Mislabeled Hematology Specimens

[dmpollock]

We occasionally are asked to do ABO/Rh testing on samples from outside the blood bank because the test results don't match what is expected.

Does anyone have a policy or opinions on this?

Is it a violation of the CLIA requirement for laboratory tests to be performed only at the request of an "authorized person?"

[L106]

Do you mean doing ABO/Rh testing because the patient's current ABO/Rh doesn't match previous records, or doing ABO/Rh testing on a sample because other lab test results do not match recent results (ie: hemoglobin or other specific test)???

I am not opposed to testing the ABO/Rh of a sample collected for another lab department as part of an internal investigation. However, if it involved an ABO/Rh dicrepancy from previous testing or records, it is our policy that a new sample should be collected and testing repeated, and the BB supervisor should be consulted for further testing and evaluation of the findings.

I interpret the CLIA requirement to mean that only an "authorized person" (ie: physician, physician's assistant, etc.) can order a test to be performedand reported, but I don't think that prohibits us from repeating a test or performing some additional tests as part of follow-up to investigate a problem. Also, the medical staff can approve automatic "reflex testing" to be done without an additional order from the physician. (ie: If the Antibody Screening Test is Positive, it is appropriate that the Antibody Identification be performed.)

[dmpollock]

These are cases where a hematology result is discrepant from what the expect: the tech thinks that the phlebotomist drew the wrong patient.

[Malcolm Needs ☆]

I interpret the CLIA requirement to mean that only an "authorized person" (ie: physician, physician's assistant, etc.) can order a test to be performed and reported, but I don't think that prohibits us from repeating a test or performing some additional tests as part of follow-up to investigate a problem. Also, the medical staff can approve automatic "reflex testing" to be done without an additional order from the physician. (ie: If the Antibody Screening Test is Positive, it is appropriate that the Antibody Identification be performed.)

We do all sorts of tests on samples to identify the specificity of the antibody(ies) with no reference to a Consultant. This includes instigating such things as a DL test or tests for mixed cold and warm AIHA. However, if our findings suggest an unusual combination of antibodies, a combination that would give us problems supplying blood (say an anti-U+Jka), an antibody against a high incidence antigen, we would take our results to the Consultant for him or her to report to the hospital.

The same applies if we find a DL positive or a mixed cold and warm AIHA, as these require medical input for correct treatment.

We would also consult with a Consultant if we think that other tests, performed by other Departments, such as HLA or platelet antibodies in the case of multiple stillbirths, could be the answer, but again, the need medical input.

In this way, we are very lucky, because we are "given our head", but have excellent Consultants to back us up (and to suggest other tests that we may have overlooked).

:):)

[Malcolm Needs ☆]

These are cases where a hematology result is discrepant from what the expect: the tech thinks that the phlebotomist drew the wrong patient.

In such a case, we would always order a new sample. If this showed that a mistake had been made, we would immediately tell our Consultant, who would tell the patient's Consultant, and would contact our Quality Department to initiate a Root Cause Analysis.

It may be no fault of the phlebotomy staff (doctor, nurse, phlebotomist); it may be someone who is using somebody else's identity (in which case, the police may become involved).

[dmpollock]

There are a lot of potential problems with the practice. In the mean time, I have made a first draft of a request form (attached).

If the hospital fires a phlebotomist based on these results, the blood bank tech could conceivably have to testify in a wrongful termination lawsuit. I have heard of a recent case where an allegedly terminated employee filed a case appealing denial of workman's comp.

The results could also be involved in a malpractice suit if patient care was affected by wrong lab results.

There might even be an issue of informed consent if the ABO/Rh was not necessary for diagnosis or treatment of the patient's condition.

I also would be concerned to have a mechanism to prevent these results from accidentally becoming part of the Blood Bank record, or the specimen accidentally being used for a crossmatch later.

BBT_Request For Blood Typing Of Suspected Misidentified Sample.pdf

[lateonenite]

"If the ABO/Rh types match, this does not mean that the samples were collected from the person indicated on the label."

So we are bothering because? Good Lord.

We request a recollection; doubt as to the specimen viability (e.g. contaminated) or identity is an automatic "No result" (i.e. we cancel the test or put in a code that answers the test with mislabelled, etc). We might, and rarely, but might leave the results in after speaking to the person responsible for the collection and attach some kind of comment - but really, not.

In Australia, an authorised person is only ever a doctor; some of our nurse practitioners (equivalent a physician's assistant, I think) may be able to order a selected panel, but it must be co-signed by a medico within 14 days of the request. It has to do with the ability to bill the government for Medicare (not the same as the US version). Jump in here please, Malcolm, if I have this wrong, but I think our Medicare is similar to the NHS in a number of respects.

[AMcCord]

We might, under some circumstances, do it for an internal investigation. It is just a 'quick and dirty' type. It never goes on the patient record, it never goes on Blood Bank's records, it is never a redraw specifically for a type. It would be done on a specimen or specimens in hand whose results are discrepant, usually wildly discrepant. It is never ever used for firing someone or as 'proof' for disciplinary action. It is simply used to solve a mystery. (And it usually doesn't help much since most folks are going to be O Pos or A Pos anyway. My lab manager is a chemist so take that into consideration.)

[clmergen]

We do the same as AMcCord.

[Malcolm Needs ☆]

"If the ABO/Rh types match, this does not mean that the samples were collected from the person indicated on the label."

So we are bothering because? Good Lord.

We request a recollection; doubt as to the specimen viability (e.g. contaminated) or identity is an automatic "No result" (i.e. we cancel the test or put in a code that answers the test with mislabelled, etc). We might, and rarely, but might leave the results in after speaking to the person responsible for the collection and attach some kind of comment - but really, not.

In Australia, an authorised person is only ever a doctor; some of our nurse practitioners (equivalent a physician's assistant, I think) may be able to order a selected panel, but it must be co-signed by a medico within 14 days of the request. It has to do with the ability to bill the government for Medicare (not the same as the US version). Jump in here please, Malcolm, if I have this wrong, but I think our Medicare is similar to the NHS in a number of respects.

Your first line is a point well made. How on Earth are we to tell?

To be perfectly honest lateonenite, I am in absolute ignorance as to whether your Medicare is similar to our NHS in any respects! I literally do not have a clue!

One thing where we do differ though, is that my staff will rarely ask me, let alone a Consultant (and certainly, under no circumstances a nurse, be they a physician's assistant or not), if they can use a selected panel, as they have a large amount of autonomy and, if you will forgive the language, I'm damned if I'm going to ask anyone if I can use, not only a selected panel, but also any of our selection of rare frozen red cells or reare frozen antisera. Within reason, I have total autonomy about the tests we perform (although if one of our own Consultants "demands" it, we do it [unless I can talk him/her out of it]), and do not have to answer to anyone except my direct line manager (and, if I get it wrong, my Consultant and, ultimately, the Law Courts).

The individuals undertaking these investigations are all Fellows of the Institute of Biomedical Science specialising in Transfusion Medicine, and with a particular knowledge of Red Cell Immunohaematology, and so they are trusted. I doubt if I have to shout at them more than four times a day each!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

:D:D:D:D

[L106]

We might, under some circumstances, do it for an internal investigation. It is just a 'quick and dirty' type. It never goes on the patient record, it never goes on Blood Bank's records, it is never a redraw specifically for a type. It would be done on a specimen or specimens in hand whose results are discrepant, usually wildly discrepant. It is never ever used for firing someone or as 'proof' for disciplinary action. It is simply used to solve a mystery. (And it usually doesn't help much since most folks are going to be O Pos or A Pos anyway. My lab manager is a chemist so take that into consideration.)

Ann -

This is exactly the situation / application I was referring to. Nothing get reported, etc. Depending on the blood types involved, we could use the ABO/Rh results to prove that someone drew the specimen from the wrong patient or labeled the specimen with the wrong patient's name.

However, I could not say that it would "never ever be used for disciplinary action". First error, probably not. But how many serious mistakes should a phlebotomist be allowed? How could we supervisors and managers allow an individual to continue to perform their job when we know that the individual does not follow procedures/policies? It's an accident waiting to happen.

Donna

[dherbert]

I read the original posts especially "These are cases where a hematology result is discrepant from what the expect: the tech thinks that the phlebotomist drew the wrong patient." to mean the situation where the haem lab runs a Hb test (for example) and the value has changed enough for them to ask if the same patient was bled so they ask the Blood Bank to group the suspect sample to see if the group has changed and hence the wrong donor has been bled?

If that's the situation then, yes, my lab has been asked to do this on a number of occasions as an investigative tool that is informally ordered but definitely recorded on the investigation report. Our justification is that it is in the interest of the patient to help in this way.

Cheers

[billdun]

Your first line is a point well made. How on Earth are we to tell?

To be perfectly honest lateonenite, I am in absolute ignorance as to whether your Medicare is similar to our NHS in any respects! I literally do not have a clue!

One thing where we do differ though, is that my staff will rarely ask me, let alone a Consultant (and certainly, under no circumstances a nurse, be they a physician's assistant or not), if they can use a selected panel, as they have a large amount of autonomy and, if you will forgive the language, I'm damned if I'm going to ask anyone if I can use, not only a selected panel, but also any of our selection of rare frozen red cells or reare frozen antisera. Within reason, I have total autonomy about the tests we perform (although if one of our own Consultants "demands" it, we do it [unless I can talk him/her out of it]), and do not have to answer to anyone except my direct line manager (and, if I get it wrong, my Consultant and, ultimately, the Law Courts).

The individuals undertaking these investigations are all Fellows of the Institute of Biomedical Science specialising in Transfusion Medicine, and with a particular knowledge of Red Cell Immunohaematology, and so they are trusted. I doubt if I have to shout at them more than four times a day each!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

:D:D:D:D

Is this an example of two countries separated by a common language? In Oz I suspect panel means panel of tests e.g U&E+LFT etc.

I have occasionally gone as far as extended phenotype where ABO and Rh were the same - nailed one a while ago on discrepancy in S typing.

We always alert the requestor that the sample "appears to be from a different patient" and suggest a redraw. This is recorded in the LIMS. We always point out the error to the phlebotomist and report to clinical governance but as yet, this has never resulted in formal disciplinary action

[Malcolm Needs ☆]

Is this an example of two countries separated by a common language? In Oz I suspect panel means panel of tests e.g U&E+LFT etc.

Yes, I think you are correct!

By a panel, in the UK, we mean an antibody identification panel.

Confusing this English isn't it!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

:D:D:D:D

[dmpollock]

To clarify the issue. I am talking about specimens that have nothing to do with the blood bank. What is happening is that a hematology or other lab test value is dubious. The hematology tech wants to do a mini-"forensic" investigation to see if the phlebotomist mislabeled the tube or drew the wrong patient. They have been asked the blood bank to go so far as performing extensive (and expensive!) phenotypes on these samples. They have also asked the blood bank to scoop out red cells from the bottom of a heparinized sample with a gel barrier!

The form I attached to my earlier message was a draft. It was intended solely for testing non-blood bank tubes. It has the statement "If the ABO/Rh types match, this does not mean that the samples were collected from the person indicated on the label" because I am not sure the hematology techs always interpret the results correctly.

Our most recent case was there were two sets of samples with identical labels and accession numbers, but different collect times. In this case one tube was B Pos and the other A neg, so it was clear that a mislabeling occurred. But what would happen if both tubes are O Pos, which could easily occur by chance? Would they go ahead and report out the coag results? What if both tubes had been AB Neg? Would they understand the difference in probablity between the two situations?

What I also don't want to happen is that this testing becomes part of the normal blood bank records. I would not want somebody in the blood bank to see these results, not knowing they came from a bad draw, to select platelets, for example, based on the blood type from that sample. I also would not want a situation where the bad tube is left in the blood bank, then a crossmatch is ordered and a blood bank tech uses the bad tube, not knowing it was a suspect mislabeled specimen.

Also, I want good documentation of the incident because of the possibility of it being used as evidence in a legal issue.

The CLIA issue of who can order the blood typing comes from the fact that this blood type has not been requested by the patient's physician. There could also be an issue of informed consent. Generally the patient is consenting to procedures and tests necessary for their diagnosis and treatment. The coag testing in my example would have been appropriate under the terms of consent. The ABO/Rh was not necessary for diagnosis or treatment.

As far as routine blood bank work which is actually ordered by the patient's physician, there is not an issue for add-on testing appropriate to the order. If a crossmatch is ordered, it is expected that the patient's blood will be typed and have an antibody screen. It would also be expected for the antibody ID to be run when the screen is positive. A separate order would not be needed for the antibody ID. It would not be appropriate, however, to take the blood bank samples and start running glucoses, HIV testing, or whatever you want without an order.

Edited November 4, 2009 by dmpollock

[Malcolm Needs ☆]

Sorry, I obviously got hold of the wrong end of the stick.

In the situation of which you write, I would flatly refuse to perform any tests whatsoever, and would be inclinded to report the requestor to a higher authority.

:mad::mad::mad:

[LAS]

Agree with Donna.

[John C. Staley]

If some area of the lab suspects the results from a given sample then they should request a re-collection of the sample and provide the reasons why. It is foolish to side step the obvious solution to the question. If the sample happened to be collected above an IV then all the blood bank testing available will not resolve the problem of a questionable sample. If the physician or other person responsible refuses to allow the re-collection then you have an additonal issue but that's for another post.

:bonk:

[eric1980]

I am in haematology now, and if I ever encounter a specimen which shows very different parameters in FBC (eg Hb and MCV without transfusion/bleed) as compared with that of hours ago, we will check the patient label on the specimen against the lab-printed barcode label. If it matches, then we will simply add in a comment to warn the physician to interpret this result with caution.