specimen identification problems

[umeshkumar]

Hi,

Our hospital does not have a HIS and LIS and documentation is all by manual method, in registers.

Our policy on acceptability of specimen for blood bank is as follows:

1. The 2 unique patient identifiers (Name and hospital number) should be matched on the request and specimen label.

2. The name should be complete 3 names (first name, middle name, last name)

Many times when there are minor errors in part of the name, the blood bank staff ask the nurse who brings the specimen to do the necessary correction and sign for that. Is this acceptable.

We have been telling them to be strictly reject specimen with specimen ID errors for the transfusion services.

The staff should retain the specimen and request and fill up the rejection form and send it back to the ward and to ask for a new sample and request. Is this correct?

How do you classify specimen identification errors and what action need be taken for each type of error? please answer this with respect to transfusion services.

thanks

[rravkin@aol.com]

Speaking from practicle experience, last time we did manual ID of specimen we had the phleb signature on the specimen as well as a witness, usually the patient RN or another RN; the witness would also recheck the data. As far as your practice with the RN coming to the BB to correct lable errors, I know in the past this was acceptable but I think currently BB practice is moving, or has moved away from this.

[Liz]

1. Many times when there are minor errors in part of the name, the blood bank staff ask the nurse who brings the specimen to do the necessary correction and sign for that. Is this acceptable.

No this is not acceptable.

2. The staff should retain the specimen and request and fill up the rejection form and send it back to the ward and to ask for a new sample and request. Is this correct?

Yes or you can call them and document that in your log.

3. How do you classify specimen identification errors and what action need be taken for each type of error?

You should have an "Incident Report" form with

Preanalytical

Analytical and

Postanalytical

And under each heading itemize what could happen. Then the tech just ticks that and writes the action taken.

This also gets signed by the supervisor.

[KKidd]

For us, the name must be letter perfect and the MR# number must be exact. All information is to be copied from the hospital armband. If a new sample cannot be obtained, the patient can receive group O, uncrossmatched blood until a new sample has been processed. We do not allow corrections to be made on the label. We also have a form similar to what Liz described.

[tbostock]

We do not allow corrections to any identifiers (name, DOB, MR#). If the date, time, or initials are missing, we do allow them to come and correct or add them onto the tube, as long as it is the person who collected it.

[cthherbal ☆]

We do not allow corrections to any identifiers (name, DOB, MR#). If the date, time, or initials are missing, we do allow them to come and correct or add them onto the tube, as long as it is the person who collected it.

Ours is similar to Terry's: Slip and Sample sent together. If one or the other is missing date, time or initials, we allow the other to be corrected (policy clearly states this). If they are missing on both, then no dice, we ask for a new specimen.

We do not allow patient identifiers to be corrected after the fact. We would ask for a new specimen.

Armbands: Hospital armband is used for patient identification. A separate Blood Band is used for Blood Identification (unique #).

[BANKTECH]

We do the same as Terri

[kate murphy]

We do the same, even though we are all computerized, electronic med record, auto-generated printed labels - no additions, corrections, deletions once the spec is in our paws. Spec/requisition must match.

Two signatures - person drawing, person verifying identity of patient.

One question, why do you insist on the middle name? We use Last, First name and MRN.

[Liz]

We use triple name too, i believe it is a joint commission or CAP or both requirement.

[Eoin]

Hi umeshkumar,

Liz succinctly listed the answers above and I concur entirely. You should definitely not ask anyone delivering the specimen to the lab to change anything, and they would be stupid to do so. They are opening themselves up to any subsequent litigation should there be any adverse outcome.

Rejection of specimens should improve practice, but you need to keep the metrics on this (using Incident Forms or Specimen/ Forms Issues Register as described). If it is currently at an unacceptable level, you need re-education of staff taking the samples now.

I know it is tempting for expediency sake to ignore minor errors, but then the line cannot be drawn. It has to be a zero tolerance policy to errors in any phase (pre-analytical as here).This issue is irrespective of manual system or HIS/LIS issues.

Good Luck with it.

Cheers

Eoin

Edited September 28, 2012 by Eoin
Acceptable to unacceptable level

[umeshkumar]

thanks Liz. I should have asked (rather I meant) how do you classify specimen ID errors based on the nature of the error (minor, major etc). For minor errors, can correction be allowed and so on.

[Liz]

Any labelling error reflects possible patient misidentification and a new sample is requested. We do not, others may, differentiate betwen errors so there is no minor error. Any negligence in one part can reflect a major patient mis identification. It is very strict because Patient misidentification can obviously be fatal.

We are strict and do not make exceptions, we do not work on a sample that is not 100%.

[drwajiha]

Hi umeshkumar,

Our hospital has a clear policy regarding this issue. The blood transfusion request or blood group request must have :

patient's three names on the request and sample tube.

medical record number

name and signature of the patient identifier and person who extracted sample on the request and sample tube.

date and time of extraction on sample tube

if these criteria are not fulfilled the sample is rejected.

Monthly a transfusion / blood group request monitoring report is generated and submitted to nursing director's office with copies of deficient reports.

Departmental incident report is generated for any technologist who processes any request which should have been rejected.

This has resulted in solving the problem in our hospital .

we in blood bank are dependent on the ward staff to collect from correctly identified patient. Deficiently labelled sample indicates a careless attitude of the collecting staff so we do not allow the nursing staff to come and do any changes in data under routine circumstances. Under emergency if it needs to be done the nursing staff is required to write that they take the responsibility of correctly identifying the patient . Since blood bank is a medico-legal record. nobody likes to give in writing their mistake rather they try to rectify it.

I hope this will help you have ideas to solve your problem according to your circumstances.

[kpend]

Can anyone document a reference supporting the process of two signatures...the collector and a second identifying person?

[Abdulhameed Al-Attas]

The two signatures are as follow:

1- Patient identified and blood taken by and 2- Patient and blood sample identity checked and verified by.

The supporting reference are in CAP - Transfusion Medicine Checklist.

[Auntie-D]

Why does no one list the DOB as a prime identifier? And what about people like me who don't have a middle name?

[carolyn swickard]

In our case - we never let the original specimen out of our hands once received. We allow corrections only for date/time and initials and only from the original phlebotomist - lab or RN - and they must come to the Blood Bank to do corrections. Never let them have the original specimen back! Any errors on the name or the indentifying hospital number (not the DOB), require a redraw. We tally all of this and write incident reports on it all and work with the RN educators, but still always seem to have about the same number of errors and the same type of errors each month. I really don't know why they can't transcribe the data correctly from the pt wristband, but it seems to be an intractable problem. Write your SOP to clearly reflect all of this and your techs can use it to back them up when RN and/or Drs (and our own phlebs - though them we can be tougher on!) do not want to redraw because of a careless mistake.

[aafrin]

1. Many times when there are minor errors in part of the name, the blood bank staff ask the nurse who brings the specimen to do the necessary correction and sign for that. Is this acceptable.

No this is not acceptable.

2. The staff should retain the specimen and request and fill up the rejection form and send it back to the ward and to ask for a new sample and request. Is this correct?

Yes or you can call them and document that in your log.

3. How do you classify specimen identification errors and what action need be taken for each type of error?

You should have an "Incident Report" form with

Preanalytical

Analytical and

Postanalytical

And under each heading itemize what could happen. Then the tech just ticks that and writes the action taken.

This also gets signed by the supervisor.

We also have a policy of zero tolerance as regards the full name of patient on the tube label as well as the form, but like cswickard has stated above we really don’t understand why the nurses can’t transcribe patient name correctly on the tube label. At our hospital, pre-printed, self-adhesive, patient demographic detail labels are already available in the patient file. The nurses just have to stick one label on the form, tally it with the wrist band and then transcribe the full name on the label, but even then they misspell the name. Not only is it annoying & frustrating but also results in many patients being sticked again for samples. We have a SOP and our Director – Medical Services backs us, so redraws are done, albeit grudgingly.

I have a question - what do you all do when a cord blood sample is misspelt?

In this case, we allow the nursing staff to make correction on tube, record it on the “BLOOD BANK MISLABELED SPECIMEN CHANGE REPORT LOG” and take their signature on it, so that they take full responsibility for it.

@LIZ, I am interested in the “Incident Report Form” you have mentioned. Could you please share it here with us? Thanks in advance.

[carolyn swickard]

I have a question - what do you all do when a cord blood sample is misspelt?

In this case, we allow the nursing staff to make correction on tube, record it on the “BLOOD BANK MISLABELED SPECIMEN CHANGE REPORT LOG” and take their signature on it, so that they take full responsibility for it. *****quote*****

In our case, with cord bloods, we make them label them with Mom's labels 1st because that is the only label available at the time of birth, then when the baby's labels are available from admissions (usually 1-3 hours later) - they have to match the correct Mom's label with the correct and matching baby's labels - no spelling involved here with preprinted labels. However - if Mom is missing (meaning the specimen sat around for some time without labels) and/or if Mom and Baby's names don't match - we require a heelstck redraw of baby. Only if (very rare) the mismatched names are the correct legal situation do we accept mismatched names - that is not happening much anymore as they now only allow 1 name assignment to the baby during one visit - lots of baby name changes around here.

[Malcolm Needs ☆]

In our case, with cord bloods, we make them label them with Mom's labels 1st because that is the only label available at the time of birth,

Surely, the sex must be known, in which case, in 50% of the time (give or take), the label can say MALE.