Protocols for 'identifying' unidentified trauma patients who need emergency transfusion

Page 2

ADDENDA Oct. 29, 2003

6. A colleague at a community hospital in Southern California reports that when blood is needed in an emergency before they have time to complete pretransfusion testing, they issue uncrossmatched group O Rh negative RBCs. A 'red tag' is placed on the uncrossmatched unit with the following information: 'UNCROSSMATCHED BLOOD FOR, the name of the patient, the patient identification number, patient location, unit number, initials of the CLS and the date the unit was dispensed'. They also fill out an Emergency Release form that has the patient's name, patient's identification number, unit number ABO/Rh, and product expiration date. This form is taken along with the blood to the patient's location. There is a space on the Emergency Release Form for the physician to certify that he/she is aware of the risks involved and that the condition of the patient is sufficiently urgent to warrant the release of uncrossmatched blood. The physician signs this statement and a copy is returned to the Transfusion Service. Once they obtain a sample and perform the compatibility test they generate a Transfusion Record Form. Documented on the Transfusion Record Form is a comment indicating that the unit was released uncrossmatched. Their practice is to forward the Transfusion Record Form to the nursing unit after the crossmatch is completed and have them enter the date/time of transfusion, person starting and ending the transfusion and time. This is done after the emergency situation is over. This is where they occasionally have a problem. They get complaints that it is too difficult to fill out the Transfusion Record Form after this point in time in their process.

ADDENDA Nov. 7, 2003

7. The colleague who submitted the original question for this discussion reports that after reviewing the input of colleagues, in his large trauma hospital, they have been able to convince the administration to leave on the original armband, with the “Statpack” number and the original Medical Record number. When the patient’s identity is determined, a new armband is issued, but the original emergency-issue Medical Record number is unchanged. The historical record number is communicated to radiology, blood bank, etc. to allow correlation with historical results, but the records are not merged with the old historical record number until after discharge. The reporting colleague feels that if these policies are followed, the release of blood to the correct patient will go much more smoothly. They greatly appreciate the excellent ideas and approaches contributed by the members of the CBBS e-Network Forum.

ADDENDA Mar. 8, 2005

8. A Transfusion Service Supervisor in Northern California reports that her hospital treats many patients who are admitted to their Emergency Department without knowing the patient's true identity or who need treatment so urgently that they are assigned a pre-prepared "Doe" identity. The specimen drawn for type and screen or type and crossmatch is labeled with the Doe name, a unique sequence of letters specific to Doe patients, and a unique hospital medical record number, and a bogus date of birth. When blood components are dispensed to these patients, they require the blood order form to exactly match this Doe identification. She comments that their system has worked well for many years. The patient is identified as soon as possible, and the computer system updated. However, now that more and more slips are computer-generated, it is difficult for them to get the blood order forms to exactly match the specimen label. Their department has an internal policy of giving only group O RBC donor units if the identifications do not exactly match; this is making a drain on their group 'O' RBC units.
   
   She wants to know what is the opinion of e-Network Forum colleagues in the following scenario: If the specimen is labeled: Doe, John ZY, medical record number 1234567, date of birth 01/01/1888 and the unit is requested for Peterson, Henry 9ZY, aka: Doe, John ZY, mr# 1234567, dob 5/16/47, would anyone dispense for the patient an ABO/Rh group specific RBC unit? If not, why, and what would you do instead?

ADDENDA Mar. 9, 2005

9. A colleague in Massachusetts reports that their Emergency Department has the same system as that described by the Transfusion Service Supervisor in Northern California (posting #8, above). Once the patient's true identity is determined the name is changed in the hospital computer system. If the patient had a previous record at her hospital, the MRN will be changed as well. The previous name and MRN are listed as "AKA" in the computer. As long as either the real name/MRN or the AKA name/MRN matches what they have on the tube, they will give ABO/Rh matched RBCs.
   
   
10. A colleague in Seattle believes that the practice by the Northern California hospital (#8, above) of entering a "bogus" date of birth into the computer system when treating a patient whose identity is unknown creates a system problem, and suggests that the use of a "bogus" date of birth be eliminated. She makes this recommendation not only because the updated date of birth may print in the 'AKA' of the requisition, but because she can not see any benefit in entering false information into a data field as a fourth identifier (i.e., in addition to name, medical record number and unique "Doe" letters). She points out that at present (although this might change in the future) JCAHO only requires 2 unique identifiers.
    
    Also, is there a reason that the ZY is changed to 9ZY? She would try to keep this consistent, unless it is not an identifier (e.g., room number). With the initiation of computerized requisitions, she thinks it is time to change the policy for "Doe" patients to avoid discrepancies in patient identification.
    
    
11. A colleague at a large transfusion service in the southeast USA (which employs a Cerner computer system and services a Level 1 Trauma Center) reports that their approach to identifying individuals whose true identity is unknown is to use specimen labels containing 5 different identifiers. The identifiers are:
    • "Trauma" name (a color)
    • "T-111111" number
    • Medical Record Number
    • Financial number and
    • Encounter number.
    When the patient is finally identified positively, the Trauma name is changed and the "T" number is eliminated. The medical record number may change, if the patient was previously a patient in their institution. However, the financial number and the encounter number DO NOT CHANGE and remain the same until the patient has been discharged. When the patient is identified, a new label is printed which will have at a minimum, TWO of the five original identifiers on it. These TWO identifiers do not and will not change, during the remaining hospitalization. The new label is placed on the original sample, so that the original trauma name, MR#, and Financial number are still visible. Thus, ALL Trauma samples will have two labels on them, which are the original label and the second label with the known identity on it. Each label will have a MINIMUM OF TWO PERMANENT IDENTIFIERS on them, which are identical and will not change. This gives positive identification of both the original sample and the patient, so that blood products may be issued continuously for the first 72 hours.

ADDENDA Apr. 12, 2005

12. Editor's note: The discussion in the publication by Weiskopf and colleagues may be of interest to colleagues concerned about rapid provision of blood in an emergency situation.

Page 1

Submit comments to the e-Network Forum at enetworkforum@cbbsweb.org

Ira A. Shulman, MD
CBBS e-Network Forum Senior Editor & Moderator

W. Tait Stevens, MD
CBBS e-Network Forum Editor & Moderator

Elizabeth M. St. Lezin, MD
CBBS e-Network Forum Associate Editor & Moderator