how do you categorize quality events?

[Mabel Adams]

If you have a series of mistakes do you record each mistake in the process separately or do you treat them all as one event. Let's say someone omits entering a BB armband # in the BBIS. Someone else crossmatches the units for a remote site. Someone at the remote site tags the units. Then someone else issues them. No one catches the missing band number. Is this tracked as one error or 4? My error tracking system includes an estimate of how likely the problem is to recur and I find that I would give different estimates for different mistakes in the event. I find the likelihood measure useful in assessing risk but I don't really want to look at only how often all of the mistakes would line up together again. I use a homemade system so I can change it all I want.

[Bill]

I would consider this as one event, unless the remote site is NOT part of your organization (system).

[Eagle Eye]

same this should be one event....and look at each step to see if there was a system problem...if no counsel each staff involved...

[David Saikin]

I think each occurrence is an event - linked to the primary non-conformance of no BBID#. We had a saying at one place where I worked - "If you have a series of multiple checks you better make sure the 1st one works." Because everyone expects it to, the checks further down the process become lackadaisical. I had a tech switch BBID#s on 2 patients. The only place it could be caught was at the transfusion . . . 3 of 4 units were transfused to the incorrect ID#. It was caught by a night nurse during the armband check. Fortunately both pts were O+. The comment from one of the nurses who mis-transfused: "Those numbers always match, we don't check them!"

[Dr. Pepper]

Hi Mable - As with David, in my system we would treat them separately. A while ago we had one tech XM 2 units but switch the bag tags. Another tech who issued it didn't notice the unit # discrepancy. Neither did 2 nurses checking the unit before hanging. Neither did a third tech releasing the remaining unit into inventory after the XM expired. Finally a fourth tech went to XM the remaining unit and saw that according to the computer it had been transfused 3 days earlier.

[Mary**]

Yikes!!!!!!!!!!!!!

[Malcolm Needs ☆]

I think I fully concur with Mary**'s comment!!!!!!!!!!!!

[Dr. Pepper]

That's not quite what I said.

[Deny Morlino]

Mabel,

I would be inclined to consider it 4 occurrances. The process you have in place had 3 opportunities to catch the error. Your description indicates 4 separate people performing a part of the crossmatch and issue. At each of these "interactions" something failed allowing an error to pass through unnoticed. Here we would be performing an analysis on each part of the process to determine if there is a process failure with the potential to repeat, and/or if some sort of process change is needed at any given step. It is much easier to perform root cause analysis on each part rather than to analyze the process as a whole and attempt to capture all of the possible interaction scenarios. It might be a good idea to consider the process as a whole as well to capture the "It never gets to this step and is still wrong thought process". Just my .02.

[John C. Staley]

Mabel, I'm with David, Deny and Phil. If I understand correctly, essentially the system broke down in 4 distinct spots, each of which should have identified the problem and each will most likely have a different reason as to why the mistake was not identified.

:eyepoppin

[Malcolm Needs ☆]

I'd try to hide it under the carpet for fear my Quality Department would get hold of it!!!!!!!!!!!!!!!! They're like bulldogs with a bone.

:devilish::devilish::haha::haha::surrender:surrender:surrender:surrender:surrender

[Mabel Adams]

Watch out Malcolm, they probably audit this site to see if there is someone that bears watching. Now you've let them know about your carpet.

[AMcCord]

That's not quite what I said.

We recently had a problem quite similar to Phil's. I'll bet he and I said very similar things! Sometimes I'd like to thump people with a large blunt object instead of counseling them...but that probably wouldn't be any more effective. You can have a great system in place, but if the people using it are not diligent at each and every step, each and every time, you system isn't going to work.

[Deny Morlino]

I interpret the "diligent" statement as the reason I am so anal with regard to the details here. No wonder I lose sleep when I am thinking about blood bank! Occupational hazard I guess.

[Eoin]

Hi Folks,

In our system this is one non-conformance. When looking at system analysis (or root cause/s if you prefer) you would discover all the errors (we would classify as"missed chances to rectify" on this one) and they would be separately listed and risk assessed, using a matrix - as you seem to be doing - to set risk. James T. Reason of the University of Manchester Swiss Cheese theory showing how the barriers line up and the mistake goes through is interesting. Often errors are a number of small errors combined to create a bigger problem. If there are a lot of errors like this, it may be worth doing a "FMEA" (failure mode error analysis) on the system to see how robust it is.

It never ceases to amaze me that just when you think a system is error proofed, along comes somebody to prove you wrong.

Cheers

Eoin

[John C. Staley]

This is part of the reason I am so opposed to overly complex systems. We seem to feel the with every problem we need to build another layer of protection when in fact that usually causes additional problems. Human nature tends to make us a little more reckless when we know there it a safety net below us. Imagine how having 6 or 7 additional safety nets below is going to bolster your confidence and reduce your concern! My mentor, probably the best blood banker I ever knew, encouraged me to adopt the KISS theory in all things. For those of you who a puzzled, that stands for:

Keep It Simple Stupid! Many of you have been on this forum long enough to have seen me post many times that complicating a process does not improve it. I stand by that philosophy and encourage others to at least consider it.

[Mabel Adams]

I agree, John. I hate adding another check or something to make someone feel safer when it just complicates things and doesn't really accomplish anything. But what do we do with people who may do a process only once every few months? How do you protect the process from new people and rusty people. They are the most likely to overlook some small thing that starts the holes in the Swiss cheese lining up. I'm pretty sure it doesn't help much to just tell people to 'be more careful'.

[Mabel Adams]

Eoin, I am somewhat familiar with all of the things you mention, but would like a reference that would really teach me how to use those tools. Can you recommend anything?

[John C. Staley]

I agree, John. I hate adding another check or something to make someone feel safer when it just complicates things and doesn't really accomplish anything. But what do we do with people who may do a process only once every few months? How do you protect the process from new people and rusty people. They are the most likely to overlook some small thing that starts the holes in the Swiss cheese lining up. I'm pretty sure it doesn't help much to just tell people to 'be more careful'.

Mabel, the only way to remove all human error is to remove all humans from the process. I realize that in our business any error is considered unacceptable but that does not mean they will not happen. All we can really hope for is to minimize the frequency the occur and the impact when they do. There will be the rare occasion that telling them to "be more careful" is really all you can do. I realize that this is an unpopular philosophy in our ranks but there really is only so much you can do as long as the human element is involve.

After reading Phil's last post again I think I could really like his system and would seriously reconsider my previous thought that treating this as 4 seperate errors would be the best approach.

[Liz]

How does the administration react? What is the disciplinary action??

Because if you have a system in place and all the checkpoints failed this is negligence not system failure! Do you suspend, official warning? treat equally in the action you take...?

[Dr. Pepper]

In my case, written warnings on the lab end (don't know what nursing did) and a root cause analysis from the hospital QA. No great revelation about preventing a repeat other than "pay attention". This was a great example of the swiss cheese incident model, where you line up a bunch of random slices of cheese and once in a while all the holes line up in a row. I go back to David's splendid observation earlier in the thread about complacency: people sometimes overlook things because "the numbers always match". Almost always.

By the way, I'll bid adieu to the BBT crew for a few weeks. I'm leaving for Zambia tomorrow to help put on a weeklong workshop on lab QA issues and then play tourist for a few days.

[Malcolm Needs ☆]

By the way, I'll bid adieu to the BBT crew for a few weeks. I'm leaving for Zambia tomorrow to help put on a weeklong workshop on lab QA issues and then play tourist for a few days.

Good luck with your excellent cause - an enjoy!

[DPruden]

I try to find electronic solutions to human error problems because the best trained most diligent tech will still once in a blue moon make a serious error. Could you make the armband number a required field that the tech could not get by without entering something? We are switching to the armbands with barcoded numbers because we were having clerical error while entering the number/letter combinations.

[Liz]

..... We are switching to the armbands with barcoded numbers because we were having clerical error while entering the number/letter combinations.

Of course you are having those problems, ... and hopefully the mechanical barrier will be smarter than them and not be by-passed :bonk:we are acquiring the hand held bedside barcode reader for transfusion... it shouts and gets out red flags if the unit is not meant for that patient.

[Barbarakym]

My error tracking system includes an estimate of how likely the problem is to recur and I find that I would give different estimates for different mistakes in the event. I find the likelihood measure useful in assessing risk but I don't really want to look at only how often all of the mistakes would line up together again. I use a homemade system so I can change it all I want.

I would like to know your system. Could you send me a note or form on how you do this? My email is kym993@gmail.com

Very appreciative. Second year as supervisior and there is nothing like that in place at this hospital. Now I do counsil on errors and keep record by tech but I don't really have a system check in writing? Only when I get several people who have the same reason for something do I look at the process. I would likea more formal review.