Specimen mislabeling

[jalomahe]

We had an incident occur as follows:

Two specimens were collected by nursing and properly labeled by nursing using labels provided by registration (contains all pertinent pt identification information and then some).

Both specimens come to the transfusion service. Orders are located in the lab computer system, received and labels printed.

While attaching the lab labels to the specimen the tech is interrupted (multiple times) and in the course fo labeling switches labels with tubes so that label A is attached to specimen B and vice-a-versa. Even though the tech who labeled the tube left the name and id number from the original label exposed the error was not caught by the other tech performing testing that shift.

Consequently one of the pt's received an ABO incompatible unit of blood (it was caught and stopped after 100cc infused). Pt survived. FDA notified etc.

We are now at the point of coming up with a way to keep this from happening again. We have a policy in place that states that at the time a new label is being attached that both the patient name and id number on BOTH labels MUST be MATCHED, then the label may be attached to the tube. However, our supervisor feels that the FDA will want us to come up with something new to prevent this from happening. By the way this is the first time I can remember this happening in our labs and I've been here for over 20 years.

Does anyone have any other policy, procedure or method they use when putting a new label on an already labeled specimen?

Sorry this is so long and thanks in advance for any ideas, help.

[Nancy L.]

Wow, I'm sorry that you have this nightmare to deal with. Scary. It certainly sounds to me like your policy should be sufficient. Unfortunately, no process is 100% human error proof. A policy/procedure must be followed for it to be effective. Would it be possible to institute a self audit tracking the relabeling process. This would provide an opportunity to 1)reiterate the importance of proper relabeling and label checking throughout the testing process and 2)provide documentation of retraining and compliance on the part of the techs. This may be enough to satisfy the FDA that you are taking steps to make sure that your policy is and will be followed. It's unfortunate that neither of the techs paid enough attention to catch the error. While it is totally unacceptable and must be avoided at all costs, it is also easy to understand being distracted. We work in an environment where we are consistently needing to deal simultaneously with multiple complicated issues coming at us from all directions. I'm sure the techs involved are as distraught as anyone. Good luck!

Nancy

[Lcsmrz]

We don't relabel the primary specimens or pour-off the plasma, but I hear it's a necessity for some automated equipment. In the General Lab, we leave the original name showing if relabeling is necssary, so each tech along the way can double-check it themselves -- and everyone double-checks labels routinely! While this is a good system, it's not perfect, since the wrong person can still be drawn, so we've been concentrating on the bedside ID and labeling process.

Don't assume that once in 20 years is a good record -- you could have just been lucky or you have a low volume! Look closely at your system for weaknesses and see if they can be reduced or eliminated. JCAHO / equiv will require a Root Cause Analysis be performed, and there are several good techniques in the quality world for doing that.

[John C. Staley]

I'm curious, I don't understand why you are applying the second label in the lab? Is it so a bar code is available for your testing analyzer?

I have always found it difficult to consider great changes based on a single, one time event. The FDA will likely accept a retraining session if you can show that this is not a pattern but a one time event. I would not rush into making significant changes as my first response to this occurrence.

[Likewine99]

As a BB supervisor I have first hand experience with the FDA related to a clerical error that occurred in our BB. We had a very good Root Cause Analysis team and we made sure to include techs who work in the lab but not in BB, sort of a fresh pair of eyes to really examine the old, defective process.

Your supervisor is probably right, youi will need some way to capture who applied the label to the tube, who resulted the test and you will need lots of documentation on the process changes and staff re-education.

We instituted a post analytical clerical check using printouts from the analyzer and screen prints from the computer. When done at the manual bench, with gel cards, the gel cards are saved for the clerical review. We do the clerical check on the shift and immediately if the pt is to receive blood. The non-blood bankers were trained in the clerical process for those nights and w/e when you may be the only blood banker in the house.

I'm sorry this had to happen to your pt, your lab and your techs. We found the RCA process to be extremely helpful and almost 4 months after our incident we really do feel more comfortable knowing someone is checking our results and it really has made it safer for our pts.

[jalomahe]

We handle approximately 3000 specimens per month with the specimens coming from off site locations as well as in house locations. Depending on where the specimen was collected and by whom the labeling varies. The second label is attached after the original label is verified by a blood bank tech. The second label is necessary in order to provide a barcode that the analyzer can read. The barcode can also be used with a handheld scanner at the manual benches when accessing the patient in the LIS.

A root cause analysis was done and the error was deemed to be human error secondary to multiple distractions. So how do we create a 'fix' for that? I don't see that we can. That is why I'm trying to convince those over me that the

FDA would find it acceptable if we did re-education and monitering.

[jhaig]

I would think re-education and training should satisfy the FDA since this error was a one-time problem and not a recurring issue. You may also evaluate how the tech that did the work had to deal with so many interruptions to begin with. Not being familiar with your department's setup, I'm not sure of where the distractions came from. Is there a way to screen phone calls so that techs doing actual benchwork do not have to answer the phone? Does the bench tech also perform multiple duties, like issuing blood etc. while performing bench work? I know that in these times that less people are stuck doing more work, but maybe focusing on limiting outside distractions could be beneficial.

[Bill]

Have you considered creating a "test" in the computer titled 'Label checked' with a yes or no answer. Put this test with all other tests for the patient. This creates a reminder to double check label at the time of resulting tests in the computer. Good luck

[Cathy]

I think I would focus on generating the '2nd' label earlier, like at the time of the draw and avoid the re-labeling all together. We are not automated so I'm not sure if this is even a possibility.

[Dr. Pepper]

If the application of the second label is to get a LIS-type label with bar code, specimen # etc onto the tube, is there a way to generate these on the floors so they may be used instead of the registration label? Any relabeling or aliqotting generates another step in the process where mislabeling errors can occur, and you could avoid this step.

I like the Bill's idea of adding a "Label checked?" test to your type/screen battery. We have a "Blood type verified" test in ours to document whether a second typing is on record for the patient, and don't give out non-group O RBC unless the typing has been verified. This might have prevented your ABO-incompatible unit from being issued. Our preop type and screen battery also has informational tests for pregnant/transfused in the past 3 months and date of surgery, which we find very useful.

[LAS]

After encountering various labeling issues, making various process changes, delivering various educational and training sessions, we ultimately developed a progressvie disciplinary policy for what we consider "patient ID errors". This covers mislabels, reporting one patient's manual diff for another, or any time anything belonging to one patient was matched to another patient. A single incident in a rolling twelve month period warrants a documented verbal warning, a second incident in a rolling twelve month period warrants a written warning, a third incident gets the staff member probation, and a 4th incident results in termination. We take ID error/mix-up issues very seriously and although the excuse was always how busy we were, since implementation of this policy we are no less busy but the error rate has gone to almost zero. Our staff members have been extremely supportive of this policy.

No one likes to go there, but once policies and processes have been examined and refined, personal accoutablity must be assumed.

[Malcolm Needs ☆]

One of the ways around this is the use of Radio Frequency Identification (RFID). This is by no means inexpensive, may not be idiot proof, but is as near foolproof as I have seen. The sample label is produced at the point of phlebotomy with a hand held device by scanning a bar code on the patient's wrist band. The label is produced with eye readable patient identification and a copy of the patient's bar code. This barcode can be read by the automation in the Blood Bank and there is, therefore, no requirement to relabel.

Pre-printed "addressograph" labels, as I think are described by JALOMAHE, are more or less "banned" in the UK by the British Council for Standards in Haematology (BCSH) Guidelines, whilst RFID labels produced at the point of phlebotomy are positively encouraged - but they are expensive. I am certain, as has happened before, they will become cheaper in a very few years. After all, I can remember when abacusses were expensive, let alone, calculators and computers!

One thing that does worry me is that you say, JALOMAHE, that this is the first such occasion in 20 years. This may be the first time that such a situation has resulted in clinical sequalae, but do you know how many times it has happened when, for example, two group O, D+ patients' samples have been swapped, resulting in a "safe" (and I use the term in its loosest sense) transfusion?

[Dr. Pepper]

I concur with Malcom regarding the incidence of misidentifications- we only see the tip of an unknown iceberg. We can detect some in the lab by historical ABO discrepancies or delta failures in heme and chemistry, and sometimes nursing may realize they've got results on a patient who had no tests drawn, but the rest, the ones that don't not make sense (excuse the double negative) remain undetected and unquantifiable.

[GilTphoto]

This is where the use of a secondary BB armband and label comes in handy. We use Typenex armbands. If using this secondary number that's checked before transfusion, the armband number wouldn't match, so the transfusion wouldn't be started.. The armband number from spec A would be put on the compatibility slip for patient B. When a nurse comes to pick up blood, they must write the armband number on a copy of the signed consent with a copy of the order to transfuse. We compare the number before issuing.

[skopti]

Our computer system generates barcode labels with accession numbers at the time the order is placed, these labels print at the nursing unit and they are used to label the tubes at the time the specimen is drawn.

[TeachBB]

In our lab, the tech ordering the test passes the LIS label with the specimen and the requisition to the testing techs. The testing techs verify tube, requisition and LIS label and then the label is placed on the specimen. The reason we place this label on the tube is to associate the Name and MRN (already on the tube and requisition)with the Accession # that was generated with the ordering of the test.

[LABKING]

I would think re-education and training should satisfy the FDA since this error was a one-time problem and not a recurring issue. You may also evaluate how the tech that did the work had to deal with so many interruptions to begin with. Not being familiar with your department's setup, I'm not sure of where the distractions came from. Is there a way to screen phone calls so that techs doing actual benchwork do not have to answer the phone? Does the bench tech also perform multiple duties, like issuing blood etc. while performing bench work? I know that in these times that less people are stuck doing more work, but maybe focusing on limiting outside distractions could be beneficial.

A collection record paper for recording double checks at the bedside and double checks of the relableing in the lab should exceed the FDA expectations. Have a place for initials for each specific check, name, number, DOB, etc . And it would be good to train the checkers to initial after each check, like check the name, initial, check the number, initial, check the DOB, initial. You really will "get-it" once you try it.

We fear that distractions will only increase as more automation and computer upgrades evolve.

Onetwomanytesttubes:eek:

[Malcolm Needs ☆]

In addition to jhaig, certainly in the UK, and more so in smaller establishments, there is only one person covering the Blood Bank at night (and sometimes Haematology too - and in the really small establishments, all of Pathology). I don't know if this is true in the US as well. In such situations, interuptions are frequent, usually by bleep, and you are obliged to answer the bleep as soon as possible, just in case the situation is more urgent and life-threatening than the one you are already working on..............but I do agree with what you are saying. Interuptions can completely break a person's concentration, and this is when errors can occur.

[eeagan]

This incident also points out the need for a second specimen to verify the ABO/Rh. At our hospital, we do not issue any blood product until there are 2 ABO/Rh types in our computer system and the types agree. Granted, this does not detect patient identification errors at any level if both patients have the same ABO/Rh. It also doesn't avoid patient identification errors at the phlebotomy level if both tubes were drawn at the same time but submitted to the blood bank at different times. But it has detected labeling errors that happened at the phlebotomy stage and in the blood bank.

[Malcolm Needs ☆]

Hi eeagan,

I can understand this for routine samples, but what happens if you have real emergency? Presumably you issue group O RhD Negative cellular products and, possibly, group AB plasma products?

I'm thinking of an exsanguinating patient in Accident and Emergency.

[eeagan]

Yes, when blood products are urgently needed before both specimens can be processed, we give O negative RBC and AB FFP. It's not often that we have to do this. But we do not hesitate to offer the emergency release (uncrossmatched) O negative RBC.

[Trek Tech]

This is where the use of a secondary BB armband and label comes in handy. We use Typenex armbands. If using this secondary number that's checked before transfusion, the armband number wouldn't match, so the transfusion wouldn't be started.. The armband number from spec A would be put on the compatibility slip for patient B. When a nurse comes to pick up blood, they must write the armband number on a copy of the signed consent with a copy of the order to transfuse. We compare the number before issuing.

This is how we address the issue as well. If the patient has an armband on with a unique number and that number is on the unit as well as the tube this problem would probably not have happened. I know many people hate the second armband idea but I have seen (in the past 23 years) at least 3 incidents prevented by the second armband. Two were nursing related and one was lab related.

Edited May 21, 2009 by Trek Tech
word left out

[Malcolm Needs ☆]

This seems an excellent method of identification - and a good deal less expensive than my suggestion of RFID!

[kmh76]

In our Blood bank our phlebotomy team uses computer generated labels for each test. For Blood bank The phlebotomist labels each tube (we collect 2 for Blood bank) with the label for Type/Screen or Type/Crossmatch so when the specimen comes into blood bank we do not need to relabel anything. We have an automated instrument and the instrument reads the barcode with patient name and accession number. We also use the hollister blood bracelet system so the phlebotomist will put an R band number on the tube. This R band number must match the set of R band numbers they return to us. There are times when phlebotomy will stick the wrong patient and if we have a previous type on file we will catch it. If not then it becomes a big issue. I think it would be very confusing for us to have to relabel the tube when it arrives in Blood bank.Also if we do not have a previous type on the patient we do a second type check on another tube. Hopefully we will be able to avoid errors made by phlebotomy.