Lab Mistake - Sad Case

[exPCM]

---

Members don't see this ad.

http://www.newsday.com/news/local/longisland/ny-licanc1003,0,6442227.story
I am curious to find out what different institutions are doing out there to prevent these type of errors.

 

[yaah]

This kind of thing probably happens all the time, but it only has a large impact on clinical management some of the time. In this case, it seems likely that two different biopsies (which probably were similar grossly) were mixed up, and one turned out to be benign, one malignant. Now, if they were flipped and both were benign, no one would likely ever know. Or if they were flipped but the histories or specimen grosses were a bit different, it might have been caught.

We have had mislabeling issues that get caught at various stages - some by the accessioning people doublechecking their work, some by alert histology techs, some by us. I doubt you can ever 100% avoid errors like this, unfortunately. You minimize chances by having rigorous procedures that you carefully follow - in the case of this biopsy it seems the rigorous procedures were probably there but they weren't followed.

There are many types of carelessness that can lead to errors like this.

For some reason this always seems to happen (I remember a few other cases too) on breast biopsies! Why is that? The cynic in me feels that part of the problem is that breast biopsies are treated with more "reverence" than other types of biopsy for some reason. And paradoxically when you are "more" careful with certain procedures they can become more prone to error. I think everybody has experience with this. Oftentimes it is with a biopsy on someone who is a "VIP" where normal specimen accessioning, processing, etc, are bypassed to make sure that a specimen gets through quickly and accurately. And then because the normal procedures are bypassed, something gets screwed up.

But why is it always breast biopsies? In our institution whenever we have problem cases like this where mislabeling errors happen and are caught at some stage of the game, it is invariably a breast biopsy.

 

[Matte Kudesai]

!!

 

[Matte Kudesai]

This is why universal RFID systems need to be employed.
http://www.rfidjournal.com/article/articleprint/2955/-1/1/

At our institution we stagger specimen... A breast biopsy is never accessioned sequentially with another breast specimen. You do breast then GI then prostate then skin etc....

This is a pain in the arse for the accessioner but helps to prevent switching a sequential cancer case with a benign one. A mix up scenario is as follows: a third shift tech prelabels slides to save time. He/She cuts one paraffin block and picks it up the section on the next patients (prelabeled) slide. The slide is stained and ends up in front of a busy pathologist who does not have time to scrutinize every single slide at the level of proper labeling. An absolute disaster ... a correct diagnosis is made for the wrong patient.


I think there should be universal RFID that follows the specimen throughout its travels... This includes integration of bar coding with everything from the clinicians office through processing, cutting, staining, and slides. A bar code and RFID system that will not let you even cut the paraffin block unless the numbers line up (with an activating scanner for the microtome). I also think this should all be linked to an intelligent history database that processes bits of pertinent history (preset by infromatics) about the patient so a barcode/RFID scan of the slide will bring up history, RADS, chemo, micro blah blah blah in an orderly way for the pathologist to read. Pathology processing, storing, and inventory of tissue in much more similar to a WALMART operations model than anything else in medicine.

 

[djmd]

Having worked someplace where all lab sambles were bar coded, I will tell you that this does not prevent those kind of errors...
It may reduce them, but it also allows for a new error, wrong bar code.

Plus when the system goes down it becomes really hard to work with...

 

[Matte Kudesai]

!!

Having worked someplace where all lab sambles were bar coded, I will tell you that this does not prevent those kind of errors...
It may reduce them, but it also allows for a new error, wrong bar code.

Plus when the system goes down it becomes really hard to work with...

We bar code all the clinical specimen and it works great.
I'm talking about a comprehensive RFID system for biopsies. This is a very reasonable proposition and it will eventually happen.

 

[djmd]

We bar code all the clinical specimen and it works great.
I'm talking about a comprehensive RFID system for biopsies. This is a very reasonable proposition and it will eventually happen.

Ask your clinical lab people if they have ever had the bar code system crash.

And what keeps the phlebotomist from drawing the wrong patient into the wrong bar code labeled tube?

I'm not saying a system like that couldn't prevent errors, but any blood banker or clinical lab pathologist will tell you that complexity does not protect, computerization will not protect. Human error can and will find a way.

 

[Matte Kudesai]

Ask your clinical lab people if they have ever had the bar code system crash.

And what keeps the phlebotomist from drawing the wrong patient into the wrong bar code labeled tube?

I'm not saying a system like that couldn't prevent errors, but any blood banker or clinical lab pathologist will tell you that complexity does not protect, computerization will not protect. Human error can and will find a way.

I'm with you in a Crichton sort of way....

Actually laboratory automation which includes bar coding has led to decreased errors and increased efficiency. I agree that systems can go down but with the volume places are dealing with there does not seem to be better option.
http://www.beckmancoulter.com/resou...DiagLit/ClinLitList.asp?ProductCategoryID=CLA

I believe in elegant simplicity coupled to redundancy. There is a way to do this without building a rock that you cannot move.

Pathology specimens (biopsies) can and should be followed from the place they were taken and traceable through the system. Frankly I am tired of crappy hand written labels and multiple processing points where error can be introduced.

Pathology collection, processing, storage , archiving, is not rocket science and can be technologically the most superior system in medicine. We should be leading the field in preventing laboratory errors by standardizing and implementing technology already being used by multiple manufacturing industries.

 

[yaah]

It sounds good in theory, but it still doesn't necessarily help as much as you think. It could get grossed and put in the wrong cassette (which bar codes are less likely to help you at). There are a number of steps in the process which can be fouled up. It could also get put on the wrong slide. People could misread the numbers, etc etc etc. And there will still always be the problem of floaters, carryovers, etc. Every potential error can be lessened by specific steps and doublechecks and other things, but of course the more steps and procedures you add, the more potential for each step to fail because they become redundant, tedious, etc.

It is not as simple as bar codes and automation. Sometimes more automation leads to different kinds of preventable errors because of the "infallibility" of bar codes. We stagger specimens too, most places probably do. If you start trusting computers too much the next thing you know then the computer is telling you what to do against your better wishes, and you listen, and before long the computers launch a global nuclear holocaust and we have to fight back using slide rules and pencil/paper.

A pathology lab can get it right 99.9% of the time and every 1000th specimen will get screwed up (in our lab that is about 60 specimens a year!). That is unacceptable. But it is hard to prevent.

 

[Matte Kudesai]

It sounds good in theory, but it still doesn't necessarily help as much as you think. It could get grossed and put in the wrong cassette (which bar codes are less likely to help you at). There are a number of steps in the process which can be fouled up. It could also get put on the wrong slide. People could misread the numbers, etc etc etc. And there will still always be the problem of floaters, carryovers, etc. Every potential error can be lessened by specific steps and double checks and other things, but of course the more steps and procedures you add, the more potential for each step to fail because they become redundant, tedious, etc.

It is not as simple as bar codes and automation. Sometimes more automation leads to different kinds of preventable errors because of the "infallibility" of bar codes. We stagger specimens too, most places probably do. If you start trusting computers too much the next thing you know then the computer is telling you what to do against your better wishes, and you listen, and before long the computers launch a global nuclear holocaust and we have to fight back using slide rules and pencil/paper.

A pathology lab can get it right 99.9% of the time and every 1000th specimen will get screwed up (in our lab that is about 60 specimens a year!). That is unacceptable. But it is hard to prevent.

Yeah you are correct. I am thinking more RFID than barcode. Its a "smart bar code" that more or less lets you wirelessly track things within the system. If used correctly when the biopsy is made it is placed in(or is part of) a container that is then tracked. This will automatically be linked to block labelers, slide labelers, staining trays, processing trays, grossing stations, yabba dabba dabba as an ID code. The same way parts for cars, computers, etc are tracked. Pharma is beginning to implement this as well. It is unlikely that "the Terminator" or "the Matrix" scenarios will happen simply by using computers appropriately.

I mean.. crap... people were freaked out about ordering immunostains through the computer and wanted to to deliver a hand written accession only a few years ago. Pathologists can be serious Luddites with an attitude.

At every step of the way we will kick and scream until other specialties take over and rape our profits... this will happen with digital slides, molecular, undsoweiter.

Das ist genug.......

 

[djmd]

Yeah you are correct. I am thinking more RFID than barcode. Its a "smart bar code" that more or less lets you wirelessly track things within the system. If used correctly ...

That is the issue. If used correctly....
Every process has a start, and it is not like the OR nurses are incapable of putting the wrong RFID chip with the wrong specimen.

The system is only as good as its weakest link.

Or more accurately, if a airplane engine has a failure rate of 0.1% then what is the failure rate in a single engine aircraft?
now what is the engine failure rate in a multi-engine aircraft...


Yes, the RFID can be used as additional error check, but even error check systems can produce errors...

 

[LADoc00]

This is why universal RFID systems need to be employed.
http://www.rfidjournal.com/article/articleprint/2955/-1/1/

At our institution we stagger specimen... A breast biopsy is never accessioned sequentially with another breast specimen. You do breast then GI then prostate then skin etc....

This is a pain in the arse for the accessioner but helps to prevent switching a sequential cancer case with a benign one. A mix up scenario is as follows: a third shift tech prelabels slides to save time. He/She cuts one paraffin block and picks it up the section on the next patients (prelabeled) slide. The slide is stained and ends up in front of a busy pathologist who does not have time to scrutinize every single slide at the level of proper labeling. An absolute disaster ... a correct diagnosis is made for the wrong patient.


I think there should be universal RFID that follows the specimen throughout its travels... This includes integration of bar coding with everything from the clinicians office through processing, cutting, staining, and slides. A bar code and RFID system that will not let you even cut the paraffin block unless the numbers line up (with an activating scanner for the microtome). I also think this should all be linked to an intelligent history database that processes bits of pertinent history (preset by infromatics) about the patient so a barcode/RFID scan of the slide will bring up history, RADS, chemo, micro blah blah blah in an orderly way for the pathologist to read. Pathology processing, storing, and inventory of tissue in much more similar to a WALMART operations model than anything else in medicine.

Agree with above. Everyone should familarize themselves with this.

 

[dermpathdoc]

This is why universal RFID systems need to be employed.
http://www.rfidjournal.com/article/articleprint/2955/-1/1/

At our institution we stagger specimen... A breast biopsy is never accessioned sequentially with another breast specimen. You do breast then GI then prostate then skin etc....

This is a pain in the arse for the accessioner but helps to prevent switching a sequential cancer case with a benign one. A mix up scenario is as follows: a third shift tech prelabels slides to save time. He/She cuts one paraffin block and picks it up the section on the next patients (prelabeled) slide. The slide is stained and ends up in front of a busy pathologist who does not have time to scrutinize every single slide at the level of proper labeling. An absolute disaster ... a correct diagnosis is made for the wrong patient.
I think there should be universal RFID that follows the specimen throughout its travels... This includes integration of bar coding with everything from the clinicians office through processing, cutting, staining, and slides. A bar code and RFID system that will not let you even cut the paraffin block unless the numbers line up (with an activating scanner for the microtome). I also think this should all be linked to an intelligent history database that processes bits of pertinent history (preset by infromatics) about the patient so a barcode/RFID scan of the slide will bring up history, RADS, chemo, micro blah blah blah in an orderly way for the pathologist to read. Pathology processing, storing, and inventory of tissue in much more similar to a WALMART operations model than anything else in medicine.

I agree with the staggering option but what if you are a lab with focus on one type of specimen???i.e you do only prostates??? Or do only dermatopathology??? staggering is useless in such a practice....
We have had instances where a physician mislabeled two gastric biopsies with the name of one patient and insisted the second biopsy belonged to the second patient... do you accept the second biopsy on that insistance??? I frankly told the surgeon we had one option only and that was to send both specimens out for dna typing to prove which specimen was what....and that we would inform both patients about the mixup up....he went ballistic of course. My response was...that responsiblity for proper labeling fell on the surgeon's shoulder. We cannot sit there in the OR suite and make sure his staff labels the bottles correctly and that was HIS job. Once the specimens arrive in our lab...and they are erroneously labeled we can reject at will and if he did not want to pay for the DNA testing to properly identify the specimen.. then we would throw Both specimens in the trash and submit a report to each patient explaining why.The main issue was both specimens had the same date and time of collection as well as as same patient name and were recieved in the same bucket. both were antrum biopsies for gastritis.
That is not to say we have not made errors in my lab....certainly many years ago two core biospies for prostates got mixed up..one with cancer and one without..that was a real nightmare and was resolved using DNA testing and two very understanding Urologists....since it was OUR error we absorbed the cost of the DNA testing.

 

[yaah]

There are all kinds of errors that can happen that even the best systems won't always help. We had a prostate core problem in which the PA discovered a single core (or core fragment) at the grossing station at the end of the day, after which she had grossed 15 sets of biopsies. So we (I) had to go through every histo slide the next day and correlate it to gross and try to figure out if we could match it up. As you might expect though, every slide had a core on it (none were missing). Some blocks had multiple core fragments in them, but anyone who does pathology knows that you can gross one core and end up with three separate pieces of the core on the slide the next day. We didn't have any blocks where we were missing a core.

The point that dermpathdoc makes about surgeons is interesting, it happens here too, although more often with mislabeled bloods and CSFs.

Most clinicians will have the same response - "just run it anyway." But the problem is it isn't truly their problem (legally) if there is a mixup, it is the lab's problem if we run it and report the result. It's very easy for a clinician to say which specimen is which, or whatever, because it isn't their problem!

 

[Matte Kudesai]

I agree with the staggering option but what if you are a lab with focus on one type of specimen???i.e you do only prostates??? Or do only dermatopathology??? staggering is useless in such a practice....

Question: In practices that do only skins, or only GU, why not provide the groups with your own containers and instruct them on labeling properly. Does that work?

Can you reject poorly legible and improperly labeled specimen? I realize that when you are running a business you have to be flexible and allow some latitude to your customers. However, I am frankly sick and tired of the crap the surgeons or clinicians scrawl on a container. It demonstrates that they have no respect for our profession. You get crappy histories and often have to decipher a coded message to figure out why the bx was taken. That is why there should be some sort of tracking, coding, marking system that circumvents at least this part of the reason why we get sued.

If pathologists demand more respect from the clinicians in terms of giving us something we can read both tissue wise and label wise is that wrong?

We are frequently treated like glorified technicians. I truly believe the reason for this is that many of pathologists do not have the cojones to stand up to pushy pricks that try and force our hand after giving us garbage to read. As my favorite attending says "garbage in garbage out."

 

[dermpathdoc]

Question: In practices that do only skins, or only GU, why not provide the groups with your own containers and instruct them on labeling properly. Does that work?

It works only to a point...as you mentioned there are those docs that label specimens with illegible writing if they label them at all. Or those who simply put an A or B on the bottle...then we find the tissue looks like B is A or vice versa...(recent case..A)biopsy of sole of foot and one of B)chin...chin was in A and Foot in B....I make a note stating specimens appear to have been transposed in the bottles and to correlate with clinical....

There is never an easy way to do this....and the docs often accuse US of messing up thier specimens.....My staff is trained to immediatley inform me of any label discrepancies.... grossing is stopped until Doctor's office is informed...but there is always somehting that slips thru....we are only human after all...and I can tell you from personal experience bar coding does not help all that much.... they find a way to put the wrong bar code on the wrong bottle

 

[kronickm]

Probably a dumb question but, they would do a double mastectomy based only upon a single biopsy sample? No imaging, nothing that would cause someone to raise an eyebrow?

I must say though there was a case at my hospital where someone who had already rejected a single kidney transplant, recruited another donor. Apparently the results of the cross match (which failed) were not made available to the transplant team. The kidney was removed from the donor and transplanted.

 

[yaah]

Well, obviously this woman who had the mistaken mastectomies had a biopsy. So there was something suspicious enough to warrant a biopsy. But oftentimes patients do proceed directly to mastectomy after a needle biopsy shows cancer if they fit certain criteria (family history, etc) and they elect for that choice.