Engineering controls/barriers to prevent errors

[kirkaw]

I have been reading about error prevention and creating an environment conducive to risk assessment and self reporting. Our institution uses the Just Culture algorithm. I am curious as to how many others are using this. One of the premises of this program is that 'to err is human' and that to expect perfection is unrealistic. However, it is incumbent upon management to create a system that is as fail-safe as possible. So where is the line between personal responsibility and system failure? 2 situations that I have encountered recently were: 1) expired antisera was used and 2) no crossmatch was done subsequent to an emergency release. The only way this is caught is on work review. How far do folks go to try to put stop-gap measures in place to prevent errors? The form error has only happened once in 2 and half years but the 2nd one has happened 2 or 3 times for sure.

Thanks for your input!

[Dansket]

With implementation of an automated testing instrument, I am no longer concerned with use of expired reagents.  Using a blood bank computer system, issue of uncrossmatched blood puts the specimen on the pending list which is required to be checked by staff begin/end of each work shift.

[AMcCord]

How often does your blood bank release uncrossmatched blood - infrequently or frequently? Is it the same tech or different techs making the error? If it's the same tech and has happened more than once, I would say it might be a training issue or a disciplinary issue. If it's infrequent and different techs, it's most likely a training issue. We struggle with this because it doesn't happen often enough for any one tech to become truly familiar with the process - definitely a training issue for us and we are going to start training multiple times a year. 

[kirkaw]

Thank you AMcCord. You are correct in that we issue uncrossmatched blood <1x/month. It is different techs that are forgetting the subsequent crossmatch. I did an 'emergency release' training session including contest a year and a half ago and it went really well. The actual emergency release process is working much more smoothly. But the follow-up is still lacking.

 

Dansket, is every test you do automated? The expired reagent we used was 'rare' antisera so we still do that in tube.

[David Saikin]

Unless you have a policy forbidding it, outdated rare antisera may be used as long as controls are run and valid results obtained.  If you are concerned about your rares and outdates you would need to develop a process on review of dating and removal/replacement of such when the time comes.  That kind of thing happens, esp if you are dealing with generalists - if it is in the refrig it must be ok to use (because who ever is running the BB ALWAYS takes care of that kind of thing - is the attitude I see).

[AMcCord]

Unless you have a policy forbidding it, outdated rare antisera may be used as long as controls are run and valid results obtained.  If you are concerned about your rares and outdates you would need to develop a process on review of dating and removal/replacement of such when the time comes.  That kind of thing happens, esp if you are dealing with generalists - if it is in the refrig it must be ok to use (because who ever is running the BB ALWAYS takes care of that kind of thing - is the attitude I see).

 

It certainly does seem like there is an expectation that somebody else will always be Mom (or the maid). I've been trying to cure them of that for years with limited success.

 

When it's critical, as David says, you have to come up with a policy/process to prevent those kinds of problems. Generalists have to know a lot about a lot of things and it's hard to remember everything, especially when they don't work in an area often. There is slippage. The system has to provide a safety net to catch the slips you can anticipate.

[R1R2]

I appointed a lab associate to "reagent czar" after too many incidents of no indate antisera available.  He checks the fridge regularly and knows when rare antisera is expiring and orders more when necessary. 

[KKidd]

It certainly does seem like there is an expectation that somebody else will always be Mom (or the maid). I've been trying to cure them of that for years with limited success.

 

When it's critical, as David says, you have to come up with a policy/process to prevent those kinds of problems. Generalists have to know a lot about a lot of things and it's hard to remember everything, especially when they don't work in an area often. There is slippage. The system has to provide a safety net to catch the slips you can anticipate.

 And I thought I was the only one who called herself Mom ( her techs children at times) .  When we perform antigen typing, the lot number and exp. date must be recorded on a worksheet.  Also, once a month rounds are performed in each department.  One item on the checklist is to make sure that no outdated reagents are available for use. 

As for the emergency release of blood, I have had several situations recently when the patient was transported before a blood sample was obtained.  I docimented this information in the computer.

[Auntie-D]

We get round the retrospective crossmatch by removing some tag line and plasing it in a coombs tube labelled with the unit number. We wand the bag and generate a barcode label that the coombs tube is labelled with. If a retrospective xm is needed we can just snip the tag line into the barcoded tubes and pop it on the analyser.

[DPruden]

If you use a paper log for emergency uncrossmatched blood, you could add a column for 'crossmatch completed date/time' or something to that effect.  If you are using your computer system to perform the emergency release, I would see if there is a way that the computer can order the crossmatch, some systems can do this easily.

[tbostock]

Every error needs to be evaluated in order to prevent it in the future.  You can only hold the tech responsible if the procedure/process is solid and:

the same tech keeps making the same type of careless errors

there is a deliberate attempt to NOT follow proper policy

 

Without the use of computer systems to prevent such types of errors listed above, you have to create a manual workaround.

[AMcCord]

Terri is abolutely correct. I find that a good worksheet (we do not have a computer) helps completion of less common processes like emergency release and antigen typing. If there is a blank to fill in, they are much more likely to document completely than if they are just supposed to remember. For antigen typing there are blanks for reagent lot#, OD, manufacturer, there are blanks for QC results, there are blanks for patient/donor test results and check cell results, blanks for the required patient identifiers - whatever it takes.

 

We do use the LAB LIS to generate reports for blood bank orders and for product pickup slips (beats a typewriter  ). For down time, I've got a form that covers all the necessary information. If we need to do an emergency release and the patient is not registered in the HIS or if there is not enough time to use the LIS report, we can use the down time slip for that, too, because it includes blanks for the info we need for that process.

 

Sometimes it makes me crazy that the generalist don't remember things that I think they should, but I try to be a realist (which is hard for an anal retentive blood banker). Instead of going postal, I work on a form and a process map and that helps keep them on track.

[John C. Staley]

One more suggestion.  Get the staff involved in coming up with a solution.  Oddly enough they can come up with some amazing ideas and when it is their idea the "buy in" is almost automatic.  No one wants to make a mistake, especially in Blood Bank and they will get very inventive when coming up with ways to keep themselves from making mistakes.  BUT, keep in mind, complicating a process never, ever made it better.  

 

There, enough philosophical drivel for one evening.   

[pstruik]

John is so right about not adding complexity. In our laboratory's MHRA inspection the poor man tasked with rating our efforts at quality having spent a lifetime in the business kept returning to his mantra - keep it simple.

His experience allowed him to cut through our layers of complications to find the necessary instructions so having an outside and knowledgeable being to cast their eyes over your end product is also helpful.