I am interested in how other facilities handle specimen labeleing errors. Do you allow corections to mislabeled or incompletely labeled specimens?

If you do how do you document the errors?

I have alway been very strict on labeling errors rejecting anything recollectable submitted with incorrect or missing information.

However, several nurses, including the CNO, have stated that other facilities have allowed them to relabel specimens if they "certify" the corrected information. Thoughts anyone?

Several years ago, I heard Kathleen Sazama (both a physician and a lawyer) speak about errors. I was particularly impressed by the assertion that an incorrectly labeled specimen was 40% more likely to have "wrong blood in tube". I think that a labeling error should always lead to a recollect.

Agree 100% Marilyn - and if anyone's opinion should be heard on this subject, it is Kathleen's.

I am with you on this MJ. If the specimen can be recollected, we have them do so. If it cannot, we are EXTREMELY careful to assure the specimen is relabeled correctly.

Without any hesitation we DISCARD the specimen. we ask for another collect. If it is not possible we check if there is a sample drawn at a different time in an other secction (and properly labeled). Never allow them to take back the sample and never allow them to re-label it.

We do not allow relabeling of tubes. It causes problems because the rest of the lab uses an affidavit form to document when they have allowed relabeling of their tubes.

We do not allow them to correct any patient identifiers (name, DOB, medical record #, or BBID #). We do allow them to correct date, time, or their initials.

I agree that they can write write their initials (and be accountable) BUT not touch the patient's triple name and MR#. the "Date" I wouldn't allow, one chap (oh sorry 11 yo doctor in a white coat) had a sample in his pocket for a few days, when asked why he said I was waiting for her to require blood. What!!!!!!!!!!!!!!!!!!!!!!!

An AABB inspector recommended this long ago, that date, time and initials are not considered "identifiers" and as such can be allowed for correction. Any identifiers should NOT be corrected. My personal opinion is that any error on a Blood Bank tube should result in a redraw, but this was our "compromise" with nursing...I am not usually a compromiser, but I lost the battle on this one.

Several years ago, I heard Kathleen Sazama (both a physician and a lawyer) speak about errors. I was particularly impressed by the assertion that an incorrectly labeled specimen was 40% more likely to have "wrong blood in tube". I think that a labeling error should always lead to a recollect.

This is a quote I trot out every time I'm having a discussion with our lab manager about accepting outpatient blood bank samples that have the name spelled wrong. I don't always win, but I give it a good try.

We reject any discrepancies - no exceptions , no sob stories accepted - patient difficult to bleed etc etc

reject 5-6% samples received every month

Thank you all for your input, I will be showing this to my transfusion commitee. I still think any error in specimen labeling should be discarded. I will do my best, wish me luck.

No corrections to identifying information EVER! Corrections to date, time, and ID of the phlebotomist are OK. However, if anything is missing entirely, the specimen is discarded. The reasoning: the tube was obviously not labeled completely before leaving the bedside, end of story. And yes, at times some people get very upset about it.

Better somebody upset, than a patient injured or dead.

The contributing factor to the death of a patient with chronic renal failure and on dialysis was a specimen that was incorrectly labeled (not labelled immediately at "machine" side) and the Blood Requisition form incorrectly completed (No previous transfusions ticked off)

Conclusion: Zero tollerance. Do it right the first time.

In cases where a second sample could not be obtained, my instruction was to crossmatch and issue group O

I completely agree, we reject all specimens that have errors with any patient identifiers. I am attaching a great article that I trot out anytime I need to defend my position. It is a Transfusion article from 1997. [ATTACH]518[/ATTACH]

JohnsHopkinsmislable article.pdf

Our ER department sends a patient sample to the BB incase the MD adds on a T/S or a XM. The request comes after the sample has been brought to the BB. Therefore, the sample and the request are coming seperately, not at the same time. I read std 5.11.1 as they must come together. Can anyone clarify this or add to it?

AABB Std 5.11.1 states:Requests for blood, components, tissue, and derivatives and records accompanying blood samples from the patient shall contain sufficient information to uniquely identify the patient, including two independent identifiers. The transfusion service shall accept only complete, accurate, and legible requests.

They should arrive together. But then again, there are centers who recieve the request on the computer. I look forward to replies.

Gotta love semantics!!! I don't see any where that indicates that the sample must physically accompany the orders/requests just when they do they must match. I think a case could be made for the word accompanying meaning "in association with" in this particular case.

Gotta love semantics!!! I don't see any where that indicates that the sample must physically accompany the orders/requests just when they do they must match. I think a case could be made for the word accompanying meaning "in association with" in this particular case.

I emailed the dilema to AABB. I will post their reply when it arrives.