Is there any difference between venous and arterial Hb/Ht? 

I could not find any data from newborns, but this study on healthy (young) volonteers showed only a slight diff: https://www.ncbi.nlm.nih.gov/pubmed/11553055

This correlation study in adults in intensive care also suggest a good correlation: http://www.thejh.org/index.php/jh/article/view/231/186

We only take venous blood (or arterial from UAC), but good to keep in mind that capillary hemoglobin levels are  ~10% higher than central values.

We also use the PINT cut-offs.  However, I want to call people's attention to the problem is knowing what device you are using to measure H/H

 If your institution has POC or on-unit Hgb/hct available (for example on an I-stat or similar device or on a blood gas analyzer in the ICU) you need to ask your lab what the accuracy of those devices is compared to a formal H/H in the main lab, particularly at the low end of the Hgb range where you will be making transfusion decisions.  The chemistry for some of the POC devices is influenced by what else is in the sample.  For example the I-Stat (the POC device I am most familiar with) only calculates a Hct, not a Hgb (it just divides the Hct by 3 and reports that number as the Hgb).  The Hct it 'measures' is corrected for Na (but not other electrolytes which may affect conductivity - the method by which Hct is calculated/measured), there is no correction for leukocytosis and the machine does not adjust for low TP or hyperlipidemia (but there are published manual corrections you can use).

There is also the problem that these devices are validated for their correlation with a gold standard over either a normal range or a 'clinically relevant' range.  Unfortunately, accuracy is really not relevant over most of that range.  What we, in the ICU, care about is accuracy specifically down at the transfusion threshold.  I don't really care if my pt's Hgb is 14 or 14.5 but I might care a great deal if it is 7 vs 7.5.  To my knowledge, there is no such data for the I-Stat (nor any other POC device that I know)

Dear colleagues, many thanks for useful suggestions and remarks. 

 

On ‎9‎/‎17‎/‎2018 at 8:59 AM, bimalc said:

There is also the problem that these devices are validated for their correlation with a gold standard over either a normal range or a 'clinically relevant' range.  Unfortunately, accuracy is really not relevant over most of that range.  What we, in the ICU, care about is accuracy specifically down at the transfusion threshold.  I don't really care if my pt's Hgb is 14 or 14.5 but I might care a great deal if it is 7 vs 7.5.  To my knowledge, there is no such data for the I-Stat (nor any other POC device that I know)

But what do you think about Hb provided in Acid-base analysis by Radiometer ABL? Anyone use it as guidelines for transfusion? Rarely we do this especially in VLBW infants with obvious bleeding.  Thanks

3 hours ago, Andrej Vitushka said:

But what do you think about Hb provided in Acid-base analysis by Radiometer ABL?

We have a ABL900 that is managed by our hospital lab department and they say that the analyses are validated. Meaning that we should trust the values (whether it is S-electrolytes, lactate or hematocrite) as if they had run the test in the "big lab machine".

So, I'd say that we generally trust our ABL
(as with all tech, it sometimes fails...)

9 minutes ago, Stefan Johansson said:

We have a ABL900 that is managed by our hospital lab department and they say that the analyses are validated. 

It is very useful information, Stefan. Thank you. I would ask our lab people about validation. 

On 9/13/2018 at 4:08 PM, Andrej Vitushka said:

Hamed, thank you! Is there any difference between venous and arterial Hb/Ht? 

Hi @Andrej Vitushka thank you for your question and discussion, it opened up a lot of thoughts. Sorry for seeing your question so late, @Stefan Johansson kindly answered. Thanks Stefan.