CAP Inspection Question TRM.30550

[Mary]

A new question dated 12/29/04 states"Does the facility have a documented program to ensure that the risk of pretansfusion sample misidentification is monitored and subjected to continual process improvement?" The NOTE goes on to say:Approximately 1 in 1000 specimens for pretransfusion testing are mislabeled. Multiple methods exist to reduce incidence including 1) documenting the ABO on a second sample collected at a seperate phlebotomy; 2)identification by a mechanical or electronic identification sysem; 3)issuing only Group O, 4) documentation of ABO from previous records. Some of these are not practical, especially if you are not using the computer crossmatch and some would be impossible in an emergency. We have always used a seond ID band and limited the procurement of the specimens only to those who have bee trained properly.

I would be interested in hearing the opinions of others.

Mary 7/26/05

[David]

I spoke with CAP a few minutes ago about this very question, among others. Their explanation is that the items in the note are not the only solutions, just some suggestions. The intent of TRM.30550 is that you have some system to ensure correct labeling, and that you monitor how well you do. We use a unique armband, similar to Typenex, and track error rates. The CAP representative said that would meet this requirement.

[Mary]

Thanks, I was also going to call them.

[Jane]

David,

I was wondering how you track your armband error rates- do you just record any errors that you see in blood bank or do you go on the floors and audit to make sure the right band is on the right patient, etc.?

We use typhenex and hardly ever see any problems since our phlebotomists draw these.

[SantaFe Jane]

Mary,

My take on that question is that a legitimate "mechanical" method is the use of a separate bb armband. Monitoring the process would be having a system in place that will allow others to notify a supervisor when a mislabeling incident occurs. A method I'm familiar with is one where we filled out "Problem Identification Reports" which our supervisor kept in a file to prove that the issue was being monitored. Written in our labeling procedure was the part about how and when phlebotomists were trained. THe new HCLL software from Mediware allows tracking right at the screen when a specimen is marked as collected. THis becomes a great reporting tool for an inspector who can pull up a history of mislabeled specimens. Hope that helps.

[John C. Staley]

We utilize a single armband system (Biologics). About once a year our lab Qa folks undertake an audit by visiting every patient in the hospital once per week to make sure they have an armband attached. They aren't here this early so I don't have access to the data but when I get it I'll share it with you. I do remember that it got a lot better after the first couple of audits showed administration how poorly we were doing and the nurses were intensively retrained. From a patient ID standpoint the most dangerous situation in the hospital is a nurse with a pair of scissors.

[David]

David,

I was wondering how you track your armband error rates- do you just record any errors that you see in blood bank or do you go on the floors and audit to make sure the right band is on the right patient, etc.?

We use typhenex and hardly ever see any problems since our phlebotomists draw these.

Currently we are just tracking the errors that appear in BB. We have a fairly large number of our specimens drawn by nursing, and about half the errors are theirs. Thanks to the JCAHO emphasis on patient identification and our impending visit from them, we are getting unprecedented attention to this issue from nursing. The next step will be going to patient rooms and see what's really on the body.

[David Saikin]

I am curious - if half your error rates are due to nursing, who is responsible for the other half? And more importantly what have you done about it? I find it almost incomprehensible that samples are mislabeled by lab personnel. My personal feelings are that drawing the incorrect patient is grounds for dismissal (but I seem to be a dying breed in that respect).

[David]

To us "mislabeled" means everything from no time of draw to misspelled names to incomplete collector information. We are long past getting WBIT's from anyone. In the past 5 years in this position I can recall that happening once (nurse draw), and we handle 70-80 specimens per day. The phlebotmist's 'error' rates are posted monthly and their supervisor handles it as part of their performance review. And, yes, we would dismiss them if they put the wrong blood in a tube.

[Linda McWilliams]

Called CAP today and they agreed that what we are doing at my facility is okay- we retype (repeat forward type) with every specimen.

The key word they said is "may" have to use mechanical / barcode methods.

This is not required yet

[QDeb]

Hello to all,

The hospitals in our system that use a independent "Blood Bank" bracelet in addition to the hospital patient ID band, are required to retest the front type of the sample, preferable by a different individual. If there is not a second or "independent BB bracelet" for ID then a second specimen, collected at a different time from the initial specimen is used to perform a blood type to validate the initial blood type. If the patient has historical results or has an autologous unit that can be used to validate the current results then that serves as the second check. This is our current process. Monitoring of the process has been in effect for a long time. The processing (specimen receipt) area checks specimen labels to be sure that the information in complete, if incomplete it is rejected and is counted as a mislabeled specimen. This data is tallied and monitored (lab services - not specific to Transfusion Services). Other ways to suggest mislabeled specimens: receipt of duplicate specimens (received close together), receipt of specimens for testing when blood is already set, receipt of specimens with a handwritten label (when there is no problem with the computer/printers), Testing - current results don't match historical results, testing requested when not indicated (i.e. immune globulin screen). The incidents when the mislabeled specimens actually reach the Transfusion Services and are determined to be wrong patient (not clerical - missing/wrong information) are presented to the Hospital Quality Executive meeting (Senior management/board member level) on a quarterly basis (they expect action from those areas where the incident occurred)... we are currently working on a barcoded system for Patient Identification.

Remember they are looking for a process to prevent, catch and monitor, and show actions for improvement and then evaluate the effectiveness of that action regarding mislabeled incidents. Look to see what you are already doing and write it down. I am sure that you have procedures for specimen collection and acceptance, you have training, you have competencies, you have performance evaluations, you have unacceptable specimens (from time to time), and you may even find one or two of those specimens that were collected from the wrong patient (I'm sure that those get quite a bit of attention) - how do you deal with them?

[conwaysbb]

The key here is the original question "Does the facility have a documented program to ensure that the risk of pretansfusion sample misidentification is monitored and subjected to continual process improvement". CAP then gives us examples. Yes, these are only examples and are are not intended to describe all the possibilities that are out there or that we are already doing ourselves!!! This does not require that you repeat the specimen in duplicate, request a second specimen on first time patients or that you have a system in place like the Typenex system!!!

Our facility has changed the process from a laboratory based phlebotomy team to a nursing based, floor oriented collection process. This has taken the ability to discipline individuals out of our hands and into nursings's hand. As part of our review process for correct labeling we have zero tolerence for errors in specimen labeling. The only 2 individuals that can make exceptions are the manager or the Blood Bank Director and I belive only 3 specimens were allowed to be accepted with minor errors in 2005 so far and these were for date or times missing on either the specimen label or confirmation slip.

As CAP requires the full name of patient, identification # (we use Medical record #), identification of individual drawing specimens, date of collection and time of collection (AABB does not require time) we require this on both the specimen label and confirmation slip which accompanies each specimen. Our laboratory computer system uses very small bar coded labels which often truncate names so we do not allow this type of label on blood bank specimens. However, our hospital information system produces a longer bar coded label which was validated to be effective for names up to 62 characters long. This label was created for Pharmacy use because of addressograph problems in pharmacy orders (i.e. cutting off names, MR# and legibility of info). We just piggybacked on this label use as these were the exact same issues we were having. In fact most of our rejections are because the floors are using the laboratory computer bar coded label instead of Hospital computer bar coded label.

We audit the receipt and review of specimens prospectively as all specimens are manually logged on an audit sheet. when logged we document whether specimen was eaccepted or not accepted, reason why specimen was not accepted, date of receipt, shift received and location specimen was drawn. We use this log to also monitor non labeling rejections, i.e hemolysis, contamination, wrong tube drawn, ect.... This information is tabulated monthly and specimen error rates are reported to Laboratory PI committee by both labeling error and non-labeling errors broken down by floors. These files are e-mailed to all nurse managers, nurse directors and the VP-for nursing. Usually when there is an increase in unacceptable specimens from a particular floor, the nurse manager asks us what shift it happens and we can give the nurse manager this information. If during our testing we identify a specimen that has a different blood type from our historical record, this information is sent to the nurse manager immediately for follow-up. We also check to see if any orders were received in Chemistry/hematology . If so we do a blood type on those specimens to see if they to were drawn from wrong patient.

The point of this information is that we do not retype all specimens, we do not require 2 different specimens before transfuing patient, and we do not require a second identification band, but what we are doing meets the INTENT of this requirement.

I believe the CAP made an error when writing their NOTE and having just those particular examples. I know I have been called by 2 other BB supervisors from smaller facilities in a panic because they thought they now needed to draw two separate samples from patient prior to transfusion.

[Eagle Eye]

No time required on the specimen label. CAP made that change in october 2005.

REVISED Checklist Questions Effective 10/06/2005

TRM.40230 Phase II

Are all blood samples used for compatibility testing labeled at the time of specimen collection with the patient's first and last name, unique identification number, and the date of collection?

[Ardele]

Re-typing from the same tube is just going to verify an error - if an error occured. You will not prevent "wrong blood in tube" At my facility, we are having two people do patient verification and both sets of mnemonics on the tube and in the LIS. We 'ran' this by CAP for this TRM.30550 and received their okay in meeting this requirement. Two people signing the tube may be two phlebotomists, a phlebotomist and nurse, a nurse and anethesiologist from the OR, etc. We have had good compliance with the two signatures, especially after rejecting a specimen without 2 sets of mnemonics.

[AUDREY]

We utilize a single armband system (Biologics). About once a year our lab Qa folks undertake an audit by visiting every patient in the hospital once per week to make sure they have an armband attached. They aren't here this early so I don't have access to the data but when I get it I'll share it with you. I do remember that it got a lot better after the first couple of audits showed administration how poorly we were doing and the nurses were intensively retrained. From a patient ID standpoint the most dangerous situation in the hospital is a nurse with a pair of scissors.

Truer words were never spoken!!

[estiner]

I would be careful about "retyping the same specimen" as fulfilling this requirement. A neighbor hospital just got sighted for doing this very thing. And they used a separate BB armband. Both practices were deemed to not fulfill this requirement.

I feel the safest practice is to verify the blood type with a history or a 2nd specimen from a different venipuncture or else give type O until you can confirm the type. We have had this policy in effect for almost 2 years now and it is not as cumbersome as many thought it would be. We are a 210-bed hospital.

[Joanne P. Scannell]

I disagree ... and that citation was short-minded, not addressing the actual intent of the requirement.

Again, the requirment is for a PROCESS ... if it works for you and you don't mind all the doubted specimens/mislabeled specimens/questionable hospital bands/extra footwork/extra sampling/delays/questionable patient identification at transfusion time, then I can't say anything about your process if I'm inspecting you, can I?

Time will tell which system works ...

[Mary**]

:DI agree with you. You can inspect us anytime!